Banat and Baka: Geography, History and Ethnographic Cartography

The present thesis sets out to examine the ways in which areas of complex populations are illustrated by thematic, or ethnographic, maps, and to draw attention to some of the techniques by which these maps achieve their effects. It focuses primarily on ethnic cartography as practised in two adjacent territories on the northern edge of the Balkans the Banat and Ba!ka.

It will trace the gradual improvement in technique until the mid "#th century, and argue that some of these advances have been forgotten, so that the recent revival of the discipline has been characterised by maps of a technically inferior standard to some of those produced more than half a century ago.

It is felt however, that an appreciation of the ethnic cartography of the regions will be impossible without a good grounding in their geographical and historical background. The scene is therefore first set by a discussion of the geographical and climactic characteristics of the two territories, and a summary of the history that has been played out here, from the medieval $ungarian kingdom to recent days. This is intended to create a rich backdrop against which some of the disputes may be more clearly understood.

The

next

section

examines

the

development

of

ethnographic

thematic

cartography within general cartography and its growth, particularly within %entral and &astern &urope as a political tool. The main methodologies will be described and evaluated in turn. 'n examination of the application of the various methodologies to the Banat and Ba!ka follows. ' representative selection of mainly (erman, $ungarian and )omanian maps of various scales are referred to and examined for evidence of tendentiousness in their design. *

There follows a discussion of local toponymy and an introduction to the substantial +lace ,ame )egister, which forms a supplement to this work. The (I-, which has been developed as a part of this research, and is the source for many of the maps herein, is not included in this presentation, but its nature, limitations and future potential are discussed.

GEOGRAPHICAL AN

ECONO!IC IN"L#ENCE$ ON BA&'A(

$E%%LE!EN% IN %HE BANA% AN

)*nction and o+er,ap

The Banat and and Ba!ka .$ungarian, Bnt, Bcska/ (erman, Banat, Batschka) are two adjoining but separate territories on the fringe of south0central &urope which have, for most of the twentieth century, been shared by )omania, 1ugoslavia and $ungary. Their geographical boundaries are almost entirely natural, comprising principally of rivers, but have frequently been in conflict with their administrative boundaries. %onsequently, a range of territorial configurations and alternative names are recognised, reflecting the historical ebb and flow of political power in the region. Indeed, the interest of the Banat and Ba!ka resides substantially in their being located at the junction and overlap of a medley of different types of frontiers a fact, which has contributed to their fluid and colourful histories. It has been conventional to treat the "

Banat and and Ba!ka before *2"# as discrete units in their own right. however, much of the literature passes over or obscures the point that these territories have but rarely corresponded to political units, even when entirely within the frontiers of a single state. Between *34" and *2*2, for instance, when both regions were wholly inside $ungary, there was indeed an exact correspondence between the megye .county5 of B6cs0Bodrog and the B6cska. There was also an exact match between Toront6l, Temes and 7rass80-9:r;ny megyek .pl.5 and the geographical Banat, but the three did not have any special collective status .see map "5. Both the Banat and Ba!ka have more usually been broken up between smaller units, or been incorporated into larger wholes. &ach territory, however, has a particular history, and the sources conventionally identify each of the pair as discrete units and distinct from each other.

<

='+ %>,T&?T.gsm

=ap Trianon

Na-ing Con+entions in the %e.t

Banat, Baka

$ereinafter, except where specified, the term Banat shall be understood to be coterminous with the whole territory of Temes, Toront6l and 7rass80-9:r;ny vrmegyek .counties5, as defined in *2*#. +ost0Trianon, the three parts shall be referred to as the Romanian Banat, Yugoslav Banat, and Hungarian Banat respectively. The term Baka .-erbo0%roat .Batin5 spelling5 is used in this dissertation as a general, non0political term for the southern termination of the Canube 0 Tisa corridor. Dith $ungarian spelling .Bcska5 it will be used to refer to the historical $ungarian counties of B6cs and Bodrog, and their combination, B6cs0 Bodrog, until Trianon/ and the reconstituted county of *2@*0@@. The Hungarian Bcska shall mean the part of the B6cska that remained with $ungary after Trianon. The Yugoslav Baka will be used to refer to the section of the B6cska detached from $ungary by the Treaty of Trianon .i.e. that portion of the Vojvodina located between the Canube and Tisa.

/o0+odina

The term Vojvodina, except where specifically noted, will be used to cover the 1ugoslav Ba!ka and the 1ugoslav Banat jointly, but should not be taken automatically to include the Srem district .-rijem, -9er;ms;g, -yrmien, -irmium5, which has since the -econd Dorld Dar formed a part of the 1ugoslav political unit, nor the Baranya triangle, which was only associated with the Banat and Ba!ka between the world wars.

P,ace1na-es +lace names will, as far as possible, be rendered in the administrative language of the day, unless a standard form exists in the &nglish language. The capital of the Banat, whose multiple name forms barely E

differ, will be referred to by a compromise standard, Temesvar. >therwise we would have to refer to it as Temesv6r until the >ttoman conquest, TimiFvar under Turkish administration, Temeswar between *4*E and the incorporation of the Banat into $ungary .*4E35. De would then revert to Temesv6r until *2"# except during the existence of the Woj odscha!t Ser"iens und #emeser Banat .Temeswar5. 'fter Trianon .*2"#5 the )omanian version, TimiFoara, would prevail. In other cases and where appropriate, a name will be accompanied by other0language forms on its first appearance.

Dhere -tandard &nglish forms are available, river names will appear without diacritics. I shall refer throughout to the Canube .Cunav .G5, Cunarea .)5, Cuna .=5, Conau .C55/ the -ava .-96va .=5, -au .C55 and Crava .Cr6va .=5, Crau .C55/ the Tisa .Tis9a .=5, TheiH .C55, the =aros .=ureI .)5, =ariJ .G5, =arosch .C55, the Temes .TimiI .)5, TimiJ .G5, Temesch .C55/ and the Bega .Begej .15 B;ga .$55. =ore minor rivers, and other physical geographical features .mountains, marshes etc.5 shall be referred to, so far as this is appropriate, in the majority language of the country in which the feature is located today.

Basic Physica, Geography

Bo*nds and Area

The Banat is a conventional label for a territorial entity whose existence, despite a remarkably chequered administrative history, is acknowledged by -erb, )omanian, (erman and $ungarian, and there is close agreement on its geographical bounds. The Canube, Tisa and =aros embrace the Banat on three sides. Civided from the Ba!ka on the west by the southward flow of the Tisa .$ungarian, Tis9a5, the Banat is separated KnaturallyL from -rem and -erbia, and Dallachia by the broad stripe of the Canube as far as >rFova. Mrom here, the BanatLs boundary runs north0eastward along the %erna valley, then among the +oiana0 4

)usca mountains towards the Transylvanian Iron (ates +ass near Bistra and thereafter northwards to +ojoga on the =aros. The BanatLs northern border follows the river =aros from +ojoga downstream to 'rad and on to its confluence with the Tisa by -9eged. Thus defined by Tisa, Canube, %erna and =aros rivers, and +oiana0)usca =ountains, the Banat covers "3,A## km", an area slightly smaller than Belgium .<#,A*# km.5.

The Ba!ka abuts on the western border of the Banat, from just south of -9eged to the TisaLs confluence with the Canube. The Ba!kaLs southern and western boundaries follow the Canube upstream from here to the confluence with the Crava. Npstream from here to Baja, the border is very confused and in part follows an old bed of the Canube .which is gradually drifting westwards5. Bikewise the Tisa border, which has not always been modified when river0control has changed the course of the river. Mrom here, the northern boundary of the historical B6cska, which before *2*2 corresponded with the $ungarian county of B6cs0Bodrog, struck out eastwards across the Canube0Tisa corridor, passing just north of the regionLs major town, -ubotica .-9abadka5, and G6noshalma before closing the loop again by the bank of the Tisa below -9eged. historical Ba!ka covered an area of *#,<A4 km ". km"5. The Taken together, the

Banat and Ba!ka are of almost the same si9e as the ,etherlands .@*,3E<

)ivers do not only define boundaries, however, they also connect, and with improving canalisation and control, together with improvements in communications, their power to isolate communities has gradually waned. In the case of the Banat within )omania, for instance, it has become practicable to combine administratively within GudeOul .county5 'rad the whole breadth of the local stretch of the =ureF valley, and, in 1ugoslavia, the $aneva%ki Rit has been detached to Belgrade. Both of these cut into the rough square of the historical Banat, as defined by the &natural' river frontiers, yet despite their aesthetic attraction and longevity, there is no good reason to view these as inviolate. They have, 3

indeed, never been so, as we shall see, and represent just one of many historical territorial configurations. They also overlook the BanatLs conflicting KnaturalL internal division into riparian lowland, heathland and mountain, each of which tracts are distinct and have their corresponding continuations beyond the historical Banat.

%he Dlvidk Concept In addition to the terms Banat, Ba!ka and Pojvodina in their various spellings, there exists the regional $ungarian appellation ()lvid)k .-outhern )egion5, with a unique string of associations and a long, complex history. In the =ediaeval $ungarian kingdom, the C;lvid;k, sometimes also referred to as the *lvid)k or V)gvid)k, included the $ungarian counties of 7eve .7ovin5, Toront6l, B6cs .Ba!5, -9er;m .-rem5, Palk8 .Pukovar5, +o9sega .+oQega5, and four KbanatesL of northern Bosnia .'ocsis, 23345. The $ungarian defeat at =oh6cs .*A"E5 was succeeded by the break0up of the kingdom, and until the +eace of +assarowit9 .*4*35, imperial Turkish and $absburg frontiers would ebb and flow across the centre and south of the country. 'fter its capture, the Banat was held as a $absburg %rown territory until the late *3 th century. &ven after this point, a $ungarian civil administration would not reach south to the Canube until *34"0<, when the military frontier districts of the Banat and Ba!ka were finally abolished and integrated into $ungary.

Curing the *2th and early "#th century, the term ()lmagyarors+g .southern $ungary5 covered B6cs and Bodrog counties, Toront6l, Temes, 7rass8, and -9:r;ny. This term clearly states a political claim, and may not be used to describe the area since *2*2. C;lvid;k appears to have remained an informal term, capable of multiple interpretations, but also implying that the region is south of something central, and thus not one of the $ungarians core territories. Ba,ogh .*2#", maps5, combines the above six counties with %san6d and 'rad counties. %e-esy, however .*2@", Kd)li harraink valto+sair,l' .%hanges in our -outhern Mrontiers5 p. @45 defines the C;lvid;k as including three territories 0 Temesvid;k, 2

B6cska, and -9er;ms;g .+o9sega, Per:ce and -9er;m counties5 until the latter was transferred by act of +arliament in *3E3 to %roatia. ,o mention is made of %san6d and 'rad. Mrom *3E3 until *2"#, there was therefore B6cska. a brief correspondence between C;lmagyarors96g and TemesyLs narrow C;lvid;k 0 both comprised of just the B6n6t and

'fter Trianon, C;lmagyarors96g shifted northwards to within the new borders, while C;lvid;k entered more common circulation as a general label for the territories detached from $ungary to the 7ingdom of -erbs, %roats and -lovenes .1ugoslav Banat, 1ugoslav Ba!ka, Baranja Triangle, =edjimurje, +rekmurje .=urak:9, =uravid;k5. The )omanian Banat dropped out of the frame, becoming absorbed into an Kexpanded TransylvaniaL concept, which embraces all the territory acquired by )omania from $ungary .Transylvania, the )omanian Banat and the %riFana 9one5.

Curing the -econd Dorld Dar, the 1ugoslav Banat 0 under (erman occupation 0 apparently also dropped out of the definition, which concentrated for the duration on the territories actually recovered from 1ugoslavia .Nyigri , .ed.5 - viss+at)rt ()lvid)k nem+etis)gi k).e , Budapest, *2@*. $a0ti ()lvid)k /01/2/0113 Budapest, *2345. K()lvid)kL, like its northern equivalent K 4elvid)kL, became explicitly associated with revisionist, expansionist policies . NI II <4E5. 4elvid)k was an old $ungarian term for the territory that had become -lovakia. In *2<3, the term was redefined to embrace just that strip of territory, which $ungary had recovered by the vivisection of %9echoslovakia .see %e,eki, (r8f +6l 5agyarors+g 6j hatra /0783 novem"er 93 - magyar !elvid)k nem+etis)gi t)rk).e3 Budapest, *2<35.

Cue to its association with revisionism and the territorial expansions of the thirties and forties, the term ()lvid)k fell out of favour in the *#

aftermath of the -econd Dorld Dar.

The +aris peace treaties had In the

quashed $ungarian border revision, and %ommunist internationalist sentiment theoretically rendered the whole question obsolete. interests of co0operation, $ungarian sources therefore used the official 1ugoslav term, Pojvodina .sometimes using the old $ungarian spelling, Pajdas6g .7ocsis, *23<, *22*, *22@, *22A, =irnics, *22<.5, sometimes coupling the two 0 KPojvodina0Pajdas6gL.

C;lvid;k and Pojvodina0Pajdas6g are today often used somewhat synonymously although, properly speaking, they do not correspond with each other. Pojvodina .0Pajdas6g5 refers to the 1ugoslav political unit Pojvodina, which embraces Ba!ka, the 1ugoslav .administrative5 Banat and -rem. The recognised meaning of C;lvid;k today .7ocsis5 corresponds territorially to that of the inter0war period, i.e. the territories detached to the 7ingdom of -erbs, %roats and -lovenes by Trianon .the 1ugoslav Ba!ka, Baranja Triangle, =edjimurje, +rekmurje .=urak:9, =uravid;k5. The 1ugoslav Banat has now drifted back into the frame5. The Kviss9at;rtL .returned5 or Kviss9acsatoltL .re0annexed5 C;lvid;k refers specifically to the period *2@*0@@ and excludes the 1ugoslav Banat. The revisionist complex of earlier periods has however been superseded by a more sanguine acceptance of the status quo.

In the 7ingdom of -erbs, %roats and -lovenes, the 1ugoslav Banat and the 1ugoslav Ba!ka were associated initially with the Baranya triangle .likewise detached from $ungary5, but later with the -rem district .between the Canube and -ava rivers5. The )omanian Banat meanwhile became associated with the %riFana, a strip of territory to the north, which ran along the $ungarian frontier, and ultimately became absorbed into the new concept of Kextended TransylvaniaL. This included all the territory attached to )omania from $ungary in *2*2.

**

'lthough they have been divided between three states since *2*2, the traditional view of the Banat and to a lesser extent, that of the Ba!ka, as discrete units has survived. Mor reasons that will become apparent, the Banat has a strong tradition of particularism. &vidence that this has survived until today is provided by the existence of a collaborative Study :rou. !or the Banat Region between the Nniversities of TimiFoara .)omania5, ,ovi -ad .M) 1ugoslavia5, -9eged .$ungary5 and TRbingen .(ermany5, of which the present author is an associate member. The presence of four main national groups each with distinguishable ethnic territories has nevertheless fostered the emergence of divergent views of history. The interest of these four groupsL respective specialists often wanes beyond the geographical bounds of their own ethnic group. It is in an effort to bypass these limitations that we have chosen to treat the Banat and Ba!ka in tandem. +airing Banat and Ba!ka also goes against 1ugoslav and )omanian visions of the regional community, as neither of these visions includes all of both territories. 's detailed elsewhere, the pairing of Banat and Ba!ka also does not correspond to a particular $ungarian tradition. 'lthough all four nations recognise the terms and agree on the territories they indicate, there is no nationalist viewpoint that binds the two territories. 'cademics from any nationality can agree, however, that what links them is an extraordinary complexity of populations, unsurpassed perhaps anywhere else in &urope. &ach of the four main ethnic groups .$ungarian, )omanian, -erb and (erman5 which are represented in the Banat and Ba!ka has, or has had, near0absolute dominance over certain parts of the regions. The extent of these areas, the degree of group concentration, and of local ethnic dominance varies both over the landscape and over time. The (ermans, for example, who in the early decades of the twentieth century, accounted for approximately a quarter of the population of the Banat and Ba!ka and were highly concentrated in four main 9ones, are now reduced to a mere sprinkling, even in the largest settlements of the )omanian Banat.

Landscape
Banat

i+isions in the Banat and Baka

*"

The

Banat

comprises

several

sub0units,

each

with

their

own

distinguishing features, and which are in some instances sharply differentiated from one another. =any of these sub0units embrace parts of the Banat and continue beyond its borders into neighbouring regions. Topographically, the Banat may be divided into five basic land0types the Bower +lain, the Npper +lain, the &(eli"latska $e;ara' sand 9one, a hilly 9one and, lastly, the Banat $ighlands.

The lower plain .about 3# to *## metres above sea level5, which runs mainly alongside the =aros, Tisa and Canube, also extends inland along and outward from the Temes and Bega rivers as far as MageO and %aransebeF respectively. =ost of the lower plain was formerly a confusion of marshes and choked watercourses. The majority of the land is extraordinarily featureless S 7ohl, a nineteenth0century traveller remarked that &all is smooth, unru!!led, and !lat as the ocean during a dead calm' . ' small KbubbleL in this landscape just south of Trenjanin .Be!kerek5, lies in the path of the Temes and Bega rivers and, before effective canalisation, used to form a significant barrier to drainage, a further cause of inundation in times past.

Mrom north of Trenjanin, a barely perceptible rise .to an altitude of ca. 2# 0 *"# metres5 brings one to the beginning of the Npper +lain, known in (erman as the Banater Heide .heathland5. This 9one, which runs in a broad crescent north0eastwards via 7ikinda towards Bugoj is almost equally featureless, but is famed for the great fertility of its black soil, which permits of several harvests a year, and commands a high price per hectare. =ost of this land ended up in the hands of (erman colonist0farmers. -uccessive $absburg regulations reserved much of this region for (erman colonists, whose loyalty to the crown was not in doubt, and who were deemed likely to bring the best return from the high fertility of the land. >ften a pre0existing non0(erman population was resettled to make room for the colonists, which did little for harmony among the nationalities.

Gust west of PrJac in the southwest Banat is the only distinctly elevated part of the western Banat. The (eli"latska $e;ara, a thinly inhabited region of *<

sandy desert formed during *3th century by Kirrational e<.loitation' of the forests, rises to a height of "@2 metres. 'ttempts to reforest the region have not been entirely successful, though nowadays sand can only drift when a strong and persistent wind .the 7oJava5 blows from the south. The dryness and infertility of this district contrast sharply with the loess plateau to its north. .Bugar, in 'bonyine0+alotas, 'hmetoviU0Tomka et al., 2*0"5.

*@

Bandscape Civisions =ap

*A

' narrow hilly belt in the central Banat, with its own distinct settlement0 pattern, is succeeded by densely wooded low mountains, rising quickly towards the southeast. The slopes around the southern towns of PrJac and Bela %rkva .1ugoslavia5 are cultivated with vines and herald the fringe of the higher ground to the east.

The mineral0rich Banat $ighlands .Banater Bergland5 climb to their highest point .*@@E m.5 at -emenic, near (VrWna .Dolfsberg5, Brebu ,ou .Deidenthal5 and PVliug .Mran9dorf5. The valleys of the Temes and %erna rivers separate the Banat highlands from the -outhern %arpathians proper. >n the borders of the Banat and Transylvania, the latter approach ","## m in height.

%onsiderable precipitation is generated here, the runoff rapidly gathering in the plain, but draining only slowly onward. The combination of high precipitation in the mountain0districts and the broad inland extent of the lower plain, with its very gradual gradient, expose wide areas to the constant danger of inundation.

Baka

The Ba!ka consists geographically of two principal parts. The uplands comprise the southernmost extremity of the sandy Canube0Tisa interfluve, and extend from Baja .2# m above sea level, the nearby >lomhegy being the highest point in the Ba!ka at *4@ m5 to -ubotica, the principal town of the north Ba!ka, .**@ m5 and south to the line Be!ej0Prbas0-ivac. Below this line, the average elevation drops to 2# metres, and as low as 3@ m in the southeast. The only higher ground here is a small plateau around Titel .*<#m.5. In earlier times, flooding on this lower ground was the rule rather than the exception, and considerable drainage was necessary before colonisation of could be begun.

*E

%he In5,*ence o5 the Ri+ers In any territory, the possibilities for locating new settlements are defined by the quantity and quality of local water resources. Nnless supplemented by more distant sources, local water resources also limit the potential for population growth, for agriculture and for industry. ,either the Banat nor the Ba!ka can be described as arid, but both territories suffer from great fluctuations in the spatial and seasonal distribution of water, as well as major variations in water quality. -ettlement along the major rivers has long been constrained by extensive and regular flooding, while inhospitable mountains in the eastern Banat have limited settlement there. The coincidence of the historical borders of Banat and Ba!ka with of the Canube and Tisa has its origin in the formerly untamed nature of those rivers and of their main local tributaries .Crava, -ava, =aros, Temes, Bega, Ber9ava etc.5. These rivers were prone to rising regularly out of their beds and wandering over wide and hence sparsely inhabited 0 but potentially extremely fertile 0 areas. 1et, while major rivers define their boundaries, their interiors are criss0crossed by relatively weak rivers, prone to stagnation. Barge areas of both territories have only been made habitable by virtue of centuries of works along these watercourses. ,owadays, the long0term viability of settlement is threatened by pollution of the rivers and of the shallow .phreatic5 aquifer, and by rapid depletion of the artesian aquifer.

The massive Canube and Tisa and their main tributaries have had a particularly decisive effect upon the settlement0patterns in the Banat and Ba!ka. 'pproaching -9eged from the north, the once tyrannically meandering Tisa already carries almost the entire runoff from the -lovak and -ub0%arpathian Nkrainian valleys, and has subsumed the -omeF .-9amos, -amosch5 from northern Transylvania and the %riF .7:r:s, 7reisch5 rivers from the Bihar =ountains.
"

Its watershed extends over

*A4,*3E km , the fringe being defined by %arpathian, Tatra, =atra and Bihar =ountains and the spine by some of the flattest country in &urope. The river is therefore extremely susceptible to seasonal variations in its flow, presenting the most serious problems in spring, when the alpine *4

snows melt.

' few minutes before it marches through the centre of

-9eged, it links arms with the =aros, which drains all -outhern Transylvania .except the >lt valley5.

The seasonal threat these two rivers posed to -9eged was greatly exacerbated until the later *2th century by the fact that at their confluence, the waters of the =aros and Tisa converged head0on. &specially at times of high water, the smaller =aros, deprived of an outlet, was forced back up its channel and had to find alternative routes. The extent of the resulting marshland in *4<3 can be seen on Migure *. The =aros was re0channelled before *32#, but the depression of the original bed is still clear today .Migure "5. The further course of the Tisa between here and the confluence with the Canube is subject to practically no natural constraints only a small plateau at Titel rises much above the river, and this the author has seen from miles away. It has been reported that over a <# year period, the variance between absolute maximum and minimum water levels, as measured at -enta station, has been as much as *#.#4 metres . NI , *2@A, @3"5. Murthermore, over the *A# km. from =artonoJ .42 m.5 to the Canube .4A m.5, the Tisa drops by only four metres while expanding and contracting to between @# metres and <A# metres in width. .Petersen et a,, *2<<, "2<5

*3

)iver ,et

*2

Before it subsumes the Tisa near Titel, the Canube has already been swelled by the Crava and the great -ava and smaller Temes soon join in at Belgrade. Thus within the space of a few do9ens of miles, between 'patin and Belgrade, the Canube gathers into itself the precipitation from %roatia, -lovenia and almost the entire %arpathian Basin. The valley of the Canube lies at an altitude of 32m at Baja in the Ba!ka, 4Am at its junction with the Tisa, and before the building of the Iron (ates Cam, the river at -viniOa was still at a level of EE metres. >nly here did the gradient began to drop more markedly. The width of the river 0 again, before the Iron (ates construction 0 expanded and contracted substantially over this distance, from *A# metres at the 7a9an (orge to ",*A# at (reben. Its path carries it over a number of shallows and narrows 0 most noticeably in the 7a9an 0 Iron (ates (orge . Nico,6esc*1 P,op7or et a,, The Iron (ate %omplex 'tlas, p. *A*5. The storage capacity of the Canube valley down to the Iron (ates substantially exceeds its capacity for discharge downstream. In ancient times, before the Canube cut its way through, the +annonian +lain was one vast inland sea 0 so at times of high water, it acts as a giant reservoir until the flood can pass through the rapids. The very considerable potential for energy generation offered by this natural phenomenon was exploited by the joint 1ugoslav0)omanian Iron (ate project.

)eturning upstream, the average level of the Tis9a falls only @ metres .from 42 m to 4A m5 over the *A# km from =artonoJ to the Canube 8Petersen et a,, *2<< "2<5. Nntil recent times, therefore, a rise in the level of the Canube at the confluence meant almost inescapably inundation along the course of the Tisa. Its waters, unable to run off downstream, would back up as far as %songr6d, rise out of the riverLs bed and submerge a huge area. This made life near to the rivers very ha9ardous .see Kvereinte =ra!t gegen die #hei>&, "asse, and $ch-idt, *<30@#5 and confined settlement to the more elevated land. -ome areas have remained practically uninhabitable up until recent times. The great fertility of the river valleys nevertheless exerted a strong attraction, and much of the population of the western Banat and the Ba!ka has "#

historically congregated as close as considerations of safety allow to the banks of the rivers. =artonoJ and Be!ej for example, were built within tight meanders of the Tisa.

'round TimiFoara, between -9eged and 7ikinda, and between PrJac and 'libunar, there had been extensive permanent marshes since ancient times. Curing the latter years of the Turkish occupation in the *3 th century, war damage and population decline resulted in wide expanses of fertile land being simply abandoned. Dhen the $absburg armies captured Temesvar in *4*E, the imperial administration extended to the Banat a policy of colonisation and economic development already begun in the Ba!ka. Between *4*3 and *4"#, the modernisation of the fortress of Temesv6r was the first priority. 's a means to this end, the choked rivers of the surrounding area, which had degenerated into an incoherent amalgam of marshland and watercourse and flowed along many shallow, contorted channels, were separated, controlled and canalised. %hannels flowing in and around Temesv6r were gathered northward to become the Bega, while those to the south of the city were gathered yet further south, to form the Temes. The Bega now flowed through Temesvar/ the Temischel, whose name had applied to the whole waterlogged valley, disappeared, while several miles to the south, a new river, the Temes, was created. ' second stage followed in *4"30<3. Initially, TemesvarLs insatiable demand for timber had been satisfied locally, as the area between the Temes and Bega had been well wooded when the $absburgs had arrived. By the late *4"#s, this resource had been depleted, and the result was that drift0canals had to be constructed to bring the densely wooded upper reaches of the Bega into the loggersL grasp. It was later found necessary to dig two small canals to reconnect the two. >ne permitted flood waters from the Bega to be diverted into the Temes to preserve the city from flood, the other diverted the Temes into the Bega when the latter 0 Temesv6rXs primary trade route to the &mpire 0 was running low. By *44E, the extent of the permanent marshes and the area threatened by seasonal flooding had been considerably reduced, but much remained to be done. Building in the "*

flood plains continued to be fraught with danger until at least the late nineteenth century and these remained more thinly inhabited than the interiors of the regions. The problem of the entire Tisa watershed was finally at last subjected to an overall strategy ith the !oundation o! the &Society !or the Regulation o! the #isa' in /81?3 >ver the period *3A# to *34A, under the auspices of %ount Istv6n -9;cheny, regulation shortened the riverXs course by @E< kilometres .<"Y, from *,@"2 km to 2EE km5. By the end of the *2th century, **# XshortcutX canals had been built, along with <,### km of embankments, A,<## km of drainage canals and A# pumping stations .=unteanu, in A9onyine1Pa,otas, Ah-eto+i:1 %o-ka et a,(, /08?)3 ' serious risk of extensive flooding every spring nevertheless persisted. The city of -9eged was practically destroyed by a severe inundation in *342. In *332, according to -omeFan .*2<2 *45, a massive 3,<E2,<34 hectares were inundated.

+rogressively more modern maps show both the natural dynamism of the Canube and Tisa, and the long process of river0control in operation, with new courses developing, dead arms, ox0bow lakes and marshes gradually shrinking while embankments and canals proliferate. The river0courses are becoming increasingly well defined. $absburgs, $ungarians, 1ugoslavs and )omanians have each in their turn struggled against the rivers, erecting massive flood0banks alongside rivers major and minor, and progressively extending the canal network. 1et modern maps are still characterised by the great lurches of the Canube and Tisa, and even the smaller =aros, after its plunge from the heart of the -9;kely country, still weaves its drunken way when it reaches the lowlands above -9eged.

It may be well understood that the few stable crossing points of these wild rivers have figured large in the regionsL military history. =oh6cs, site of the fatal $ungarian catastrophe of *A"E and a major Turkish defeat in *E34, is located at a crossing point on the Canube, though just outside the Ba!ka. -enta, where the Turks were crushed in *E24 is on ""

another, between the Banat and Ba!ka on the Tisa. Important fortresses were located at Belgrade, -emlin and +etrovaradin Banatska and Ba!ki +alanka and >rsova. 7arlowit9 and +assarowit9, which commanded the opening of the =orava valley, one of the principal historical routes to and from the Balkan interior5, were venues for treaties of great significance in *E22 and *4*3 respectively. Both were south of the Canube and thus symbolised that control of major crossing0points had passed into the hands of the $absburgs. 'fter defeat in the war of *4<30@*, the southern frontier of the $absburg &mpire reverted to the line of the Canube, and would remain there until *2*3. 'stride one of the easier north0south transit routes across the huge marshes, and an important staging0post between Buda and the Balkans, the fortress of Temesvar, the chief city of the Banat, long benefited from its inaccessibility and its protective belt of marshes. &vliya %elebi, who visited in *EE#0E@, aptly described it as a Ktortoise in the swampL, and pronounced it conquerable only by famine .)ieser, *22", A#5. The city successfully resisted many sieges and in *A24 S then in Turkish hands S had in fact been saved from starvation by a timely flood.

Co,onisation o5 the Lo;er Gro*nd 's the drainage of the wetlands proceeded, so colonisation was extended into progressively lower ground. In some cases settlements proved untenable, and had to be abandoned or relocated. The most striking example is perhaps offered by a group of seven communities .=arienfeld, &lisenheim .Belo Blato, ,agyer9s;betlak5, 7:nigsdorf, 'lbrechtsdorf, (iselahein, Ivanovo and (yurgyevo5, founded by the (eutsch2 Banater @n!anterie :ren+regiment in *3E3 in the +an!evaUki )it .+antschower )ied5, extensive marshes at the confluence of the Canube, Tisa and Crava. ' series of devastating floods, beginning in *3E4 and recurring almost every year for a decade, practically destroyed this last officially backed colonisation of the $absburg period. Iwanowo .'lexanderskirchen5 became -6ndoregyh69a, then Ivanovo. 7:nigsdorf, near >povo .Zp6va5, failed entirely, being Kcompletely submerged to its highest pointsL on ,ew 1earXs &ve *3E2, half of its **" houses having "<

already been swept away. 'lbrechtsdorf, (yurgyevo, &lisenheim and (iselahein were also all abandoned. =ariendorf had already had to be abandoned in 'pril *3E2 due to flooding, and remained under water for most of the year. )enewed flooding in =ay *34# decided the community to apply to Pienna for a transfer to a higher location. This was eventually accepted, but it was only in *34E that the last houses were abandoned to the waters .A*ssch*< der Ge-einde =oi,o;it>, 23?2@.

"@

+ancevo

"A

Gohannisfeld .at 3* metres5, founded on some of the lowest0lying lands in *3#E between Temesvar and Be!kerek., was situated in the narrow corridor between the Temes and Bega rivers and, &like all other communities in the Banat, has had to struggle against natural catastro.hes and e.idemics' . In *3A@, the Bega overflowed its banks and inundated all the fields. *3E<, in contrast, brought a drought. *3E2 brought a renewed inundation and in that year the villagers harvested from canoes and rafts[ To make up for this, however, fish could be caught &"y the "asket!ul'. Inundation and drought have &o!ten recurred to the .resent day' .Banater Post "#th Meb. *224. KBanater Artscha!ten stellen sich vor& ,r. A4, Gohannisfeld5.

In contrast, 7Rbekh69a, now an inconsequential village in the $ungarian Banat facing Baba Peche .)omania5 and )abe .1ugoslavia5 may serve as an illustration of the gradual taming of the hydrography. 't the time of the $absburg conquest, most of this corner of the Banat was under water. Dhen (ermans were settled in the village in the late *2th century, a minor watercourse .the +og6ny ;r5 still determined its outline, and acted as a drainage channel, and had saved the village in *3AA when floods had surrounded the village. 'n inspection today reveals only the vaguest depression in the ground to suggest where this may have been. The =aros river, a few miles to the north, and the Tisa to the west, are now so effectively controlled that it has been deemed safe for the people of 7Rbekh69a to fill in and even cultivate their KmoatL. Mor miles around the land is perfectly flat, and, apart from the roadside ditches, there is no longer any hint of local protection from flooding. 1et, in fact, the danger of inundation has even today still not been completely banished from the region. In early Ganuary *22E, the =aros once again rose over its constraining banks, causing widespread damage. Dhile much of %entral &urope remained fro9en, an unseasonably warm spell in central )omania released a sudden surge of run0off from the snowy %arpathians. Cecelerating into the flatter land of GudeOul 'rad, the river rose and eventually spilled over its constraining banks.. Bodrogul Peche, immediately to the west of 'rad, and on the opposite .northern5 bank to "E

Bodrogul ,ou, was \completely cut off from the outside world\. &stimates of the total cost of the inundations in GudeOul 'rad .where **,@A" hectares ."3,"3E acres5 were flooded5 ran to <# billion Bei. This included *.< billion Bei of agricultural damage and a daily loss of "A# million Bei due to the forced closure of "@ oil0wells. +etrol stations, telephone and electricity networks also suffered considerable damage estimated at <# billion Bei for GudeOul 'rad alone 8Banater Aeit*ng:. K=reis -radB C.idemie2 :e!ahr vor"eugenB Bei :urahont sind 9D von 97 Brunnen verunreinigt&3 Temesvar, *#th Ganuary *22E5.

"4

%hree $carce Reso*rces in the Lo;er Banat 1 $tone, =ood and "resh =ater

>ne of the peculiarities of the lower Banat and Ba!ka 0 and here the distinction between lower lying ground and the Banat highlands must be emphasised 0 is a great scarcity of stone. %elebi states that all streets and lanes in Turkish Temesvar were built from wood, as &in this !ertile 333 land there as not a stone to "e !ound, not even the si+e o! a "eanE L >nly the old $unyadi citadel 0 the tortoiseLs KheadL 0 was built of stone. The rest of the fortress and city, including the streets, were built entirely of wood F&vliya, - e y h ] a t n k m e P <3A ff.5.. ' $absburg bombardment of the timber0built city in *4*E appears to have been sufficient in itself to bring about its surrender. The suburbs had been burned down, and the interior of the fortress reduced to matchwood .-9entkl6ray, **45. Beyond the confines of Temesvar, the contemporary building norm among the -erb and Plach majority of the population, as reported by the $absburg (raf $amilton in *4<@, was for a poor roof of straw or tree0bark to be erected over a framework of stakes thrust into the ground, the walls formed by a coating of mud. -ome, he claimed, lived simply in holes in the ground. -uch dwellings could have no permanence, unlike the stone0built -axon towns of rocky Transylvania, and the population was consequently &+um #ransmigriren sehr geneigt' , .$amiltonLs Ghorogra.hia Bannatus #emessiensis , *#v, **r, in =o,5 *22A5. -ome elements in the population long retained this mobility, presenting landowners and tax collectors with recurrent problems and leading to regular abandonment of settlements.

The deficiency in stone is confirmed by G( 7ohl, a nineteenth0century traveller, who records the frustration of travelling in the Banat after rain in the early *3@#Xs. Darned against the sticky condition of the main road north0west from Temesvar, his coachman made the mistake of taking a minor road. Pery soon, they were struggling their wheels were &333no longer distinguisha"le as heavy mud, in heels, "ut a..eared !our thick, solid "alls o! as discovera"leE' 'll "3 hich, literally, no trace o! a s.oke

around was &one thick .udding o! mud' and 7ohl regretted the complete absence of any stones one could use instead. $e reproduces an anonymous otherLs characterisation of roads in the Banat as &stri.s o! "og enclosed "et een t o ditches'3 $e then noted, however, that &the tough slime o! a Banat .lain soon dries, and then "ecomes hard as stone' .'oh,, <"3025. ' British military map from *2@@ notes that along the Temes and Bega, movement is still &restricted "y marshes and atercourses', and that their lower courses were &im.assa"le in e<ce.t =.I..T5 >verprint inter hen !ro+en' .* A##,### &urope .'ir5 -heet ,.&. @@^*3 Beograd *st edition5. 'll but one road passing through Pelika

7ikinda were in *2@A termed passable only in dry weather .,IC, III, A*45. Indeed, the botched evacuation of (ermans from the 1ugoslav Banat in -eptember *2@@ was badly hampered by a few daysL heavy rain.

Cespite an apparent abundance of wooden buildings, it is also true that the lower0lying parts of the Banat and, after an initially generous resource was depleted, the Temesvar area, soon suffered from a severe lack of timber. (raf $amiltonLs report to Pienna makes this clear, describing it as a &de!ect' of the Banat on the lower land, and especially pronounced towards the Tisa. The cityLs predatory demand for building timber and fuel had to be satisfied by recourse to the forests further east. In the boggy west, the population made do by using peat for fuel and could not be induced to use more costly timber brought down river from the densely wooded Banat mountains.

In addition to stone and wood, another item in deficit in the lowlands of the Banat, despite the superabundance of moisture, was water fit to drink. >n the lower ground, the extent and age of the stagnant marshes meant that deep wells had to be dug before a clean water supply could be secured. -outh0east of Bugoj, _tuica .(erman name, &bendorf5, located at *2E metres above sea level and therefore above the worst of the contamination of the wetlands, nevertheless had to be supplied by two wells sunk to depths of A@ and A4 metres. Before pumps were "2

installed in the *2th century, these had to be pulled by hand. They remained the principal source of water until _tuica was linked to the main water supply in *24# . Banater Post, Ath >ctober *22E5. Nntil this century, the bad water was a constant source of cholera epidemics were a regular occurrence, often arising as a consequence of floods. Gohannisfeld .*,E@E inhabitants in *3A* . Hand"uch der Woj odscha!t Ser"ien und des #emeser Banats *3A@.55, lying between the Temes and Bega rivers downstream from Temesvar, suffered both inundation and cholera in *3<*, and a second epidemic in *3<E claimed E@ lives within four weeks .Banater Artscha!ten stellen sich vor FH?)3 Iohannis!eld3 Banater Post, =unich, "# Meb *2245. Curing the Ganuary *22E floods in GudeOul 'rad .see above5, the village of +escari .%ommune (urahonO5, suffered the contamination of "# of its "< wells, giving rise to renewed fears of cholera or typhus 0 which was last encountered in the Banat twenty years before. +escari was but one of many villages still reliant for its drinking water on wells rather than on mains supply 8Banater Aeit*ng:. K=reis -radB C.idemie2 :e!ahr vor"eugenB Bei :urahont sind 9D von 97 Brunnen verunreinigt&3 Temesvar, *#th Ganuary *22E5.

<#

BA !ost Insa,*9rio*s C,i-ateC The combination of regular, devastating floods, the marshes and the shortage of solid building materials locally .to the consequent detriment of the quality of housing5, with the poor quality of drinking0water, induced a poor general standard of health among the population, amply reflected by the widely0quoted colonist saying &dem Crsten den #od, dem J eiten die Kot, dem (ritten das Brot' . 7ohl reported that &#he stench o! the stagnating hot Banat summer F&!or aters, com"ined ith the sultry, heavy air o! a eeks together there is sometimes a .er!ect

calm333') could "ecome over.o ering, particularly in Temesvar. Curing his visit to the city in about *3@#, &the inner !ortress o! the to n !elt like a "aker's oven333 there eek333' .7ohl, "435. as not a "reath o! air to "e had3 A! the t o ere in hos.ital in one thousand soldiers o! the garrison, nine hundred

In the wake of the disastrous Turkish war of *4<30@*, the plague entered the Banat, provoking widespread fear if not large numbers of deaths. The effect of the epidemic on the morale of the colonists was underlined by the habit of the boggy ground to give up its dead, which simply could not be buried at an adequate depth. It was eventually decreed that the corpses of plague victims must be cremated. -ubsequently, the $absburgs introduced a SanitLtskordon along the Canube, restricting passage into the &mpire of goods and persons deemed to be plague0 carriers. This was observed in action by 7ohl during his passage through >rsova in the *3@#s. &Ko here had @ heard the su"ject o! health so constantly discussed as at Arsova, and indeed throughout the Banat'3 By *3A<, the +reyer, =ayor of Temesvar .whose fatherLs first wife and five children had died during an epidemic in the Banat during the *42#s5, was able to state confidently that the city of Temesvar had outgrown its reputation by virtue of the $absburg programmes of river0control and land reclamation .+reyer, *22A, *"@5.

<*

The Banat and Ba!ka have remained difficult environments, however. ' series of villages .the 4ran+is+eische Waldsiedlungen Mran9dorf, Dolfsberg, Dolfswiese, Deidental, Bindenfeld, Bu9iasch5 founded in the *3"#s in the -emenic =ountains of the -outhern Banat, where timber and fresh water did not present problems were soon abandoned and, when re0established, barely survived, because of harsh winters and an exceedingly short growing season .$ch-idt, *22*5. >ne of the villages .Bindenfeld5 presently has a population of one. =ontenegrin colonists to the 1ugoslav Ba!ka after the end of the -econd Dorld Dar experienced the harsher side of that region. 'mong all the post0war colonists, they suffered the most heavily from contagious diseases, tuberculosis and gastric ailments. This was brought on by a Msura"ondance dMeau et dMhumidit)M combined with Xchangements e<essi!s de climatM3 This may be explained partly by the contrast between the life of the mountain0 dweller with that of the inhabitant of the plain, for which the =ontenegrins were not prepared. They were accustomed to an abundance of stone and wood and clean water. In the Ba`ka, none of these were in plentiful supply . They had to get used to living in brick houses with a high level of humidity .which, in some villages, reached up to 3#Y in all houses5. Murthermore, Mil dut "oire de lMeau de .uits de "as2 !ond, souvent si malsaine NuMon doit la !aire "ouillir .our la rendre .ota"leM3 =any of the =ontenegrins died before the coloni9ation succeeded, and several of the survivors returned to the south. *

%he Condition o5 the Ri+ers in the Banat %oday

,owadays, while flooding has not been entirely banished, the quality of the waters flowing in the rivers is of more concern than the waywardness of the rivers themselves. Dhile the Canube has recovered somewhat thanks to careful management, heavy industrialisation and poor environmental controls have degraded many of its tributaries to a disastrous extent. Dithin the Pojvodina, @24 water pollution producers are registered, including <"E industrial, **< agricultural and <# * Pasovic, p, *E4 <"

municipal. <"* do not treat their effluent at all, and for most of the rest, treatment is inadequate. >f the A"A settlements, only <# have sewerage systems completed or under construction, covering a mere "#Y of the population. *EE of these @24 water polluters are located in the 1ugoslav Banat, representing the oil, chemical, petrochemical, food, textile, leather, paper, packaging, metalworking industries.

The Tisa is polluted within $ungary, locally attributed to )omanian industry and carelessness. By its 1ugoslav section, it is severely degraded. The construction of the Be!ej (ate in *24<, while permitting water upstream to be drawn off and used, led to more rapid silting in the channel. The bed of the river is consequently rising and the embankments have constantly to be upgraded. The water quality has also degraded, resulting in algae blooms and fish0kills. The =aros also suffers from )omanian industrial pollution, particularly from the $unedoara area.

Dhile three major rivers 0 Canube, Tisa and =aros 0 flow along its margins, the rivers wholly within the Banat have an insubstantial flow and suffer from extreme levels of pollution. -everal of them .e.g. >ld Bega, +lovni Bega and Tlatica5 are so bad they cannot even be used for irrigation. Canube0Tisa0Canube canal forms a diagonal cut across the major internal rivers, and draws their polluted waters off to the south0 east, away from Belgrade. 't times of low water, when the concentration of pollutants is highest, the CTC replenishes the shrunken rivers and dilutes their poisons.

The >ld Bega, whose very slight gradient favours stagnation and inundation, was canalised between *33E0*324. It is presently extremely polluted .class IP5, being characterised as an Xopen sewerX .+.A<5. 'fter passing the CTC, it is again badly affected by Trenjanin. The ,avigable Bega, the canal which passes through the centre of Temesvar, has become an open drain for the cityXs waste0waters. Both branches of the <<

Bega are extremely polluted when they reach the CTC canal, and are again subjected to heavy pollution by the Trenjanin area. The Temes, dammed at its source .Trei 'pe5 suffers from pollution on a similar scale to the Bega. Mrom being an anglerLs paradise, it degraded badly between *23A and *22#, and this degradation has accelerated since then. The Tlatica .'ranca5, which follows an old course of the =aros, by which it is still influenced via subterranean streams, has a low volume of water passing along a very wide bed at a very gradual gradient. ' couple of canals have been built linking it to the =aros, and pumping is necessary when the Tisa is high. The BWr9ava is in many ways typical for a Banat river. Its swift upper course, steep and narrow, and dammed at several points. gives way to a slower alluvial breadth in the plain. The %erna .in the south0east5 has been dammed twice for hydroelectric power.

Cana,s

In the 1ugoslav Banat, the KCTC Dater =anagement %o.L is responsible for the maintenance of the major canals. Their remit covers drainage, irrigation, dilution of polluted streams and the supply of water for industry and settlements. In several places, ox0bow lakes left behind by the meandering rivers have been retained for fishing and tourism. Their principal concern is the CTC canal, but also the 7ikinda %anal, controlled sections of the 1ugoslav Banat rivers and related canals, along with pumping stations, locks and dams. But many communes also maintain extensive networks of dykes and canals +landiJte looks after A#* km and ,ovi Be!ej, @24.A km.

The brackish nature of the rivers of the plains, combined with their high levels of pollution, causes water for irrigation and domestic use to be drawn principally from the ground waters. In the 1ugoslav Banat, the best quality water is to be found in the extensive artesian aquifer which is to be found at depths of between A# and A## metres. The quality of <@

water is not uniform throughout, and often contains methane and free hydrogen sulphide, which gives it Kan unpleasant smell and odourX. Dater from greater depths is warmer and shows a higher degree of mineralisation.

Nntil the *24#s, the majority of wells deeper than *## metres had sufficient natural pressure to bring water to the surface unaided. There has been widespread drilling of artesian wells due the fact that the water can usually be used without purification, and the majority of settlements in the 1ugoslav Banat have A 0 *# or more artesian wells for water supply and irrigation. This level of exploitation has led to a negative water balance in the aquifer, and it is currently dropping at an average rate of <#0@# cm ^ year, and its reserves will soon be depleted. +umps have had to be installed and some wells have dried up. ' start has been made on exploitation of sub0artesian levels.

Dater from the KphreaticL .shallow5 'quifer .p. @"5 is very variable in quality across the 1ugoslav Banat, particularly along the Tisa and Canube, along the fringes of the loess plateau and in tectonic depressions, and has locally suffered severe degradation due to fertiliser leeching, industrial waste and the use of depleted shallow wells for the disposal of municipal and industrial waste0water .at a rate of A0*#,### m< ^ day5. +hreatic waters cannot be used without purification. 'dditionally, only 4Y of shallow0aquifer water can be used for irrigation without risk of soil salinisation. "

2 This section reliant on: ivan Bogdanovi et al. Hydrological Situation in the Yugoslav part
of the Banat ; Boo Dalmacija et al. , Quality of Waters in the Yugoslav Banat in Abonzine et al, Rodica Munteanu Banat hydrographic Network and its Main Artificial Modifications. (West University of Timioara). ovi !a", Timioara, !zege", #$$%.

<A

%he Nat*ra, B,essings o5 the Banat

The Banat does, however, have some redeeming features, and these are of such magnitude as to have encouraged large0scale state programmes of colonisation and a constant stream of fortune0hunters. The fertility of the lower0lying land is such that two or even three harvests a year can be made without recourse to fertiliser, and an occasional disaster due to storm, flood, drought or frost can thus be more readily borne. The production0level of Banat and Ba!ka wheat became gradually more significant during the *2th century and was priced highly due to having had a higher gluten content than that grown in (ermany. Its importance was undermined by the opening of the %anadian and N- prairies, although a high yield meant that it still remained competitive. addition to the traditionally important raising of horses and cattle. The region also supported significant oats, mai9e and pork production in

Dhile the Ba!ka and the lower Banat suffered from an almost total lack of mineral ores, the Banat mountains, treasure0house. in contrast, were a veritable =ining had been done in these mountains since the

time of the )omans 0 perhaps the Cacians. Bavarian and -tyrian miners were operating in the XBanater BerglandX of the $ungarian kingdom as early as the *<th century, and the Turks had continued the tradition, though production was not great. 'fter the $absburg conquest of the Banat, (eneral =ercyXs Banater Bergeinrichtungs2=ommission .*4*45 revived mining activity in the mineral0rich highlands by introducing skilled Tirolean, -tyrian and Bohemian labour. The tremendous wealth of mineral resources inspired the $absburgs to invest heavily in the province and even subordinate it directly to the crown for over half a century. The mineral deposits, and the rich agricultural land towards the north0western corner of the Banat, became magnets for successive waves of colonists. The settlements they swelled themselves later went on to provide further colonists to other parts of the Banat. By *4<<, half of the $absburg &mpireLs copper production came from the Banat. Dhile the disastrous Turkish war of *4<30@* forced the substantial <E

abandonment of mining in the Banat, the discovery of massive hard0coal deposits later in the century again boosted the regionXs importance. By the nineteenth century, the 'ustro0$ungarian Staatseisen"ahngesellscha!t .-T&(5 had acquired from the crown virtual monopoly in area of "A square miles, stretching from Bogschan .BocFa5 in the north to ,eu0=oldova .=oldova ,oua5 by the Canube. Dhile (ermans were the chief beneficiaries of the $absburg policies, acquiring along the way the richest lands and the greatest capital assets, the Banat was not colonised by (ermans alone. Indeed, without the contributions of -erbs and )omanians, among others, the (ermansL remarkable feats of river control and economic organisation would certainly have been far harder to achieve.

The overwhelming majority of the population was )omanian the capital and profit however, were firmly in (erman hands. By the Mirst Dorld Dar, the Banater Bergland had become the most significant heavy industry 9one in south0east &urope, with important centres at )esica .)eFiOa5, >ravica .>raviOa5, -teierdorf, and bjmold8va .=oldova ,ouc5. The most important mineral deposits locally were coal, iron, sulphur, copper, and lead, gold and silver. ' late *2 th century -T&( document lists "A other minerals that were extracted commercially . Beschrei"ung der Banater (omLneO).

<4

-T&(

<3

+opulation growth in the industrial towns was powerfully stimulated by the discovery of new deposits, and by advances in technology which it possible to extract known deposits profitably. >raviOa boomed early in the eighteenth century but was soon overtaken by )eFiOa, and subsequently by -teierdorf in the mountains, once a tortuous mountain railway had been built there and it became possible to exploit and transport its enormous hard coal deposits.

The settlement geography of the Banat is dominated by water, mainly because there is too much of it. Nntil the mid eighteenth century, about a third of the western Banat was permanently flooded, due to the coincidence of very gentle gradients and the confluence of many of &uropeLs major rivers within a few miles of the BanatLs borders. -ettlement near the rivers has thus historically been extremely ha9ardous. Temesvar, the capital of the Banat, grew up at a convenient and well0protected river0crossing, but being surrounded by marshes and suffering from a dearth of fresh water and solid building materials, was home to a sickly population. These conditions were widespread throughout the lower Banat and reached even into the mountains.

In sum, characteristics of the physical environment have historically determined the location, si9e, and rate of growth of settlements in Ba!ka and Banat. Cespite several large waves of colonists, with the exception of the (erman0)omanian mining towns, few onetime villages have consequently been able to develop into significant poles of attraction. =ost of the *3th century foundations have remained essentially rural farming communities. Mew have even expanded much beyond the bounds established at the time of their foundation. The earlier stages of the planned settlement of the western Banat and the lower Ba!ka resulted in a patchwork environment, with a number of large villages .usually generously spaced, but sometimes abutting an older settlement5, each locally significant but none so regionally. 's the Banat and Ba!ka were predominantly agricultural regions, later colonists 0 <2

needing ready access to their allotted land 0 were mainly funnelled towards new foundations which tended to be much smaller, and squee9ed into lesser0used corners. Cespite the region being a destination for hundreds of thousands of colonists since the beginning of the eighteenth century, the net natural increase of the population has remained low.< 'fter an initial burst associated with the first couple of generations of the $absburg colonists, equilibrium was soon reached, set by the limitations of the physical environment in combination with the si9e of the original land holdings and by personal economic expectations. -urplus population from the more populous villages migrated to newer foundations, to the towns or abroad, southern )ussian being a common destination. Indeed, by the late nineteenth century, the generally prosperous (erman rural population in many of the larger villages was facing a long0term decline, due to shortage of land and the prevalence of a Xone0child0systemX. Nntil the later *2 th century, the development of an urban way of life was limited to the few old towns that had survived into the $absburg period, and to the industrial 9one in the mountains.

<
$.=.-.>. #ransylvania and the Banat3 Bondon, *2"#.

@#

A !INI$%RA%I/E HI$%ORD

In an area of conflicting territorial claims, the definition of where any territory actually begins or ends is itself often a matter of dispute. >ne need mention only =acedonia, where the rivalries of (reek, -erb and Bulgar have served to cancel each other out and propagate the growth, within an indeterminate territory, of a native =acedonian identity which is neither (reek, nor -erb, nor yet Bulgar, and certainly not Turkish. The present inquiry looks at a territory at the other end of 1ugoslavia, and includes its historical extensions into )omania and $ungary. Dhile there is little evidence of a specifically KPojvodinanL identity, the area is another classic frontier 9one, and has historically been subject to repeated changes of frontier and population. &ach successive change of rule has brought in its train disastrous results for one part or another of the population. PWe !ind ourselves here in a classic territory !or do n!all and ne starts3 Gonstruction and destruction succeed each other as naturally as the seasonsP .@ . ip,ich, *2A"5 The area under consideration, a battlefield for $ungarians, >ttomans, $absburgs and -lavs for two and a half centuries has, since *2*2, been an apple of discord between 1ugoslavia, )omania, $ungary and (ermany.

The ebb and flow of various administrations over the territory has given rise to a number of conflicting viewpoints among the main protagonists on what constitutes the regional KcommunityL. These include the concepts of K-outhern $istorical $ungaryL .that of Budapest5, Pojvodina .that of Belgrade5, K&xtended TransylvaniaL .shared by $ungarians and )omanians5 and that of the Canube -wabians. There are, of course, not simply four main schools of thought, with minor ones representing the viewpoints of the less numerous national groups. There is rather a great &-Ja, hier sind wir im lassischen !ande der "nterg#nge und neuen $nf#nge% $uf&au und 'erst(rung
wechseln in nat)rlicher *olge wie die Jahres+eiten% Q Translation by Bru'e (it'hell .

@*

range of writing from the balanced to the outrageous and differing in ethnic viewpoint, period and enlightenment. -ome differences in opinion arise naturally between the various language0groups due to imperfect communication, and to their witnessing similar chains of events from different perspectives.

's the present author has no family involvement with any of the regionXs national groups, whatever biases may be detected in the following can safely be said to have arisen during the process of research itself. These will have been nourished in proportion with the balance of the sources he has been able to consult and understand, which, limited by availability and language knowledge, are proportionately first in (erman, then $ungarian, with some in &nglish or Mrench and a very few in )omanian. Banguage difficulties unfortunately prevent the author from commenting on -erbo0%roat and )omanian literature, except such as has been published in &nglish or other translation. $istorians are always products of their cultural environment, and this is true also of the present author. But he may be able to avoid some of the pitfalls into which the authors of some of his sources have plunged 0 whether wilfully or innocently, or in the blind service of the state. Dhere embittered territorial disputes are concerned, it seems that few historians have been really able to overcome or completely ignore the inherited romanticised traditions of historical myth, or the pressures of patriotism, and emotions of conquest or defeat confound objectivity and detachment.

P-s a rule, the victor had conNuered the Banat "ut recently and sought to justi!y, or at least disguise, this act o! !orce through historical P!actsP3 #he de!eated, in contrast, had not yet overcome the .ain, and sought to justi!y or !urther their e!!orts to recover the lost territory "y a similar use o! historical P!actsP3 #he truth is o!ten mid ay "et een these P!actsP, ignored, as "eing Pjust not askedP FRieser *22", @A5
H

) Translation y Bruce Mitchell. Taken in conte!t" This #uotation refers to the $ungarian %
Romanian disputes" ut can e just as easily applied to any other situation.

@"

!edie+a, H*ngary

In the mediaeval $ungarian kingdom, two regions covered the territory understood today as the Banat. The K #emesvid)kL or K#emeskR+L was an ill0defined territory bounded by the Temes, Bega, =aros, 7:r:s and lower Tis9a rivers, and comprised the counties Toront6l, Temes, 7rass8 and 7eve. Np until *<<#, the Banat of -everin served as a form of marchland east of the Temesk:9. The exact borders of the counties and B6ns6g can no longer be reconstructed with any certainty, though the B6ns6g reputedly stretched from the highlands to the line of the river >lt. Curing the period of the $ungarian 7ingdom, the county boundaries ebbed and flowed substantially. 't times, Temes extended well into Toront6l, 7rass8 into -9:r;ny and Bittle Dallachia .>ltenia5, whilst Toront6l was also encroached upon by parts of 7eve and %san6d .to the northwest5. 't one time, the northern parts of 7rass8 belonged to 'rad and Tar6nd counties, its southern parts to -9:r;ny and its western fringe to Temes, 7rass8 being Kso reduced as to scarcely merit the name of countyL .-9entkl6ray *342 E5.E 7rass8 and -9:r;ny were in time, like B6cs and Bodrog to the west, to become associated as a pair. The retrospective view that the counties Toront6l, Temes and 7rass80-9:r;ny as constituted between *34< and *2*2 corresponded to KhistoricalL precedents is therefore only partially justified, and one can certainly not talk of a millennium of continuity.

Nnder the 'ngevin dynasty .*<#303#5, Temesv6r, the main fortress0town of Temes %ounty and of Temesk:9 as a whole, occasionally played host to the $ungarian king. Between *<*A and *<"<, during the closing stages of a long civil war, the inaccessible city even served as the main

ETranslation by Bruce =itchell. @<

base of %harles )obert until he was able to impose his rule over the rebellious nobility. 'fter %harles )obert transferred his entourage to Pisegrad, Temesvar and southern $ungary remained heavily militarised. (radually increasing Turkish pressure in the southern Balkans stimulated a northward flow of the Balkan peoples, especially after the collapse of -tefan CuJanXs -erbian 7ingdom in *<AE. Bulgars, who had occupied the area before the $ungariansL arrival, had again been settled in the area of 7rass8fd from *<<<, and went on to found a series of daughter colonies into the *3th century. 't the same time, the $ungarian kingdom expanded southward to absorb ,6ndorfeh;rv6r .Belgrade5.

Temesk:9 grew again in importance with the >ttoman advance. Ioan $uniade, or $unyady G6nos .both )omanians and $ungarians claim him as one of their own5, from the an immensely wealthy second0generation was created %ount of Temes with immigrant south,

responsibility for the defence of the realm. Nsing Temesv6r as his base, he undertook a long series of campaigns against the Turks, culminating in a great but fatal victory at Belgrade in *@AE. The fall of the ,orth -erbian Cespotate in *@A2 sparked off a major -erb exodus, led by -erb nobles, into the land south of the Temes river and reinforced existing -erb enclaves between the rivers =aros and 7:r:s. 'fter a period of adjustment and a disputed succession, $ungary settled into its Kgolden ageL during the enlightened rule of king =atthias %orvinus .*@A30*@2#5, but quarrels within the nobility persisted. Dhatever unity may have survived =athias was shattered by the dreadful brutality of the +easantsL Dar .*A*@5. 't its climax, the peasantsL leader (y:rgy C89sa besieged Temesvar, but was decisively defeated by the timely arrival of G6nos T6polya, Poivod of Transylvania. The rising culminated in a series of singularly barbaric executions at Temesv6r.

Deakened and divided by the events of *A*@, $ungary was unable to withstand the whirlwind which next struck. The young -uleiman .the =agnificent5 had just become -ultan and in his first campaign, *A"*, @@

took Belgrade. By *A"E, the kingdom was in turmoil again. 7ing Budwig II was unable to assemble an army sufficient to meet the threat posed by the ambitious -ultan, and both he and the majority of his army were killed at the Battle of =oh6cs. The two southernmost counties of $ungary, -9er;m and Palk8, with the fortresses of -lankamen, +etrovaradin, Ilok, >sijek and Pukovar, were annexed to the >ttoman &mpire. The $ungarian disaster at =oh6cs left two rival claimants to the crown, T6polya on the one hand and Merdinand of $absburg, 'rchduke of 'ustria on the other. This left $ungary at the mercy of both $absburgs and >ttomans and by *A@*, the country was effectively divided between them, with Transylvania playing an uncomfortable balancing act off to the side.

Ha9s9*rg E Otto-an Ri+a,ry

Dithin the Turkish territory, %entral $ungary was transformed into a +ashalik, under a Beglerbeg. Transylvania became a -andjak under a nominally independent prince, while the Banat was translated into a -andjak of TimiFvar. The Ba!ka was transformed into a -andjak of -eged. 'll retained some independence, and while considerable military forces were retained in the Banat, recruited substantially among -erb refugees, support for the -ultanXs planned *A@@0A expeditions was limited to logistics .Berindei, /einstein, "@0A5. In *A@20A*, efforts to unite the former $ungarian territories under the $absburg crown provoked >ttoman military action. In *AA", having been repulsed the previous year, the Turks returned in larger numbers, *E#,### men and *E# guns arriving before Temesv6r in Gune. The garrison, of ","*# men, consisting mainly of $ungarians, with the rest made up of -paniards, (ermans and Bohemians .Preyer , @@5 held on for a month before finally succumbing on "2th Guly. The fortresses of Bippa, %san6d, -9ent0=ikl8s, and %sakov6r subsequently fell swiftly into Turkish hands. It is claimed that whatever sedentary $ungarian population still remained in the Banat at this date had disappeared by the end of the *AA#s .'ocsis *232 @A

"#5.

't some point before the end of the *Eth century, a Bugoji0

%aransebeFi Banat, with own Ban, was detached from the -anQak of TimiFvar and subordinated to the rulers of Transylvania, thus being placed under indirect Turkish authority .Preyer , A<, footnote5. TimiFvar itself and the Temes district were placed under a Beglerbeg .uniting civil and military functions5 who was a senior, Ktwo0horse0tailsL officer. The incumbent was not allowed to feel secure, however, and there were 4# changes in *E@ years.

Dith the continual decline of the $ungarian royal house in Transylvania, >ttoman influence prevailed over most of $ungary and throughout Transylvania, and would continue to do so to a diminishing degree until *E22. $ungarians nevertheless persevered in their attempts to break free from both them and the $absburgs. Mrom *A2@03, for example, B6thory Tsigmond, +rince of Transylvania, pursued the struggle for independence substantially in the Banat. $e incited the Ban of Bugos to invade the TimiFvar Banat in *A2@ and raise local Bulgars, -erbs and Plachs against the Beglerbeg. %apturing Becskerek, he repelled four attacks before being overwhelmed. The next year, Tsigmond himself invaded the Banat and captured Bogs6n and Persec, Macset, 'rad, Bippa and %san6d. In both *A2E and *A24, he unsuccessfully besieged TimiFvar, on the second occasion being frustrated by floods. In the following year, he held %san6d against a Turkish siege and won a significant victory before Bugos.

Though the Beglerbeg of TimiFvar was always a senior official, the (overnor .-tadholder5 seems to have been of a different mettle, and in *E#@ and *E#4, two unpopular (overnors were driven out by the cityXs exasperated population 8Preyer , AE5. ,either was reinstated by the +orte, which raises interesting questions about the relationship between the rulers and the ruled. ' new series of wars, between various leading $ungarian nobles, raged fitfully in and around the Banat from *E#" to *E#3. @E

Though recurrent devastation of parts of the Banat and Ba!ka was wrought by all of the various combatants 0 including $ungarians 0 the Turks took good care at least of TimiFvar. The city was a crucial staging0 post for Turkish campaigns against Pienna and Transylvania, and profited from trade with the several >ttoman armies which passed through. It was the TurksX most secure possession in $ungary and their most pivotal, as illustrated by their saying Kthey who capture Buda acquire a city, but they who capture TimiFvar win a country .Rieser , p. @@,
quoting from T. Trcpcea/ #imiSvar 2 :hid #uristic, BucureFti *24A5 .

It served as

jumping0off point for any campaign against the +artium or Transylvania, and protected communications with Buda. The garrison consisted of *#0 *",### men, with "## cannon, and the fortifications were constantly modernised. Information on the condition of TimiFvar in the later decades of the >ttoman administration rests substantially on the report of &vlija %elebi, who stayed in TimiFvar between *EE# and *EE@. 'ccording to his report, the Ktortoise in the swampL was built entirely of wood, with whitewashed oaken walls fifty to sixty feet thick whose maintenance was entrusted to seventy villages. The circuit of walls could be walked in an hour, so the city itself cannot have been very large, however .#imiSoara ?DD, TimiFoara, *2E2. 's reported by )ieser, @2. 'lso
excerpt printed in "asse,, HorstF $ch-idt, )ose5: -n (onau und #hei>B Banater Tese"uch3 Bandsmannschaft der Banater -chwaben5 *E5. %elebiXs

report is confirmed by a description by $. >ttendorf from *EE4 . A(P( Petri ,

*2EE. 's reported by Rieser , @25

)ieser quotes, without offering a date, an

assertion in KTimiFoara 4##X, that in the fortified part of the city, *,"## houses were located in four quarters, along with baths, cafes, schools, mosques, monasteries, and over @## shops. Dood0clad streets linked the city to ten suburbs with a further *A## houses .#imiSoara ?DD, quoted in
Rieser , A#5.

The same source describes TimiFvar as a Kcorner of +aradiseL

.Rieser A#5. ' late >ttoman text describes the +rovince of TimiFvar, before the KPienna 1earL .*E3<5 as very rich despite its proximity to the border, including several prosperous towns

@4

and villages.4 The $ayducks, who inhabited some of them, were famous for the quality and si9e of the herds they would drive to western markets. It may also be pertinent to point out that fiscal records relating to fishing on the Canube only begin during the Turkish period, with the revenue going directly to the -ultan or -anQak0Beg . Airo0e+iG, )eFiOa,
*22A5.

The reign of Beopold I .*EA30*4#A5 of $absburg saw the tide turn decisively against the >ttoman &mpire The first major defeat of Turkish land forces was inflicted at -9entgottard in *EE@, and in *E3<, the >ttomansL grand effort against Pienna ended in catastrophe. The siege, led by the (rand Pi9ier 7ara =ustafa, brought about an otherwise unlikely coalition of %hristian princes an international relief army led by the +olish king Gan -obiewski descended on Pienna. -hortly before joining battle at the 7ahlenberg, 'hmedpasha, Beglerbeg of TimiFvar, made himself infamous and his -ultan hated by leading the merciless slaughter of <#,### captives irrespective of age or sex . Preyer A35. ' disastrous defeat followed and the shattered Turkish forces fled to the border town of )aab .(yer5. Curing the ,ew 1ear, $absburg and allied forces drove Turkish forces from Budapest and out of central $ungary. By the late *E3#s, $absburg armies had sei9ed the Banat highlands, establishing Derschet9 as its administrative centre and, aided by a -erb rising, had penetrated as far as >ld -erbia and Bulgaria. Cespite victories at ,iJ and historic 7osovo, overstretch combined with the reforming efforts of a new (rand Pi9ier to produce a Turkish recovery. The reorganised Turkish army drove to the -ave and Canube and, in *E2#, Imperial indecision and tactical wooly0headedness permitted the >ttomans to recover Belgrade. The rebellious -erbs had now to face the vengeful -ultan unaided, and a large number of -erbs led by +atriarch 'rsen III of Ipek fled north from K>ld -erbiaL .i.e. 7osovo5. with the retreating $absburg armies. Traditionally, he was accompanied by <#0
% . *or more information see +,er -./e von Temes'h/ar0 1rrinneringen an 2a3fer 4as'ha "en 5lteren, aufgezei'hnet von seinem !iebelbe/ahrer 3Ali67 unter (itarbeit von 8. Te9ly :bersetzt, eingeleitet un" er;l<rt von =* 8reutel7 >?smanis'her @es'hi'htss'hreiber, B" #A) @raz, Wien, 8.ln, #$B#, !.%#f). *ootnote in Rieser, )A).

@3

@#,### families, but some authorities cast doubt on this figure 8$ch-idt, )ose5 23HI J2@ -erbs were accepted within the expanded bounds of the $absburg &mpire and granted extensive liberties in return for forming the &mpireLs first line of defence against the >ttomans. )enewed hostilities resulted in a $absburg victory at the Battle of -lankamen .*2.3.*E2*5, where the (rand Pi9ier was killed, and another further south, at ,iJ. The first (erman colonists, artisans, were settled in +eterwardein fortress by *E2@ .Die Donauschwaben .335. *E24 saw a temporary conquest of -arajevo and, after initial Turkish successes in the Banat, their crushing defeat at Tenta .on the Tis9a5 by +rince &ugfne of -avoy .+rin9 &ugen5, despite the $absburg armies being hampered by limited revolts and even defections of whole units to the Turks. The +eace of 7arlowit9 .now -remski 7arlovci5 of "E th Ganuary *E22 formalised a still unstable situation, wherein the $absburgs were confirmed in possession of the Ba!ka and central $ungary, Transylvania and -lavonia, along with the peripheral towns of the Banat . Petersen et a,(, *2<< <#*5. In order to secure the communication of TimiFvar for the Turks, the $absburgs nevertheless had to relinquish >rFova and =ehadia, %aransebeF, Bugoj and Bippa, along with some other territory held since *E32. -erbia and Belgrade were retained by the +orte, which prevented the -erb refugees in $absburg territories from returning home.

Dith the $absburgs transfixed by the Dar of the -panish -uccession between *4#* and *4*<, and the Dar of Independence in $ungary and Transylvania .*4#<0**5, centred around )6k8c9y Merenc9 II and the K7urut9L militia, and the >ttomans increasingly concerned with )ussia, an 'ustro0Turkish equilibrium began to emerge, roughly along the lines of the Canube, Tisa and =aros rivers. The historical trans0Canubian connections of the Ba!ka had been furthered by the disintegration of the old $ungarian kingdom and, when the province came under $absburg administration, these links had become dominant. 'fter *E22, they would never quite be broken, despite all that the )6k8c9y rising could achieve. In *E23, B6cs megye .county5 was, despite -erb resistance, @2

nevertheless re0established, and the 'rchbishop of 7alocsa reinstalled as >bergespann ..$un. Mdisp6n .eqv. Bord Bieutenant55 Petersen et a,( <#*5. *E22 saw the re0establishment of Bodrog megye in the Kas good as uninhabitedL north0eastern part of the Ba!ka 0 which had not been part of mediaeval Bodrog .Ibid5. >nly a few descendants of the original $ungarian landowners were able to realise their claims, and most of the land was in consequence taken over by the state, and distributed as estates to loyal servants. The Tis9a and Canube banks were allocated to various -lav0dominated =ilitary Mrontier units, facing -lavonia, -rem .-yrmia, -9er;ms;g5 and the Banat. The interior towns of -ombor and -ubotica were considered exclaves of the military frontier in the county lands. The development of the two countiesL civil administration was thus hampered by denial of appropriate civic centres, proximity to the border and the defining voice of the military authorities. Murthermore, they were also dominated by the Pienna Ho!kriegsrat and the privileges of the -erb militia. Minally, the absence of a $ungarian nobility with local tradition weakened the possibilities for a fully developed civil administration. -erb militia were also established along the line of the =aros, the northern limit of the Banat, partly to frustrate $ungarian 0 Turkish co0operation against the $absburgs. ' powerful fortress was built at 'rad and garrisoned by (ermans. &fforts to introduce a new administration for the parts of the Banat still under $absburg control after the +eace of 7arlowit9 .*E225 were frustrated by the continuing Turkish hold over TimiFvar. ,evertheless, %atholic Bulgars were settled in Pinga during the reign of Beopold I .d. *4#A5, some -erbs from +atriarch IpekLs following were settled temporarily in the south. There was additionally some settlement from across the Tisa, and a small number of ex0military (erman invalids were also settled in Pinga.

The formerly flourishing economy of TimiFvar and its dependent villages .-ee K(er TR e von #emesch ar') collapsed during this period, due to the heavy demands of the Turkish military for supplies, the depredations of the various armies and militias disputing possession of the land, and the flight of population. The campaigns of *E34024 were fought A#

substantially over the possession of the Banat and, consequently, the population of the province virtually disappeared. The $absburg armies applied K-corched &arthL tactics, particularly around TimiFvar, which they had isolated between -eptember *E33 and >ctober *E2#. >sman 'ga, an >ttoman interpreter from TimiFvar captured in -ummer *E33, described the environs of the city thus Kabout three hours short of TimiFvar we camped in a forest, as most of the )aXcyc .non0=uslim population5 of the province of TimiFvar had even at this early date fled in all directions, and their villages lay there abandoned. The whole area was as though extinct and also most unsafeL .. (er :e!angene der :iauren., <@5. Mootnote in )ieser, A*5. Criven by scarcity of food, wolves were even attacking people around TimiFvar . (er TR e von #emesch ar, "35. Np until *E2#, a proportion of the Krespected Islamic citi9ensL remained in the city, but by *4#3, these had been supplanted by mere KrabbleL and refugees .Ibid. E35. =uch of the territory was in consequence chronically under0cultivated. The repeated destruction of the 'ustro0Turkish0$ungarian wars had compelled those living in the theatres of war to adopt a mobile pattern of existence, and thus in the Banat, Ba!ka, =oldavia, Dallachia, inner $ungary and even the Pienna Basin, established arable communities dissolved and were supplanted by a pastoral population.

In *4#A, after the beginning of the 7urut90rising, about 3,### $ungarians took refuge in the Temesk:9 from the $absburgs and -erbs of the =aros marchland. They settled mainly in the Bugoj0%aransebeF Banat, still attached to Transylvania, but some in the Turkish0occupied territories .B*ch-ann5. The >ttomans did what they could to encourage this KcolonisationL until *4**. In the Ba!ka, however, the 7urut90rising brought about the general destruction of the $absburgsX various -erb colonies. Being allies of Pienna 0 having defeated the 7urut9 at $alas 0 they attracted the attention of )6k8c9y Merenc9 II himself who, in *4#@, devastated the Ba!ka and drove them to -rem and -lavonia . =eid,ein, *AE. $>entk,Kray, **45. In *4*#, 7unbaja, the last surviving settlement of the old $ungarian Ba!ka, was destroyed by 7urut9 militia . =a,>, <E5. A*

The recurrent warfare led to a gradual concentration of the population in a few, fair0si9ed settlements, while the smaller, more vulnerable, outlying settlements were progressively abandoned. &xtensive territories thus came to be extremely thinly inhabited, and acquired over time the $ungarian epithet Kpus9taL .desolation5. Bandowners apparently initially retained both title to and occasional use of their vulnerable lands, and exercised these from the relative safety of the towns. -urviving settlements thus became associated with pus9tas. In *4"#, for example, Baja included twelve pus9tas within its bounds, of which only four were cultivated .=a,>: LH5. The limits .(er., :emarkungen5 of these pus9tas, while originating in a lost past, being set during the mediaeval period by royal gift and purchase retained an influence over the later colonisations. Curing the *3th century process of colonisation, (emarkungen were often amalgamated, as new the villages were larger than their mediaeval forebears, but served nevertheless as a guide. >nly in the -chwgbische $eide .in the north0west Banat5, where much land became available only after the *3th century drainage0schemes, does one find regularised bounds .see Billed, -6ndorh69a5.

The end of

the 7urut9 rising

brought the full

incorporation

of

Transylvania into $absburg $ungary .*4*<5, and coincided with the K+ragmatic -anctionL. This passed the succession to 7arl PILs .*4**0*4@#5 daughter =aria Theresa, and sowed the seeds of future conflicts. =eanwhile, after >ttoman successes against )ussia .*4**5 and Penice .*4*A5, the -ultan refused to acquiesce in a $absburg demand to restore the status quo ante. The war that followed ended disastrously for the Turks. In early 'ugust *4*E, &ugfne won a great victory at +eterwardein and marched straight for TimiFvar. The siege began in mid0'ugust. It was defended by *3,### men, including A,### $ungarians under the command of Imre Th:k:ly, onetime leader of the 7urut9 rebels and the TurksX increasingly unfortunate candidate for the +rincipality of Transylvania .Rieser, A*5. Initially, Kwhen requested to surrender, the +asha who was defending the place replied that he knew perfectly well that he couldnXt win, but that he felt it his duty to contribute to the A"

)enown of +rince &ugfne by making his victory more arduous and gloriousL. .!agris: <#@.5 +rince &ugfne himself accepted that his capture of TimiFvar had not been achieved without inflicting heavy damage the bombardment of the timber0built city appears to have been sufficient in itself to bring about its surrender on *<th >ctober. 3 The suburbs had been burned down, and the interior of the fortress

3Better from +rince &ugfne to the 7aiser. KDhat moved this numerous and formidable garrison to
surrender can, so far as one can ascertain, be sought only in the continual cannonade and bombardment, which enkindled an indescribable fear and allowed no rest either day or night in the narrow alleys and their wooden houses, where many people were killed and wounded. 'part from this, no explanation can be offeredL. .-9entkl6ray, **45 Translation from the (erman by B=5. The generous terms offered to the +asha, detailed in full on -9entkl6ray *"402 may have offered a further incentive.

A<

Temesvar in *4*E

A@

reduced to matchwood.. Rieser, A*5. Mive days later, the Turks were permitted to leave the city. The ,ew 1ear saw a further $absburg victory at Belgrade. The +eace of +assarowit9 .+oQarevac5 that followed on "* Guly *4*4 confirmed earlier $absburg gains and additionally detached the whole of the Banat from the >ttoman &mpire. Mive districts of northern -erbia and Bittle Dallachia .>ltenia5, also came under the $absburg &mpire.

Nnlike the Ba!ka, the new province was extremely rich in minerals and its key strategic position across several passes made its immediate re0 attachment to the restive $ungarian territory undesirable. >n <#th Cecember *4*E, just two months after the capture of TimiFvar from the Turks, &mperor Beopold therefore declared the creation of a =ron2 und =ammerdomLne #emeser Banat , subordinated directly to Pienna. Nntil *4*3, the Banat highlands continued to be viewed as constituting a distinct and separate -everiner Banat. Thereafter, the term Temeser Banat was applied to the quadrangle enclosed by Canube, Tisa, =aros and %erna rivers .Cipling, *2E45. Nntil *4<2, $absburg maps of the Banat included the north -erbian provinces, but they were not included within the term and it appears they never acquired a separate identity of their own within the $absburg &mpire, as though their possession were accepted as being temporary. 2

>ne of +rince &ugfneLs most successful subordinates, (eneral (raf %laudius Mlorimund !ercy, had already been appointed civil and military governor of the conquered Banat in ,ovember *4*E. $e was to hold this post until *4<< and to have a decisive effect on the future development of the province. >ne of his first concerns was to set up the KBanater CinrichtungskommissionL to prepare the development of the new province. This was replaced in *4*3 by a &Tandesadministration des #emes arer Banats' , subdivided into more than a do9en administrative
$ !ee, for instan'e, the &Mercy%'arte6 #%2C D #%2)7 +2harte von "em Temes/arer Bannat samt "enen ihm in'or9orierten ) ,istri'ten, so zum 8.nigrei'h servien geh.ren6 >Banat im Eahre #%2C)7 Briffaut Theatre "e la guerre "ans le Bannat "e Temesvar, #%CB)

AA

districts which would remain largely intact for the duration of the %ameral Banat. ' KBanater Bergenrichtungskommission L was also set up to further the exploitation of the mineral0rich highlands and was, in its turn, replaced in *4"< by the Banater >berbergamt, based in >rawit9. >n the instructions of the Pienna $ofkammer, =ercy was to repopulate and cultivate the new province and stimulate manufacturing and mining. $e was chiefly responsible for the first of the Sch a"en+Uge .immigrant waves of (erman5, and thus for the introduction of about *"0*A,### colonists between *4"" and *4"E. =any (erman settlements were founded in the south of Banat .see below5, and substantial progress was made towards the drainage of the Banat through the building of the Bega K-hipping %analL between *4"3 and *4<<. The name of =ercy is also associated with the first detailed map of the Banat, prepared between *4"< and *4"A, published in Pienna in *4"3 by military cartographer Haring . The Banat and its mineral wealth were reserved entirely for the state until *4<E, when the first opportunity arose for private persons to acquire mining rights, albeit the state protected its monopsony over the minerals extracted and processed. Dhile the (erman coloni9ation was principally concerned with economics and with the securing of communications, the (ermans were not destined to serve in the actual defence of the province. This was allocated to the mainly -erb militia which, in *4"@, was established along the Tisa0=aros .the right banks of those rivers, facing the Banat5 stretch of the military frontier, and part of the Canube section. In *4<A, however, an insurrection broke out among the -erbs, and belatedly the $absburg administration began to consider involving the (ermans in their own defence.

In *4<3, 'ustria over0hastily declared war on the >ttoman &mpire, under pressure from %9ar +eter the (reat to comply with a treaty of mutual assistance. -mall0arms were issued hurriedly to (erman villagers in the south of the Banat, and along the =aros, but lost time could not be so swiftly made up, nor soldiers so easily made out of artisans and peasants. By the beginning of *4<3, the Turkish reputation for brutality AE

had persuaded most to desert. ' Turkish army broke into the Banat from the south0west .the Kdisaster of >rsova and =ehadiaL5, but declined to venture forth in strength much beyond 7ar6nsebes, despite a brief foray along the Canube towards +ancsova. 's the (erman villages founded thus far had been so located as to secure the principal lines of communications, denuded of defence, these were ravaged by Turkish and associated .principally Plach5 forces. 'lmost the entire (erman population fled from the south most who did not were killed or enslaved. ' mild outbreak of plague further terrified all sections of the population. Dhen the war was brought to an end with the +eace of Belgrade .*4<25, 'ustria was humiliated by the loss not only of Belgrade itself, but also northern Bosnia and the north -erbian districts with the lucrative =aidanpek copper0mining district, which placed the Banat once again directly on the frontier. The island fortress of ,eu0>rsova .'da 7aleh5 was also surrendered to the Turks and in this case a Turkish population was to remain until at least the *2<#s. The failure of civil defence had contributed largely to the collapse of the Kkarolinische 'nsiedlungL .coloni9ation carried out during the reign of 7arl PI .*4** 0 *4@#5 and after the war, the (erman settlement0area of the south Banat would never be re0established on the same scale. .

A4

'aro,inische Ansied,*ng

A3

-ubsequent to the disastrous Turkish war and with the accession of =aria Theresa .*4@# 0 *43#5, the $absburg military frontier was reorganised. The west and south of the Ba!ka, being covered by the continued 'ustrian possession of -lavonia and -rem, remained under $ungarian county administration. 't the Tisa0Canube confluence, a -erb0 dominated military district .the K #schaikistendistriktL5 was allocated to the Canube flotilla. The dissolution of the obsolete Tis9a0=aros military frontier, rendered obsolete already by +assarowit9, was begun in *4@*. This stimulated a major -erb migration, partly into expanded military districts along the Banat Canube frontier, where a large number of -erb villages were founded, and partly to )ussia. 'gents of the %9ars, recruiting for the colonisation of K,ew )ussiaL, were particularly successful competitors for the footloose frontiersmen. The migration of the -erbs meant that the lower Tisa valley had to be substantially re0settled.

*4A* brought the establishment of civil administration in the K%ameralL Banat .excluding the military 9ones along the Canube5 and the raising of -ombor and bjvid;k .,ovi -ad5 in the Ba!ka to the status of )oyal Mree %ities. -9abadka .-ubotica5 under the name =aria0Theresiopel, acquired a similar privilege only in *442. The last vestiges of the =aros military frontier were dissolved in *4A", but it was not until *4E@03 that a new -erb military frontier was formalised along the Canube. The mining estates were still administered separately within the districts, and their inhabitants referred to as 5ontanisten in contradistinction to the =ameralisten elsewhere these labels had no bearing on nationality or language. ' new -dministrations.rLsident, the -panish0%roat Pillana0 +erlas, was appointed in *4A<. Nnder his leadership, the -erb 0 (reek cattle breedersL lobby became dominant and, locally, a complete break was made with the former policy of favouring colonisation by (ermans and the transformation to an intensive, agrarian landscape. The resulting friction with central $absburg authorities lasted throughout +erlasL long tenure .he was removed from office for KcorruptionL in *4E35 and, while

A2

=aria Theresa reigned alone .Goseph II was appointed co0regent in *4EA5, the $ofburg was unable to exert much influence in the Banat.

>ne result of the -ilesian Dar of *4A3 had been the mortgaging of the %ameral province to the Diener -tadtbanco and the bank resisted any longer0term initiative that might imperil the provinceLs present revenues. 's a great part of those revenues derived from the -erbsL extensive pastoral economy, upon which the food economy of the $absburg &mpire was coming to depend, the bank opposed colonisation on the almost uninhabited praedien .similar to the pus9tas of the Ba!ka and inner $ungary5 where most of the cattle0rearing was practised. +raedien were mainly located in the waterlogged western part of the Banat, and included completely abandoned and derelict settlements from the pre0 $absburg period, settlements devastated during the more recent fighting, and partly inhabited villages, as well as dairies, waterlogged meadows and marshes. The reasons for abandonment of a settlement might be varied and included the direct impact of war/ epidemic/ tax evasion, or desertion. The bankLs insistence that the extensive cattle pastures remain untouched by colonisation underlined the urgency of canalisation and drainage of the BanatLs substantial marshes .completed *4E4025, which increased the amount of land actually available. The (erman colonisation of the -econd K-chwaben9ugL .*4E<0*44#5 was directed substantially towards these new lands. Dhen Goseph II KsanitisedL the state finances by donating his inherited lands to the state .*4E35, he was able to pay off the Diener -tadtbanco, opening the way to a more progressive, and intensive, scheme of development.

In the declining years of =aria Theresa, the history of the Banat as a cameral province draws to a conclusion, and a $ungarian county administration, which had been introduced before the end of the previous century in the Ba!ka , was at last established. In *443, the BanatLs provincial

E#

%heresianische Ansied,*ng

E*

administration was dissolved and replaced by a &Gommissio Regia @ncor.oratia'. Its function was to oversee the transfer of the cameral Banat, excluding the Banat military frontier districts set up between *4E@ and *4E3, to $ungarian county administration, guide its political and legal integration and manage the sale of the %ameral estates. The counties of Temes, Toront6l, and 7rass8 were resurrected in *442. The reign of Gosef II .*43#0*42#5 was characteised by the so0called KIose.hinismus' , which coloured and sought to strengthen the hand of central authority against that of provincial administrators. This included Goseph first elevating Temesv6r to the status of )oyal Mree %ity .*43*5, and then abrogating the counties and unifying the Banat and Ba!ka in a KTemeswarer 7reisL .*43A5 so as to benefit the %ameral colonisation mechanism. Gosef II is reported to have made several incognito tours through the Banat .$ch-idt &(as Banat und die Banater Sch a"en' in
"asse,, $ch-idt: -n (onau und #hei> *4A5. The new administration was

very unpopular with the $ungarian nobility and repealed in *42#, the previous KTheresianL system being reintroduced. GosephLs reign also brought the third and final K-chwaben9ugL .*43"0*4335. Dhile the first K-chwaben9ugL stimulated mining and industry and the second concentrated on increasing the physical population, the main objective of this colonisation was the improvement of agriculture by introducing colonists from areas where the most modern methods prevailed. GosephLs centralist innovations so irritated the $ungarian nobility that almost all had to be reversed during his reign, and his successor, Beopold II .*42#0*42"5 abandoned the policy altogether and declared $ungary to be a Kfree and independent kingdom under its own lawsL .NI II ""5.

In *434, a new 'ustro0)ussian war with the >ttoman &mpire broke out and a rising organised among the northern -erbs. ' $absburg army was, however, defeated at -latina in the eastern Banat. Turkish forces advanced to near Temesv6r and *@< settlements were ravaged, including the most important mining centres. The Turkish booty included <E,### persons, although this time, most of the (erman settlers were E"

evacuated in time, so a repeat of *4<3 was avoided. Dhen the Turkish forces were driven out of the Banat in *432, they were never to return. The +eace of -viJtov .*42*5, forced on 'ustria by the pressure of the )evolutionary Mrench wars, and Britain, brought hostilities in the 'ustrian theatre to an end. The Banat was not to see warfare again until *3@3. To the south, however, the >ttoman &mpire was increasingly caught up in the -erbian struggle for independence, and gradually lost its hold over the Canube frontier. >nly the preoccupation of )ussia with the ,apoleonic invasion of *3*" permitted the -ultan to suppress the 7arahorhe rising of *3#@0*3*<. The ultimate success of the >brenoviU rising may be attributed in part to the final defeat of ,apoleon .*3*A5 liberating the %9arLs hand in dealings with the >ttoman &mpire. Dith the retreat of the Turks, the military frontier became less of a necessity, but the SanitLtskordon, which had supplemented it along its length from *44#, remained in place )osephinische Ansied,*ng

E<

despite its restrictive effect on trade. ' glimpse of the SanitLtskordon in action at >rsova, and the real fear of the plague on both sides of the frontier, may be grasped from the detailed description in 7ohl. Cespite its effect, however, imports to the $absburg &mpire from the Balkans exceeded imports by a factor of five to one .Bampe, <@5.

The Banat and the neighbouring Ba!ka have, at least since the $absburg conquest, been lands of almost perpetual colonisation. )egular change has been one of the few constant factors. %olonists have been brought from every part of &urope, even from todayXs Mrance, -pain and Italy, but the demand for new blood has never been satisfied. In the eighteenth century, the Banat in particular became infamous for the toll it extracted from the colonists. +lague, malaria, recurrent war with Turkey 0 all hit the newly arrived colonists perhaps more heavily than the more established locals, already inured to the hardships of life in the region. 1et colonists still came. &migration and the abandonment of the familiar are themes that arouse strong emotion and are not undertaken lightly. Nnemployment, lack of future prospects, high taxation, religious persecution and the ha9ards of warfare all too often contributed to loosen individualsL bonds to the homeland and, coupled with the illusory freedom of the frontiersman, impelled them and often their family to move abroad. In south0western (ermany, as in Ireland, the tradition of sub0division of inherited plots had contributed to the alienation of a proportion of the rural population and to the drift to the towns. The surplus British, Irish, Mrench, -panish, +ortuguese and Cutch urban populations were drawn off to the ,ew Dorld and other colonial possessions. Mor the citi9ens of the $absburg &mpire, after the split with -pain, and of +russia, there was no such extra0&uropean outlet, but for that a relatively thinly populated east.

Mrom the late *4th century, $absburg 7aisers, newly possessed of extensive, but thinly inhabited and therefore vulnerable territories in the south0east, being aware of social imbalance and land0famine in south0 western (ermany and, constrained by economic and military E@

imperatives, had themselves issued invitations of varying generosity for colonists. The embroidered versions of these invitations broadcast by the 7aisersL agents, coupled with misery at home and supposedly greater opportunities for land, farming and trade, had persuaded would0be colonists to leave home in their thousands. It is worth noting that the greater part of the *3th century (erman emigration came not from the smaller kingdoms and the petty principalities of the central (erman space, but from the medium0si9ed territories of Baden, DRrttemberg and the )hineland +alatinate, where life was continually threatened by the encroaching Mrench 7ingdom. This was particularly true during the predatory reign of Bouis ?IP 0 and from Bavaria, which had been a major victim of the Thirty 1earsX Dar. The impossible confusion of ThRringen, which might be expected to have produced Beopoldinsche'nsiedlung

EA

great numbers of colonists due to the impotence of the numerous petty princes, barely figures in the list of source territories for Banat colonists. The colonisation of the Banat was however far from mono0ethnic in nature although $ungarians were initially held at armXs length for political and partly confessional reasons, )omanians and -erbs were major beneficiaries of $absburg policy besides the (ermans and both played central roles in the Banat economy. Bulgars, (ypsies and other more minor groups were also welcomed. This does not mean that there were not frictions between the various ethnic groups. In (erman literature, one often comes across recriminations against the major population0groups 0 principally the -erbs and )omanians 0 for a lack of enthusiasm for the major construction and mining works. But the $absburg way of life was in conflict with their pastoral ways, so they could not be realistically expected to contribute to their own demise. Mriction between (erman and -erb would strongly influence the process of colonisation throughout the *3th and *2th centuries.

The patchwork quality of the $absburg lands was accentuated by the nationalisms stimulated by the Mrench )evolution. 'ustriaLs preoccupation with the )evolutionary Dars and her numerous defeats at the hands of ,apoleon, permitted her constituent nationalities more freedom to observe their own particularities. The short0lived Illyrian +rovinces .*3#20*"5, for example, stirred previously dormant awareness of a common -outh -lav heritage and proclaimed the possibility of a common political future.

The $ungarians had also benefited from the humbling of Pienna, and $ungarian Banguage Baw of *3"4 served notice that this was henceforth to be the official language of +arliament. 'fter decades of pressure on Pienna, the $ungarian language finally replaced Batin as the official medium of communication in *3@# for all offices of state. 1et the =agyars resisted any other national groupLs political aspirations, even those of the %roats, their longest0serving allies. It was made a crime to EE

mention Illyria in public and, in *3@<0@ it was decreed that $ungarian would become the official language of %roatia by the end of the decade .NI II "A5. Thus the evolution of a particularist =agyar $ungary served to focus the attention of the non0$ungarian0speaking communities on their own nationalisms, and to undermine the doctrine of the Bands of -t -tephenLs %rown .see van Beisen, *2@< &der ungarische Reichsgedanke Ste!ans des Heiligen&5. ' perceived denial of the opportunity for self0 expression by the increasingly exclusive and assimilatory =agyar nature of the state and its administration stimulated a self0awareness and resistance from the other nationalities. In regions such as Banat and Ba!ka, where these made up the great majority of the population, the temperature gradually rose to dangerous levels.

In the tumultuous year of *3@3, the anti0$absburg =agyar revolution 0 initially supported by the Pojvodina -erbs 0 was shadowed by an anti0 =agyar -erb rising in the Banat and Ba!ka, led by their >rthodox metropolitan .Bampe @E5. >ne of the chief aims of the ultimately unsuccessful rising was the creation of a -erb0dominated territory within the $absburg &mpire under the jurisdiction of a -erb Pojvoda .civil governor5. The -erbsL right to elect their own Pojvoda had been conferred by Beopold I .*EA30*4#A5 and confirmed by subsequent 7aisers, but never translated into reality. Curing the year0long struggle, during which the -erbs of the Banat and Ba!ka were involved in long negotiation with the $absburgs for concessions in return for armed support .7;9, *2@"5, -erbs and %roats .under Gela!iU5 fought side by side for the 7aiser. The $ungarians, meanwhile, received at least passive support from the (erman mining communities of the south0eastern Banat, which had been impressed by the feats of $ungarian engineering, particularly the K-9;chenyiL road from >rsova to Zmoldova along the 7a9an stretch of the Canube valley . Hro-adka @A, Ba*-ann @*5. ' (erman civil guard was raised in )esica after its occupation by the $ungarian honv;ds and sought vainly to resist )omanian counter0 insurgents. The (ermans of the plain, and particularly of Temesvar, however, remained loyal to the $absburg crown, and the city withstood a E4

*##0day siege before being relieved. The $ungarian KinsurgentsLK final defeat was at ,eu0Beschenowa, a few kilometres distant. >n *" th 'ugust *3@2, the $ungarian surrender was signed at Pil6gos .$ellburg5, just outside 'rad. 7ossuth began his exile from $ungary on *E th 'ugust, crossing the Canube near >rsova after concealing the crown of -t. -tephen in the riverbank.

%he =o0;odscha5t $er9iens *nd %e-eser Banat 'fter the $ungarian defeat, the new emperor Mran9 Goseph II in ,ovember *3@2 acknowledged the services of the -erbs and %roats by uniting Banat and Ba!ka with part of -irmium in a D oj odscha!t Ser"iens und #emeser Banat .see map5. The =ilitary Mrontier, and thus a large proportion of the &mpireLs -erb population, was still excluded, and (erman was made the official language of administration. The large )omanian population of the eastern Banat further diluted the -erbsL influence within the Dojwodschaft. 'n enumeration of the population was made in *3A@. The published data reveals the total population by settlement, but not the linguistic or national breakdown. The closest language0data was from *3@#, which reported linguistic majorities .M;nyes5. The two datasets may be compared on the following map, but a note of caution is necessary. The two datasets, firstly, are fourteen years apart and, secondly, are either side of the *3@302 revolution, which brought great destruction and disruption to the population. There are bound to have been great differences in the populations on which the two surveys reported.

E3

*3@# i *3A@

E2

The autonomy of the Dojwodschaft was severely restricted by Pienna and even members of the local (erman communities were kept away from positions of responsibility. The -erbsL political ambitions were in no way satisfied by this political fiction, which served only to make the rule of Pienna more direct .La-pe, @E5. 'n imperial diploma of >ctober *3E# abolished the Dojwodschaft -erbiens und Temeser Banat, restoring the $ungarian county administrations of Temes, Toront6l, and 7rass8 in *3E*, although they retained a K(ermanL administration until the 'usgleich of *3E4. (y:9d .*2@", A35 terms the lifetime of the Dojwodschaft as a period of Kforeign occupationL.

Cue

in

part

to

increasing

defections

of

:ren+er,

or

:ranicari

.frontiersmen5 to -erbia, the military frontier was abolished in *34*, with the exception of the K #schaikistendistrikt' , which was retained until *34"0 <. 7rass8 now became 7rass80-9:r;ny county and, like Toront6l, Temes and B6cs0Bodrog, was extended to the banks of the Canube. It is this administrative arrangement that $ungarians popularly term KhistoricalL. Mrom now until *2*2, the terms Banat and Ba!ka lost any political significance, and were no more than geographical names. Nnder successive $absburg administrations, $ungarian settlement in the Banat had been restricted while (ermans, -erbs, )omanians and others had been welcomed. Curing these decades, Budapest sponsored an intensive $ungarian colonisation of the regions and the $ungarian population increased rapidly, a trend assisted in no small measure by magyarisation of the minorities. %ommunications were strengthened with Budapest and the interior of $ungary, but not developed towards the south.

-erbia, growing into independence south of the Canube, drew many of its leaders and intellectuals from the $absburg Banat and Ba!ka .Pojvodina5. These regions had become the cultural centre of the -erbs 0 ,ovi -ad acquiring the name K'thens of -erbiaL .Bampe, <45 and remained so even after the attainment of independence. Pojvodina -erbs had a special disparaging term for those -erbs born south of the Canube 4#

0 i.e. within Balkan -erbia 0 ged+a .=agris, (anu"e, "235, while they themselves were known in the south as .reani 0 Kthose from across the riverL . Bampe @35. ,evertheless, more united the Pojvodina and Balkan -erbs than divided them and many .reani moved south and engaged themselves in the construction of -erbia. There was, therefore, considerable dismay among the -erbs as a whole when 'lexander 7arahorheviU .*3@"0A25 was compelled by the fears of both 'ustria and )ussia to adopt a policy of neutrality in the 'ustro0$ungarian conflict of *3@302. (rowing distrust between the 'ustro0$ungarian &mpire and -erbia stemmed in great part from the relationship between the Pojvodina and the Balkan -erbs, the intellectuals of the former encouraging the martial exploits of the latter. $absburg efforts to impose %atholicism or the Nniate creed during periods of occupation, and rapacious economic exploitation when Pienna had the upper hand .e.g. the =aidanpek copper mining complex5 undermined any trust the -erbs may have had for the $absburgs. Irritation over the long failure to provide the promised Pojvoda and the contrived fiction of the Woj odina Ser"iens und #emeser Banat was compounded by $absburg economic domination of -erbia. The occupation of Bosnia0$er9egovina after the %ongress of Berlin .*3435 and its outright annexation in *2#3, together with 'ustrian occupation of the -anQak of ,ovi +a9ar until *2#2 intensified these feelings. $absburg and -erbian economic needs ran across each other, -erbia seeking access to the 'driatic because of a crippling dependence on the $absburg market for her products .mainly cattle and pigs5, 'ustria the route to the deep Balkans. The outstanding -erbian and =ontenegrin successes in the Mirst and -econd Balkan Dars .*2*"0<5 blocked the 'ustriansL last route to the south by the division of the -anQak, and the inspiration of the $absburg -erbs was such as to threaten the tranquillity of the &mpire.

The steadfast resistance of the -erbian army until >ctober *2*A and then its extraordinary retreat across the mountains to %orfu, and the occupation of the defeated country served to unify -outh -lav sentiment. The wartime losses of the -erbs amounted to one half of the men of 4*

military age .Bampe, *#45. The initial success of the Brusilov offensive in Gune *2*E had persuaded )omania, despite the defeat of -erbia, and the danger of Bulgaria, to throw in her lot with the 'llies and on *4 th 'ugust *2*E signed a secret treaty with the &ntente powers. The catastrophe of the -omme had just begun, and the British and Mrench were desperate to secure some relief from the deadlock on the Destern Mront by any available means. In recognition of )omaniaLs desertion of the Triple 'lliance and declaration of war against the %entral +owers, the &ntente, come the peace negotiations, would support their allyLs claim to )omanian0inhabited territories of the $absburg &mpire. This was held to include the whole of the Banat. )omania was required not to sign any separate peace with the %entral +owers. Maced with total defeat and occupation of )omania, however, Brctianu .+rime =inister of )omania5 did just that in =ay *2*3. Ceclaring renewed war on the %entral +owers .,ovember 2th5 on the grounds of their violation of this Treaty of Bucharest unfortunately acknowledged that this was a valid instrument of international law .Nicho,son , *2<4 *<A5. The 'llies were therefore able to reject the -ecret treaty of *2*E as void. In the meantime the allies had accepted a 1ugoslav claim to the Banat, based on the Ceclaration of %orfu ."4 Guly *2*45, which stated that the territory of the -erbs, %roats and -lovenes included all those areas Kin which our three0 named people resides in compact and coherent massesL . Anna9ring *2AA, <5.

Banat and Baka in the B$*ccessor $tatesC 'fter summer *2*3, the %entral +owersL resistance rapidly disintegrated. +resident DilsonLs refusal to negotiate with 'ustro0$ungary unless 1ugoslav and %9echoslovak unity were recognised instantly invalidated belated $absburg moves towards federacy. The former $absburg territories 0 the Pojvodina, -lovenia, %roatia and Bosnia0$er9egovina 0 were stampeded into an over0hasty union with -erbia and =ontenegro, under the rule of the -erbian king. Dhile dreamed of by intellectuals for some time, the exigencies of the immediate end0of0war situation caused many very difficult issues to be addressed with unconsidered haste 4"

.Bampe, *#30**#5. The creation of the 7ingdom of -erbs, %roats and -lovenes was declared on @th Cecember *2*3, but federalist hopes were soon dashed in the face of -erb centralism encouraged in part by border disputes with Italy. The disintegration of the $absburg &mpire left $ungary particularly vulnerable to territorial claims, as about half of the population of Khistorical $ungaryL was non0=agyar. 's the first $ungarian government was supplanted by a revolutionary %ommunist cabinet under B;la 7un, at a time when the entente powers were involved in intervention against Bolshevik )ussia, the chances of a fair hearing for the $ungarians were diminished. 's it was, $ungary lost approximately half its population and two0thirds of its territory to the Ksuccessor0statesL 0 %9echoslovakia, the 7ingdom of -erbs, %roats and -lovenes, and )omania. &ven 'ustria was rewarded with $ungarian territory. These losses were to create of $ungary a nation0state, but considerable $ungarian minorities would be left within her new neighbours.

-erbian forces reached Temesv6r on *2th ,ovember *2*3, and had soon occupied the entire Banat and begun to install a -erbian administration. Mrench protests resulted in a -erbian withdrawal from the eastern Banat, which was then occupied .Ganuary *2*25 by Mrench troops pending the decision of the peace conference. 'lthough the )omanian case in +aris was done no favours by the abrasive Brctianu, Tak; Inoescu .leader of the )omanian %onservative Cemocrats5 had met with +asiU, the 1ugoslav +rime =inister, and negotiated a division of the Banat between the two countries. BrctianuLs antipathy towards Inoescu, however .he was not included in the +aris delegation5, ensured that )omanian01ugoslav rivalry over the Banat would continue, and even led to threats of renewed hostilities. The Banat border with )omania was nevertheless one of 1ugoslaviaLs least problematic. -ubsequent condemnation of the +aris +eace Treaties stemmed in part from dissatisfaction among the victors about the new borders and how these were arrived at. Dhile the defeated nations had no reason at all to cheer their new territorial bounds, disputes between the victors 0 Italy and 1ugoslavia for example 0 showed up the pressures under which the Persailles diplomats were 4<

working. ,icholson, a young diplomat at +aris, decries the conferencesL &a..alling dis.ersal o! energy' .Nicho,son, A5, its &ama+ing inconseNuence Fand) com.lete a"sence o! any consecutive method o! negotiation or even im.osition' . In a letter to his father ."A th Mebruary *2*25, he also speaks of &the im.ossi"ility o! e<tracting !rom the lies ith hich e are surrounded any real im.ression o! hat the various countries and nationalities honestly desire3 With the time allo ed to us e can only come to more or less em.irical decisions, and the only justi!ication !or such decisions ill "e the honesty o! intention in hich they are conceived'3 )omanian and -erbian negotiators for the Banat issue were first brought together in +aris in Mebruary, and talks continued acrimoniously, with various border solutions being offered by Britain, the N-', Italy and Mrance, the last of which was reluctantly accepted by BrVtianu in Gune. )omanian forces began to enter the eastern Banat on "*st Guly *2*2. The Treaty of -fvres .*# 'ugust *2"#5 clarified the border, but eleventh0hour Mrench objections resulted in Persec .PrJac5 and Meh;rtemplom .Bela %rkva, Biserica 'lba5 being reallocated to the 7ingdom of -erbs, %roats and -lovenes. &ven then, further minor adjustments were made 0 probably to compensate )omania for the loss of these two towns 0 and were codified in the %onvention of Belgrade ."@th ,ovember *2"<5.
*#

The division of the Banat

followed ethnic lines as closely as possible, but only with regard to the distribution of the -erb and )omanian population. >ther groups 0 and particularly the (ermans 0 were ignored. $alf0hearted talk among the (erman population of an autonomous Banat at the end of the war was rapidly submerged by both )omanian and -erbian ambition. The (erman effort was soon restricted to trying to persuade the +eace %onference not to divide the Banat, but to present it whole to one or the other party. Dhen it became known that Temesv6r had been awarded to )omania, this became the (ermansL favoured country. In the event, the -fvres0 Belgrade border cut the (erman settlement0area quite in half.

#A The 2onvention transferre"0 ForvGthD2sene an" !zerbD2sene, 8is;omlHs, Isombolya, Jregfalu, 8arGtsonyiliget, KLvGr, agyzsGm an" -a'mGs to =omania >9o9ulation 'a.CA,AAA) 7.the a'Muisitions of the 8ing"om of !erbs, 2roats an" !lovenes in'lu"e" IsGm, 4Gr"Gny an" (H"os >9o9 'a. B,AAA)

4@

Nltimately, the four counties, except for small remnants of B6cs0Bodrog and Toront6l, were detached from $ungary. >f B6cs0Bodrog, twelve settlements from Bajai j6r6s and ten from B6csalm6si j6r6s remained with $ungary. %hief among these was the town of Baja .Mrankenstadt5, %s6talja and G6noshalma. The rest of the Ba!ka, including bjvid;k .,ovi -ad5, -9abadka .-ubotica5 and Tombor .-ombor5, was transferred to the 7ingdom of -erbs, %roats and -lovenes .1ugoslavia5. ' small slice of the Banat .*Y5, comprising a compressed hinterland to the city of -9eged .Ces9k, 7Rbekh69a, Tis9as9iget, bjs9entiv6n, et al.5, was allowed to $ungary. =ost of this territory was divided between )omania and the 7ingdom of -erbs, %roats and -lovenes .*#,A## km ", <4Y5. The latter acquired most of Toront6l, a sliver of Temes .including the towns of Persec and Meh;rtemplom5, and one settlement .Ndvars96ll6s5 from 7rass80-9:r;ny. In total, this state acquired from $ungary .exc. %roatia0 -lavonia5 "#,AA* km", with a population of *,A*#,324.

)omania acquired about two0thirds of the Banat .*3,E## km ", E"Y5, comprising a fraction of the former $ungarian county of Toront6l and most of Temes and 7rass80-9:r;ny. Temesv6r .TimiFoara5, the BanatXs capital, fell to )omania, as did the major towns Bugos .Bugoj5, 7ar6nsebes .%aransebeF5 and >rsova .>rFova5, and the Banat uplands in their entirety with the vital mining areas of )esica .)eFiOa5 and >ravica .>raviOa5. )omania also acquired the whole of Transylvania and the K+artiumL .'rad, Bihar and -9il6gy counties5 from $ungary. In total, )omania acquired *#<,#2< km " with a population of A,"AE,@A* persons .$ungarian census of *2*#5. In the )omanian part of the Banat, the new 1ugoslav border made a revision of the old $ungarian county 0 and, in some cases, settlement0boundaries necessary. 7rass80-9:r;ny was split into IudeVele .counties5 %araF and -everin, while the )omanian portions of Temes and Toront6l were combined into TimiF0Torontal. Dithin K$istorical $ungaryL, 'rad had been a tRrv)nyhat,sgos jog6 vros .Tjv., +rivileged Mree Town5 and the megyes+)khely .ms9h., %ounty Town5 of 'rad vrmegye .county5, which extended north from the line of the river =aros. The old city was located on the right .north5 bank of the =aros, 4A

and, together with the settlements of 'rad v6rmegye, is excluded from consideration in the present work. The satellite town of bjarad .'radul ,ou5, on the other hand, had developed more recently on the southern bank, and formed the centre for an Kjjaradi jrsL .district5 within Temes v6rmegye. In *2"A, this became $lasa .district5 'radul0,ou in GudeOul 'rad, with the addition of several settlements from the adjacent former $ungarian lippai j6r6s, and was thus administratively detached from the )omanian Banat. Mor the purposes this study, the border of the Banat shall, however, continue to be viewed as the line of the =aros and, for the period *2"A 0 *22", so far as possible, data for the settlements of GudeOul 'rad south of the =aros shall be included and, of those in GudeOul 'rad, for these alone.

Dithin the 7ingdom of -erbs, %roats and -lovenes, the Banat, Ba!ka and KBaranja TriangleL were informally considered as one unit .the KBBBL5. Though suggested by common political traditions, the poor communications between the Baranya Triangle and the Ba!ka hindered an integration of the territories. The wildly meandering and practically uncontrolled Canube was unbridged and without even a ferry along this stretch. (ermans from the port of 'patin fleeing the Ba!ka in *2@@ had to pass northward and cross at Baja .Mrankenstadt5, nearly @# km inside Trianon $ungary.

't the time of the 1ugoslav census of *2"*, the 1ugoslav Banat was divided into eleven sre+ovi .districts5, plus the )oyal Mree Towns of +an!evo and PrJac. The Ba!ka meanwhile fell into twelve districts and three )oyal Mree Towns .,ovi -ad, -ombor and -ubotica5. ' decree of "E 'pril *2"" .effective from *2"@5 divided the 1ugoslav Banat between Belgrade oblast .district5 and +adunavlje oblast 0 south of the Canube and based on -mederevo. Petersen et a, ascribes this .p. "3*5 simply to a desire to place the (erman population of the 1ugoslav Banat within an >rthodox -erb majority. The 1ugoslav Ba!ka was united with the Baranja Triangle to form ,ovi -ad >blast. 'dministrative reforms 4E

accompanying the introduction of the )oyal Cictatorship in *2"2 established grand regions based on the major rivers of the 7ingdom in place of the earlier KnationalL divisions .i.e. -erbia, %roatia etc.5. This arrangement was intended to establish a more rational, economic base to the country, and to dislocate the various national oppositions by breaking them up among a number of administrative units. The KBBBL was transformed into the Cunavska BanoviJte by the association of the Banat, Ba!ka and the Baranja triangle with heavily -erbian -rem and -umadja. +an!evo and -emun .with strong (erman communities5 were combined into (reater Belgrade. The weight of the $ungarians and (ermans was thereby reduced, and the preponderance of -erbia within the 7ingdom of 1ugoslavia simultaneously lessened. This solution was not to last with the 'nschluss, the third )eich had become an immediate neighbour of 1ugoslavia. The continuing bitter acrimony between the stateLs constituent nations and the increasingly dangerous &uropean climate forced a -erb0%roat K S.ora+umL .'ugust "Eth *2<25 which combined the Savska and +rimorska Banovine with some portions of others into a new autonomous Hrvatska .%roatian5 Banovina. The new arrangement was soon swept aside by the events of the -econd Dorld Dar.

"ro- the $econd =or,d =ar to %oday

>n the (erman0Bulgarian0$ungarian invasion of 'pril *2@*, the state of 1ugoslavia was dismembered. ' (erman0 Italian occupation demarcation line ran down the country from north to south. =acedonia came under Bulgarian occupation and administration, and 'lbania was extended into the plains of 7osovo. The Calmatian coast was annexed by Italy. -lovenia was divided between Italy and the )eich, and the ,C$, or KIndependent -tate of %roatiaL .including Bosnia0$er9egovina5 established as a wayward 'xis satellite. In the interpretation favoured at the time in Budapest, the proclamation of the ,C$ .2 th 'pril5 dissolved the state of 1ugoslavia, thus rendering void the Treaty of &ternal Mriendship that 44

Budapest had signed with Belgrade in *2@#. In the ensuing KvacuumL, the $ungarian army .Honv)d5 crossed the border and occupied the Ba!ka and Baranja Triangle to restore another portion of $ungaryLs KhistoricalL frontiers. The twenty0two settlements of rump $ungarian Ba!ka were fused with what had become the 1ugoslav Ba!ka, and the county of B6cs0Bodrog was re0established. The 1ugoslav Baranja .the Baranja Triangle5 was reunited with the $ungarian county of Baranya. $ungarian forces also occupied the =urak:9 and =uravid;k, territories acquired by %roatia in *2*20"#, an act that soured $ungarian0%roat relations for the remainder of the war. The old term K ()lvid)kL .see above5 was re0 invented as a collective label for the recovered territories, which were formally incorporated into $ungary in Cecember *2@*. $itler had initially given $orthy to believe that the 1ugoslav Banat, too, would revert to $ungary, but vehement )omanian objections, and a threat to invade the territory themselves to forestall any $ungarian move, led to a (erman rethink and $ungarian forces were halted on the line of the Tisa. =uch as with divided Transylvania, the ultimate fate of the 1ugoslav Banat was left in question, as a spur to the performance of both (erman satellites.

-erbia was truncated and, despite the introduction of a puppet regime under (eneral ,ediU in 'ugust *2@*, remained under the rule of the (erman military. This rump -erbia included the 1ugoslav Banat, but the situation here was particularly ambiguous.. The presence of a substantial Volksdeutsche population in the 1ugoslav Banat provided a pretext, and some manpower, for direct (erman rule. =ilitary administration 0 0 was superseded in =arch *2@" when ,ediU appointed Cr -epp Ganko, Volksgru..en!Uhrer of the (ermans in the 1ugoslav Banat, head of a new civil administration officially subordinated to -erbia, but enjoying considerable autonomy.** The borders of the Banat remained sealed, however, even towards -erbia. Initially, there was not even a postal service .NI II, <3#5 . 'lthough the other national groups of the Banat

##2onfirme" "uring intervie/ /ith Ferr * Bin"er, 8onre;tor a. ". !9re'her "er Eugosla/ien"euts'hen an" Auslan"s;orres9on"ent f:r "ie ,euts'he 4resse in Eugosla/ien #$&#D& >an" 9ersonal frien" of ,r Ean;o), !e9tember #$$).

43

were involved to varying degrees in the running of the province, there was no mistaking the leading role of the the arrival of the )ed 'rmy in -eptember *2@@. Volksgru..e, nor its subordination in turn to the Third )eich. This state of affairs lasted until

,ationalism leads one to crassly oversimplify and to classify oneself and others according to \nation\, to erase grey areas and to draw indelible mental boundaries between one perceived \nation\ and another. This was carried to its logical conclusion by the ,a9is, and the Conauschwaben, being of (erman stock, were inescapably drawn into the resultant maelstrom. BerlinXs interest in the (ermans abroad .Volksdeutsche5 had brought catastrophe upon %9echoslovakia and +oland, and, despite the distance from the )eich, the Volksdeutsche 5ittelstelle was also heavily involved with \education\ for the Volksdeutsche in )omania, 1ugoslavia and $ungary. 's the (erman authorities took pains to identify the (onausch a"en, and to check their loyalty to the )eich, they duly ensured that non0(ermans would also come to view them as an indivisible block. BerlinXs decision not to evacuate the Conauschwaben &Heim ins Reich' /9 , whilst still exploiting them as a source for recruits against TitoXs partisans .which act, with regard to -erbia and the 1ugoslav Banat, some (onausc ha"en today claim to be contrary to international law5 had fatal consequences. 'ccording to an as yet unconfirmed verbal report, (ustav $alwax, an elected representative of the 1ugoslav (ermans, went to $itler in the Dolfsschan9e $] in &ast +russia and tried to argue the point. $e was drafted into a punishment batallion on the spot, and within a month of the start of Barbarossa, was dead.

The actions of the 4th -- =ountain Civision K+rin9 &ugenL were viewed in 1ugoslavia at the warXs end as the embodiment, in a few, of the desires
#2 Un"er this talisman, the @otts'hee @ermans ha" been move" to usur9 !lovenians, Bessarabian an" Balts to usur9 4oles. What began as an astoun"ingly /ellDfun"e" an" Norganize" venture >at least /hen it 'on'erne" the O9re'iousO @erman 'olonists an" not those evi'te" to ma;e room for them), "egenerate" "uring the /arDyears into a series of lastD"it'h eva'uations of Pol;s"euts'he an" 9roD@erman grou9s before they /ere overrun by the =e" Army.

42

of the majority of Conauschwaben. -ure it was that the early drafts of Conauschwaben were of a largely voluntary nature, but once the hotheads had been drawn off, the -- had to resort increasingly to force. It is also not to be forgotten that brutality breeds brutality, and there was plenty of that from all sides in the 1ugoslav conflict. In any event, an ordered evacuation of the 1ugoslav Banat having failed almost completely due to official stonewalling and peasant unwillingness to depart, the great majority of the Conauschwaben of 1ugoslavia fell into the hands of the +artisans *<. Dhile the war was dying away the +artisans began to exact a terrible revenge upon their prisoners. This is perhaps best illustrated by the *2@3 1ugoslav census. Dhile the *2<* census reveals <A3,E#@ persons of (erman mother tongue in the Pojvodina .including -rem5*@, in *2@3, only @*,@E# are recorded in the whole of -erbia*A. >f these, the largest single group, A,423 persons, were recorded in the prison0camp for Conauschwaben at 7niUanin .)udolfsgnad5, a place which received thousands more inmates than it ever released.*E =ost Conauschwaben did, however, actually leave 1ugoslavia alive, crossing the border into $ungary and thence to 'ustria and the two (ermanys. 'fter the *2@3 census, the number of (ermans again declines sharply.

#C !'herer, Anton >Frsg.)0 Unbe;annte !!D@eheimberi'hte :ber "ie 1va;uierung "er !:"ost"euts'hen im ?;tober un" im ovember #$&& so/ie :ber "ie 9olitis'he -age in =um<nien , Ungarn, "er !lo/a;ei, im !erbis'hen Banat un" im OUnabh<ngigen !taat 8roatienO. @raz0 !elbstverlag, #$$A.

*@ klmnopqrsp tupuvruvsp. tupwnxwyuxn zn {lonwrznxl|uv v }pulo~lm lvs, zn znwv| n <*0III0 23L2 nvwl t rp {nxnvwnm v nrnxn0 }lunvnm, oll zn zyuvwpmp. kopxwp tupuvruvsvol klmnsopursp llopuvxwl, lnop, *2@A.
#) Although the authors "e'lare themselves a/are of the ten"en'y for some O@ermans staying in YugoslaviaO >for ho/ longQ) to "e'lare themselves to be Fungarian. 8onaRni rezultati 9o9isa stanovniStva o" #) marta #$&B go".. !tanovniStvo 9o naro"nosti, 8nL. $. !avezni Iavo" za !tatisti;u, Beogra", #$)). 4age T-U. #V 4lan of =u"olfsgna" W 4anRevo to/nD'umD'am9 from 8.ller, Feinri'h0 O4anRevo, "ie Totenliste von "er "euts'hen Bev.l;erung "er ,onausta"t 4ants'hova im BanatO un9ublihe" >as at !e9tember #$$)) a''ount. With 9ermission of the author.

3#

Mor the Conauschwaben, the post0war situation in )omania was somewhat different. )elations with the state0forming nation had since *2*2 been considerably easier. Mor a start, unlike in 1ugoslavia, there was only one to take account of. -econdly, the )omanians were concerned to prevent a unity of purpose between the two major (erman groups 0 Conauschwaben and -iebenbRrger -achsen 0 and the largest minority, the $ungarians. Thirdly, )omania had not been invaded by the )eich as an enemy state. It had rather, like $ungary and Bulgaria, been forced into a position from which it had no option than submitting itself to an odious alliance. Two )omanian armies had been launched against the -oviet Nnion on the day Barbarossa began, and with the reconquest of Bessarabia already achieved, were foolishly sent ever further eastward to disappear 0 along with the $ungarian "nd and the Italian 3th armies 0 into the snows around -talingrad. )omanians, like their comrades in arms 0 fought bravely in a dishonourable cause, something they still seem unable to face up to. By 'ugust *2@@, despite further great sacrifices, it was clear that all was lost. Dhen the astonishingly successful coup of the "<rd was pulled off, a few days confusion reigned while all sides strove to orientate themselves to the new state of affairs. Dhile there was little love lost between )omanians and the $ungarians at this turn of events, there does seem to have been much confusion and some distress at the rupture between the )omanians and, particularly the -iebenbRrger -achsen, whose ancestors had acquired broad privileges from the $ungarian king as early as *""@ .the 'ndreanum5. -ome thousands of (ermans fled westward *4. In early -eptember *2@@, a probe towards TimiFoara by very limited (erman forces was accompanied by the evacuation of about <E,### of the 3#,### Conauschwaben in the )omanian Banat *3. There does not appear to be
#% $ol(tr)ger" $ans0 ,ie 1va;uierung "er ,uts'hen aus or"siebenb:rgen, !athmar un" "em =um<nis'hen Banat ab !e9tember #$&& im -i'hte "er 8riegstagebu'hes "er Ferresgru99e !:";raine. Un0 Der *ngarndeusche" $eimat(eitung der Deutschen aus *ngarn . #%. Eg >#$%C), 2#,22,2C,2&7 #B. Eg >#$%&), r #,2.

*3 +rincipally from the following settlements .(erman names5 Northern gro*p $at9feld, Bowrin, Benauheim, (roHjetscha, (ertianosch, Triebswetter, >stern, Tschanad/ so*thern gro*p: Nlmbach, Gohannisfeld, (uilva9, ,euburg, 'urelhausen, -ackelhausen, Tschene, Beregsaul, Ceutsch -kt. =ichael.. (okumentation der Vertrei"ung der (eutschen aus Ast25itteleuro.a . 3*

any evidence of a settling of scores such as was witnessed in the Pojvodina. To be sure, tens of thousands of Polksdeutsche were handed over to the -oviets for forced labour, and there were many fatalities among those sent. Mamilies of those who had served in the -- were probably better off out of the country, but generally, fear was not so ever present as in TitoXs land. The latterXs split with -talin however, brought about a sudden shock for the inhabitants of the border09one, and most particularly for those who inhabited the rich arable lands of the Banater $eide.

In *2A*, "2,"2A persons of mixed origin were deported out of the Banat, mainly to the inhospitable BVrVgan steppe. These included persons who had come illegally to the Banat after the war, but their proportion in the whole is uncertain. Dhat becomes very clear, however, when the number of deportees from each settlement is portrayed cartographically, is that most came from the rich farmlands in the $eide. The choice of settlements can hardly be said to reflect the disposition of the -outh -lav population of )omania which, considering the heroic struggle of the +artisans against the wartime (erman occupier, might have been expected to show pro01ugoslav sympathy. >n the contrary, the -erb settlements along the Canube remained practically unmolested, while the (ermans, who had no reason for Titoist sympathy, seem to have drawn the most attention. >ne can only suppose that the authorities in Bucharest saw this as a suitable opportunity to weaken at last the privileged position of the Conauschwaben, and to acquire for )omanians an equitable share of the fertile heathland. I have no figures regarding subsequent )omanian immigration to (erman farmsteads, however. 'fter some years, Conauschwaben were allowed back to their farms, but many henceforth had to share with immigrants from other parts of )omania. Though the whole BVrVgan episode directly affected only a proportion of the Conauschwaben in )omania 0 and not only (ermans Band III B (as Schicksal der (eutschen in RumLnien . Bundesministerium fRr Pertriebene, MlRchtlinge und 7riegsgeschgdigte. Bernard (raefe, Berlin, *2A4. 3"

were deported 0 it served to emphasise that the )omanian regime was now firmly in control of their destiny. (radually, however, the Conauschwaben and -iebenbRrger -achsen of )omania came to accept that such a control was not to their liking, their resolution doubtless eased by the existence of a relatively free and economically very successful Dest (ermany. The :rundgeset+ of that country made it theoretically their second home, and agreements between the Bundeskan9ler and %eaucescu, by which the Mederal )epublic agreed to pay thousands of Ceutschmarks for every RumLniendeutsche KreleasedL converted this for many to a reality.

3<

*2A*

3@

'fter the demise of %eaucescu, restrictions on ethnic (erman emigration from )omania were eased and the remaining population rapidly diminished. The *22" census records, for the )omanian Banat, a total of @@,*2E (ermans. The *2@* census, for the same area, recorded "<3,2@3. Dhereas in *2@*, 4" towns had over *,### (ermans *2, by *22", only six such settlements remained"#. In )omania, there has been no massacre of the (erman population. It has just wasted away. In the course of an interview"* with the =ayor of Biebling %ommune, =r =ichael &rk .a (erman5, I was told that only E# of BieblingXs (erman population remained, the exodus having accelerated enormously after *232 "". Those who remained were all quite old, and there was not a single (erman child in the village. 'll the younger folk had left for (ermany.

't the end of the -econd Dorld Dar, the pre0war international borders were restored. Baranja was allocated to %roatia and subordinated to >sijek which was, administratively and in terms of communications, an infinitely more sensible centre than either -ombor or ,ovi -ad. The *2 TimiFoara .<#,2@#5, )eFiOa, Gimbolia, Bugoj, 'nina, 'radul ,ou, (iarmata, Biebling, +eriam, -VcVla9, Mratelia, Bovrin, Biled, Teremia, =are, Tomnatic, %VrpiniF, Mreidorf, PariaF, Benauheim, (uttenbrunn, Ceta, -Wnnicolau =are, +eciu ,ou, Becicherecu =ic, CudeFtii ,oi, Bacova, (rabaO, Iecia =are, ,iOchidorf, BulgVruF, -Wnandrei, %aransebeF, >raviOa, Bipova, Gamu =are, Carova, -Wnpetru (erman, =Vureni, >rOiFoara, %iacova, _andra, (ottlob, Iohanisfeld, MrumuFeni, %omloFu =ic, )ecaF, _agu, %enadu =are, =oraviOa, TVdVreni, Nivar, -Wnnicolau, (erman, %enadu Pechiu, Bensecu de -us, >rFova, -Wnnicolau =ic, +etroasa =are, %arani, MWntWnele, _ag, Teremia =icV, (VrWna, -tamora (ermanV, ,eudorf, =aFloc, 'luniF, +iFchia, -Wnmihaiu (erman, Poiteg, Brebu ,ou, ColaO, Iecia =icV .total, *2*,2#E5.

"# TimiFoara .*<,"#E5, )eFiOa, Bugoj, %aransebeF, Gimbolia and BocFa =ontanV.

2# ,e'ember #$$V.

"" =omanian 'ensuses #$&#, &,ABV7 #$%%, $B27 #$$2, )&. 3A

Ba!ka and 1ugoslav Banat were amalgamated with -rem to form the K'utonomous +rovince of PojvodinaL, a part of the )epublic of -erbia. There was discussion of a division of the Ba!ka between -erbia and %roatia along a line -ubotica 0 ,ovi -ad, but Cjilas decided against this, as it was felt better to retain the whole Ba!ka under the aegis of Belgrade because of the importance of the territory in the post0war inner colonisation programme .La-pe ""45. Bocal government was subject to kaleidoscopic changes in the period *2@E 0 *2A*. In the Banat and Ba!ka, many of the combinations of the immediate post0war period were connected with the eviction of the Conauschwaben and the re0 occupation of their villages by +artisan veterans and other war0affected persons .see appendix5. The number of districts and communes in 1ugoslavia as a whole changed dramatically every year, the *2@E figure of @#4 and **,AEE respectively dropping to <E# and 4,*#@ by *2A* .Po*,son, in Ho55-an, "<*5. These figures continued to drop as administrative units were amalgamated or otherwise grew. =otivations for such change may have included Kempire0buildingL and rivalries among local politicians .ibid "@25. In some cases, the colonisation of the former (erman areas of the Pojvodina was also accompanied by amalgamation of proximate settlements .e.g. Banatsko Peliko -elo5. In *2AA, *#4 districts and *,@42 communes remained, while in *2EA, the figures had dropped sharply to "@ and A@3 respectively. This trend was connected with the policy of decentralisation to the commune level, which required that communes be large enough to be viable economic units .ibid "@A5. Interestingly enough, in some parts of 1ugoslavia, this policy resulted in the communes . o.;tine5 expanding to the si9e and into the capacity of the increasingly obsolete districts . sre+ovi5. The A@3 communes of *2EA, were Kin the majority of cases nearly identical with the @#4 districts of *2@4. They were not only focussed on the same urban centres, but included 3#Y to 2#Y or more of the villages found in the prewar administrative unitsL.. Po*,son, in Ho55-an, "A*5. Mor the period *2A< to *22*, the 1ugoslav Banat and Ba!ka were divided into <E0<3 districts, up from "* in *2@3.Today, 1ugoslav scholars differentiate between the administrative .1ugoslav5 Banat as recognised within the Pojvodina .Banat in Pojvodina5, and the geographical .1ugoslav5 Banat. 3E

The Banat in Pojvodina consists of sixteen districts, including *4E settlements and E@3,E** inhabitants .*22* census5. The (eographical 1ugoslav Banat also includes the $anevaki Rit .with E settlements and A2,<4E inhabitants .*22* census5, belongs to the city of Belgrade and is attached administratively to central -erbia . 'ioMe+, N*rOe+, P,a+Ma, *22E, awaiting publication5. The latest chapter in the administrative history of the 1ugoslav Banat and Ba!ka began in *23E, when the Pojvodina lost its autonomy and was absorbed bodily into -erbia. It is still, however, recognised as a unit distinct from -erbia .%ensus of *22*5. The wars of partition in former 1ugoslavia spared the Pojvodina from direct involvement, although tragic Pukovar lies on the Canube facing Ba! district in the Ba!ka. The province has however been a favoured destination for refugees .#NHCR %ensus of )efugees and other Dar0 'ffected +ersons in the Mederal )epublic of 1ugoslavia, Belgrade *22E5.

The )omanian administration has in the Banat by and large retained the settlement boundaries it inherited from the $absburg &mpire, which is fortunate, as modern maps showing the lower0level administrative boundaries are extremely hard to come by. ' number of community amalgamations have, however, taken place since *2*# .see 'ppendix5.

GudeOul 'rad was, at some as yet uncertain date, enlarged southward to embrace the whole of the =aros valley, by the attachment of elements of +lasa BirchiF .GudeOul -everin5 and +lasa Bipova, +lasa +esac0+eriam, +lasa -anicolaul0=are and +lasa Pinga .all TimiF5. 't the time of the *2@3 census, GudeOule %araF and -everin were separate administrative divisions, while TimiF had been re0christened TimiF0Torontal. ' new framework was introduced in -eptember *2A#, comprising "3 provinces .regiunes5 and *44 districts .raionuls5 and @,#A" communes across the country .down from A3, @"* and E,"@3 respectively5. The provinces were whittled down to *3 in *2A" and *E in *2AE. The Banat was covered by K)egiune TimiFoaraL. *2E3 saw the abolition of both provinces and districts. <2 counties were established, divided directly into communes. 34

The )omanian Banat today is divided between five counties

TimiF,

%araF0-everin, 'rad, =ehedinOi and $unedoara. The first two are situated entirely in the Banat, whilst the other three counties are only partly within the region. -imilarly to the case over the -erbian border, there is in )omania an administrative definition of the Banat, which comprises 'rad, %araF0-everin and TimiF counties and extends over "@,24#.A km ".

33

%HE!A%IC CAR%OGRAPHD 1 AN IN%RO #C%ION

=aps,

or

Pgra.hic

re.resentations

o!

the

cultural

and

.hysical

environmentP .Cent *22E "5, are probably amongst the most widely used of all physical communicative tools. The first known maps originated in the Ath or Eth centuries B% .Bagrow *2EE, in Cent *22E "A5 in the modern world, we see them every day, in the papers, or on the television. They may show us how to get to a guesthouse, where to find the cheapest carpets, which cities are hosting 4rance &08, or how far -'s army has advanced into B's territory. Dhen we give someone directions we often draw them a map 0 or, if we donXt, we tend at least to visualise one. =aps are everywhere. The genre is almost endlessly flexible in theme and scope, and a map can be used to show anything from the layout of electronic micro0circuits, to the bottom of the oceans, to galaxy clusters at the limits of the known universe. But what all maps of whatever genre have in common is the objective to portray the locational context of objects and their spatial relationships. ,owadays the cartographer has also been freed of the tyranny of paper0si9es, as his maps may be stored as microfiche, slides or as digital data. Dith digital maps, always bearing in mind the limitations imposed by the scale and the level of generalisation accepted in the creation of the original, one can 9oom in or out to capture or print the required part of the map at an appropriate level of magnification.

The continuing popularity of maps of all types stems in part from their ability to express the equivalent of several pages of data and dense text in pictorial form. They can therefore serve as a form of shorthand for those unable or unwilling to plough through such texts, and can bring across a complex message with great weight and immediacy. =onmonier claims that P.eo.le trust ma.s, and intriguing ma.s attract the eye as ell as connote authority3P .=onmonier *22E 345. In cartographic form, data relates easily to the outside world, and may, by the studied 32

application of graphical and artistic license, be made very attractive. But are we right to trust them 'nd can one say that any map is \true\

=aps are usually prepared to illustrate accepted knowledge, plans, or a point of view. They may offer a \snapshot\ view of one point in time, or seek to show an evolution of events over time. They are, in other words, suitable vehicles for education, or for persuasion. -choolchildren are made aware of the contemporary extent, form and characteristics of their native country and its neighbours partly through maps in their textbooks, designed for that very purpose. 's the children progress, they meet other maps plotting the respective statesX relative territorial ebb and flow over history, and encounter a range of other maps in daily life, such as weather maps on the television. In later life, they may elect to publish maps of their own. They may become historians, and use maps to illustrate key events in a complex period. >r economists, and wish to depict relative regional wealth within the state. 'n astronomer may use one to persuade the reader of the \Big Bang\, or of \black holes\. They may be in business and their objective commercial. 'n advertising map may show off a firmXs central location to emphasise its accessibility or its branch network to draw attention to its nation0wide coverage. 'n estate agentXs sketch0map may emphasise the good connections, or the remoteness, of a property to make it attractive to different clients. -imilarly, when the author is operating in a political environment, and most particularly when concerning disputed territories, his or her motivations will determine the form of the map.

It is therefore particularly important to be aware that maps, even in their most basic forms, are not objective statements of fact. They are composed, and reflect, within the limitations of the authorXs skill, his or her ambitions. \... 5a.s, like s.eeches and .aintings, are authored collections o! in!ormation and also are su"ject to distortions arising !rom ignorance, greed, ideological "lindness, or maliceP .!on-onier *22E "5. The final map does not arise through an irresistible logical process. It is the product of a series of decisions, some of which are imposed by the 2#

available technology, or considerations of cost, while others result from the free choice of the author. P- single ma. is "ut one o! an inde!initely large num"er o! ma.s that might "e .roduced !or the same situation or !rom the same data3 333 5a. authors can e<.eriment !reely ith !eatures, measurements, area o! coverage, and sym"ols and can .ick the ma. that "est .resents their case, or su..orts their unconscious "iasP .!on-onier *22E "5.

=aps may be divided into two sub0classes thematic maps which Psho

general0purpose,

or

topographical, maps which represent the physical environment, and the s.atial distri"ution o! a .articular geogra.hical .henomenonP or theme . ent *22E "5. Topographical maps are guides of relative location and, until the eighteenth century, geographers were preoccupied with the topographical mapping of the earth. Thematic cartography only came into its own once the topographical world map had been \filled in\ . ent *22E E5. Thematic maps are Pdesigned to demonstrate .articular !eatures or conce.ts3 @n conventional use this e<cludes to.ogra.hic ma.sP F !enyen *24< "2*5 In the case of our electronic microcircuit, a general0purpose map would show the location of each element and their spatial relationships within the whole. Babelling would provide little more than the name of the individual objects. ' thematic map of the same circuit board could be designed to show current0flow, or resistance, or selected other values, to clarify the functions or inter0relatedness of the individual parts.

Thematic maps, which display the \structural characteristics of some particular geographical distribution\ .Ro9inson *24A 20*@5 or depict \distance and directional relationships, patterns of location, spatial attributes of magnitude change\ . ent *22E E5 consist of a geographic base map and a thematic overlay. The overlay may be classified in many ways, but perhaps the most fundamental division is into qualitative and quantitative sub0types. The key difference is that quantitative maps are based directly on numerical data, while qualitative maps are not. 2*

]ualitative maps show Ps.atial distri"ution or location o! kindP . ent *22E 45 They indicate the extent of a variable in geographic space, but supply no data on quantity, or proportion. They are easy and cheap to produce, and are therefore more common and popular than quantitative maps. They are also potentially more misleading, because of their independence from actual data. ]uantitative maps, in contrast, are Pa sym"olised generalisation o! the in!ormation contained in a ta"leP .)enks, *24E **0*25 and are based not only on presence, but also on the numerical strength of that presence. +ut another way, they record spatial aspects of tabular data, from which their thematic overlay is directly derived. This data may be in the form of actual values, percentages or ratios, or be classified into interval groups.

Thematic map0types 00 either Nualitative or Nuantitative, may be classified in any number of different ways. %lassifying by s*90ect -atter, one may define maps of the landscape and atmosphere/ geological maps/ population maps/ economic maps, political maps/ maps of the municipality/ advertising maps .adapted from !on-onier P $che,, *2335. By 5*nction , they might be classified as educational, journalistic, persuasive, advertising, or propaganda. =ost relevant to the present inquiry are population maps showing location, extent, density and increase^decrease of population, language, nationality or confession. This family of maps may fall into educational, political, journalistic or propaganda0map categories, and will be referred to collectively as ethnographic maps. The process of their preparation, ethnogra.hic cartogra.hy, is therefore a sub0element of thematic cartography, and may be defined as the spatial representation of historical language0, confession0 or nationality0 related data in cartographic form.

Ethnographic Cartography 11 $o*rces

2"

Thematic cartography as a whole has an extensive literature, but despite the interest and significance of ethnographic cartography for the political process, there has been only one major publication on the latter topic in the &nglish language 00 P5a.s and $oliticsP .Dilkinson *2A*5. This is limited to a consideration =acedonian problem and of is ethnographic cartography of disappointing in that the the display0

methodology of the actual source0maps is overlooked. Mor ease of comparability, all the original maps have been standardised to one scale, si9e and schema. Cr. =artyn )ady of --&&- has recently completed a paper on P-ustrian 5a.s o! the Bulgarians in the Kineteenth GenturyP, which highlights the significant role of 'ustrian ethnographic cartographers in the splintering of south0Balkan ethnic identities and the frustration of Bulgarian ambitions on =acedonia .)ady *22<5.

Mor methodology, one needs to turn to sources in other languages. ' *2<4 memorandum by +6l Teleki and 'ndr6s )8nai is perhaps the earliest attempt to analyse in depth P the RQnai problems *2<45. facing the ethnographic cartographer .%e,eki 'nother two

$ungarian publications followed during the -econd Dorld Dar, but these are taken up almost entirely with ridiculing )omanian and defending $ungarian maps .$taats;issenscha5t,iches Instit*t *2@#, *2@"5. +ost0war, thematic cartography as a whole, and ethnographic cartography as one of its parts, has been examined in great depth by Derner Ditt .=itt *24#5. -imilarly indispensable is the methodological examination by Dilfried 7rallert .'ra,,ert *2E*5. $olger Mischer .*22*5 subjects thirty0two *2th and "#th century mainly $ungarian maps to a rigorous appraisal, counterbalancing the tendentious but largely justified $ungarian publications of *2@#^" .in Hosch P $ee;ann *22*5. ' *22* dissertation by the present author examines the methodology and fairness of 2@ mainly $ungarian, (erman and )omanian ethnographic maps of Transylvania and the Banat .=itchell *22*5.

2<

%he P*rpose o5 !aps

=aps are usually prepared to illustrate accepted knowledge, plans, or a point of view. They may offer a \snapshot\ view of one point in time, or seek to show an evolution of events over time. They are, in other words, suitable vehicles for education, or for persuasion. -choolchildren are made aware of the contemporary extent, form and characteristics of their native country and its neighbours partly through maps in their textbooks, designed for that very purpose. 's the children progress, they meet other maps plotting the respective statesX relative territorial ebb and flow over history, and encounter a range of other maps in daily life, such as weather maps on the television. In later life, they may elect to publish maps of their own. They may become historians, and use maps to illustrate key events in a complex period. >r economists, and wish to depict relative regional wealth within the state. 'n astronomer may use one to persuade the reader of the \Big Bang\, or of \black holes\. They may be in business and their objective commercial. 'n advertising map may show off a firmXs central location to emphasise its accessibility or its branch network to draw attention to its nation0wide coverage. 'n estate agentXs sketch0map may emphasise the good connections, or the remoteness, of a property to make it attractive to different clients. -imilarly, when the author is operating in a political environment, and most particularly when concerning disputed territories, his or her motivations will determine the form of the map.

It is therefore particularly important to be aware that maps, even in their most basic forms, are not objective statements of fact. They are composed, and reflect, within the limitations of the authorXs skill, his or her ambitions. \... 5a.s, like s.eeches and .aintings, are authored collections o! in!ormation and also are su"ject to distortions arising !rom ignorance, greed, ideological "lindness, or maliceP .!on-onier *22E "5. The final map does not arise through an irresistible logical process. It is the product of a series of decisions, some of which are imposed by the available technology, or considerations of cost, while others result from 2@

the free choice of the author. P- single ma. is "ut one o! an inde!initely large num"er o! ma.s that might "e .roduced !or the same situation or !rom the same data3 333 5a. authors can e<.eriment !reely ith !eatures, measurements, area o! coverage, and sym"ols and can .ick the ma. that "est .resents their case, or su..orts their unconscious "iasP .!on-onier *22E "5.

2A

%opographica, and %he-atic !aps

=aps may be divided into two sub0classes thematic maps which Psho

general0purpose,

or

topographical, maps which represent the physical environment, and the s.atial distri"ution o! a .articular geogra.hical .henomenonP or theme . ent *22E "5. Topographical maps are guides of relative location and, until the eighteenth century, geographers were preoccupied with the topographical mapping of the earth. Thematic cartography only came into its own once the topographical world map had been \filled in\ . ent *22E E5. Thematic maps are Pdesigned to demonstrate .articular !eatures or conce.ts3 @n conventional use this e<cludes to.ogra.hic ma.sP F !enyen *24< "2*5 In the case of our electronic microcircuit, a general0purpose map would show the location of each element and their spatial relationships within the whole. Babelling would provide little more than the name of the individual objects. ' thematic map of the same circuit board could be designed to show current0flow, or resistance, or selected other values, to clarify the functions or inter0relatedness of the individual parts.

Thematic maps, which display the \structural characteristics of some particular geographical distribution\ .Ro9inson *24A 20*@5 or depict \distance and directional relationships, patterns of location, spatial attributes of magnitude change\ . ent *22E E5 consist of a geographic base map and a thematic overlay. The overlay may be classified in many ways, but perhaps the most fundamental division is into qualitative and quantitative sub0types. The key difference is that quantitative maps are based directly on numerical data, while qualitative maps are not.

]ualitative maps show Ps.atial distri"ution or location o! kindP . ent *22E 45 They indicate the extent of a variable in geographic space, but supply no data on quantity, or proportion. They are easy and cheap to produce, and are therefore more common and popular than quantitative 2E

maps. They are also potentially more misleading, because of their independence from actual data. ]uantitative maps, in contrast, are Pa sym"olised generalisation o! the in!ormation contained in a ta"leP .)enks, *24E **0*25 and are based not only on presence, but also on the numerical strength of that presence. +ut another way, they record spatial aspects of tabular data, from which their thematic overlay is directly derived. This data may be in the form of actual values, percentages or ratios, or be classified into interval groups. Thematic map0types 00 either Nualitative or Nuantitative, may be classified in any number of different ways. %lassifying by s*90ect -atter, one may define maps of the landscape and atmosphere/ geological maps/ population maps/ economic maps, political maps/ maps of the municipality/ advertising maps .adapted from !on-onier P $che,, *2335. By 5*nction , they might be classified as educational, journalistic, persuasive, advertising, or propaganda. =ost relevant to the present inquiry are population maps showing location, extent, density and increase^decrease of population, language, nationality or confession. This family of maps may fall into educational, political, journalistic or propaganda0map categories, and will be referred to collectively as ethnographic maps. The process of their preparation, ethnogra.hic cartogra.hy, is therefore a sub0element of thematic cartography, and may be defined as the spatial representation of historical language0, confession0 or nationality0 related data in cartographic form

Ethnographic Cartography E $o*rces

Thematic cartography as a whole has an extensive literature, but despite the interest and significance of ethnographic cartography for the political process, there has been only one major publication on the latter topic in the &nglish language 00 P5a.s and $oliticsP .Dilkinson *2A*5. This is limited to a consideration and is of ethnographic disappointing in cartography that the of the =acedonian problem display0 24

methodology of the actual source0maps is overlooked. Mor ease of

comparability, all the original maps have been standardised to one scale, si9e and schema. Cr. =artyn )ady of --&&- has recently completed a paper on P-ustrian 5a.s o! the Bulgarians in the Kineteenth GenturyP, which highlights the significant role of 'ustrian ethnographic cartographers in the splintering of south0Balkan ethnic identities and the frustration of Bulgarian ambitions on =acedonia .)ady *22<5.

Mor methodology, one needs to turn to sources in other languages. ' *2<4 memorandum by +6l Teleki and 'ndr6s )8nai is perhaps the earliest attempt to analyse in depth P the RQnai problems *2<45. facing the ethnographic cartographer .%e,eki 'nother two

$ungarian publications followed during the -econd Dorld Dar, but these are taken up almost entirely with ridiculing )omanian and defending $ungarian maps .$taats;issenscha5t,iches Instit*t *2@#, *2@"5. +ost0war, thematic cartography as a whole, and ethnographic cartography as one of its parts, has been examined in great depth by Derner Ditt .=itt *24#5. -imilarly indispensable is the methodological examination by Dilfried 7rallert .'ra,,ert *2E*5. $olger Mischer .*22*5 subjects thirty0two *2th and "#th century mainly $ungarian maps to a rigorous appraisal, counterbalancing the tendentious but largely justified $ungarian publications of *2@#^" .in Hosch P $ee;ann *22*5. ' *22* dissertation by the present author examines the methodology and fairness of 2@ mainly $ungarian, (erman and )omanian ethnographic maps of Transylvania and the Banat .=itchell *22*5.

$e,ection o5 "eat*res
' map is a scaled and generalised representation of certain selected characteristics of a chosen space. ,o map can show everything, and in the preparation of every map, decisions therefore have to be taken as to what is material and what can legitimately be ignored. PWhen e acce.t the idea that not all the availa"le in!ormation needs to "e .resented, that instead in!ormation must "e selected !or .articular .ur.oses, then 23

the ma..ing task "ecomes the identi!ication o! relevant elements \ .=e,t-ann , *242 "A5. It follows that what the map shows is not necessarily as revealing as what it does not. The appropriate si9e of the map, the geographical space, administrative level, scale, map projection and the variables to be represented each has to be selected, and each has a powerful influence on the message communicated by the map. The language.s5 used for the labelling can also have an impact. Cecisions relating to si9e, geographical space, area of reference and scale will each act separately and decisively on the range of possibilities for the others and upon the visual impact of the final product.

!ap Pro0ection

The latitude^longitude system, created with mariners in mind, relates locations to each other not by distance, but by bearing. &Tatitude2 longitude is a geogra.hic re!erence system, not a t o2dimensional F.lanar) co2ordinate system333 5eridians converge at the .oles "ut se.arate or diverge as they get closer to the eNuator3 So the length o! one degree o! longitude varies de.ending u.on the latitude at hich it's measured3 4or e<am.le, one degree o! longitude at the eNuator is /// kilometres FW0 miles) in length, "ut the length o! one degree o! longitude converges to +ero at the .oles3 Because degrees aren't associated ith a standard length, they cannot "e used as an accurate measure o! distance or area .E$RI, *22@ 4.25. It is therefore not possible directly to obtain areal0world co0ordinates, distance between points, or areas, from a latitude0and0

22

projections

*##

longitude map, although these may be derived by mathematical formulae.

' map that uses a simple latitude^longitude graticule therefore distorts reality. $owever, no other method has yet been devised which does not. The process of transferring the graticule onto a two0dimensional grid is called .rojection, and there are at least *3 different systems in use .Illustr. 35."< PKo !lat ma. can match the glo"e in .reserving areas, angles, gross sha.es, distances and directions, and any ma. .rojection is a com.romise solutionP .!on-onier *22E *<5. The difficulties in this process are perhaps best exemplified by the case of (reenland. The familiar =ercator projection .dating from *AE25 conspicuously shows it about the same si9e as -outh 'merica, which in reality is eight times its si9e. >ther projections, which avoid this gross distortion in area, are subject to other problems, particularly directional, distance and shape distortions. &ach has its own strengths and weaknesses, and each is more or less appropriate for particular scales and even for particular parts of the world. The 'lbers &qual 'rea %onic for instance, is particularly suited to mapping territories covering several time 9ones like the N-' and )ussia. =aps of extra0terrestrial space are even more problematic, as they must portray not only the surface of a \sphere\, but its \interior\ as well.

Geographic $pace

>ne of the most fundamental decisions taken by the cartographer is the extent of the geographic space for the map. ' map may be, after all, of K(ermanyL, or Kthe +hilippinesL, or K)ussiaL, but where do these territories 2C +lo e" Mercator"Trans,erse Mercator" - li#ue Mercator" .pace - li#ue Mercator" Miller /ylindrical"
Ro inson" .inosoidal 0#ual Area" -rthographic" .teregraphic" +nomic" A(imuthal 0#uidistant" 1am ert A(imuthal 0#ual Area" Al ers 0#ual Area /onic" 1am ert /onformal /onic" 0#uidistant /onic 2.imple /onic3" 4olyconic" Bipolar - li#ue /onic /onformal 24oster enclosure to Dent" 56673.

*#*

begin or end +articularly where maps have a political message, the choice of the geographic space represented on the map has a distinct bearing on the message sent by the final product. Dithin the %arpathian Basin, perhaps the best example is provided by a pair of $ungarian ethnographic maps of Transylvania produced . BKtky, 'og*to;ic> P %e,eki *2@# .x"55 at the height of a \map0war\ between $ungary and )omania. Dhile technically the maps were of the highest order, and offer some lessons for today, the chosen geographic space was not just Transylvania, but Transylvania within the pre0*2*2 borders of $ungary. Dhile this was perfectly defensible, there was a purpose. Two maps were published, the titles of one including the phrase KTransylvania and the &astern +arts of $ungaryL. >ne was based on the $ungarian census data from *2*#, the other on census data from the $ungarian, )omanian, 1ugoslav and %9ech censuses from *2<#0<2. Cespite the political changes after the Mirst Dorld Dar, both maps show the same territory and the same boundaries. The fact that in the second period, Transylvania was part of )omania and the Banat shared between )omania and 1ugoslavia in the second map is not highlighted. The new international boundaries are printed in grey and to the uninformed or careless, there is little to differentiate the two maps. The subdued frontiers and the mauve representing the )omanian population dissolve into the background, unlike the striking red representing the $ungarian population. These factors combine to visually bind the dense $ungarian population along the $ungarian0Transylvanian border to the solidly $ungarian plain. The Transcarpathian parts of )omania were excluded from both maps, further strengthening the visual impact of the $ungarian population0group and depressing that of the )omanians. The )omanian answer was to republish the $ungarian map, but with =oldavia and Dallachia appended to Transylvania and &astern $ungary removed .see above5. 's these other parts of )omania were overwhelmingly )omanian inhabited, the impression created by this map was the direct opposite of that conveyed by the $ungarian pair, particularly because the $ungarian colour0scheme was also reversed. 'n example of the )omanian map is held in the =ap Bibrary of the British =useum. This )omanian map itself earned a damning criticism in *#"

RumLnische Tandkarten und ihre =ritik , $taats;issenscha5tk,iches Instit*t, Budapest, *2@" .hereinafter referred top as ReL*i'5, as it totally ignores the presence of minorities in =oldavia, Dallachia and Cobrudja.

$ca,e and Genera,isation

' certain amount of generalisation is inevitable with any map whatever the scale, and such generalisation may be thought of as wilful distortions of reality. ' large map0scale permits more details to be shown on the base0map and for their position and si9e to be reported more accurately. 't progressively smaller map scales, objects on the ground shrink into invisibility, meaning that if they are to be indicated at all, they will have to be shown at more than life si9e. )oads are a prime example. In the N7, the * "A#### maps of the >rdnance -urvey >utdoor Beisure series, show K'L )oads as a * mm wide line, equating to a real world road width of "A metres. The * mm width is retained for K'L )oads on both the * A#### Bandranger metres .equivalent road width A#m5 and the * "A#### )outemaster series .equivalent road width "A#m5.

The common tension between detail and overview is particularly present in thematic cartography. The ideal scale for a thematic map will depend upon the objective for that map, but where data is available at settlement level, it would seem to be unhelpful to conceal this from the map user. ' map based on settlements will have to be drawn at a reasonably large scale. Dhile this will aid clarity, it might bring complications if the area being mapped is extensive. ' number of separate map0sheets may become necessary, frustrating an easy estimation of the wider picture, which may be one of the objectives. In the context of large states, such as the $absburg &mpire, this difficulty will assume substantial proportions. This is admirably illustrated by the \'dministrative =ap of the $ungarian -tate, supplemented with the *2## %ensus =other0Tongue )esults... produced by the $ungarian central -tatistical >ffice in *2##. The overall objectives, to portray the *#<

contemporary ethnographic situation over the whole $ungarian kingdom at settlement level, forced the publishers into a cleft stick. The joint objectives of overview .i.e. small number of sheets5, large scale for clarity, and settlement level data could not equally be met. They settled on a moderate scale of * <E#,###, but even so thirteen giant sheets were needed to cover the kingdom .'R>ponti $tatis>tikai Hi+ata, *2##5.

Cespite their larger scale, the individual mapsheets for 7rallertXs * "##,### KVolkstumskarteO& series were considerably smaller and more manageable, but the price for this was that @@ were required to cover just )omania. There were a further @# for 1ugoslavia and "< for $ungary .P*9,ikationsste,,e =ien , *2@#0*5. It is rare indeed to be able to plan a project so free of financial or other constraint/ one normally has to \cut oneXs map according to oneXs means\. The physical si9e of the finished product is restricted by considerations of cost, and also by realisable paper si9es for the product. Mor everyday purposes, therefore, ethnographic maps of larger areas must be prepared using a much smaller scale, which will necessitate a greater degree of generalisation. ' large map scale nevertheless brings a number of methodological benefits, which might be seen to outweigh the inconvenience of multiple, large map0 sheets. The thematic variables 0 in terms of ethnographic cartography, smaller linguistic^ethnic^confessional groups 0 may be shown in closer proximity to their actual location. Cepending on whether one wishes to amalgamate or dissect certain population groups, this may or may not be desirable. 't smaller scales, in contrast, there has to be a greater degree of amalgamation of groups and generalisation of location. The decision as to how far this is acceptable or desirable may be made not on academic grounds, but may be forced by technical considerations, cost, or political hidden agendas.

*#@

Period
' map is often supplied as a supporting illustration to a text. In these cases, the geographical space for the map will be chosen for the author by the context of the narrative. ' well0chosen map series will include a variety of scales and moments in time to illustrate the dynamics of the historical processes at work .see, for example, G,*99 #he :reat -ra" GonNuests). In many cases, however, the map itself tells the story, with text in a supporting role. It is often decided to illustrate only the maximum extent of a realm, or deepest penetration of a campaign into enemy territory, and all other considerations are subordinated to these objectives. ' date may be chosen for the staticity, or dynamism of the contemporary scene, or because the moment represented a moment of calm before a period of great change. Mor his $enguin -tlas o! 5edieval History, =c&vedy .*23< edn.5 chose to avoid the &nglishmanLs obvious year .*#EE5 and opted instead for *#4*, a year of vastly greater drama. $is maps use a carefully balanced range of outlines for the various states. The By9antine &mpire, in bold, dominates the map for *#4*, in which year the &mperor )omanos IP Ciogenes was marching eastwards to catastrophe at =an9ikert. The following map, for *#2", shows a broken By9antine husk in a finer outline, barely holding on to the shores of the Bosphorus.

*#A

=c&vedy

*#E

'dministrative Bevel

's described above, a thematic map may be designed to represent any chosen space from the microscopic to the universal, but the chosen geographical space will itself be divided into sub0units across which the data will be distributed. In terrestrial, geographical thematic maps, this \area of reference\ may be as large as continents on a world map, or as small as flowerbeds in a map of a single garden. The area of reference must be appropriate to the chosen geographical space and the scale of the map. The map can otherwise appear cramped, or too sparse. Mor ethnographic thematic cartography, the ideal level is the settlement, but sometimes the scale or geographical space demands a higher administrative unit, while at other times, settlement0level data is simply not available. -ettlement level maps of extensive areas can also become gigantic B see below. (iven a similar scale and map si9e, the visual impact of a map based on a higher area of reference will anyway be cleaner and less cluttered than one at a lower area of reference. $owever, having recourse to a higher administrative layer allows the cartographer more freedom in generalisation and manipulation of the data. Mirstly, one loses sight of the areal distribution of population and secondly, smaller groups can more easily be consigned to the catchall \others\ category.

&ven at the smallest level, boundaries divide one unit from another. Nsually, any number of alternative boundaries might be drawn, each of which will result in a different statistical landscape. This is referred to as the Kmodifiable unit area problemL. $ence, a landscape may be divided once by reference to river catchment areas, again by mountain and valley, again by contemporary economic 9ones, and again by historical counties. &ach scheme may be justifiable but each excludes the other. Nnless there be some compromise between them the choice of one precludes the adoption of any of the others.

*#4

'ny politically recognised boundary results from human will and is subject to change over time. Murther, its definition may have been motivated by expressly political considerations. Nnless mechanisms are in place to prevent gerrymandering, an administration that undertakes boundary reform presents itself with the opportunity so to define the new boundaries as to favour any population group according to whim or fancy. The results of the )omanian census of *2<#, for example, proclaimed that *E of the "< counties .within the contemporary administrative boundaries5 on the territory acquired at Trianon from $ungary showed an absolute and four a relative )omanian majority. ' $ungarian absolute majority was recorded in three. The same results applied to the <* pre0Trianon $ungarian boundaries, however, would have given twenty )omanian, ten $ungarian and one (erman majority. The changes had resulted from appending areas of dense )omanian population onto the edges of $ungarian or (erman blocks, in order to dilute their influence. The boundary0changes at the junction of Bihar, -9atm6r and -9il6gy counties .Transylvania5 in *2"A are particularly striking. rmih6lyfalva and ,agyk6roly districts, both with a high proportion of $ungarian0speakers, were detached from the former two and attached to the predominantly )omanian0speaking -9il6gy %ounty. 'll three counties obtained thereby a )omanian0speaking majority .Illustr. 45. Dhile it cannot be denied that this amounted to gerrymandering, the former $ungarian administrative structure, taking the three counties as a whole, had greatly favoured the $ungarian population. The total population of the counties according to the $ungarian census of *2*# was *,##<,E#", of whom @EY were $ungarian and A*Y )omanian. The proportion of $ungarians within the old border was EA.*Y .-9atm6r5, <4.2Y .-9il6gy5 and A".3Y .Bihar5, which average to A*.2Y. The $ungarian proportions in the same census data, on the *2"A )omanian boundaries, were @E."Y, @2.@Y and @".@Y respectively. These average to @E.#Y, an exact match with the overall proportion. To be fair, one must note that the former $ungarian administrative structure, whose roots went back many centuries and therefore had considerable historical legitimacy, had included a number of )oyal Mree *#3

%ities and +rivileged Towns, and these had long been dominated by $ungarians and (ermans due to restrictive settlement laws.

%hanges of administrative boundaries within the Banat and Ba!ka have been kaleidoscopic, and almost defy telling. These boundary changes have fortunately occurred mainly at the county^regional or district level, with most of the community boundaries, particularly in )omania, remaining intact. The most striking cases of settlement0boundary changes are the division of the K :emarkung' .outlying territories5 of =agyarkani9sa into nine and that of -9abadka into sixteen sub0units.

*#2

Bihar, -9atm6r and -9il6gy

**#

!aps as $y-9o,s

's implied above, maps are symbols which, through scaling and generalisation, stand for a reality which is larger and more complex. 's such they can acquire an iconic quality and can stand for a great range of subjective and emotional responses. To recall =onmonier, Pi! your grand duchy or tri"al area seems tired, run do n or !rayed at the edges, sim.ly take a sheet o! .a.er, .lot some cities, roads and .hysical !eatures, dra leader o! a ne a heavy distinct "oundary around as much territory as the sovereign, autonomous country3 Should anyone dou"t it, state on .a.er, itMs on a you dare claim, colour it in, add a name 333 and .restoB you are no merely .oint to the ma.3 Kot only is your ne

ma., so it must "e realP .!on-onier *22E 335. The first national atlases, of &ngland and Dales .*A425 and Mrance .*A2@5 proclaimed the unity of those states under the rule of the monarch. -uccessive land registrations in the $absburg &mpire and the resulting map0series were named after the reigning monarch "@ affirming the power of the crown over every part of the realm. In the title and in spatial balance, a beautiful engraved -erbian map from *3#A .Illustr. 25 places equal emphasis on -erbia and Bosnia. The very bold outline to -erbia, however, draws attention to the contemporary war of liberation against the >ttoman &mpire and a prominent legend at bottom left includes a pile of shields, -erbia uppermost, displaying heraldic devices for *E other, \unredeemed\ territories including %roatia, -rem, $ercegovina.

2& Theresianische Aufnahme 25879%7:3 Josefinische Aufnahme 258;9%58;:3" Franzizeiische Aufnahme 25;<9%
5;763" Francisco-Josefinische Aufnahme 25;8=%5;;83.

***

>ld -erbia

**"

-ymbols within =aps

,ot only is a map itself symbolic, but a map itself consists of symbols, each of which may be designed to convey a message. -ymbols on maps fall into two main categories replicative or abstract. )eplicative symbols are \designed to look like their real0world counterparts\ and are used P333 only to stand !or tangi"le o"jectsP . ent *22E *E5. &xamples are the >rdnance -urvey symbols for buildings, cliffs, and trees, from their * "A,### +athfinder series. 'bstract symbols, on the other hand, represent events or values. 's it is impossible except at the very largest scales to produce a reduced image of the world, one has to symbolise. ,ormally, the symbols representing linear features e.g., streams, rivers, lanes, roads etc. .see above5 will be wider than the actual features on the ground, and those for point features used for orientation, e.g. churches, spot0height markers, well0known attractions, will also obtain an unrealistic prominence. >n older, primitive, thematic maps of the Banat, various replicative symbols denoted \praedia\ or abandoned villages, and the predominant nationality or religion of inhabited villages.

+ropaganda or journalistic maps sometimes employ symbolic soldiers, or military hardware to make their point, and their orientation may be suggestive. ' fine example is \sixteen against one\, from a vociferous anti0Trianon tract .A Pesti Hir,ap *2<* 3* Illustr. *"5. $ere, a single $ungarian soldier stands in the centre of a reduced $ungary surrounded by %9echoslovak, 1ugoslav and )omanian soldiers. The *E * ratio refers to the total strength of the %9echoslovak, 1ugoslav and )omanian armed forces on the one hand and that of the $onv;d on the other. The $ungarian soldier, visibly taller than his KopponentsL looks down slightly on the KrestiveL )omanians and rather more KmilitantL 1ugoslavs. By showing only the territories of KhistoricalL $ungary and ignoring the rest of 1ugoslavia, %9echoslovakia and )omania, the map conveys the impression that these statesL armed forces are concentrated solely against $ungary. In the real inter0war world, 1ugoslavia was riven with **<

disunity and preoccupied with Italian pressure on the Calmatian coast and a dispute with Bulgaria over =acedonia. The %9echoslovak army was predominantly deployed along the long -udeten border against (ermany, and the )omanians strength was also dissipated, due to a dispute with )ussia over the status of Bessarabia and with Bulgaria over the Cobrudja. 'll were aware of and opposed to $ungarian revisionist ambitions, but none could dream of deploying all their forces against $ungary. In case one might find this interpretation fanciful, a second emotive illustration & hich .eace treaty as the !airest !rom a national .oint o! vie X'is reproduced below, comparing the fate of $ungary at the end of the Mirst Dorld Dar with that of (ermany and Bulgaria. Dhereas one in twenty (ermans and Bulgarians ended up inder foreign domination, for the $ungarians, it was eight in twenty.

**@

Iga9sagot

**A

Cata for &thnographic =aps

The central purpose of thematic cartography is to show the spatial distribution of data. The nature and origins of the data displayed, and how the map author has manipulated the data, are therefore vital concerns.

' census is carried out a government for its own ends, and not as an objective exercise .see for example Stanovi;tvo .o narodnosti *2AA, Introduction5. To understand the territory, economy and population, over which it holds sway, an administration works out a system of classifications by means of which the respective field might be more easily grasped. This tendency to subdivision is probably as old as mankind itself, and in terms of the population, a number of interests dictate how this system is constructed. The basic motivation for a state to take a census is self0information. Cata obtained from censuses permits governments to update their understanding of their citi9ens and should inform subsequent decisions on fiscal, educational, minorities and sometimes foreign policy. The recorded si9e of minorities within its borders is of political importance, as it will sometimes have an impact on external irredentist pressure, or internal pressure to enact minority legislation. $ungary and )omania, for instance, can barely ever agree on the number of $ungarians living in Transylvania.

In the Balkans, where the tenacity of the $absburg and >ttoman &mpires prevented the channelling of local nationalisms into indigenous state0 structures, nations grew up around the common bond of language. This was contingent not only upon a certain awareness of shared identity and nationhood, but also upon an understanding of their geographical location and extent. Banguages are not, however, islands unto themselves, they merge into one another via a range of sub0dialects, and one of the preoccupations of *2th century ethnographic cartographers of **E

the Balkans became the definition and recombination of these dialects with the major language groups. The cartographers of 'ustria0$ungary strove to maximise the number of languages and sub0dialects in the Balkans so as to impede the growth of wider -erb and Bulgar national identities. The opposite tendency can be detected in the works of the Balkan peoples. Mar from splintering the population, they unified as much as they could. .Rady *22< "5.

The individualXs language knowledge, confession and ethnicity are usually felt to have a bearing on degree of commitment to and identification with the state. )elative homogeneity in these terms was seen as having contributed to the early success of Britain and Mrance as nation0states. It should be stressed that none of the three can be relied on as an accurate indicator of an objective standard value 0 cultural identity 0 as no such objective standard value exists. 'ny individual operates within a limited sphere, and identity has far more to do with his or her personal network than with any possible objective standard such as nationality. -uch \standards\ are of some utility when their subject is viewed from afar, but increasingly lose their definition as one seeks to focus more closely. &very one of us belongs within a bewildering array of relationships with others 0 concentric circles, if you will 0 progressively stronger the closer they are to the \essence\ of oneself for example, culture/ language group .i.e. )omance5/ language/ dialect/ community/ acquaintances/ colleagues/ friends/ extended family/ immediate family. 1et in the context of ethnographic data, this complex is suppressed and at best three elements treated as definitive. &ach one of the three, even if conscientiously applied, reflects only facets of the individualXs identity, and none can be expected to cover the whole complex of identity, for each is a different index, measuring a different variable.

The initial decisions regarding which measure.s5 should be employed and published can be influenced by the contemporary relationship of the three and perceived political consequences. Dithin narrow time0spans **4

and areas, a close correlation may exist between them, but no overall rules of relationship valid generally across time and space can be established. %ensuses have been carried out in the Banat and Ba!ka since the $absburg conquest, initially as Tandeskonskri.tionen for the military. Cetailed ethnographic data, however, only begins to appear from the later *2th century, and continues to be published by the various administrations to the present day. -ince *33#, censuses of the Banat and Ba!ka have employed various aspects of all three principal measures to classify the population. >ften, two of these have been printed side0by0side for comparison.

' meeting point of cultures, Banat and Ba!ka contain a number of ha9ards for the census0collector, no matter which of the principal measures is applied and irrespective of how conscientiously. Mirstly, the principal measures of language and nationality apply more properly to the core areas of population0groups, none of which lie within Banat and Ba!ka. There is no specific local identity, just references to identities beyond the bounds of the region "A Those references are often indirect, however, and contain within themselves differentiation from the external identity. ,orthward migration from the Balkans brought about a -lavic dominance in the Banat and Ba!ka and, in the early modern period, they became the cultural centres of the -erbs. They remained so even after the wars of independence began in *3#@. 1et -erbia was resurrected south of the Canube, and remained there until Trianon. The -erbs of the ,orth, while identifying themselves with those of -erbia, had a special disparaging term for those -erbs born south of the Canube, (ed9a .=agris, Canube5. The (erman colonists who began arriving from the end of the seventeenth century also owed their identity to an KelsewhereL, in their case more remote and more diffuse. Initially they arrived not as (ermans, but as speakers of a wide variety of (erman dialects from various $absburg territories. >ver time, they absorbed the small bodies of Mrench, Italian and -panish colonists, and became 2) The three micro%nations Bunje,ac. .okac" /raso,ans are deri,ati,es of the /roats or .er s. **3

collectively known as Sch a"en. This reflects more local perceptions of (ermany than the actual origins of the colonists/ most came from the )hine0=osel region, not -wabia .around -tuttgart and the Black Morest5. Thus Sch a" came to acquire a different meaning along the middle Canube than in the colonistsX original (erman homeland. The small Bulgarian population owed its transplantation to the Banat to its %atholic faith. Their language on the one hand and the contrast with the confession of homeland on the other have maintained the separate identity of the Bulgars and act as an obstacle to their return. The 7rasovans of the south0central )omanian Banat have similarly retained their identity by reference to their origin ion mediaeval Bulgaria. 'nother problematic group is the Bohemian (ermans of the -outhern )omanian Banat, who do not belong properly to the Conauschwaben, and, though -udeten (ermans by descent sometimes appear as %9echs.

In parts of former 1ugoslavia, a local identity .e.g. Bokelj, -umadinac, Illyrian, Istrian and Calmatian5 is more significant than KnationalityL, and as such is often entered on the census0form. These micro0nations 0 in the Banat and Ba!ka, principally the Bunjevac and -okac plus the 7rasovans in )omania 0 present methodological problems similar to the above. The controversy surrounding the status of the Bunjevac and -okac is sharpened by their proximity to the -erb0%roat0$ungarian border. The 1ugoslav census of *22* records both groups separately, rather than include them among the %roats. The published results of the )omanian census of *22" similarly included separate columns for -axon and -wabian nationality as well as (erman, -9;kely as well as $ungarian. 'lthough this may have been a true reflection of individualsL declarations on their census forms, the fact that these groups were not combined with the (erman and $ungarian population groups in the published data was seen by some as deliberate fragmentation of the minority groups.

&ven were the local identities themselves clear, placing the individual among them is beset with difficulty. Dhile many villages have borne an **2

almost unitary aspect since their foundation .e.g. Bindenfeld, Biebling5, and many others have consisted of mutually distinguishable ethnic sub0 villages .e.g. -inmihaiul )oman, -inmihaiul (erman5, the majority of settlements, and particularly the larger towns, have had a mixed population. $ere, bilinguals, and even trilinguals, are not uncommon, and ethnic or linguistic appurtenance is correspondingly flexible. The )omanian census of *2<# even provided for cases where mother tongue was different from that of both parents. The most significant and most rapid shifts between groups are to be detected in the urban populations, where there is a greater capacity for a Xvicar of BrayL flexibility in the face of political change.

e-arcation o5 /aria9,es

The demarcation of variables for a census presents a dilemma reminiscent of the Kmodifiable unit area problemL, in that the choice of any single scheme will preclude all others, each of which might be equally valid. >nce that scheme is decided upon, the actual responses will be vetted and counted according to the sanctioned categories. This will involve the re0allocation of impermissible, incorrect or spoiled entries according to the rules. Thus the rules of the )omanian census of *2<#, while stating that the individual Kshall register under the race to which they feel bound by ties of tradition and sentimentL, included the proviso that the declaration must be plausible 0 i.e. that the individual Kmay not declare that they belong to races non0existent in )omania...L. Beyond this, there was Kno list of permissible nationalities, and registrations were accepted as made by the citi9ensL .=anuila *2<3 "5.

The range of ethnographic data published from any census is similarly the result of bureaucratic decision0making. -mall groups considered to be related may be combined, or large groups subdivided, perhaps with a view to moderating the impact of their collective identity. In the *"#

)omanian census of *2<#, where the permissible range of national groups was set by the KplausibleL perception of the individual, the ensuing fragmentation of the population was counterbalanced by allocating certain unconventional responses to better0known groups. It was not in every case made clear in the published data, but many of the nationalities shown in the final publications included such splinters. $ungarians, for instance, combined claimants for =agyar, $ungarian and -9ekler nationality. (ermans included -axons, -wabians and Tipser/ -erbs included 1ugoslavs, %roats and -lovenes/ )ussians included =uscal, while )uthenes incorporated Nkrainians, =alorussians and )usneacs. $ut9an and $ut9ul were combined, as were 'lbanians and 'rnauti .=anuila *2<3 "5

't times, the selection of data for publication has reflected contemporary foreign affairs. The )omanian census of *2@* may be offered as an example here. The )omanian census of *2<# had been processed with the assistance of calculating machinery supplied by the )ockerfeller Moundation .7rallert, *2<2, p.@2E5. +ublication was slow, and final publications were not released until *2<3, after a burst of activity, but 7rallert, in Cecember *2<2, stated that several volumes were still not available. Dhatever momentum had been obtained by this late publication could not be carried over into the work on the *2@* census, which was carried out in the context of a )omania truncated by -oviet, $ungarian and Bulgarian annexations and overrun with refugees and (erman KadvisorsL. 'lthough 7rallert .*2@<, p.*A5 claims that final results were obtained mechanically using punch0cards .perhaps the )ockerfeller Moundation machinery5, the further course of events in the -econd Dorld Dar prevented the )omanians from publishing any more than +rovisional -ummary Cata, in *2@@"E This publication gave figures only for )omanians, $ungarians, (ermans and >thers. 'fter the annexation of most of the more than one million )ussians and Nkrainians to the N--) in *2@#, the $ungarian and (erman minorities were the most important 2V Recensmntul eneral al Romniei din ,-.,! ". A#rilie. Date .umare 4ro,i(orii. Bucure2ti" 56::. *"*

politically and numerically, even after the retrocession of northern Transylvania to $ungary5. 's in other countries, the Third )eich had exploited the (erman minority as a lever against the government and, on *" =ay *2@<, Berlin and Bucharest sealed an agreement regulating recruitment of )omanian Polksdeutsche into the -- and Dehrmacht .)einerth, Jur :eschichte der (eutschen in RumLnien /077 "is /01H **@5. The si9e of the $ungarian minority had been the one of the prime causes of the division of Transylvania in *2@# .the -econd Pienna 'ward5, and relations between )omanian and $ungarian remained tense thereafter. The )eich stole the )omaniansL thunder by actually publishing district level data for the *2@* census before the )omanian government. -ignificantly, in the (erman publication, the Gews appeared as a separate category . 'ra,,ert (ie BevRlkerungs+Lhlung in RumLnien , *2@*. +ublikationsstelle Dien. Dien, *2@<5.

Aggregation o5 P*9,ished

ata

Dhere a large number of variables is published, as for example with the )omanian censuses of *2<# and *22" and the 1ugoslav census of *22*, the ethnographic cartographer is often, for the sake of clarity, compelled to combine some, and probably expand the \others\ category. There will be some conventions and precedents to follow. In the Banat and Ba!ka it is conventional practice to combine the %9echs ^ -lovaks, )omanians ^ Plachs, and Turks ^ Tatars. )uthenians and Nkrainians also form a convenient pairing, sometimes supplemented with the )ussians and occasionally the Bippovans. The -outh -lavs, predictably, raise the most problems. In some censuses, e.g. the 1ugoslav census of *2"* .and the )omanian census of *2<#5, combined -erbs, %roats and -lovenes 00 reflecting the contemporary chimera of a common destiny for the state0 forming nations 00 but it is more usual for them to be shown separately. It is not clear what happened with =ontenegrins and =acedonians in *2"* and *2<*/ they were probably recorded as -erbian0speakers. The 1ugoslav =uslims present particular problems, as they cannot be *""

distinguished at all by language0based data, and a =uslim \nationality\ had to be invented for them. In the earlier censuses, they appear simply as -erb0 or %roat0speakers. Treatment of the Bunjevac and -okac varies from period to period and also depending on the nationality of the academic. -tandardising data across a range of censuses for use on a historical range of ethnographic maps is therefore extremely problematic. The ethnographic cartographer must seek to work out a set of data0classifications which retains a high degree of differentiation, but which is sustainable across the entire territory and throughout the period under investigation.

O9ser+ations on the %hree !eas*res

!other1%ong*e

ata

%lassification according to mother0tongue, i.e. the linguistic community to which the individual belongs, is likely to give an inflated impression of homogeneity, particularly with long0established nation0states, as persons of non0majority nationality tend to become absorbed into the majority mother0tongue mileu to advance their career, >ver generations, this can lead to the complete absorption of minority groups. &ducation can be mobilised to accelerate this tendency, as was done, for example, by the $ungarian government particularly after *3E4. The $ungarian censuses of *2## and *2*# recorded whether individuals belonging to a linguistic minority could additionally speak $ungarian. The language question in censuses is not always couched in terms of the mother0tongue. Nmgangssprache .everyday language5, Bieblingssprache .favourite *"<

language5 or Cenksprache .language of thought5 have all been used .7rallert, *2E*, p.*##5.

Con5essiona, data

In the $ungarian censuses, and the inter0war 1ugoslav censuses, mother tongue was supplemented by confession, or the religion and denomination to which the individual adhered. In combination with mother0tongue, this gave a more accurate guide to the individualXs identity, and permitted the differentiation of, for example, the %atholic -erb Bunjevac and -okac, the %atholic Bulgar 7rasovans and, most tellingly, the -erbo0%roat speaking Bosnian =uslims. In some cases, where state census data is unavailable, church records may offer an alternative source of information on one or another population group. These may provide a fairly accurate picture of the si9e of a group where religion and nationality coincide, )omanians, such as among the -erbs, or

Nationa,ity data

,ationalit"4 as a means of classification gained ground in the context of the new national states created by the +aris +eace Treaties after the Mirst Dorld Dar. This tended to compensate for the homogenising tendencies of the dominant mother0tongue classification, of which the best0known example may be the $ungarian0speaking Gews of Transylvania. 'lso hidden to some degree were the -9atm6r -axons, whose numbers were depressed in $ungarian censuses by virtue of the fact that they mostly spoke $ungarian .7rallert *2<2 A#*5, and the (ypsies, for the same reason. >nly with the )omanian census of *2<# 2% 0thnic appartenance 2Romania" 56=>3? 0thnic -rigin 2Romania" 56:53? 0thnic Nationality 2@ugosla,ia"
56:;3? Nationality 2otherwise3.

*"@

was the si9e of these three groups fully appreciated. ,ationality is, however, also problematic, primarily because it stems from a concept 0 the \nation\ 0 which has proved itself capable of almost infinite re0 interpretation and is probably the most mythical and least understood of human constructs. &specially in terms of censuses, nations 0 more than languages and confessions 0 are creations, malleable to the designs of the administration of the day. ,ationality as a defining property of the individual is also purely subjective. Inasmuch as it is derived from the statement of the individual, it measures the will to identify with a group, rather than actual belonging, or a testable capacity such as speech. It can therefore reflect the processes of assimilation and opportunism even more rapidly than mother tongue. Dhile the will of the hard core of any group may never falter, this is less likely among the peripheral members, whose will may change with the wind. This may be demonstrated by comparison of the (erman Bestandsau!nahme of ,ovember *2@# with the )omanian census of 'pril *2@*. In *2@#, <@3 out of EA@ settlements in the )omanian Banat recorded a higher (erman population .Volksgru..e publication5 than the *2@* )omanian census .provisional summary data5. The difference totalled <*.*A# persons. or *<.*Y of the *2@* figure. Both were based on ethnic origin. ' special case is that of the 1ugoslav census of *2@3, in which many (ermans chose to be recorded as $ungarians out of considerations of personal safety . Save+ni Javod +a Statistiku *2AA ?BI5. =ore modern examples may be offered by the exponential growth of a (erman minority in the former >stgebiete in +oland after *232, or the tendency of %roats .and Bunjevac and -okac5 in north0west Ba!ka to identify themselves as less0conspicuous 1ugoslavs in the *22* 1ugoslav census.

$*--ary

%ensus data, therefore, even before it reaches the desk of the ethnographic cartographer, is the result of political dialogue. It will have been produced not as an aim in its own right, but for the purposes of government, to inform *"A

subsequent policy. Tensions with neighbouring states may have an influence of which data are inquired after and which ultimately published. >f the three principal ethnographic measures .religion, language knowledge and ethnicity5 none indicates an objective standard value. &ach has its strengths and weaknesses, their uses and abuses. They are best used in combination. In the Banat and Bacska, all three have been used in various guises since *33#. The complex mix of ethnic^linguistic groups, a number of local variations to larger, extra0regional identities .-erb, (erman, etc.5 and a considerable linguistic and cultural flexibility among the people greatly magnify difficulties which exist even in areas which are relatively unitary.

The cartographer must be aware that the categories he is presented with by the censuses themselves are themselves political constructs, and be prepared, where necessary, to combine them, or to make estimates of the relative strength of their major sub0groups. ,o set of data0 classifications can be wholly satisfactory, as it will by nature be a compromise. $e must also be alive to the effects of changed administrative borders over the period, and, at the deepest level, to the degree of reliability of the data itself. The viewer, in his turn, must be aware of these concerns, and be prepared to refer to the data behind the map. This will permit an understanding and informed criticism of the decisions of the cartographer.

*"E

%HE %ECHNIS#E$ O" E%HNOGRAPHIC CAR%OGRAPHD

-ince its origins in the late sixteenth century .-cultetus, 7arte der Bausit9, (:rlit9, *A2<5 the mapping of ethnographic data may be said to have been characterised by more dynamism and drama than any other branch of cartography .7rallert *2E* 225. The discipline grew out of, reflected and fostered the growth of national consciousness, so areas like the swathe of territory from &stonia to the deep Balkans, where political developments were often closely connected to ethnicity, generated a wealth of ethnographic maps. The territories of the multi0ethnic Tsarist, $absburg and >ttoman &mpires became the \classical\ territories for the discipline. The first thematic map with ethnographic data for each settlement appeared around *3##."3 The discipline did not, however, really take off until the first publications based on official data 00 an *3A* map of &uropean )ussia by the (eographical -ociety of -t. +etersburg, based on a \+opulation0)evision\ of *3<@. 'n ethnographic map of the 'ustrian =onarchy soon followed, based on the *3A* census
"2

. The first

comparable map for the >ttoman &mpire .-ax *3435 was produced with the %ongress of Berlin in mind, but was published too late for consideration<#. Dith the dawning of the idea that political and ethnic boundaries might best coincide, the discipline became heavily politicised, unleashing from time to time full0scale \map0wars\ between the champions of the existing states and the various aspirant nations. These have generally accompanied, and outlasted, KconventionalL wars as being less costly, and have both communicated war0aims and sought to influence the subsequent peace processes. 't Berlin in *343, +aris in *2*20"#, and +aris again after *2@A, the cartographers were in demand. In the *22#s, Bosnia has kept them busy. The periodically intense
2B 8orabinsz;y, E.(.0 Atlas =egni Fungariae 9ortalis. eue un" vollst<n"ige ,arstellung "es 8.nigrei'hs Ungarn auf -T Tafeln in Tas'henformat. 1in geogra9his'hes othD un" F:lfsb:'hlein f:rs gemeine -eben. Pienna, year not given, arou" #BAA. 2zoernig, 2.0 1thnogra9his'he 8arte "er Jsterrei'his'hen (onar'hie. #0#,)B&,AAA. Pienna,

2$ #B)V.

CA !aX, 2.0 1thnogra9his'he 8arte "er euro9<is'hen Tur;ei un" ihrer ,e9en"enze zu Anfang "es Eahres #B%%. 2a. #0&,AAA,AAA, Pienna, #B%B >=a"y OAustrian (a9s of the Bulgarians in the ineteenth 2enturyO)

*"4

investment in these \wars\ has sometimes stimulated cartographic innovations which subsequently rubbed off onto other branches of thematic cartography .7rallert *2E< 225.

Technical

and

methodological

advances ethnographic

in

censuses, data,

resulting

in the

increasingly

comprehensive

encouraged

ethnographic cartographer to refine his own methods. Improvements in printing methods meanwhile promised the faithful reproduction of the more complex schemes that this data required. $owever, the indistinct nature of the basic data and the emotional content of nationalism combined to frustrate these ambitions, even when considerable technical problems were overcome in the production. ' map may display the equivalent of many

*"3

-'?

*"2

pages of dense text, and can depict ethnically very complex regions with apparent clarity, yet it may be every bit as tendentious, and its very immediacy can be a trap set for the unwary. 'uthors of such maps often exploit this quality quite deliberately. Dith questionable base0data, considerable technical difficulties in drafting and printing, and political intrigue all involved in its production to some degree, every ethnographic map is a potent brew. In a volatile atmosphere of imperial decay and frustrated nationalisms \the methodologies of nationality0 and language0 maps were developed, refined, debased and perverted according to the pressures of academic research or political necessity\. &ach methodology has is own advantages and disadvantages, and any map based on any one of these will pass the data through a filter peculiar to that methodology. This makes the field one of the most interesting and problematic in thematic cartography .7rallert *2E<5.

The conscientious ethnographic cartographer has to acknowledge that, as discussed above, his source data is itself skewed and substantially a reflection of the Deltanschauung of the authority behind the census. But the data is only the first difficulty that must be addressed. The cartographer also has to overcome a series of methodological problems while preparing ethnographic maps.

These are principally $ow to achieve a balanced portrayal of areas with mixed population/ $ow to portray isolated groups/ $ow to reflect the density of population, particularly with regard to the disparity between cities and the countryside/ $ow to treat thinly, or uninhabited regions.

. Translation by Bru'e (it'hell.

*<#

 The selection of an appropriate geographical sub0layer for the thematic map .after 7rallert *2E* *#@5.

=any methodologies have been developed to address these problems. &ach is appropriate to a particular range of circumstances, but each may also be employed for deliberate distortion of the resulting map. Bound hand and foot at every stage by emotion and politics, the ethnographic cartographer has acquired an armoury of techniques and stratagems from which to make his selection, depending upon the result he wishes to achieve. In sum, the contentiousness of ethnographic cartography stems from vague definitions within the data portrayed, on the characteristics of the various available approaches and, at a deeper level, in the choice of filter or combination of filters that authors tend to set between the data and their readership.

%HE !E%HO OLOGIE$

The methodologies of thematic cartography may be sub0divided any number of ways, and no one classification is inherently right. The best that may be said is that some classifications may have more utility than others. The most fundamental division might be said to be between those maps that give a rough guide to the spatial distribution of variables .qualitative maps5, and those which are mathematically based on actual data .quantitative maps5 for those variables.

*<*

S*a,itati+e !aps
]ualitative maps, being based neither on discrete territorial sub0units, nor on numerical data, are the easier to prepare, the more widespread and also the more problematic. They generally record only the spatial distribution or location of kind, the presence or absence of a single variable .ethnic^linguistic^confessional etc. group5, in vaguely defined areas, ignoring the presence of any other variables. There is no indication of quantity, or proportion. They will often be ostensibly based on official data, but there is no clear relationship between that data and the construction of the map. 's such they carry little weight. The reader is referred to Illustrations ' and B, both of which purport to show the extent of a particular population within a territory. The first shows )omanians in the Banat. Territory supposedly inhabited by )omanians is coloured in red, the rest of the map remaining white. 1et there does not appear to be a close relationship between the distribution of the coloration and data on the )omanians as per census results. 'lmost all of the southern part of the 1ugoslav Banat is shown as )omanian, considerably in excess of their actual distribution. There is no effort to differentiate between thoroughly )omanian and mixed0population areas, nor is there any differentiation between densely and thinly inhabited areas. 's is often the case with )omanian maps, the most thinly inhabited areas are counted as being )omanian inhabited .which is reasonable5 and coloured with exactly the same intensity as urban areas .which is not5.

*<"

=olin -eisanu

*<<

The main problem with the second map, -eiFanuLs &the (istri"ution o! the Romanians', is that it accords with its title only beyond the frontiers of )omania. It behaves completely differently within the country. Dhile beyond the frontiers, areas of KcompactL and KdenationalisedL )omanian settlement many greatly exaggerated and some altogether spurious<* are individually shown, contemporary )omania is shown as one undifferentiated whole, with no ethnic holes showing. 1et the )omanian census of *2<# demonstrated that, in many parts of the country, such as the Banat .A@.@Y )omanian5, or Transylvania .A4.AY )omanian5 the Cobrudja .@@."Y )omanian5 or Bessarabia .AE."Y )omanian5 or Bukovina .@@.AY )omanian5<", )omania was far from being a homogeneous nation0 state.

<*. *our islan"s of +'om9a't6 =omanian settlement are sho/n in a 9art of (oravia, /hile the
2ze'hoslova; 'ensus of #$CA ha" referre" to only #) in"ivi"uals. The greater 9art of Transnistria, an" 'onsi"erable 9arts of the further U;raine, are sho/n as =omanian. 1ight substantial grou9s are sho/n bet/een the FungarianD=omanian frontier an" the Tisza, /here, in #$#A, only #&,AAA >A.%Y) of a total 9o9ulation of #,$CA,AAA /ere =omanianDs9ea;ers >$ /agyar S+ent 0orona ors+1gaina $%$&% 2vi n2ps+1ml1l1sa: ,%r2s+: $ n2pess2g f(&& adatai (+s2ge 2s n2pese&& pus+t1 , telepe s+erint , &2. ;.tet. ##. rZsze. (agyar !tat. 8.zlemZnye;0 Bu"a9est, #$#2). C2 3ecens4m5ntul 6eneral al 7opulatiei 3om5niei din 8-% 9ecemvrie $%'&. 4ubli'at "e ,r !abin (anuila, ,ire'torul =e'ens[m\ntului @eneral al 4o9ulatiei . Polumul UU0 eam, -imb[ (atern[, =eligie. Ti9[rit la (onitorul ?fi'ial, Um9rimeria ational[, Bu'ureti, #$CB. 4ages )B VA.

*<@

S*antitati+e !aps

]uantitative maps may be sub classified as being based on data occurring in areas, on points or along lines. They record the spatial aspects of numerical data, and it may be said that they exist to illustrate the transformation of tabular data into spatial format. They are Ka symbolised generalisation of the information contained in a table<<. ' quantitative map may illustrate the absolute value of a variable in an area, or its value relative to another variable, area or period, thus facilitating a wide range of analytical tasks.

]uantitative maps may also be divided neatly into absolute and relative types. 'bsolute maps represent real0world data values .e.g. number of persons, tons of coal, etc.5. )elative maps, on the other hand, show the relative strength of one or several variables across the constituent parts of the map. ]uantitative maps may show a single variable, or show several.

S*antitati+e !aps 5or %he Chorop,eth !ap

ata 9ased on Areas

The most familiar types of quantitative area0based maps are better known as choropleth, or area0coloration maps. The thematic overlay is divided into areal units, each of which represents a constituent part of the thematic whole. Thus a country may be shown as counties, or a city as electoral wards. &ach sub0unit is processed according to its value in the data0range being mapped. Palues, which may be in the form of absolute or relative values, are portrayed by allocating a specific shade or colour to particular data ranges.

CC Aenks" +eorge B.C 2ontem9orary !tatisti'al (a9s D 1vi"en'e of !9atial an" @ra9hi' Ugnoran'e. Ameri'an 2artogra9her C, #$%V0 99.##D#$.

*<A

-ingle0Pariable %horopleth =aps

In a Ksingle0variable absolute choropleth mapL, the colour accorded to the area reflects the absolute value of the variable. The more common Ksingle0variable relative choropleth mapL shows the overall extent of one variable only, but distinguishes the proportion of the group in the total population by means of a graduation0scale. >ften this type of map shows the variable only where it forms the

*<E

)omanian delegation

*<4

majority of the population. 'lthough it appears to offer a scientific approach, it is easy to doctor both the graduations and the shadings used to the produce a misleading effect. The )omanian delegation to the +aris +eace %onferences of *2*20"# presented an extremely questionable, and in fact crassly falsified .=itchell *22* AE045, set of maps of this model .Illustr.*2, "#5. Their K)umgnen 9wischen den 7arpathen und der TheiHL <@ claimed to be based on the district level results of the $ungarian census of *2*#, yet a close examination reveals that this claim is at best inaccurate. The $ungarian %entral -tatistical >ffice reprinted this map in )eBui7, along with a set of revised maps in the same style. >ne represented how the )omanian map should have appeared had it really been based on the *2*# census. ' second highlighted Kthe districts and towns whose data on the map published by the )omanians are obviously incorrectL. ' careful comparison of these claims with the actual results of the $ungarian census of *2*# confirms that the criticism is correct in every point concerning the counties %sk, Ndvarhely, $6roms9;k, Brass8, ,agy07RkRlle, 7is07RkRlle, and -9il6gy. In other words, the )omanian map deliberately falsified the results for the thirteen highlighted districts in these counties. The proportion of the )omanian urban population shown for C;s, -9amosjjv6r, =arosv6s6rhely, Cicses9entm6rton, &r9s;betv6ros, 7;9div6s6rhely, -epsis9entgy:rgy, Brass8, -9il6gysomly8, Tilah, ,agyk6roly, -9atm6rn;meti, ,agyb6nya, Melseb6nya and ,agyv6rad was also distinctly exaggerated. It cannot be mere coincidence that all these districts lie either along the disputed outer frontier of )omania, or among the principally $ungarian or (erman districts of the -9eklerland. 1et even had the data been conscientiously and correctly transferred to the map, the choropleth colour0ladder was deliberately loaded to enhance the visual weight of the )omanian population.

. The 'o9y /as obtaine" via ,r. *is'her from a Fungarian 9ubli'ation. Ut 'orres9on"s eXa'tly /ith the 'o9y 9ublishe" in the =omanian Sie&en&)rgen' of #$&C.

C& =e9rinte" in0 OSie&en&)rgenO7 Unstitut f:r rum<nis'he @es'hi'hte in Bu;arest0 Bu;arest, #$&C an" 3um#nische :thnographische !and arten und ihre 0riti . !taats/issens'haftli'hes Unstitut. Bu"a9est, #$&2.

*<3

>ne is accustomed to seeing colour ladders where the middle shade corresponds to the median value in the data. $ere, however, the scale is severely distorted. Cespite a superficially generous seven value0ranges being given, the top range covers )omanian absolute majorities, and the other six diminishing degrees of minority. To this should be added that the top two are mutually indistinguishable and the third can barely be distinguished from these. The next two are again nearly identical. The result is that we have only four distinguishable groups <#Y and above, <#Y to *#Y, *#Y to *Y, *Y and less. 'reas where the )omanian made up only <#Y of the population are therefore hardly distinguishable from those where they were in a majority .see particularly ,agy0 and 7is0 7Rk:lle counties in eastern Transylvania5. ,one of this, perhaps, would matter, if the map had not been used to support territorial claims at Persailles. The lopsided nature of the )omanian map was clearly demonstrated by the publication by to two further companion maps by 7-9$, both using the )omanian methodology. >ne showed the $ungarian population, the other the non0)omanian population. In effect, the latter should be a direct reciprocal to &RumLnen + ischen den =ar.athen und der #hei>' that it is not clearly displays the tendentiousness of the methodology. The great wide stripe of )omanian population from north to south in Transylvania is compressed and non0)omanians are shown to overwhelmingly dominate the entire +artium and eastern Transylvania.

=ulti0variable %horopleth =aps and the +roblem of $ow to -how =inorities

'ttempts to display the extent of many population0groups, whether majority or minority, may be far more elaborate, and has attracted considerable attention from ethnographic cartographers over the decades. It can be difficult to obtain a satisfactory result, and the ingenuity of the cartographer can be stretched. The shades representing the range of variables, and the graduations, have to be chosen with great care, and there is always a trade0off between empirical accuracy and the overall \look\ of the map. *<2

In a multi0variable absolute choropleth map, the absolute values of several variables 0 not necessarily all 0 would be portrayed by means of area coloration. In order to succeed, this configuration would call for the value range of each variable to be represented by a distinct set of colours or shades. 1et the colour0ladders for the different variables would have to be comparable at equal values. 'dditionally, multiple variables could only be shown if a system of hatchings were used. The practical difficulties in this type of map are such that present author has not come across any published examples. ' multi0variable relative choropleth map also shows several mutually distinguishable variables. In its simplest and most common form this shows each variable only where it forms the majority of the population. In this case, areas where no one population group are in the majority may be left blank.

*@#

)onai ,ationalities

*@*

=anuila

The portrayal of minorities is one of the most challenging methodological problems, as mentioned above. ' number of different approaches have been developed, including the use of symbols, or hatchings. The K&tnographische 7arte )umgniensL by Cr -abin =anuila, director of the )omanian statistical service, showing the results of the *2<# census, is an example of the symbol approach<A. ,ine nationalities .)omanian, $ungarian, )uthenes, Bulgars and Turks, (ermans, )ussians, Gews and (agau95 were shown, differentiated into three degrees of majority .relative, absolute A*Y 0 32.2Y and totality 2#Y 0 *##Y5 and four degrees of minority .*#0"#Y, "#0<#Y, <#0@#Y and @#Y0A#Y5. &ach nationality is accorded a colour0ladder for nations forming majorities. This scheme is fine in principle, but the graphical means chosen for the portrayal was poor. The seniority of =anuila, and the detailed nature of the data from the census, led one to expect a more advanced product. The three degrees of majority are dealt with in similar fashion choropleth area coloration of varying intensity while the minorities are represented by undersi9ed symbols overprinted on the majority backdrop. ' triangle, circle, square or rhomboid represents the various degrees of minority. But all the symbols are the same si9e, and there is no obvious hierarchy of value. The significance of the rhomboid @#0A#Y minority is severely undermined by two factors, neither of which can be accidental. Mirstly, an almost identical rhomboid is liberally sprinkled beyond the national boundaries to represent K)omanians in the neighbouring countriesL. Their large concentrations clearly suggest a low unit value/

this appreciation cannot but be carried across to the rhomboid within the country. This is made the more likely by the fact that the two rhomboids are printed together, and set aside from the other minority symbols .*#0 "#Y, "#0<#Y, <#0@#Y5. The @#0A#Y symbol is additionally in the optically least weighty location of the page.

C) Bu'harest, #$&#. Base" on a metho"ology "evelo9e" by the @erman 4rof. 4aul -anghans.

*@"

The map attracted a damning, and mainly justified criticism from the $ungarian %entral -tatistical >ffice .7-9$5, which was accompanied by an illustration of how a different methodology would have brought about a radically different map. %oncentrating their criticism on the handling of one particular county =aros0Torda .which is itself perhaps a questionable technique5 the 7-9$ printed the =anuila version of the county beside a district0level proportional pie chart version. The weakness of =anuilaLs is immediately clear. The county in question had been subject to boundary reform, which had secured 0 by the narrowest of margins a )omanian majority. The )omanian census of *2<# reported on both nationality and mother tongue data. In terms of nationality, the population comprised @A.3Y )omanians and @".EY $ungarians, so a relative )omanian majority. The closeness of the figures is underscored by the fact that mother tongue data showed a relative $ungarian majority .$ungarian speakers @A.2Y, )omanian speakers @A.AY5. Cespite the closeness of these data, the )omanian population is represented by area coloration and the $ungarian population by a small rhomboid. Pisually, the $ungarian population is thoroughly diminished. 's the )omanians constituted at least a relative majority in all but three counties after *2"A, the overall effect of this methodology was bound to enhance the impression of a homogeneous )omanian nation0state, an effect that can only have pleased Bucharest.

*@<

=anuila

*@@

cs

' more recent outing for this methodology was published by Tolt6n cs .,em9etis;gek a t:rt;nelmi =agyarors96gon *3@# k:rRl, Budapest, *23E5. $e again uses area0coloration for majorities and symbols for minorities, but he uses the various symbol shapes to represent national groups and the si9e of the symbols to represent proportion .*#0"AY and "A0A#Y5. This is certainly an improvement on =anuilaLs approach, for there is a correlation between the proportional strength of the minority and its visual impact. 1et the symbols representing the various nationalities are not really visual equivalents .i.e., a rectangle, a circle and a square do not really convey a similar visual weight as an x or a i sign5. The disparity between the impact of symbol and area0coloration is again very apparent. This can be is particularly clear in the case of Bihar %ounty. The county is highlighted and shows a $ungarian majority with a )omanian minority of "A0A#Y. If we were to refer to the except from a proportional symbol map for *2## we would see that the population of the county falls into two distinctly different parts. The $ungarian .red5 population concentrates in the west, in large villages and towns around the city of ,agyv6rad. 2*.*Y of the urban population of Bihar county in *2*# was $ungarian speaking, and only A.EY )omanian speaking<E. The )omanian .grey5 population on the other hand, overwhelmingly dominates a good two0thirds of the county, but are distributed between a huge cloud of small villages. The *32# census recorded "*22@# .@".EY5 of a total population of A*E4#@ as )omanian, "3<3#E .A@.2Y5 as $ungarian. <4 ' variation on this method is to use outlines in the minoritiesL colours for mixed areas, rather than symbols. This is even less successful, as may be simply demonstrated by reference to the following graphics.

CV Ea;abffy 1lelmGr, 9.) C% Balogh, ;a n2pfa<o /agyarors+1gon=, Bu"a9est, #$A2 9. B)$.

*@A

'cs

*@E

Bihar

*@4

&dges

*@3

7ocsis

' considerably refined descendant of this methodology was developed by 76roly 7ocsis, of the $ungarian 'cademy of -ciences M:ldraj9tudom6nyi 7utat8 Int;9et in Budapest. $is settlement level Banguage =ap of Transylvania, *23E .Budapest, *2325 also uses area coloration where a mother tongue enjoys an absolute majority .six languages and KothersL5 and symbols for linguistic minorities .seven languages, no KothersL5. The symbols for minorities .a circle for @2.2Y0"AY and a triangle for "@.2Y0AY5 are graduated to represent absolute numerical strengths of fewer than A##, A#*0*###, *##*0"###, "##*0@### and over @###. -ettlements where no one language is in a majority are shown colourless, but with hatching. KNninhabitedL territories, for which no definition is offered, are extracted from the territories of adjacent settlement, and shown blank. ,o fewer than A2 towns and cities are shown as proportional pies .see below5 around the edges of the map, their urban districts remaining blank within the map itself. The cities of TimiFoara, 'rad and )eFiOa now have a population vastly greater than any other settlement in the Banat. &xtracting them to the edge of the map permits the populations of the neighbouring settlements to be shown without difficulty. The main difficulty with 7ocsisL map is the sources. $e claims the map to be based on the census of *33#, *2*#, *2<#, *2@*, *2AE and *244, the results of natural population development, and migration between *2AE and *23E. 1et there are no indications of how these various sources were used to bring about the results we see on the map. 'lso, while one can make rough estimations of the whole population where minorities are indicated, there is also no indication of the density of population in settlements where no minorities exist.

*@2

7ocsis

*A#

7ogutowic9

+erhaps the most sophisticated and successful of all the choropleth maps covering our area, and to which 7ocsisL map owes some of its methodology, is however, already nearly eighty years old. Its extremely effective design suffered only from the $ungariansL perennial failing of allocating the visually striking red to their own and an optically close orange to (ermans. Though one of the flood of ethnographic maps examining the territories of the 'ustro0$ungarian &mpire in the aftermath of the first Dorld Dar, the ethnographic map by 76roly 7ogutowic9 seemed to escape the tendentiousness which plagued most of its contemporaries. 's elsewhere, the majority .seven languages are differentiated5 is portrayed with surface coloration, and the minorities with symbols. 'djacent areas with no minorities, and similar other characteristics, are combined and their internal boundaries dissolved.

Dhere 7ogutowic9, even today, leaves his rivals standing is the next level of sophistication. The hue accorded the majority varies with population density. &ach majority hue has four levels #0"A, "E0@A, @E0EA, and over EA inhabitants^ km". Cistinct symbols, again in the colour applicable to the particular languages, represent the minorities. These have two layers of sophistication their various shapes indicate the actual si9e of the minorities .in the ranges *#*0"A#, "A*04A#, 4A*0*,A##, *,A#*0<,A##, <,A#*04,A## and 4,A#*0*A,### inhabitants5, and their si9es indicate the density of population .values as above5. These symbols actually decrease in si9e as the density increases, to avoid the crowding and displacement that would occur in towns or small, dense, minority0rich areas if this were not the case.

Dhere this map scores so highly over 7ocsisL 0 its nearest modern equivalent 0 is that one can work out from the symbols both any actual minority population and its density, which might well be different from *A*

that of the majority, or from any other minorities. 's with 7ocsis, one cannot work out the si9e of any majority population, but one can nevertheless observe the density.

The methodology has a number of consequences not seen on other maps. The reduction in the intensity of area coloration for thinly inhabited areas keenly affects both )omanian and $ungarian populations, particularly in Transylvania. Dhile no uninhabited 9ones are shown, the palest shading is pretty close to white, so one might be excused for thinking such areas empty. 's the shading represents "A persons ^ km" or less, one might be forgiven. The )omanian crescent0stripe across western Transylvania is shown to be extremely thinly inhabited in parts, and most of the -9eklerland, on which $ungarian territorial aspirations hinged, was also revealed to be thinly inhabited, as was Transdanubia .$ungary west of the Canube5. It may be that this was politically uncomfortable within the revisionist inter0war climate, and this may account for the map having had so little influence on later maps.

*A"

7ogutowic9

*A<

$atchings

' different method for showing several variables within one area is to employ hatchings. This approach was already developed by the time of -axLs map of the >ttoman &mpire .*3435. -tripes in the colours or shades representing the various groups are used in areas of mixed population. By this means, -ax was able to give a good impression of the complex situation in Bulgaria, but it is not really possible to quantify the population groups illustrated. Indeed these are the strength and weakness of the technique. It is often used where minorities live scattered within ill0defined areas, or where credible and accurate settlement level data is not available. &ven where such data at is available, hatchings are often used merely to convey an impression of mixture rather than concrete information, and this is reinforced by the common tendency to print hatchings unconfined by any administrative boundaries. In many cases, where the processes by which the data is converted to the actual patterns on the map is not clear, there is little to choose between such maps and qualitative maps as described above.

1et the map by +alovics demonstrates that hatchings may be used very effectively to convey complex information within set boundaries. The map is based on district0level data from the *2*# census, and boundaries between districts with similar populations have been dissolved. Thus large homogeneous areas are combined whereas areas of great diversity are shown separately. 's might be expected, the greatest confusion of colour occurs in the Banat and Bafka. By a judicious use of colours, he manages to illustrate no less than *" variables, in proportions of over 2#Y majority .uninterrupted colour5, 2#04AY, 4A0A#Y, A#0"AY and "A0*#Y minority. The only illustrated national group that could fairly claim to have been hard done by is the Bosnian =uslims, who appear as white, and might therefore be counted as absent. 'lso, in reproduction, the black for the %9echs and the dark blue for the (ermans very nearly completely merge. >n the original, all the other colours work admirably, both individually and *A@

in combination and no one colour really dominates.

<3

'n examination of

the hatchings0scale nevertheless leads to some confusion, as of the five proportions that may be illustrated on the map, only three appear on the key.

CB Un fa't 'olour re9ro"u'tion is a maLor issue in its o/n right, an" nee"s to be a""resse" seo9arately.

*AA

+alovics

*AE

The +rism =ap

In a prism, or <0C choropleth map, the absolute value of the variable for each area is represented by the projection of the area to an appropriate height. If desired, the proportion of the variable within the total population value for the area can be expressed by shading its surface. In other words, the height of the prism map columns may represent the number of =ontenegrins in a district, while the shading might represent the proportion of =ontenegrins within the total population of the district. 'lternatively, a multi0variable prism map may be used to display the full range of variables in an area, each variable being represented by an appropriate percentage of the height of the column.

Dhile the prism map is intuitively attractive, the methodology has some inherent limitations. Mirstly, taller columns at the \front\ of the map will obscure smaller or equal columns behind, and this may hide important information. The effect of this may be constrained by careful choice of four variables within the methodology height0scale, eye distance, rotation and tilt. The height0scale relates to the column height per unit of value. The eye distance refers to the apparent closeness of the map to the viewer, the rotation is the divergence from north0south orientation and the tilt is the divergence from birdsL eye view. 'ny of these can be adjusted to achieve the desired effect and one can end up with a rather unfamiliar look to the resulting map. The single0variable prism map of =ali IoJ district, showing the population by settlement as recorded by the 1ugoslav census of *22*, is a case in point. 's the two towns Prbas and 7ula dominate the district, with @A*E2 .EA.3Y5 of the total population of E3A4E, their two columns would obscure those for =ali IhoJ, BovUenac and MeketiU, were the map viewed from the conventional south. This type of map also suffers from the fact that, to be optically correct, the columns towards the back of the map to be scaled smaller to take account of perspective. It is not usually made clear whether the cartographer has taken this into account during the cartographic process, or whether the *A4

map user themselves are expected to make adjustments during the viewing process.

*A3

=ali Idos

*A2

%he /a,*e19y1Area !ap, or CartograThe cartogram is an excellent way of conveying complex statistical relationships. The method requires that the data forming the theme of the map be transformed into areal space. >ne unit of the variable therefore equates to one unit of area on the map. =odern cartograms tend not to overlay a topographical base0map. They themselves are both base map and thematic map, the statistical areas being so configured as to resemble 00 however loosely 00 their geographical shape, and ^ or so arranged as to reflect their spatial inter0relationship. This type of cartogram may be referred to as an Xareal transformationX type. Being but incidentally concerned type with geographical can thus verisimilitude, display pure Xareal transformationX cartograms statistical

relationships, without being confined into geographical space.

Dhere absolute values are mapped, the actual resultant areas .statistical districts5 will vary accordingly in si9e, and often in shape, producing novel visual relationships. -uch cartograms are $owever often difficult to construct, and their aesthetic appearance is not always pleasing. Dhere an attempt is made to retain the shape characteristics of the districts, an appreciation of the statistical value they represent will be hard to make, unless the viewer is very familiar with their original form.

*E#

&cono0%artograms

*E*

Balogh

If one is modelling percentage rather than absolute values however, the cartogram can be relatively simple, as all districts may be standardised to a common si9e and shape for *##Y. >ur first example of this methodology comes from the turn of the last century and represents .with some inaccuracy5 the mother tongue of the *33# population of the districts and counties of contemporary $ungary<2. 'll districts are shown as identically si9ed squares, divided into one hundred parts, and nine language0groups .including KothersL5 are differentiated by area0coloration, in units of a quarter of one per cent of the districtLs population. The methodology does not differentiate between the urban and rural populations, but gives a precise indication of the proportions of the various language0groups within the population of each district. The large towns and cities are not shown. ,evertheless, in terms of visual weight, the urban populations benefit because they are shown as a proportion of the total population of the district within which the town or city falls, and are not limited on the map to the area covered by the urban districts to which they belong. $ad the map been at settlement level, with each set of *## squares then representing the population of each settlement, then towns would have been set at the same level as villages. 's a district contained around ten settlements of disparate si9es, the look of such a map would have been quite different.

)elatively uninhabited 9ones are shown as blank KdistrictsL, without population coloration. 's administrative districts covered all parts of $ungary, the whole population will have been shown within the coloured segments of the districts, and these blank mountainous KdistrictsL serve to fill in the map to approximate more closely to the countryLs actual geographical shape. They do also make a political point, however. 'part from these blanks, the map actually says nothing at all about density of population, a subject that was hard fought over in the Transylvanian map0
C$ BGlogh 4Gl, Bu"a9est, #$A2.

*E"

wars. 7olo9sv6ri G6r6s, for example, includes the Transylvanian capital 7olo9sv6r .7lausenburg, %luj0,apoca5. The districtLs population at the time of the *33# census was A4,*A4. ' mere *<,AE3 persons inhabited the neighbouring ,6dasmenti G6r6s, but both districts were shown as the same si9e.

*E<

Balogh %artogram

*E@

'n alternative is the Xdistributional transformationX type cartogram. Nnlike the areal distribution cartogram, this does use an underlying topographical base0map, but retains the direct relationship between data and areal representation on the thematic overlay. It is perhaps closest in spirit to the density0dot map, displaying units of colour to represent units of absolute data, but with the programmatic addition of coalescence in high0density areas. Dhere the colours representing data for multiple areas overlap 00 like two adjacent towns 00 one may be offset slightly so that the full extent of both may be shown.

Teleki

+erhaps the most renowned Xdistributional transformationL type cartogram of $ungary is the mis0titled &thnographic =ap of $ungary on the basis of +opulation Censity .=agyarors96g n;praj9i t;rk;pe a n;psrs;g alapj6n5, by the later $ungarian +rime =inister, (r8f +6l Teleki .*2*25. This may stand as an example of where methodologies, while appearing to be objective and scientific, are chosen for very unscientific reasons, and may indeed be applied with considerable cunning.

The methodology of the Teleki cartogram is mathematically very appealing, and allowed the author to claim that the entire population of $ungary, as recorded by the $ungarian census of *2*#, had been represented on the face of the map. It is, unfortunately quite impossible to verify this as, the quite apart from the arithmetic, the correspondence between the location of the blocks of population represented and that of the blocks of colour which were their chosen symbols is somewhat loose. There are also no internal administrative boundaries on the map. Teleki chose to represent the *2*# census in the following manner each *## persons of a particular mother tongue would be represented by * mm " of a colour accorded specifically to that mother tongue. The * mm " of ink would be placed on the map at the location where the persons were *EA

recorded. Dhere a particular mother tongue is represented by less than *## persons, the appropriate proportion of one mm " of the colour in question will be relocated, leaving a gap, and amalgamated with the nearest block of that colour. There is only one unit of measure of population *mm" *## persons. The scale of the original map being * *,###,###, this equates to an average population density of *## persons per square kilometre, arithmetically beautifully neat. This map made a big impression at the time, and is still seen in $ungary today.

The technique chosen appears to the uninformed to be entirely objective, but this actually far from being the case. 't the time, the average population density of $ungary was considerably below *## persons per square kilometre and the figure for Transylvania was around fifty persons per km". The

*EE

teleki

*E4

Teleki method will require that about half of its area appear blank. In a country where population density throughout the land varies within a narrow range, such an approach would probably yield satisfactory results. 1et in the case of $istorical $ungary, there was a very wide range of differences .ranging between "3.3 persons^km " in %sk v6rmegye .county5 in Transylvania, to @,A<3 persons^km " in Budapest. 'dministrative boundaries offer no barrier to the ebb and flow of \displaced statistics\, and so it becomes impossible to reconstruct, or to criticise, the precise procedure followed. >ften, local settlement patterns had an ethnic aspect. In the county of 7rass80-9:r;ny .Banat5 for example, the population was distributed between no fewer than <4* settlements. 3@Y of these .<*#5 had a )omanian majority, and "4# of these had less than ",### inhabitants. >nly *# settlements had more than <,### inhabitants. The county was indisputably )omanian in character and densely dotted with these small )omanian villages. 1et, because the population density averaged out at @".* persons^km ", the Teleki method requires that these coalesce and the population is \scraped away\ from the mountains upon which many )omanians depended for their livelihood and lumped together in the valleys. The mountains become \uninhabited\. The overwhelmingly =agyar counties of %sk, $6roms9;k and Ndvarhely, on the other hand, show a different pattern of settlement. Dith average population densities of "3.3, <3.* and @".< persons^km " respectively, the three counties of the -9;kelyf:ld were among the most sparsely populated counties of $istorical $ungary. In contrast with 7rass80-9:r;ny with its micro0villages, however, the population in %sk in particular was concentrated in a far smaller number of swollen villages with populations of A,### and more each. This difference in settlement pattern goes completely unnoticed in the Teleki map, however. The few settlements of the upper =aros and >lt valleys are spread out, the visual effect being much the same as the gathering in of the )omanians into the valleys of 7rass80-9:r;ny.

. >#B$A 'ensus, from Balogh 4Gl, 99.BA&D2A). )B =omanian settlements ha" fe/er than )AA inhabitants, #C& less than #,AAA. BV /ere smaller than 2,AAA 9ersons, /hile 22 ha" bet/een 2,AAA an" CAAA inhabitants.

*E3

The case of cities in general, and Budapest above all, is even more striking, albeit its influence on the map as a whole is completely obscured. =uch as the single inflexible index of *## persons ^ km " was unable to accurately represent the diffusion of )omanian hamlets and villages in 7rass80-9:r;ny, or the handful of oversi9ed villages in %sk, its representation of Budapest is at once geographically inaccurate and extremely tendentious. Dhat was effectively a towering mass of square millimetres of coloured dye was spread about the countryside nearby, filling many areas .mainly with $ungarians, of course5 which were in reality sparsely populated. The colour for the actual population of this area has therefore to be shunted outwards, beyond the 3,3#<mm" .population in *2*# 33#,<4* persons5 which is accorded to the population of Budapest. This metropolitan population is \rolled flat\ across the face of the map, giving the completely misleading impression of a densely inhabited $ungarian heartland. The point is that the $ungarians are represented as a compact mass, whereas the )omanians and other minorities are everywhere fragmented among them. The -lovaks around B;k;scsaba are reduced to a number of geometric figures, while the $ungarians around form a compact mass. &ven in the valleys of Transylvania, the method works to the benefit of the $ungarians. The dense urban population, at the time substantially $ungarian and (erman, gets ironed out into the valley floors .first, it would seem, while there is still room[5. >nly then, it seems, are the more fragmented populations 00 usually )omanians and other minorities 00 settled downwards to link up with them. It is very instructive to compare the solid red mass of $ungarians with their portrayal on the 7ogutowic9 map described above. If it is recalled that the strongest colour on 7ogutowic9 is for densities of EA persons per km", and yet covers no more than a third of the main $ungarian area of habitation, it is plain that some jiggery0pokery would have been necessary on Teleki, for the whole of that area to be covered where the index of density was *## persons per km ".

*E2

' development of TelekiLs methodology may be seen on KThe PojvodinaL, a map by Moreign >ffice )esearch Cept, dated 'ugust *2@@, which is held in the Bibrary of the )oyal (eographical -ociety, Bondon. The data are estimates of the nationalities in the 1ugoslav Pojvodina, derived from the language data from the *2"* census and the confessional data from the *2<* census. The basic geographical backdrop is given by the administrative boundaries for *2<3. Dithin these boundaries, small rectangles representing roughly *,### persons of any one of seven national groups are arranged. Censity of population is shown to some extent, as where the number of rectangles of the default si9e is insufficient to fill a district, the remaining space is left blank. Dhere, on the other hand, the population is quite large, such as around 'patin, the si9e of the rectangles is compressed so that they should all fit within the district. It is not clear why so large a default si9e was chosen that compression was made necessary. The cities and major town are shown as sectored circles, except -ubotica, which was identified in the census with the whole of the extensive K(emarkungL. Dhile there is some variation in si9e, there is no direct relationship between circle si9e and population. This may be confirmed by reference to the *2<* population figures, which appear at the bottom left of the map. ,ovi -ad, with a population of E<,23A, has a circle smaller than that for Temun ."3,#4@5. &xcepting this peculiar oversight, the mapLs methodology is quite attractive.

*4#

M>)C

*4*

' factor that has dogged all the maps mentioned above has been population density. This is a matter of some importance wherever it varies widely and is reflective of other characteristics within the population. &ven today, this goes unnoticed. The >rdnance -urvey .'ttica5 Interactive 'tlas of (reat Britain .*2235 perpetuates this error into the digital age. Their map of K&thnic =inoritiesL shows minority residents as a proportion of total population by local 'uthority according to the *22* census of the N7. There was an undercount of approximately one million for this census, mainly from the cities and perhaps most of whom were from the ethnic minorities. These are therefore not represented on the map. The map takes no account of population density, and the significance of the pattern revealed on the >- map would pass by anyone unfamiliar with the settlement pattern in the N7.

The band of densest settlement, passing northward roughly from Bondon to Biverpool and Beeds, coincides closely with the ethnic pattern. The thinly inhabited but extensive rural areas, extending from ,orfolk to the top of -cotland and including the Dest %ountry and much of Dales, are represented with exactly the same weight as the densely inhabited cities. The effect is to sharply diminish the optical weight of the ethnic minorities in Britain. ' different approach is necessary to represent relationships more accurately.

*4"

Ordnanace $*r+ey

*4<

%he $y-9o, -aps

Dhereas we have so far seen the symbol used mainly as a palliative to the ills of the choropleth map, a whole family of thematic maps takes the symbol as the principal numerical indicator of presence. The family consists of two main branches the dot0density map and graduated dot0map, and the pie chart and proportional pie map.

%he ot1 ensity !ap The principal purpose of the dot density map is to display the density of a variable across a given space. The method, when applied conscientiously, involves the placing on the map of dots of a single set value to represent the occurrence of a variable, as close as considerations of scale allow, to its actual location. The technique is generally applied to large areas, rather than points, and to objects which occur in large numbers. Two examples are offered, showing wheat growing and production in &urope for *2*<.

The thematic layer of a dot density map per se is most effective when both the si9e and value of the individual dot is set low, in proportion to the overall map si9e and total data value. This will result in a \bloom\ of dots in the densest areas, while keeping coalescence to a minimum. Beyond the accuracy of the dotsL geographical placement, two factors 0 dot0si9e and dot0value 0 decide the effectiveness of the thematic dot0density overlay. The si9e should be such that the dots neither dominate, nor are lost, on the map. The value of the dots should also be set to harmonise with the dot0si9e. ' further consideration is that a KbloomL of dots representing a minority should not obtain an optical weight exceeding that appropriate to the proportion of the minority within the population as a whole

*4@

Dheat Cot

*4A

Prbas Cots

*4E

'lexandershausen

*44

)onai dot

*43

The function of dot0density maps is to convey patterns and impressions rather than concrete data about specific areas, so what can happen is for the map to be prepared at a large scale and subsequently reduced for publication so that the patterns predominate over the individual points. The manual process is slow and painstaking, but still far superior in result to computer production. The latter works arbitrarily and simply sprinkles the dots randomly within the statistical areas of the map. Nnless these areas are extremely small, the map cannot correspond to the settlement pattern. This may be observed on a sample dot0density map of the Prbas district in the Ba!ka, prepared using =apPiewer using data from the *22* census of 1ugoslavia ' manually prepared map would have reflected the high degree of centralisation of the population in the settlements. =any of the villages founded by the $absburgs in the region had an exceptionally tight and regular street grid, and most retain these to the present. ' flight over the Tis9a valley today presents a landscape view barely changed in a hundred years. This is the dominant pattern in the Ba!ka and western Banat, where many villages have scarcely grown beyond their colonist foundations The Banat has a richer variety of settlement types, each with a distinct pattern. @# Teleki and )8nai .*2<45 demonstrated that great spatial variations in densities can exist even within the bounds of single villages
&A These in'lu"e0 Faufen"orf, Anger"orf !treusie"lung !'h/armsie"lung !tra]en"orf Fang"orf Talsie"lung @eh.ft TGnyasie"lung Weiler"orf Ieilen"orf >s'attere" village) >village built aroun" a village green) >s'attere" settlement) >"ense settlement) >linear village) >hillDvillage) >valley settlement) >farmstea") >farmstea" settlement 9attern) >hamlet) >straggling village)

*42

The mid0way point between the dot0density and the symbol map proper is the hybrid graduated dot0density map. $ere, as in )8naiLs KCistribution of +opulation and -ettlements in the Territory of %entral &uropeL, six symbol si9es are used, albeit only five are quantified in the key. By extracting the largest cities, it was possible to retain the classic bloom look of the dot map. It is, however, a methodological step away from the simple dot map.

!aps o5 Point1Based ata, or Area, aggregated at Points

ata Ass*-ed to 9e

In symbol maps, as opposed to dot0density maps, the story is actually told by the symbols themselves, rather than by the pattern they make, although this distinction may sometimes be fine. The point or symbol method goes back at least to the *3th century, with, for example, a series of decorative icons differentiating (erman and other villages in the =ercy0 map of *4"<, and the later &#emesvar Bannat a"getheilet in seine (istrict und $rocess, -nno /?W/'. 't this early stage, icons representing data of a general nature .e.g. Plach or >ld Believer (erman Pillage5 were drawn by hand on the map according to the actual location of that data. In many cases, the symbol as rendered in the key is only a guide for what to look for on the map. -ome &#eutsche Gatholische Yrter' on the latter map, for instance +eriamosch, Pinga .Theresiopel5, -aderlak, 7aran .=ercydorf5, Beschenova and Mreidorf, are all variants on the icon in the key. There appears to have been an attempt to adjust the si9e of the symbol in line with the localitiesL populations, but this cannot be confirmed. Dith the passage of time, the si9e, shape or colour of the icons began to be varied in direct proportion with the data, but often the chosen si9es could not be accurately represented, particularly where printing equipment was primitive or in cases of hand0drawn maps. The more flamboyant icons were therefore gradually dropped in favour of the circle, square and other

*3#

simple geometric shapes, which were easier to print and whose respective values were also easier to read.

In modern usage, symbols may be used to represent numerical data which either occurs at points, or which is areal, but may be assumed to occur at points. The distinction depends upon the scale of the map. >n a map of the N-', for example, data on cities could be said to exist at points, while at county level, the same data would be thought of as areal. It could nevertheless in both cases be treated as if the data existed at a point.

'mong the first modern symbol maps covering the Banat and Ba!ka were those accompanying Balogh +6ls
@*

book, referred to above. The whole of

contemporary $ungary was covered at a scale of * <##,### by three series of eight giant maps. The first reported on settlements, the second, language areas and the third, confessions, as recorded by the *32# census.

The first map shows every single settlement as reported by that census, represented by a small square. Mollowing the system in his text, each settlement is graduated into one of sixteen si9e brackets, from less than A## inhabitants to more than 2#,### Budapest .4##,###5 has a seventeenth bracket to itself. There are actually only three levels of symbol, however .small community, large community and town5. Beyond this, the si9e of the settlement is indicated by notations entered on the map by the settlement symbol. Binguistic majorities .at least seven plus KotherL are distinguished5 are indicated by patterns within the square, while minorities and their proportions are shown by further notation beside the settlement symbol.

@*' n;pf6jok =agyarors96gon. Budapest, *2#". *3*

Balogh

*3"

The amount of information on these maps is ama9ing, but a number of factors reduce their utility. Mirstly, the scale is such that an overview of the whole country S or even large parts of it S is really not possible. Dith that in mind, one would look to obtain concrete information at a lower level. But the combination of monochrome printing and a detailed orographical underlay with the host of notational signs rather confuses the picture. 'nd at the lowest level, the standardised nature of the data permits one to work out proportions in the population, but not group si9es.

These difficulties can best be illustrated by reference to excerpts from the maps themselves. If we look at >rsova, at the downstream end of the Banat, we find that it is classified as a town with between <#,### and @#,### inhabitants, the majority of whom were (erman speakers. "#0 <#Y of the population belonged to linguistic minorities including )omanians, $ungarians, >thers and -erbs. 's we have seen elsewhere >rsova was an important town. The other settlements are considerably smaller, the largest being Tsupanica .*,###0",###5 and no less than five being villages of under A## inhabitants. >rsova is rather lost. The second excerpt shows a rather more densely inhabited part of the Banat just east of Peliki 7ikinda. Being one of the favoured destinations for colonists, the area is remarkably rich in minorities, for an area barely fifty kilometres from east to west. There are settlements with (erman, $ungarian, )omanian, -erb and %roat majorities, and a fair mix of these, and KothersK as minorities throughout. Dhile Peliki 7ikinda ."#,###0 <#,###5 does stand out, there is no distinguishable difference between Tsombolya .2,###0*#,###5 and the nearby %roatian 7ecsa and 7lari settlements, each of which has less than A## inhabitants. Dith the large amount of notation, labelling this part of the Banat, which is characterised more than most by the pattern of double settlements .-erb and %roat 7lari and %seney, -erb, %roat and =agyar %ernya, %roat and )omanian 7ecsa, as well as 7is and ,agy T8s9eg and G;csa5 has resulted in considerable crowding.

*3<

BaloghLs second map series concerned language areas. Dhere two or more settlement share a common majority language, they were deemed to form a language area, and are drawn within a common boundary. The ()lvid)k regional map is bounded by three massive language areas -erb to the south, $ungarian to the north and )omanian to the east. The settlements on the boundaries of these areas are shown, but all others drop out, clarifying the picture. Islands of other languages splinter these areas. In relatively simple areas, the scheme works well, but, as might be expected, it fails to deal adequately with the 7ikinda area, which is massively splintered. That Balogh had great difficulty in drawing the boundaries of the language areas is demonstrated by the extent of the Bega melleki magyar medence. Dhile genuinely being close together these $ungarian settlements are all small and interspersed with other0 language settlements. Mour settlements with very small $ungarian minorities, vi9. =ollydorf, %s:s9telek, )ogendorf and -9erb Ittebe sit at the heart of the basin but in order to preserve the K$ungarianL basin, all four are included within its boundary. It should really have gone round them, but would then have disintegrated. The three -erb majority settlements of 'racs and =elence .each 4,###02,### inhabitants5 and Basahid .@,###0A,### inhabitants5 should certainly have been a -erb island, rather than being split up as they were. Dhile his first map0series does go some way to illustrating the population density, the language area maps S because only the settlements at the area boundaries are shown S lose this perspective. In Transylvania, the perennial problem of the uninhabited areas is revealed again. The mountains just west of >rsova are some of the highest and most inhospitable in the Banat and, at the time of the *32# census, there were no permanent settlements there. ,evertheless, Balogh allocates the entire area to the )omanian language area, on account of the )omanian shepherds who spent the summers there. Dhile it is clear that such areas should not have been allocated to any other language area, it is a moot point as to whether they should not have been treated completely differently. )omanian cartographers have almost exception followed BaloghLs example and claimed the mountains for their own.

*3@

' simple ethnographic map of the Banat was produced by +rof. 'C 'tanasiu in *2*2. This represented the linguistic majority of every settlement individually identified by the $ungarian census of *2*#. The key states merely that the map was created by reference to Balogh +6l and unidentified Kofficial statisticsL. There are only three si9es for the settlement symbols .which are coloured with the majority language5, small, large and town. The towns are surrounded with a KcollarL whose bands represent the proportion of each language in the population, but though they vary slightly in si9e, the collarsL si9es do not correspond to the townsL populations. The symbols for most of the villages where minorities are known to have lived are accompanied by small quadrangles, often subdivided two or three ways. These are absent for mono0ethnic villages like Bindenfeld and most of those in the Macset valley. They vary enormously in si9e and format throughout the map and while it is a fair assumption that they are connected with the presence of minorities, there is no explanation anywhere as to their meaning.

*3A

'tanasiu

*3E

+au

*34

=a9ere

*33

The +ie %hart

Pirgil +auLs &Harta !i+icZ, .oliticZ Si etnogra!icZ a Banatului' from *2<E gives evidence of some development, but not much. $is map distinguishes six nationalities and KotherL and represents each settlement in the whole of the Banat .)omanian, 1ugoslav, and $ungarian parts5 by a pie0chart, whose sectors are coloured according to the nationality. Dhile the data for the )omanian Banat claims to be from the *2<# )omanian census .results not published until *2<35, there is no reference to the data used for the 1ugoslav or $ungarian parts. ' summary of the results for the whole Banat from the $ungarian census of *2*# may mean that these data have been used, despite the availability of 1ugoslav census data for *2"* and .partially5 *2<*, and $ungarian data for *2"*. There is a derisory effort to reflect the density of population. Three symbol si9es are shown. ' circle of ca. A0mm diameter represents communes with less than *,### inhabitants, one of 40mm for settlement with a population of *,### S ",A##, and a circle of 3.A0mm diameter serves for every settlement above ",A## S including TimiFoara. That this is absurd may be simply demonstrated. The area of a Amm circle is E.*4 mm" . This gives a value of 2HT persons per mm", for the largest settlement .*,### persons5 in that group. Mor the next group, the area .*".* mm"5 and the highest value .",A##5 gives a value of "#3 persons per mm". The area of the largest circle is *4.3 mm", but the highest value it has to represent .TimiFoara5 is 2*,A3#. The scale for the symbol representing TimiFoara is 4,2J4 persons per mm". The result of choosing such similar symbol si9es has been to greatly reduce the impact of the larger settlements. This completely undermines the superficial exactitude of pie charts, as the sectors simply do not correspond across the si9e categories.

It cannot be that +au was ignorant of the techniques being practised elsewhere. The splendid -chwalm maps of the Banat and Ba!ka .see below5 had been published only three years earlier and examples must *32

have seeped into the Banat. There is a particular purpose to this decision. The )omanian settlement pattern favoured small, scattered settlements, such as those in the upper Bega valley near Macset. De have already discussed the overwhelming prevalence of micro0villages in the )omanian dominated county of 7rass80-9:r;ny in the context of the Teleki map. The villages of almost all the other nationalities were generally much larger, very often topping <,### persons. The towns were still dominated by non0)omanians .only "EY of TimiFoaraLs population were of )omanian mother tongue. This map represents a deliberate attempt to exaggerate the )omanian proportion in the BanatLs population.

It may be that +au took his inspiration from the

PHarta etnogra!ica a

#ransilvanieiP by ,. =a9ere .IaFi, *2#25. =a9ere followed Balogh in inadequately distinguishing between the various si9es of settlement, but did make some improvements on his scheme. $is symbols are graduated more visibly, and he uses pie0slices instead of notation, but there are only two symbol si9es. ' simple circle represents a settlement with a population of under *,###/ a circle with a dot in the centre shows a village with a population of *,### S A,###, and a circle with a ring inside represents villages of over A,### inhabitants. Towns are shown by squares that are not much larger than the circles. The excerpts, from ReL*i', show how this system fails to convey the real population density, again to the benefit of the )omanians.

It is remarkable that 'tanasiu, +au and =a9ere number among the very best )omanian productions. =ost of the rest are significantly worse in many respects, and rely on poor methodology, indistinct printing and inaccurately transcribed data.

In *2@#, $ungary and )omania were in bitter dispute over Transylvania. )omanians were keenly aware of impending territorial claims from the *2#

-oviet Nnion, $ungary and Bulgaria and ethnographic cartography was one of their few defensive weapons. =ost )omanian maps of the period were therefore impregnated with their fears. $ungarian academics on the other hand, were justifiably proud of their traditions in ethnographic cartography and poured scorn on their neighboursL publications. Dhile the $ungarians used their technical know0how to enhance the visual impact of the $ungarian population group in Transylvania and the Banat, the )omanians seemed to revel in the opposite approach, choosing outdated of simply bad methodologies to suppress groups other than )omanians. The editors of )eBuik consequently went so far as to proclaim that they had not found Kany development, improvement, or consummation in )omanian maps since =a9ere .*2#25, and that any development had been retrogradeL. @" They pointed out that, in the current climate, ethnographic maps were serving as documents and were being referred to by negotiators .the reference to )ibbentropp and %iano is pointed5 in preference to great academic tomes. They concluded that their main discovery had been that the failings in the )omanian maps they has examined Kwere not isolated cases, not coincidences, not errors, but attempts, systematically applied on all )omanian maps, to deceive the map0userL. Dhether one agrees or not with this conclusion, the contrast between the best $ungarian and the best )omanian work is very striking. ' =agyar llam k:9iga9gat6si t;rk;pe .Budapest, *2#5,

Dhile predating =a9ere, and 'tanasiu and +au by several decades, benefits from a vastly superior methodology. &ach settlement in the $ungarian state is represented by a proportional pie0symbol, coloured according to the language data results of the *2## census. &ach circle is drawn to the same scale, "<.< persons per mm " of colour, thus reflecting density of population equally across the whole state. The key is singularly clear, comprising two circular symbols. The one carries equally si9ed pie chart slices in the colours representing the language groups, the other comprises a regular annotated scale showing the si9es for
&2 =e-ui8, 9. &&

*2*

circles representing particular population si9es and a radial scale showing percentage divisions of a circle. >n the map itself, where pie charts overlap, the larger are placed below the smaller so that none are obscured. Thus, while cities S particularly Budapest 0 do grow exceedingly large, they appear on the Kbottom layerL of all the points, which slightly diminishes their impact. >ne can derive from the map the population si9e of every settlement in contemporary $ungary, the proportions S and the absolute strength 0 of the seven most significant .plus KotherL5 language groups. >ne can also get a good impression of the density of population and the overall linguistic pattern across broad areas .as mentioned above, the large scale inhibits taking this further5. To date, to the authorLs knowledge, no single )omanian map has come close to this level of achievement.

*2"

7o9iga9

*2<

-chwalm

The proportional pie chart was fully developed years before +au produced his ethnographic map of the Banat. 's part of the enormous &Hand Rrter"uch des :ren+2 und -uslandsdeutschtum' project .Breslau *2<<5, $ans -chwalm designed two beautifully crafted maps. $is &das (eutschtum des Banats' and Kdas Ceutschtum der BatschkaL stand out head and shoulders above the crowd. The standard of printing alone makes the maps worthy of attention, clear and crisp, almost above criticism. The maps are monochrome, except for the (ermans .shown in red5, which focuses the mind more centrally on the avowed subject of the map S too many colours can oppress. -chwalm differentiates nine groups plus KothersL in the Ba!ka and thirteen plus KothersL in the Banat. The pie chart segments for all but the (ermans are identified by hatchings, yet all are sufficiently different that identification is easy. The key clearly specifies the sources used, and the basis .mother0tongue or nationality5. Both maps are provided with graticules, and are drawn to the same scale. 'n informative geographical backdrop is also provided, which cunningly does not interfere with the reading of the thematic overlay. The territories of settlements with an absolute (erman majority receive area coloration, in red, but this is not overwhelming, and the shade for a (erman relative majority is quite pale. The density of population is clear. The only substantive problem is that the scale for the pies is not the same on both maps. >n the Banat map, * mm " represents "2 inhabitants, while the Ba!ka map, * mm " represents E4. It is apparent that an objective for the author was to avoid symbol overlap. This was achieved on the Banat map .the only exceptions being three small villages close to Bugosch, and ,eu 'rad and &ngelsbrunn5 by the extraction of the symbols for Temesvar and 'rad to the edge of the map. $ad the same scale been used on the Ba!ka map, the main settlements S Baja, =aria Theresiopel .i.e. -ubotica5, -enta, -ombor ,eusat9 .,ovi -ad5 would have overrun most of the map.

*2@

-chwalm

*2A

The K7leingeldL 'pproach

'n alternative to the pie chart is the so0called K7leingeldL or small0change method, which can be used equally well to illustrate single or multiple variables. This approach utilises a set of symbols of similar form but increasing si9e to represent a sequence of fixed values .e.g. *#, A#, *##, *,### persons5, which are arranged around the centroid of the settlement or area.. =ultiple variables may be displayed by the 7leingeld method simply by according to each variable a unique colour or pattern .hatching5. Thus in mixed settlements, a combination of symbols of various si9es and colours^hatchings gives the absolute value of each population0group. This methodology was applied to the Banat and Ba!ka twice in *2@#, each time to great effect.

The first set of maps was closely related to the $ungarian claims on Transylvania, which were addressed by the -econd Pienna 'ward of <# 'ugust *2@#. They were also published in the map folder accompanying an impressively bound propaganda volume entitled KCrd)ly', which was published by the $ungarian $istorical -ociety on *st 'ugust of that year. >n these maps, some restricted to Transylvania and Destern $ungary .see 'rea of reference5, others showing the whole of historical $ungary, there were two symbol si9es, representing *## and *,### persons. These si9es permit even tiny scattered settlements to be shown, though is not entirely clear what methodology was adopted for the minute shepherding populations in the mountains. The coalescence of larger dots at the chief cities extents over more than the actual urban area on the ground, which leads to a certain exaggeration of their weight. Budapest .on the maps extending that far west5 is treated differently, being represented by a large, but translucent pie chart symbol. In terms of symbol methodology, this map series is admirable. Its chief problems .colour and area of reference5 are discussed elsewhere.

*2E

Mrom *2@#0@* come the three Polkstumskarten by Dilfried 7rallert. The three series cover contemporary $ungary, )omania and 1ugoslavia, each at a scale of * "##,###, and with a comparable thematic overlay, based on the latest available census results for these countries. The 1ugoslavia map differentiates no less than seventeen nationalities by reference to the language and confessional results of the *2<* census. %learly distinguishable symbols differentiate six population si9e units .*#0"A persons, "A0A#, *##, A##, *,## and *#,###5, which are coloured individually according to the nationality of those represented. -ymbols are not subdivided between nationalities each nationality is separately represented by the KcoinageL necessary to represent their number. The symbols gather by nationality around the centroid of each settlement, and there is no overlap. Dhile the largest symbol represents *#,### persons, major cities are represented by a conventional pie0diagram. The absolute value of the largest symbols is printed within the pie. The provenance of the

*24

7rallertLs data is incidentally, a matter of great interest. Dhile on research in (ermany, $ungary and M)1 in *22E, it came to the authorLs attention that a mystery surrounded the records for the 1ugoslav *2<* census. -ome said that all records were destroyed during the BuftwaffeLs bombing of Belgrade in *2@*, but (ermans and $ungarians both published maps based ostensibly on these data.@< 'ccording to the widow of Dilfried 7rallert, who worked closely with him at $u"likationsstelle Wien, however, the records were recovered from Belgrade and shipped to Pienna. >thers told that they disappeared at the end of the war, and Cr Gosef Dolf at the
Instit*t 5Ur dona*sch;V9ische Geschichte *nd Landesk*nde, %U9ingen,

reprted that the entire archive of the $u"likationsstelle Wien was impounded by the British 'rmy, and transported away in the back of three or four lorries. ,o0one seems to know where they ended up. This census is of some significance, as it was the last official census in 1ugoslavia before the dislocations caused by the -econd Dorld Dar. The 1ugoslav government only ever published only partial data, and it has not hitherto been possible to check the claim that 7rallertLs visually excellent KVolkstumskarte von Iugosla ien' is indeed based on official data.

Isorith-ic -aps, 5ro- data a,ong a ,ine

'fter area0 and point0based data, the third principal form for mapping is that which may be said to exist along a line. The most familiar are contour lines on topographic maps, which represent lines of constant height. In ethnographic cartography, such \isolines\ may represent population densities, rates of growth, levels of economic activity etc. ,o examples of isorithmic ethnographic maps on Banat and Ba!ka, or the surrounding areas have yet come to light.

@< *or a Fungarian eXam9le see0 $ viss+at2rt B1cs a 2s Baranya nem+etis2gi 2s &irto politi ai
t2r 2pe% Az ^Lvi"Z;i agrGrDhivatal GtvetelZvelZs a birto;9oliti;ai ;Zr"Zse; tanulmGnyozGsGval megbizott miniszteri tGra;.zi bizottsGg. KLvi"Z;, & (ay #$&#

*23

Art and Ethics in !ap-aking

It is clear that areas such as the Banat and Ba!ka, with great complexity and mobility of population, must pose a serious challenge to the ethnographic cartographer, and serve as a demanding testing0ground for his or her inventiveness, and honesty. It has been demonstrated that a good map comes about not simply as the result of a series of rational decisions, it is composed, like text or music, such that all the component parts work together in harmony to reinforce the desired message. The practice of cartography P333 as a .ro!ession involves more than the mere learning o! a set o! ell2esta"lished rules and conventionsP .Cent *22E *45. Coes this mean that cartography is a science, or an art It clearly belongs to both, as a combination of technical skill and draughtsmanship with aesthetics, and rhetoric and propaganda. Technical skill only is required P hen a ma. uses a scale such that o"jects may "e re.resented in their true .ro.ortionP. .&ckert, *2445. 'rt in cartography begins when the scale and si9e of the map prevents this, but when present, it is rarely art for its own sake. It is the art of the cartographer as propagandist, whether for himself or his company, for his city or his state.

The art in maps is not however, pure sophistry. 'esthetics are essential in the production of maps, aside from the fact that an ugly map will not sell, nor win anyone over to the desired point of view. Dithout some level of symbolisation, no map is possible and, since symbolisation involves representation of one object by another, the mapmaking process involves a degree of aesthetic awareness. ' willingness to make the map look \right\ in terms of layout, shadings, colour, etc. in other words, to seek artistic balance, is also required. PHo ho ell the cartogra.hic designer can orchestrate all the varia"les in the a"straction .rocess is the measure o! much o! an artist he or she isP .Cent *22E *45.

*22

But however well a cartographer passes these tests, he or she should ideally allow the reader the opportunity to see through the process that led to the map. -ometimes, explanations raise more questions than they answer, as was the case with the booklet accompanying an ambitious *2"3 ethnographic map of Transylvania .-tot9 *2"35. 'bove all, the cartographer should beware of Klaying down the lawL about the verity of his or her map. P-ny single ma. is "ut one o! many cartogra.hic vie s o! a varia"le or a set o! data3 Because the statistical ma. is a rhetorical device as ell as an analytical tool, ethics reNuires that a single ma. not o! the data3 ith di!!erent im.ose a dece.tively erroneous or carelessly incom.lete vie Scholars must look care!ully at their data, e<.eriment re.resentation,

eigh "oth the reNuirements o! the analysis and the likely

.erce.tions o! the reader, and consider .resenting com.lementary vie s ith multi.le ma.sP .Cent, p.*25. This criticism can certainly be levelled at the $ungarian editors of K)eBui7L, who systematically labelled the )omanian maps as KwrongL and their own as KcorrectL.

(iven a familiarity with the fundamental methods of ethnographic cartography, and a native aesthetic talent, the key test of the ethnographic cartographer is how well s^he resists the temptation to wilfully employ the available filters to further nationalist or other hidden agendas. -ome of those whose productions have numbered among the very best have failed on just this point.

%he Choice o5 Co,o*r or Hatching1Pattern

&ven when available printing technology does not itself intervene, it can be difficult to avoid skewing the visual impact of a map in favour of one or another variable. >n a map which shows areas or symbols of equal value, but representing different variables .e.g. areas where different nationalities form over A#Y of the population, or \7leingeld\ symbols5, each hatch0pattern should be of equal visual intensity, though in practice, this is very often not the case. Dhere data graduations are indicated by "##

an increasingly0saturated stepped hatch0pattern or colour .stepped0value reporting5, great care must be taken that each step be visually identifiable with the value it represents .Dilliams *2A35. >ccasionally, this rule has been deliberately ignored for political ends, as with the set of maps presented by the )omanian delegation to the +aris peace conferences of *2*20"# .see above5.

' combination of multiple variables and stepped0value reporting only rarely succeeds visually in black and white. &mploying colour makes it possible to differentiate between the basic variables while leaving the full range of hatching0patterns available for other purposes 00 e.g. for portraying density of population. &ven more so than with hatchings, though, great care must be taken with the choice, as colour is a psychologically most powerful communicative tool.

PGolours .ossess su"jective values !or the o"server and arouse in him emotions and irrational res.onse, .roducing a range o! .ossi"le reactions to a ma.3 #he .sychology o! colour is a com.ound o! tone, intensity and degree o! saturation3 Besides the dominance o! red and "lack over all other colours, dark colours are automatically associated light colours ith lo ith high and statistical values3 Golours !urthermore trigger arm colour2

su"jective emotional res.onses, leading to e<.loitation o! the contrast "et een cool and arm colours and the em.loyment o! hich the author
.

scales !or those data to

ishes to give .recedenceP

.Mischer, in $osch -eewann *22*5

's may be demonstrated by a wide range of ethnographic maps covering Banat and Bafka, dating from the nineteenth century to the post0 1ugoslavia present, colour permits of gross distortion without falsification. 'ssociation0by0colour is common. Take for example the habit of $ungarian
. Translation by Bru'e (it'hell.

"#*

cartographers to use a bright red for the =agyars. 'lmost every $ungarian map slavishly follows this tradition S even today. It was also habitual to represent the (ermans by an optically close orange, which in some cases was almost indistinguishable. This suggested a relationship between the two \principal\ nations of the 'ustro0$ungarian &mpire. +articularly disappointing in this regard is the otherwise outstanding 7ogutowic9, *2*2 .see above5. >ther variables are usually indicated by less conspicuous colours. The most notorious example of abuse of colour is probably the pair of \7leingeld\ method maps published in *2@# to accompany the $ungarian claim on Transylvania in *2@# .B6tky, 7ogutowic9, Teleki, *2@# see above5. Dhile an intense scarlet .=agyars5 and bright orange .(ermans5 stand out from the mountainous grey backdrop, the mauve chosen for the )omanians almost disappears into it .note this is particularly striking on the originals of the map under electric lighting, the conditions under which they were designed to be used5. The visual impact of the $ungarian population0group is strengthened by the display of the substantial $ungarian population of eastern $ungary. )omanian and (erman cartographers have also regularly used red to represent their respective nationalities. 7rallert and -chwalm, for example, while producing some of the most effective work on the region, employ scarlet for the (ermans, in -chwalmXs case as the only colour 00 other nationalities being depicted by hatchings. 's a token of the intensely political nature of 7rallertXs maps, the colour denoting Gews is black, visually second in impact only to red. ' recent pair of -erbian maps .-tepiU IliU, Nniversity of Belgrade, and 7icoJev, Nniversity of ,ovi -ad *22<5 also employs a very tendentious colour scheme. The -erbs and =ontenegrins are both indicated by very dark blues, and the \1ugoslavs\, who were to a great degree not -erbs, are also allocated a strong blue. The %roats stand out very conspicuously, in yellow. It was jokingly suggested to me in ,ovi -ad that these maps had been very helpful to etniks in their ethnic cleansing of the Pojvodina.

"#"

REGI$%ER O" PLACE NA!E$

Recorded Ro-anian, D*gos,a+ 8W $er9ian@, H*ngarian and Ger-an Na-e1"or-s, 2?4I 1 233T(
Incorporating Ad-inistrati+e A,,ocations and Latit*de1Longit*de Co1 ordinates o5 $ett,e-ent Centroids

"#<

The register of place0names that follows includes every settlement which has been recorded as a separate entity in any of the censuses mentioned in the >fficial %ensuses section of the -ources for Toponyms.

The main body of the register comprises three parts, corresponding to the )omanian, the 1ugoslav and the much smaller $ungarian territories. The first column gives the official name of the settlement in the most recent census .$ungarian, *22#/ 1ugoslav, *22*/ )omanian, *22"5. The next columns give the +rimary .+'5 and -econdary .-'5 administrative division to which the settlement belonged at the time of the census. -ubsequent columns show older names in the language of the state and alternatives recorded in other regionally significant languages. The administrative allocation of the settlement in selected other official censuses, going back as far as the last pre0Trianon $ungarian census of *2*#, is also given in the form of a two0 or three0digit code, for which the key is given towards the back. The final columns give the co0ordinates of the settlementLs centroid in decimal degrees of latitude0longitude based on (reenwich, and the K+rimary ICL, which serves in the main body of the forthcoming work as the unique KtagL for the settlement.

%he Bo*ndaries o5 the Research Area

The place name register is confined to the Banat and Ba!ka. The fragments of B6cs0Bodrog and Toront6l v6rmegye remaining with $ungary after *2*2 are included, and the districts of the northern Banat allocated under )omanian rule to GudeOul 'rad continue to be considered as part of the Banat. 'll settlements entered separately in the censuses of *33# 0 *22", and only those settlements which were within the historical counties of B6cs0 Bodrog, Temes, Toront6l and 7rass80-9:r;ny as defined in *2*#, or have since been founded on that territory are recorded. $orgoJ and its small neighbour, Ba!ki Pinogradi .7ir6lyhalom, B6css9:ll:s5, while in the 1ugoslav Ba!ka, are excluded, as being until *2*2 part of %songr6d county. >v!a and "#@

Bor!a, on the other hand, though included within Toront6l, have since been absorbed into the Belgrade area and have also been excluded.

The tongue of land between the Canube and -ava rivers, known variously as -rem, -rijem, -9erems;g, -irmium, -yrmien, which became part of the Pojvodina after the -econd Dorld Dar, is not dealt with in this study. Dhere 1ugoslav Ba!ka districts have been extended southward across the Crava to include &astern -lavonian settlements, the data for these settlements has been extracted. -imilarly, the KBaranya TriangleL 0 the southernmost tip of historical Baranya v6rmegye, wedged between the Canube and the Crava rivers 0 is not covered here.

Cartographic Reso*rces

The toponymic data for the *#A# settlement0units covered by this investigation has been culled from a variety of sources to ensure the highest attainable degree of accuracy. "< state censuses carried out between *33# and *22" were the prime sources @@, but these were laboriously cross0 checked and supplemented with data from a variety of large0scale maps, place0name0registers and other sources. 'll of these settlements @A have been individually identified and their latitude0longitude co0ordinates recorded by reference to several series of large0scale maps, the key ones dating from *2#", *2@#0*, *2EE and *234. It was initially difficult to locate maps of Banat and Ba!ka with an adequate level of detail. =ost recent maps naturally show either a part of the territory, as is the case with maps of particular countries, or the whole territory without any great degree of detail, as in maps of %entral &urope or the Balkans. $eld in the =ap Bibrary of Nniversity %ollege
&& Fungarian7 #BBA7 #B$A7 #$AA7 #$#A7 #$CA7 #$&#7 #$BA7 #$$A >B) =omanian, #$2A >base" largely u9on offi'ial estimates)7 #$CA7 #$&#7 #$&B7 #$VV7 #$%%7 #$$2 >%) Yugoslavia, #$2#7 #$C#7 #$&B7 #$)C, #$V#, #$%#, #$B#7 #$$# >B)

&) With the eX'e9tion of the mi'roDsettlements aroun" 2ornereva an" Bog_ltin, see belo/, an" footnote &$.

"#A

Bondon, and within the =ap Bibrary of the British Bibrary, however, are copies of Dilfried 7rallertLs Volkstumskarte von Iugosla ien .@# sheets5, Volkstumskarte von RumLnien .@@ sheets5, and Volkstumskarte von [ngarn ."< sheets5 .+ublikationsstelle Dien, *2@#0*5. They were derived from the *2@# (eneralkarte von =itteleuropa * "##,###, which was in its turn developed from the last general survey of the whole $absburg empire, the massive 4rancisco2Iose.hinische -u!name survey .* "A,###/ * 4A,###/ * "##,###5, which took from *34< to *334 to complete. In the 7rallert series, each map0sheet corresponds to an area of * degree each of longitude and latitude. The area of the present project 0 $ungarian, 1ugoslav and )omanian parts 0 is covered by twelve adjacent mapsheets enclosing the quadrangle @E<#L, <E<#L& to @@<#L, @#<#L&, as measured from the Merro =eridian.

The inclusion of a Kminute0to0minuteL scale along all four edges of the 7rallert maps permitted the present author to use them as a basis for his (I-. 'n integrated, single0minute grid was therefore drawn across copies of all twelve individual mapsheets, and the location of each settlementXs centroid catalogued with some considerable accuracy. The three 7rallert series are in fact primarily ethnographic maps, and employ K7leingeldL dots of various si9es and colours to represent the distribution of the various nationalities throughout the territories. The method employed is very effective, and furthermore conveys an accurate impression of density as well as location of population. I have not used the ethnographic data on these maps .whose provenance for 1ugoslavia is uncertain5, merely the underlying geographical data, which pre0dates the preparation of the Volkstumskarten. The dates of the three series .)omania, *2@#/ $ungary and 1ugoslavia, *2@*5 and the colour scheme .for example, (ermans, bright red, Gews, black5 impress upon one their intensely political nature. Cr 7rallert, I have since discovered, was actually not only the director of the $u"likationsstelle Wien, but also the .perhaps nominal5 head of a department in the Berlin $eadquarters of the (estapo.

"#E

&ach mapsheet nonetheless contains a superfluity of detail 0 topographical, political and ethnographic. The courses of the Canube, Tisa and =aros rivers, with their various tributaries, are elaborated by their endless meanders and ox0bow lakes, and the area of the Ba!ka and Banat are strewn with such details as +us9tas and T6nyas .farmsteads5, each accompanied by its gaggle of g;meskjt .lit. KheronL, reminiscent of the KshadufL5 wells. &ven the street layout of the over *,#A# towns and villages of the area under investigation is clearly discernible. Mor the purposes of this study, much of this is of incidental interest only, and indeed, obscures an easier reading of the maps.

1et the 7rallert map still represented a key geographical and toponymical source in the earlier stages of this project. ,ot only is every single settlement shown separately and in its appropriate location, but, where multiple names exist, or where a settlement has been absorbed by its larger neighbour .as in the case of MraOelia .by TimiFoara55, these are shown along with the official, name. The original toponymy from the ex0$absburg base map is retained in a subdued tone beside the contemporary 1ugoslav, or )omanian usage. (erman usage, where this differs from the foregoing, is given in bold, and was presumably added during the preparation of the Polkstumskarten..

Mrom this source it was possible to identify the location almost every single settlement mentioned by the various censuses analysed so far @E. The only settlements that could not be identified by this means were those that came into being or first became independent during the -econd Dorld Dar .e.g. -9;kely and %s6ng8 settlements *2@*0@@5/ new creations since the war, principally the centres of considerable post0war -lavic colonisation of Pojvodina/ two sites@4 mentioned for the *2@# )umgniendeutsche KBestandsau!nahmeL and four villages noted by $errschaft as having received small numbers of (erman colonists. @3 These latter six sites have
&V (ore than #,A)A. &% Po"i`e7 /ith #) @ermans an" (ailat >se'on" of 2 settlements of this name >/ith B)). They are only mentione" in the arti'le in + ation un" !taat67 an" U 'an fin" no other evi"en'e of their eXisten'e &B Temes'hburgD8al;.fen >#%VC)7 Temes. !eu'hen"orff >#%%2)7 !tarosso/a >#%VV)7 Wilhelminenfel"7 >#B)#)7 all TorontGl.

"#4

been discounted as being unsubstantiated. Cetailed )omanian maps were difficult to obtain, even within )omania, but a reasonable selection has been assembled. ' mid0*23#Xs )omanian road atlas @2 .which had apparently served as the base0map for %artogr6phiaXs &rd;lyi t;rk;p ;s helys;gn;vt6r A#5 permitted the identification and integration into the data of settlements that have acquired a separate identity since the war. The extraordinary rash of <2 minute settlements 0 the total population of the largest .+ogara de -us5 is recorded in the *22" census as """ persons, while that of the smallest .Cobraia5 was "2 0 around %ornereva and BogWltin .north of BVile $erculane5 has, however, been amalgamated into the larger neighbours A*.

)eference to two series of excellent 1ugoslav maps kindly supplied by the N7 =inistry of Cefence kindly supplied by the N7 =inistry of CefenceLs =ilitary -urvey has permitted the reduction of the number of unidentified settlements in 1ugoslavia to 9ero A". Both 1ugoslav series are at a scale of * "##,###, which makes comparison with the 7rallert maps simple. The three series contain three different meridians, however. The 7rallert series is based on Merro .=adeira5, the *2E#s 1ugoslav maps were drawn to the +aris and (reenwich =eridians, and the *234 maps were laid out according to (reenwich only. I have allocated each settlement digital0degree co0ordinates and a location code according to the (reenwich meridian only.

Pasile, @eneral maLor ing. ,ragomir7Pi'tor, 2ol. Ung. Baea7 @heorghe, 2ol. Ung. (ureanu7 @heorghe, 19uran0 =omania, Atlas =utier. >9erha9s #02)A,AAA) 1"itura !9ortD Turism, /ithout 9la'e, #$BQ.
&$

)A

#0)AA.AAA. 2artogrG9hia0 Bu"a9est, #$$#.

)# Batrin, Bratova, 2ozla, ,ren'ova, (a'ovite, (arila, 4_rneaura, 4lo9u, 4oneas'a, !atu, !ub (argine, Palea (inisului, 2C August >/hi'h last, un'ertain lo'ation). ?f these, ?bi`a, 2ozia, 4rislo9, 2amena, Imogotin, Ianogi, !trugas'a, Fora (are, BoLia, Ibegu, 4ogara, Uzvor, To9la, Boguri, ,obraia, 2osti, 4risa'ina, 2ra'u (are, Unele`, an" !'arioara a99ear on the ma9s. )2 >#) To9ografs;a ;arta #02AA,AAA, ,elimiRno "o9unLeno #$)B. @o". Uns'rtano to9ografs;im zna'ima iz #$C$.go", Uz"anLa PoLnogeografs;og Unstituta, Belgra"e in #$VC, #$VV. >2) To9ografs;a ;arta #02AA,AAA, ,rugo iz"anLe stam9ano #$B%. @o", !a"baL 9rema stanLu iz #$B).go", PoLnogeografs;og Unstituta, Belgra"e, #$B%

"#3

'bove and beyond the question of the location of the settlements themselves, the boundaries of the various districts .opJtina ^ s9re9 ^ commune ^ j6r6s5 and, in some instances, of the settlements of which they comprise, have been subject to enormous change. This being the case, the author has been frustrated by the considerable difficulty of finding detailed maps showing these boundaries, particularly in )omania 0 where the majority of the settlements covered by this investigation lie A<. Thanks principally to Cr -ebdk B6s9l8 at the #eleki Tas+l, -la.\tvny, =Rnyvtr )s (okumentci,s S+olglat, Budapest/ Cr 7ocsis 76roly at the 5#4Rldraj+tudomnyi =utat, @nt)+et Budapest/ 'sist. univ. Crd. )emus %reOan of the Cepartment of (eography, Nniversity of TimiFoara, )omania, and Cr -aJa 7icoJev at the Institute of (eography of the Nniversity of ,ovi -ad, I have now acquired probably the most comprehensive collection of maps covering the Banat and Ba!ka in the Nnited 7ingdom, ranging from the early *3th century to the present day. Nnfortunately, the recent maps showing administrative boundaries for 1ugoslavia and the )omanian Banat are mainly rather poor, hand0drawn productions, and where there are multiple issues, these seldom agree fully. I have therefore had to compromise, and accept that the best I can hope for are approximations of where the boundaries run.

' guide to the complex merging, division and dissolution of administrative sub0units, which has characteri9ed the regionLs modern history, is in preparation. This is as yet incomplete, but a comprehensive and completely accurate guide to all the to0ing and fro0ing, as is intended, will require many months of research in the field, further close collaboration with colleagues in $ungary, )omania and 1ugoslavia, and access to very large0scale administrative maps of the regions. A@ The digital0map series has also yet be completed, as 0 due to the difficulty in obtaining proper maps of the region 0 the line of the exact administrative boundaries are as yet available for only a few of the years in question. The question of surface0areas of the individual
)C The most 9roblemati' areas are the small settlements south of !uboti'a, an" the =omanian 'ommunes of 2ornereva an" !i'hevi`a >2araD!everin), Bethausen an" (anatiur >Timi). )& The Yugoslav "ata is at 9resent essentially a translation of the + otes6 se'tion of ,r -GszH !ebc;6s 1tnogrGf C.A version "emo. The =omanian se'tion 'omes from my o/n observations an" from footnotes to the various offi'ial 'ensus 9ubli'ations.

"#2

units thereby remains unresolved, hindering for the present any meaningful commentary on the density of population.

)etaining the minute0based grid provided by the 7rallert maps, rather than Kgrossing0upL to a five0 or ten0minute grid, permits of every settlement being accorded a unique numerical code, specific to their individual grid0reference. This has supported differentiation between the large number of tiny, proximate settlements, some of which, over the period of our study, came to merge into larger wholes. The three clearest examples, the former Borrainer villages of %harleville, -eultour and -t. $ubert, near 7ikinda in the 1ugoslav Banat, were in extreme proximity to each other, and eventually were merged to form Banatsko Peliko -elo. Caruva 0 _tefaneFti offer another illustration. In cases such as this, one of the original settlements is taken as being the core of the new conglomerate settlement, and data subsequent to the amalgamation is Kpinned toL that settlement. In the present register, the location of settlements is indicated by their co0ordinates in decimal degrees 0 based on (reenwich 0 to assist the user who might wish to locate any particular settlement on a map before them.

"*#

daruvar

"**

Though every settlement is identified by a centroid accurate to one minute of longitude^latitude, this does not mean that their exact geographical location has been plotted with equivalent accuracy. ,or does it mean that, in the digital maps being prepared in parallel with this package, distance between points, or areas, are accurately portrayed. This stems from the age0 old problem of portraying on a flat surface .map5 a reality which is spheroid. &Tatitude2longitude is a geogra.hic re!erence system, not a t o2dimensional F.lanar) co2ordinate system333 5eridians converge at the .oles "ut se.arate or diverge as they get closer to the eNuator3 So the length o! one degree o! longitude varies de.ending u.on the latitude at e<am.le, one degree o! hich it's measured3 4or longitude at the eNuator is /// kilometres FW0 ith a standard length, they

miles) in length, "ut the length o! one degree o! longitude converges to +ero at the .oles3 Because degrees aren't associated can't "e used as an accurate measure o! distance or area' .Nnderstanding (I- the 'rc0Info =ethod. E$RI, *22@5. It is therefore, not possible directly to obtain Kreal0world co0ordinatesL using latitude0and0longitude, although these may be derived by mathematical formulae.

The British =useum =ap Bibrary also provided a copy of the &Garte ethnogra.hiNue du Banat de #emesvar' by +rof. ' C. 'tanasiu .+aris *2*25. This, drawn at * <##,###, shows every single settlement recorded in the *2*# $ungarian census, named according to $ungarian practice K on a conserv) la nomenclature o!!icielle, im.os)e .ar les 5aghiars', and clearly showing the contemporary administrative divisions into Toront6l, Temes and 7rass80-9:r;ny counties and their subordinate districts .j6r6sok5. -pelling on this map was very unreliable .=arlonos for =artonos, bj +;es for bj +;cs, Tsurrova for Tsur9sova, )u6na for )udna5 and had to be carefully compared with the *2*# census. ' smaller map from the same valuable source 0 &(ensit) de la .o.ulation Ser"o2Groate dans le Banat' .+aris, undated, probably *2*25 0 errs on the side of -lavonic spelling within what was to become the 1ugoslav Banat and includes some within )omania0to0be .TemiJvar, ombolja, 7e!e, Tebelj, akova, etc.5.

"*"

Mor (erman names, the most valuable maps, besides 7rallertLs, were the work of $ans -chwalm. $is two maps &das (eutschtum des Banats' and &das (eutschtum der Batschka' supplied a wealth of (erman toponyms, almost all of which could be confirmed from other sources. These were supplemented by reference to the Hand"uch der Woj odscha!t Ser"ien und des #emeser Banats sammt der 5ilitLrgrLn+e !Ur das Iahr /8H1 and the published results of the (erman K BestandsnahmeL of *2@#, which in all probability contains some of dubious validity. These sources far from exhaust the variety of names under which these individual settlements have been recorded in the past century alone. They far rather press upon one the extent of the problem. )eference has therefore further been made to further place0name registers, ranging from an official $ungarian state &helys)gn)vle<ikon', to a donausch L"isches Artsnamen"uch published to assist Conauschwaben in claims for compensation, to such unprepossessing sources as an appendix to a &WRrter"uch der donausch L"ischen Bekleidungsge er"e' . The &(icVionar de localitZVi din #ransilvania' HH, provides a trilingual .)omanian, $ungarian, (erman5 guide to place0names in Kextended TransylvaniaL, the name of each community being accompanied by that of the local administrative centre and the county. There is no specific confirmation of the year to which this administrative divisions relate. >ccasionally, reference has had to be made to more detailed works on individual settlements, as was the case with the TimiFoara conurbation, (iarmata, and Carova^-tefaneFt. The associated data on administrative allocation has come directly from the relevant censuses. Dhere a settlement does not appear separately in a particular census, and yet is still accorded an -' code, this corresponds to the administrative sub0 unit in whose territory the settlement was located.

Brie5 Re+ie; o5 P,ace Na-es The language of the labelling on a map clearly has a major influence on its appeal, and decisions here are clearly dependent on the market for which the map is being prepared. The purpose of any map is primarily to communicate, and therein lies its worth. In purely commercial contemporary terms, a geographical map of Thailand on which all the place0names were
)) 8riterion7 2luL, #$$2.

"*<

rendered only in Thai characters would obviously, among ordinary &nglish0 speaking communities, have curiosity value but would not communicate meaningfully, and would therefore probably not sell except to certain collectors. To appeal more widely, and to communicate some phonetic value, one would have to publish the map with place0names rendered in the Batin alphabet. +ublishing both the Thai form and Batin transcription would be more informative yet, as the more adventurous map user could derive some basic understanding of the Thai alphabet from comparison of the two. &ach solution would appeal, in different ways, to different user0groups, and therefore to differing markets. &ssentially then, one has to have due regard to oneLs market.

%ommunication, however, is not limited to simple transmission of sounds. Dhile there is no space here to pursue this thread far, one may say that even within the context of a printed map, communication proceeds on a number of levels and through a range of media. The map itself is a symbolic representation of a generalised, selected, view of reality. The communicative power of this view is technically circumscribed by considerations of scale, available technologies and even paper si9es. Beyond this, the map0author will introduce other constraints, including the choice of frames, symbols, colours and fonts S and place0names.

The use of KethnicisedL place0names can give an impression of cultural connectivity with the chosen community or nation. In an ethnically mixed area or where a territorial dispute exists between neighbouring ethnies, electing to use only mono0ethnic toponyms where multiple forms exist can therefore be a conscious political act. The impression can thus be conveyed that the population of the region is mono0ethnic 0 or at least the presence of historical ethnic minorities that have their own ranges of toponyms may be thereby concealed. The choice of language.s5 for the labelling on a map is therefore a matter of some importance, and the cartographer is likely to be fully aware of this at the time of design.

"*@

It is entirely normal for place0name to have different renditions in different languages. This stems in part from differences in scripts and alphabets. There are divergent pronunciations of some characters and other characters specific to each of even such closely related languages as -panish, Italian and Mrench. Dithin the Batin alphabet group, the range of different phonetic practices diverges markedly beyond the core )omance trio, but even in &urope, Batin is only one of four main alphabets .with (reek, %yrillic and 'rabic5, each with their own peculiarities and variations. ' second reason for divergent orthography is that certain sounds do not pass easily between languages, causing adaptation of the native form in the KtargetL language. There is a tendency to adapt sounds to make them sound KrightL in the ears of the speaker. Thus Bivorno became Beghorn in &nglish, and +aris became +arigi in Italian. In the Banat, $ungarian K(yarmataL became remarkably adapted into (erman as Gahrmarkt .see below5. It is to be noted however, that all versions in all languages are prone to evolution and that there is no fixed point of reference from which to gauge divergence.

,evertheless, many place names have been given with a meaning in mind, and this meaning is often carried across into other languages by translation. Bela %rkva, within 1ugoslavia, is known in )omanian as Biserica 'lba, in $ungarian as Meh;rtemplom and in (erman as DeiHkirchen. 'll are exact equivalents S KDhite %hurchL. The (erman rendition to the traditional $ungarian %oronation %ity S -9;kesfeh;rv6r S is -tuhlweissenburg, again a direct translation. -ometimes the translation goes awry, as in the case of 'rab historians writing of the 4th century =uslim conquest of &gypt. ' strategically vital By9antine fortress at the head of the ,ile delta was named Babylon. In 'rabic, "a" means gate, so the name of the fortress was recorded as Bab al 1ona. Bater 'rabs, Kreading that Bab al 1ona had been assaulted, concluded that the gate of 1ona had been captured, and hence that the name of the fortress was 1ona .(lubb, *2E<, ""45. In the *2th century Ba!ka and Banat also, the processes of adaptation can create an entirely new and inappropriate meaning, as with TovariJevo .-rb.5. In this case, magyarisation, B6cst8v6ros, has created a pu99le, as the Klake0townL is nowhere near a lake. "*A

=any toponyms familiar to &nglish speakers recall specific personages .Dellington .,T55, events connected with their past .&aster Island S discovery5, or the roots of those who named the location .e.g. Boston .=ass.55. 'mong the heritage of overseas &nglish place names there is a rich vein of especially imperial toponyms whose very purpose was to reflect and engender pride. The Pictoria Malls paired the greatest cascade in 'frica with the ]ueen of &ngland, while Pirginia had been named for &li9abeth I. The &nglish are far from being the only people to practice this form of self0 aggrandisement it has a pedigree at least as old as the conquests of 'lexander the (reat. ' change of government has often been accompanied by a wholesale change of names as the new regime tries to legitimise its presence, honour prominent individuals, and impose itself on peoplesX consciousness. The process is common world0wide and still current .-aigon 0 $o %hi =inh %ity, Burma 0 =yanmar, and -t. +etersburg 0 Beningrad 0 -t +etersburg5.

In the Balkans, there is a similar heritage, due in part to the ebb and flow of the various states and peoples across the peninsular but the non0contiguity of political and ethnic boundaries and the instability of the former mark the region out for special attention. Dith every change of regime, a new wave of name changes was introduced countless towns and villages, streets and squares etc. have been affected since the mid0*2th century.

The reader will already be aware of the extreme ethnic complexity of the Banat and Ba!ka and of the fact that, for example, TimiFoara, the chief city of the )omanian Banat, is known to the $ungarians as Temesv6r, and to the (ermans as Temeswar, or Temeschburg. Dhat may be less well known is that this multiplicity of names for the capital is no one0off, but rather a reflection of a region0wide pattern of toponymy. The name of some settlements, like 7ikinda, assume almost identical forms in all the languages here being considered .)omanian, -erb0%roat, $ungarian, (erman5. >ther, on the other hand have several differing forms, and even completely "*E

different names, even within the same language. The most extreme example of the latter is offered by +landiJte .otherwise known as .-erb0%roat5 Peliko +landiJte Banatsko +landiJte, Ti!ifalva, Ti!i -elo, =ariolana/ Tichyfalva, =orminta, =orminty;, =ariolana .$ungarian5/ Tichydorf, Tichidorf .(erman5, Tichi .)omanian5.

It is clear that the names of some settlements may be explained by reference to some figure connected with their foundation, such as for example, =ercydorf. This village was founded in *4<@, on the site of a .raedium, 7aran .>rthgograpy from =ercy07arte5, and named after the celebrated %ount =ercy, whose &"anatische Cinrichtungs.rojekt L repopulated and restored the Banat after the devastating Turkish wars. The village received a few do9en families from his native Italy. 1et with many other toponyms .for instance 7olo9sv6r ^ %luj ^ 7lausenburg in Transylvania5, the relationship with the eponym is far less clear, and but rarely signifies that an entirely new settlement was founded at the time of the naming. >ften, names of existing settlements are changed for the benefit of contemporary figures, as in the cases of =aria0Theresiopel ^ -ubotica ^ -9abadka, and Gosefsdorf ^ abalj ^ Tsab6ly, and refer more to the contemporary political scene than to the history and contemporary ethnic composition of the settlement. -ome others 0 such as =arosnagyv:lgy ^ Palea0=are 0 refer to simple topographical fact, or to some blessing of nature .e.g. (uttenbrunn, $idegkjt, .TVbrani55. >thers again derive their names from a local crop, such as Gabuka, Toront6lalm6s, 'pfeldorf, or from the local flora 0 e.g. Buchberg, BRkkhegy, -intar. TimiFoara^Temesv6r^Temeschwar, on the other hand, is one of a number of settlements whose name refers to the location of a particular structure 0 in this case an island fortress in a river .Temischel5 that no longer exists 0 without explicitly claiming association with a particular historical figure, or ethnic group, and yet whose origin may be deduced from archaeological and documentary evidence regarding the founding of the structure. Dhere this structure is a major church, its foundation may be seen as evidence of an already firmly established civilian life, and claims that the settlement owed its origin to the ethnic group which gave rise to the structure may be justified. In the case of castles or monasteries, however, "*4

both suggest the existence of a dichotomy between nobles or evangelising monks and the surrounding population, and such claims may therefore be treated with more caution.

The $absburg colonisation of the Ba!ka and Banat during the late seventeenth and eighteenth century, however, gave rise not only to =ercyfalva, but also to a whole string of other settlements. In some cases, as for example with ,eu0-iwat9, a completely new (erman village was built, abutting onto an older settlement with an indigenous population 0 in this case the -erb settlement of -ivac, which henceforth became known to the (ermans as 'lt0-iwat9. $ans $errschaftLs controversial *2@# book &(as Banat, ein deutsches Siedlungsraum' .first edition5 includes a schedule of all the communities where (erman colonists were settled between *4*4 and *34#. In this general list, new foundations are not differentiated from expansions. ' second list focuses on the period *4EA 0 *44" .9weite -chwaben9ug5 and does make this distinction, and furthermore relates the number of houses built in the various settlements. ' shortcoming however, is $errschaftLs inclusion of non0(ermans within the ranks of the (erman colonists, like the Italians of =ercydorf and the 'lsatians and Borrainers of -oltour, %harleville and -t $ubert, whose descendants only later became (ermanised.

In the majority of cases, however, the date of e< nihilis foundation is not known, and it is futile to try to claim a particular settlement for a particular nationality. It was popular among (erman specialists of a certain persuasion to claim that the Banat was utterly desolate and practically deserted when the $absburg colonisation began, thereby magnifying the undoubted achievement of these colonists. )eference to contemporary cartography can help to clear up this question. -everal important eighteenth0century maps differentiate between inhabited and deserted villages .KpraediaL5. Mrom these maps, it is clear that the Banat was no terra deserta3 -9entkl6ray speaks of "##0plus inhabited places at the time of the conquest, but it is not clear what is meant by inhabited. If one might offer an example from the *22" "*3

)omanian census, Bindenfeld still appears as a separate entity, despite having a registered population of only one.

Though many of the (erman villagesX toponyms may be satisfactorily explained, most place names are extremely ancient, and the meaning of their names in any language correspondingly obscure. &xcept in cases of completely new namings .-atu =ic Pictor Plad Celamarina, 7aran =ercydorf5 or of translations .,ovi -ad 0 Njvid;k 0 ,eusat9/ 7ruJevlje 0 7:rt;s 0 Birndorf5, most name0forms seem to be attempts to confine an existing sound0pattern within the orthography of a language to which it might be entirely foreign. The alternative $ungarian and (erman name0 forms for 7ruJevlje 0 7rusevlya, and 7ruschewel ^ 7ruschewelje 0 bear this out. 'nd yet it would be a mistake to imagine that (ermans, $ungarians, )omanians and -erbs are the only actors who have been active on this particular stage. 7unbaja, 7uns9lls .%umans5/ Zbesseny .+echenegs5/ T:r:ktopolya, T:r:ks96kos .Turks5/ all these bear witness to other layers of population who, during their longer or shorter stays, have contributed to the enrichment, transformation and transmission of the regionLs toponymic heritage.

It seems that sometimes, a name that is easily understood is an indication, not that this is the core form, from which the other names have developed, but rather that this is the name most recently adapted to orthographical norms. +erhaps the most illuminating example is offered by (iarmata, north east of TimiFoara. The (erman name .Gahrmarkt5 seems to point unequivocally to the existence of an annual market, and the )omanian and and $ungarian name0forms .(iarmata and (yarmata5 appear to be attempts to adjust the sound of the (erman name to their orthographies. >ne might easily be seduced into concluding that this village was founded by (ermans. ,ot so[ The first documentary evidence of the settlement stems from *<<@, under the names Tamar, (ormat, (armad, (armat. This connects the settlement firmly with one of the original $ungarian tribes involved in the $onfogl6l6s, or conquest. In the mediaeval period, the name shifted into "*2

(yarmath, a play with a separate word 0 (yarmat 0 KcolonyL, or KsettlementL. The name KGahrmarktL appears for the first time during while the Banat was administered as an imperial province, separated from $ungary .*4*30*4425

=any of the regionLs other (erman place0names .some original creations, others (ermanisations of existing names5 originated during this period of intense colonisation. 'ttachment to $ungary .*4420*3@25 broadened the base of $ungarian naming, which, after the failed rising of *3@302, was interrupted by renewed (erman and -erbian influence during the lifetime of the -erbian Pojvodina and Banat of TemesL .*3@20*3E#5. Thereafter, magyarisation 0 in many cases, re0magyarisation 0 was resumed and continued until the Mirst Dorld Dar. Curing the same period, the $ungarian government also very strongly encouraged their citi9ens to conform their names to the $ungarian language. By *2*#, almost every settlement in Banat and Ba!ka had a recognisably $ungarian name0form .Illustr. *<5 albeit this was often merely an attempt to confine an existing sound0pattern within the orthography of a language to which it might be entirely foreign. 't least one community was ultimately successful in resisting magyarisation. In *2**, the inhabitants of Biebling .&ngl. meaning KCarlingL5, about <#km south of Temesv6r, successfully appealed to 7aiser Mran9 Gosef II after the decision of the $ungarian =inister of the Interior to change the communityXs name to 7edvence .same meaning5 .Cocuments reproduced in Blum, *2<E, pp.*#@0E5.

Trianon brought a new wave of name changing, as )omania and the 7ingdom of -erbs, %roats and -lovenes introduced their own brands of official names. Nsually, these differed from foregoing versions in little more than orthography .7ani9sa 7anjiQa, 7aransebes %aransebeF5. In one case, however, the )omanians dispensed with an already old $ungarian name, %sat6d .)omanian %etad, (erman, Tschatad/ -erb, atad ^ -chadat5 and renamed the settlement Benauheim in honour of the (erman litarary figure ,icholas Benau, perhaps the townLs most famous son. Curing and immediately after the Mirst Dorld Dar, a number of )omanian and 1ugoslav ""#

maps were produced to back up territorial claims against the %entral +owers. Bike their $ungarian and (erman counterparts, they made free use of own0 language labelling, which could now be used to political advantage .Cufr;noy, KCensit; de la population...L Teodorescu =eruOiu *2"#/ +oa0 BurcV, K$arta fi9icV a )omWnieiL. 's a logical reaction to the magyarisation of place0names that had preceded the war, official 1ugoslav and )omanian forms were introduced in Banat and Ba!ka, some openly political, like 7arahorhevo, or Banatsko 'leksandrovo. In the )omanian case, as there had not been any )omanian administration of the Banat for centuries, this was particularly difficult. 'n enumeration of the population in the territories acquired from $ungary was carried out in *2"#, and this was coupled with a not entirely successful attempt to standardise )omanian name0forms .7rallert *2<2 @2</ Parga *22" <40@*5. In many cases, existing (erman forms were accepted without amendment. Biebling is one of the few which remain in exclusive official use today.

+erhaps encouraged by this fluidity in the official toponymy, the (ermans of Banat and Ba!ka began to adjust an increasing number of place0names to their own tongue. This tendency was carried to an extreme by such researchers as 'nton Tafferner, who not only uncovered centuries0old (erman or (erman0sounding names that had been forgotten, but translated others and even created some which the inhabitants of the villages themselves had never heardAE5.. 's these were sometimes adopted and came into limited use, these must also be included.

The $ungarian return to the Ba!ka *2@*0@@ brought a new wave of political place naming, the post0*2*2 -erb colonists being deported with their place0 names and $ungarian colonists, and place0names such as $orthyv6ra taking their place. ' *2@" map of bjvid;k .,ovi -ad5 already has a fully magyarised street grid, e.g. $orthy =ikl8s sug6rjt, $onv;d utca, )6koc9y Merenc9 utca, 7ossuth Bajos utca, etc..Illustr.*@5. The (erman occupation of the 1ugoslav Banat *2@*0@@, saw merely the elevation of some (erman names, e.g. (roH0
=egnZyi W !'herer, #$BA, 99)D%. *ollo/ing their Tafferner6s /or; are mar;e" /ith an asteris;
)V

eXam9le,

@erman names /hi'h a99ear only in

""*

Betschkerek, to joint0official status with the -erbian .)egn;yi -cherer *23# E/ $asselblatt .$rsg.5 =arch *2@<5. $asselblatt speaks of @ towns and "4 communities that had been renamed in the 1ugoslav period, Kreceiving backL their former (erman names. The @ towns were probably Peliki Be!kerek .(roH0Betschkerek5, Bela %rkva .DeiHkirchen5, PrJac .Derschet95 and +an!ova .+antschewo5 A4. The collapse of the Third )eich and $orthyist $ungary and the communist take0over in &astern &urope introduced at once a new set of KheroesL and a new set of KvillainsL. +odgorica became Titograd, Be!kerek became Trenjanin, in deference to a +artisan leader. $orthy =ikl8s sug6rjt in bjvid;k became Bulevar =arsala Tita in ,ovi -ad. In $ungary, the village of Cunapentele was swamped by the ideologically driven -talinv6ros factory0town. In &ast (ermany, %hemnit9 became 7arl0=arx0-tadt. >rwellLs KdoublethinkL and Knon0personL concepts .*23@5, and 7underaXs K-lanskyXs hatL episode .the Book of Baughter and Morgetting5 fit the situation well. The Tito0-talin split almost certainly had a new surge of renaming as one of its results. -adly, the schi9ophrenia is still with us. 'ny visitor to $ungary for example, and particularly to Budapest since *232 will have been confused by the profusion of KnewL names. (one are the once ubiquitous reminders of -oviet domination, the Benin Boulevards, =arx -quares etc., replaced 00 again 00 by political figures from another age, conveying a new message .this time, a retrospective connectivity with %entral &uropean history via the $absburgs5. -talinv6ros, which had no older tradition to look back upon, is now Cunajjv6ros, i.e. KCanube ,ew TownL.

=egnZyi W !'herer, #$BA, 9V. Also, a short 9aragra9h in0 Fasselblatt, Werner >Frsg)0 ation un" !taat, ,euts'he Ieits'hrift f:r "as euro9<is'he ationalit<ten9roblem. Wilhelm Braum:ller, Wien7 (<rz, #$&C. +Wie"er"euts'he ?rtsnamen im Banat6, s9ea;s of & to/ns an" 2% 'ommunities, /hi'h ha" been rename" in the Yugoslav 9erio" +re'eiving ba';6 their former @erman names.6 The & to/ns /ere 9robably IrenLanin >@ro]D Bets'h;ere;), Bela 2r;va >Wei];ir'hen), PrSa' >Wers'hetz) an" 4anRova >4ants'he/o).
)%

"""

Njvidek ,ovi -'C

""<

=uch ink has been spilt over the meaning of multiple place names, and various nationally minded scholars involved in territorial disputes have invoked toponymy as a justification for their claims, seeking thereby to claim priori habitation for their respective KancestorsL A3. .-ee for example, 7nie9sa, in Bukinich .ed.5 *2@#, and +opp, in K-iebenbRrgenL *2@<5. The argument goes that if a settlement was originally founded by group ', it by rights should still belong to their descendants, an exercise which has of course, tended to produce more heat than light. 'll that either party managed to really prove is that some of the names are so old that they predate all reasonable present0day claimants. The reader is asked to bear in mind that original place0names belong to places, and to the local folk who coined them, not to peoples and nations. In any case, the existence of a place0name in a particular language of itself means little more than people who spoke that language have at one time or another had to do with that place, or that river, or mountain. ,ames survive long after their meanings have been forgotten, and long after the folk who gave them have passed away. The names of .(reat5 Britain and .Besser5 Brittany are linguistically linked, but from a time so distant .Eth century 'C5 that it means next to nothing in our age. )ather than for example, )omanian names, or (erman names, I will therefore speak of )omanian etc. usage or name0forms, and revel in the regionLs rich linguistic heritage.

%HE GEOGRAPHICAL IN"OR!A%ION $D$%E!

The rapid developments in modern multimedia software and hardware are making map0making available to a wider and in the main untrained public. This is leading to a dramatic increase in volume of cartographic output, but not necessarily to its overall standard. 'lthough the opportunities offered the conscientious cartographer by the new technology are fantastic, it is easy to get lost in the labyrinth of possibilities. The new technologies bring new
)B !ee, or eXam9le, 'nie(sa" Dst,EnC :rd2lyi f(ldra<+i nevei% Un0 1ukinich Dmre >e".)0 +1r"Zly6. A magyar t.rtZnelmi tGrsulat. # Augusztus, #$&A, an" 4opp" Dr .e,er0 9ie Toponymie Sie&en&r)gens. Un0 +!iebenb:rgen6, Bu;arest, #$&C.

""@

challenges and impose strict discipline as well. The opportunity to rapidly produce many alternative images of the same thematic map can and should be exploited, both for the education of the cartographer and for the benefit of the viewer. This is a quick and easy process nowadays, and the ethical cartographer cannot afford to overlook it. They must also be prepared for the map0user to wish to involve themselves with the process of map construction. Piewers are now advised that they &333should "e sus.icious o! systems that .resent .rocessed data as irre!uta"le !acts and !or"id inter!erence ith the narrativeMs .rogress3 #he conscientiously curious vie er needs to sto., think, "ack u., and .ro"e the data interactively3' .!on-onier , *22E *3<5.

The +lace ,ame )egister is intended as an adjunct to the discussion of the history, geography and ethnographic cartography of the Banat and Ba!ka. It is also to serve as a reference source for the accompanying (I.(eographical Information -ystem5 database from which many of the illustrations in this thesis have been generated. This package, developed using the desktop mapping application 5a. Vie er 93/7 F:olden So!t are @nc.5, shows all *,#A# settlements individually identified in the "< national censuses carried out between *3A# and *22", in their appropriate geographical location. ' collection of spreadsheets is connected via =apPiewer to the settlement centroids. The spreadsheets contain three types of settlement0level data raw population data from the censuses/ names under which the settlements have been officially recorded since *3A#/ and, within the respective country, the administrative units to which the settlement belonged at census0times. Dithin =apPiewer, digital base0 maps of Banat and Ba!ka were digitised by reference to official * "##,### topographic maps, and consultations with specialists in Banat and Ba!ka in (ermany, $ungary, )omania and 1ugoslavia suggest that they are as yet unique. 's described elsewhere, the courses of the Canube, Tisa and =aros are extremely complex 0 full of meanders, ox0bow lakes, dead arms and marshes. Mrom the available maps, it has often been hard to decide upon which river0courses to digitise, and which to ignore. ' certain amount of additional generalisation has therefore been required. ""A

The original objective was the production of a %C0)>= with a dynamic, user0 friendly and comprehensive (I- package. It would interactively illustrate processes of change in population si9e and component proportions in Banat and Ba!ka from the eighteenth century to the present day, and would be capable of considerable analysis. The methodology underlying thematic maps created within the (I- would be robust enough to serve as a benchmark for the digital and paper maps of other authors. In its present form the (I- permits the creation of a wide variety of maps, displaying data for a single village, a district, or any other administrative unit, for most of the censuses, singly or in combination. Bimited computer processing0power at first restricted progress to crawling0pace, but upgrades and persistence brought about the production of a broad range of maps. The plan has in part therefore been achieved, but has fallen short of the two key elements of benchmark status and interactivity.

+rogressively improved computer hardware gradually revealed the essential limitations of affordable graphics software. The census data has been recorded in %orel ]uattro +ro and =icrosoft &xcel spreadsheets. These are Ktwo0dimensionalL in format .i.e. have only an < and a y axis5, and have limited analytical power. Dhile =apPiewer, the chosen desktop0mapping package, produced adequate thematic maps, it had a poor spreadsheet element and was similarly unable really to analyse the data. +rogressive upgrades to =apPiewer have failed to address the major concerns, and the latest, =apPiewer <, uses the characters hitherto representing )omanian and 1ugoslav diacritics for mathematical functions, which prevents it from opening maps created with the earlier version.

In Cecember *224, the >ffice for ,ational -tatistics .>,-5 recruited the author into a unit working on a project for Ksmall0area estimation by multi0 level modellingL .referred to as K-'&+L5. Briefly, the objective is to create a robust system for estimating a wide range of variables, across the whole N7 but in areas as small as postcode units .ca. *A households5, on the basis of ""E

related data revealed by the decennial census and >,- national sample surveys. The project includes marshalling the huge .several do9en gigabytes5 datasets and geo0referencing to postcode level and (I- analysis using =apInfo. 's the methodology resulting from the research is to be disseminated, a graphical front0end application has to be created for clients. ' great variety of extremely powerful industry standard software has therefore become available. These include =apInfo and =apBasic, -'-+'% .for the *23* and *22* N7 censuses5, -+-- .for general statistical analysis5, =icrosoft Pisual Basic, 'ccess and Mox+ro. (rowing familiarity with these programmes, specific training and consultation with experienced colleagues is revealing vastly increased possibilities for the dataset, and demonstrating that the original intention is not only now realisable, but within reach.

The techniques for -'&+ will be directly applicable to the Banat and Ba!ka dataset. Indeed theoretical progress in the one project will ineluctably feed into the other. The settlement centroids have already been converted to =apInfo format, and the conversion of the Banat and Ba!ka census data into a relational database using =icrosoft 'ccess has begun. %orel Craw is proving useful as a medium for transferring existing boundary sets into =apInfo while new, and much more accurate, boundary sets will be generated by scanning recently acquired administrative maps. -canning of original census publications and interpreting by >ptical %haracter )ecognition .>%)5 software will fill the remaining gaps in the census dataset. ' %C0)>= based interrogative (I- tool covering census data in Banat and Ba!ka from *3A# to *22" should result. It will include properly digitised boundary sets for the territories and a relational database .based on 'ccess or Mox+ro5 for settlement population data. The user will be able to query the database interactively using -]B query language. (I- graphics will be supplied via =apInfo. It will include a graphical front end constructed using Pisual Basic.

""4

'++&,CI?

$ett,e-ent Co-9inations in D*gos,a+ Banat and Baka a5ter == T

Enlargements of Existing Settlements in Yugoslav Banat and Baka, 1 !"#$

Baka Pa,anka a9sor9ed $rpska Crn0a Banatski Aren0anin BaMaid Pane+o +or

Stara $alanka and Kova $alanka .*2@E5/ Kemaka Grnja .*2@45/ Banatski (u;anovac .*2@45/

5u]lja .*2@3, temporary association5/ P. BikaI .*2@3. temporary association5 Pojlovica .*2@3 temporary association5.

BNe; $ett,e-entC Creations in D*gos,a+ Banat and Baka, 23JH1?

E,e-ir

from combination of Kemaki Clemir and Sr.ski Clemir .*2@E5/ from Kovi 4utog and Stari 4utog .*2@E5/ from (onji =ovilj and :ornji =ovilj .*2@E5/ from Kovi ^ove and Stare ^ove .*2@E5/ from Soltur, Sveti Hu"ert and ^arlevil .*2@45/ from Slova%ki -radac and Sr.ski -radac .*2@45/ from 5ali #orak and Veliki #orak .*2@45/ from Sr.ska Keu+ina and Hrvatska Keu+ina .*2@45/ from Vranjevo and Volo;inovo.*2@45/ from A"oli%evo and Kova =anji]a .*2@45/ ""3

"*tog 'o+i,0 Ra+no $e,o Banatsko /e,iko $e,o Aradac Bege0ci Ne*>ina No+i Bee0 No+i 'neXe+ac

No+i 'o>arci No+o !i,oMe+o $*t0eska $i+ac /r9as Ro-ania

from Haj!eld and 5astort .*2@45/ from Beodra and (ragutinovo .*2@45/ from Kemaka Sara and Kova Sara/ .*2@45 from Kovi Sivac and Stari Sivac .*2@45/ from Kovi Vr"as and Stari Vr"as .*2@45.

$ett,e-ent Co-9inations in Ro-ania since 232I

%e-es+ar a"sor"ed333 MraOelia, %hiFoda Arad Caranse9e7 L*go0 Ora+iYa Or7o+a Re7iYa $Znnico,a* !are Bistra Boc7a !ontan6 C6+6ran Cenad Cenei Checea Co7tei* aro+a Ictar1B*dinY Po0e0ena 8!are@ -tefaneFt BudinO +ojejena de -us ""2 =ehala, (hiroda, (hiroda0,oua, +lopi, Mreidorf, 'radul ,ou and -Wnnicolaul =ic %aransebeF ,ou 'rmadia, -endel and -Vlha >raviOa )omWnI %oramnic, Gupalnic and Tufari )eFiOa )omcnI -Wnnicolau (erman >Oelu )oFu and >haba Bistra BocFa )omcnV and Pasiova Istv6nhegy %enadu Pechiu %enei .$orv6t5 %hecea .$orv6t5 %oFteiu

$Znpetr* !are %o-e7ti %*rn*1R*ieni A69rani

-Wnpetru ,ou (oi9eFti )uieni (uttenbrunn

$*9-erged a5ter B*i,ding o5

an*9e Iron Gates

a-:

>gradena ,ouc >gradena Peche +laviFeviOa TisoviOa

"<#

B%B&%'()*+,Y

"<*

O55icia, cens*ses

A*strian:

Hand"uch der Woj odscha!t Ser"ien und des #emeser Banats sammt der 5ilitLrgrLn+e !Ur das Iahr /8H13 7aiserl. u. 7:nigl. -taatsdrRckerei, Temesvar, *3A@.

H*ngarian:

- 5agyar =orona Ars+gai"an a+ 1$$13 )v elej)n v)grehajtott n).s+mlls !R"" eredm)nyei megy)k )s kR+s)gek s+erint r)s+lete+ve . II. 7:tet. >rs96gos =agyar 7ir. -tatis9tikai $ivatal Budapest, *33".

- 5agyar S+ent =orona ors+gainak 1 --3 )vi n).s+mllsaB /3r)s+B n).ess)g ltalnos leirsa kR+s)genkint , *. 7:tet. statis9tikai hivatal. Budapest, *2*". =agyar 7ir. 7:9ponti

- 5agyar S+ent =orona ors+gainak 1 1-3 )vi n).s+mllsaB /3r)s+B n).ess)g !R"" adatai kR+s)gek )s n).ese"" .us+tk, tele.ek s+erint , @". 7:tet. =agyar -tat. 7:9lem;nyek Budapest, *2*".

-+ 1 !13)vi n).s+mlls3 - n).ess)g anyanyelv s+erint, kR+s)genkint )s tRrv)nyhat,sgonkint Fideiglenes eredm)nyek)3 =.7ir 7:9ponti -9tatis9tikai $ivatal *2@<.

-+ 1 !13)vi n).s+mlls demogr!iai adatok kR+s)genk)nt FArs+ghatron kivUli terUlet)3 7;9irat. 7:9ponti -tatis9tikai $ivaltal Budapest, *22#.

"<"

-+ 1 $-3 )vi n).s+mlls3 ,r. "*. Cemogr6fiai adatok. 7:9ponti -tatis9tikai $vatal Budapest, *23<.

-+ 1

-3 )vi n).s+mlls3 'nyanyelv, nem9etis;g telepRl;senk;nt, *23#,

*22#. 7:9ponti -tatis9tikai $ivaltal Budapest, *22<.

-+ 1

-3 )vi n).s+mlls3 #ele.Ul)senk)nti adatok )s tele.Ul)hatros

t)rk).ek mgnesleme+en3 7:9ponti -tatis9tikai $ivaltal Budapest, *22<.

"<<

Ro-anian:

RecensZm_ntul :eneral al $o.ulaViei Rom_niei din 903 (ecemvrie 1 .-. +ublicat de Cr -abin =anuila, Cirectorul )ecensVmcntului (eneral al +opulaOiei . Polumul II ,eam, BimbV =aternV, )eligie. TipVrit la =onitorul >ficial, Imprimeria ,aOionalV, BucureFti, *2<3.

RecensZm_ntul :eneral al Rom_niei din 1 !1, +rovi9orii. BucureFti, *2@@.

W3 -.rilie3 Cate -umare

$o.ulaVia Rom_niei la RecensZm_ntul din 9H @anuarie 1 !$/ $rovi+orii3

(ate Sumare

-tructura CemograficV a +opulaOiei. ,umVrul Fi reparti9area teritorialV a populaOiei/ starea civilV/ RecensZm_ntul .o.ulaViei din 9/ 4e"ruarie 1 0"3 Structura (emogra!icZ a $o.ulaViei3 KumZrul Si re.arti+area teritorialZ a .o.ulaViei` starea civilZ` naVionatate` lim"a maternZ` nivel de @nstruire` !amilii3 CirecOia %entralV de -tatisticV, BucureFti, *2E#.

RecensZm_ntul .o.ulaViei 333 1 ""3 CirecOia %entralV de -tatisticV, BucureFti, *2E3.

RecensZm_ntul .o.ulaViei Si al locuinVelor din H3 Ian3 1 11. Pol. *. +opulaOiei 0 structura demograficV. CirecOia %entralV de -tatisticV, BucureFti, *23#.

D*gos,a+ian:

"<@

Ta StatistiNue d'Ctat3 R)sultats .r)liminaires du recensement de la .o.ulation dans le Royaume des Ser"es, Groates et Slovanes du 7/ Ianvier, 1 21. +ubli; par la direction de la statistique dLetat. Belgrad ^ -arajevo, *2"@.

(e!initivni re+ultati .o.isa stanovni;tva od 7/3 januara 1 21 god . *2<"

-arajevo,

Cefinitivni re9ultati popisa stanovniJtva od <*. marta 23L2 god. 7njiga II. +risutno -tanovniJtvo po Peroispovesti. CrQavna tamparija, Beograd, *2<3.

bcdefghijg klglmilmjg3 klgneonploe qe rcfeniqeocslm m tglcfucd vcwmjx, qe qenmsx ey 7/2@@@21 .1 zeymnc k{|}v~ ig reoeymned m eieoe2 tclemed, fczcy qe qplmngdg3 kopxwp tupuvruvsvol klmnsopursp llopuvxwl, lnop, *2@A.

-tatisti!ki Bilten, *E. $o.is stanovni;tva @ stanova, 1 113 Stanovni;tvo .o narodnosti @ ;kolskoj s.remi .o nasejima, /7 5arta /0?/ god33 ,ovembar, *24". -ocijalisti!ka 'utonomna +okrajina Pojvodina, +okrajinski Tavod 9a -tatistiku, ,ovi -ad,

$o.is stanovni;tva, domacinstva i stanova

u 1 $1 godini . ,acionalni sastav

stanovnistva po opstinama. -ave9ni Tavod 9a -tatistiku Belgrade, *23"

eqmi

ilgneomplog, x 1

yedgnmnilgog,

ilgneog

qeeqfmofcynm ovzpwnru

zgwymnilgog

13zeymnm

ilgneonmploe .

p|vnwpwp

kpupwp prvqvsp|vp. <. lnop, *22<.

"<A

Other Cens*s P*9,ications

$ch-idt, Andreas 8Hrsg(@( (ie deutschen Siedlungen in RumLnien nach der Bestandsau!nahme vom 7 Kovem"er /01D . Bearbeitet im Inst. fRr -tatistik und Bev:lkerungspolitik der deutschen Polksgruppe in )umgnien. $ermannstadt .-ibiu, ,agys9eben5, *2@*.

- Iugos+lviho+ csatolt terUletr)s+ek ko+s)geinek n).ess)ge anyanyelv s+erint a+ /0/D )vi mgyar )s a+ /09/3 illetleg a+ /07/3 )vi jugos+lv n).s+mllsok ala.jn F/01D3 )vi jugos+lv kR+ga+gatsi "eos+ts s+erint cso.ortos\tva)3 =. 7ir. 7:9ponti -tatis9tikai $ivatal, Budapest, *2@*.

Crd)ly

tele.Ul)seinek

nem+etis)gi

Fanyanyelvi)

megos+lsa

/8HD2/01/ .

7:9ponti -tatis9tikai $vatal Budapest, *22*.

Co--entaries on Cens*s

ata

Atnagea, Ioan( (ie BevRlkerung der Siedlungen vom (onaueng.a>, im SUden des =reises Gara/2Severin, "ei der let+ten Volks+Lhlung3 In !*se*- 5Ur Geschichte des 'reises 'arasch1$e;erin, Rechit>a, in /er9ind*ng -it de- Instit*t 5Ur dona*sch;V9ische Geschichte *nd Landesk*nde, %U9ingen, *nd der #ni+ersitVt Ba9es1Boyai, ',a*sen9*rg . =ulturraummittlere und untere (onau3 #raditionen und $ers.ektiven des Jusammenle"ens . )eFiOa, *22A.

Csa,,ner, A,5red( (ie rumLniendeutschen Bestandsau!nahmen von /07?8 und /01D . -onderdruck aus Molge <^*234, <E. Gahrgang. -Rdostdeutsche Pierteljahresblgtter. =Rnchen, *234.

"<E

E,e-[r, magyar

r( )aka955y( Crd)ly statis+tikja3 - Romniho+ csatolt eg)s+ volt terUlet lakossgnak nem+etis)gi, !eleke+eti )s kulturlis

kRys)genk)nti stati+tikja a+ /0/D )s /09D3 vi hivatalos adatok ala.jn . ' \=agyar 7isebbs;g\ nem9etpolitikai s9emle. Bugos, *2"<.

"ischer, Ho,ger: Regional"e+ogeheit nationalitLtens.e+i!ischer (aten in den ungarischen Volks+Lhlungen des 9D3 Iahrhunderts . Nnpublished )eferat, manuscript, $amburg, *22A. ( GR,,ner, Ra,5 %ho-as: (er ungarisch2rumLnische KationalitLtenkon!likt in Sie"en"Urgen im S.iegel der Volks+Lhlungen von /0/D, /09D und /07D3 -iebenbRrgische -emesterblgtter, 4. Gahr, $eft *0", =Rnchen, *22<.

Gregesc*, C(: Statistische und kartogra.hische Jeugnisse U"er die RumLnen und die [ngarn + ischen den =ar.athen und der #hei> . In \Sie"en"Urgen \/ Institut fRr rumgnische (eschichte in Bukarest Bukarest, *2@<.

Hasse,9,att, =erner 8Hrsg@: (ie (eutsche Volks+Lhlung im rumLnischen Banat` Sie"en"Urgen und das Banat in der Statistik3 In ,ation und -taat, Ceutsche Teitschrift fRr das europgische ,ationalitgten0problem. Dilhelm BraumRller, Dien. =gr9, *2@*.

'ra,,ert, =i,5ried( :eschichte und 5ethode der BevRlkerungs+Lhlen im SUdosten3 /, RumLnien3 5it "esondeer RUcksichtsname au! die JLhlung des Iahres /07D und ihre VerR!!entlichung . -onderabdruck aus KCeutsches 'rchiv fRr Bandes0 und PolksforschungL, Gahrgang III $eft <^@. Beip9ig, Ce9. *2<2.

'ra,,ert

8hrsg(@ Iugosla ienB Verteilung der VolkstUmer nach Be+irken3

Ver+eichnis der Arte mit mehr als /D deutschen Cin ohnern . Berarbeitet von der +ublikationsstelle Dien. Dien, *2@*.

"<4

'ra,,ert 8hrsg(@ (ie :liederung der BevRlkerung des ehemaligen Iugosla ien nach 5utters.rache und kon!ession nach den unverR!!entlichten -nga"en der JLhlung von /07/ . Bearbeitet und herausgegeben von der +ublikationsstelle Dien. Dien, *2@<.

'ra,,ert

8hrsg(@ (ie BevRlkerungs+Lhlung in RumLnien, /01/3 (&$&I=.

Bearbeitet und herausgegeben von der +ublikationsstelle Dien. Dien, *2@< 3

!an*i,a,

r(

$a9in:( Hinsicht3

(ie

4olgen

der

#eilung Institut

Sie"en"Urgens fRr

in

demogra.hischer

In \Sie"en"Urgen \/

rumgnische

(eschichte in Bukarest Bukarest, *2@<.

!arga, Andrei: RumLnische ethnische @denti!i+ierung . In -iebenbRrgische Bandeskunde *3 .*22A5, $eft *.

Teitschrit fRr

5agyarors+g nem+etis)geinek )s a s+oms+)dos llamok mag+arsgnak statis+tikja F/0/D2/009)B a /009 s+e.tem"er 92H3 kR+Rtt tartott ko!erencia eloadsai3 7:9ponti -tatis9tikai $ivatal, Budapest, *22@.

!irnics, 'Kro,y =ise""s)gi Sors . Morum 7:nyvkiad8, *22<.

,ovi -ad ^ Njvid;k,

$ch-idt, A( 8Hrsg 3/ (ie deutschen Siedlungen in RumLnien nach der Bestandsau!nahme vom 7 Kovem"er /01D3 Bearbeitet im Inst. fRr -tatistik und Bev:lkerungspolitik der deutschen Polksgruppe in )umgnien. $ermannstadt .-ibiu, ,agys9eben5, *2@*.

/arga, E(ArpKd: K).s+mllsok a jelenkori Crd)ly terUlet)n FIegy+etek Crd)ly *22". )s a ka.csolt r)s+ek 3 S++adi nem+etis)gi statis+tikjnak Budapest, tRrt)net)he+)3 )egio k:nyvek 0 =T' T:rt;nettud8m6nyi Int;9et,

"<3

=agner, Ernst: Ctnische und religiRse 5inderheiten in #ransslvanien nach der rumLnischen Volks+Lhlung vom Iahre /0093 In Teitschrit fRr -iebenbRrgische Bandeskunde *3 .*22A5, $eft *.

=o,5,

r )ose5: (eutsche 5inderheiten in SUdosteuro.a im [m"ruch3 (ie

Volks+Lhlungen /00D 2 /0093 institut fRr donauschwgbische (eschichte und Bandeskunde, TRbingen, *22@ 3

"<2

!aps 1 in Order o5

ate o5 P*9,iation

(er #emes arer Bannat, a"getheilet in Seine (istrict3 -u!genommen und a"geme!sen au! hohen "e!ehl ihrer hoch!Urstlichen (urchlaut den .rin+en Cugenii von Savoyen333 unter dem :ouvernement seiner C<cellence des Herrn :eneralen 4eldmarchallen (rafen 4laudii von 5erc6 /?97 2 /?9H .the K=ercy07arteL5.

Gharte von dem #emes arer Bannat samt denen ihm incor.orierten H (istricten, so +um =Rnigreich servien gehRren FBanat im Iahre /?97)3 ,o publication details.

Bri55a*t, Etienne .Bibraire Pienne5, #heatre de la guerre dans le Bannat de #emesvar ... Dien, *4<3.

#emesvar Bannat, a"getheilet in seine (istrict und $rocess, -nno /?W/3

(riselini, 7ransiscus #a"ula Bannatus #emesiensis . Pindabonce .Pienna5, *44E.

+on !i,den9erg, 'ie9ern .$auptmann beim Dallachisch0illyrischen )egiment5 =Lys3=Rn3 $rovin+ genannt #emes arer Bannat ... ,o place of issue, *43<.

Hoh-ann, Er9en 8Hrsg(@ (as #emesvarer Bannat 333, =it ):m. 7ayserl. 'llergngd. Mreyh.... ,urnberg, *42#.

+on 8\@ 'is,, )oseph *nd Ga9rie,: Situations.lan von Bacser Gomitat, darin die Vereinigung der (onau mit der #hei> mittels eines Schi!!!ahrts Ganal ge+eigt ird3 'round *3##.

"@#

S*its, )*r( Geo-( Anton Bcs )s Bodrogh tRrv)nyesen egyesUlt vrmegyek FGomitatuc Bacsiensis et Bodrogiensis articulariter uniti)3 Budapest, around *3##.

/ot[sky, Geo-( !ich. #orontl vrmegye FGomitatus #orontaliensis 5. Budapest, around *3##.

Bagosi, Ord( Geo-( )osepho =rass, vrmegye FGomitatus =rassoviensis)3 Budapest, around *3##.

Cegen, )i*so( Co-it: Geo-( )aco9*-: #emes vrmegye FGomitatus temesiensis)3 Budapest, *3#".

kfqijg, |eing, bgdgsmc, bxfeonc, fnczefc , *3#A

4rancisco2Iose!inische -u!nahme \. <E x * 4A.### sheets covering all Bacsa and Banat. 7.u.k =ilitgrgeographisches Institut, Dien, *34< 0 *334.

- 5agyar *llam kR+iga+gatsi t)rk).e3 -+ /0DD )vi anyanyelvi adotokkal kieg)s+itette a =. 7i. 7:9ponti -tatis9tikai $ivatal .Ver altungskarte des ungarischen Staates3 Versehen mit den mut ters.rachlichen -nga"en von /0DD vom k:niglichen ungarischen statistischen Tentralamt5. * <E# ###. Budapest, *2#.

Ba,ogh, PK, - n).!ajok 5agyarors+gon . .a5 PIII mell;klet - n).!ajok3 d)lvid)ken3 .b5 ??II mell;klet - nyelvhatrok )s nyelvs+igetek3 - d)lvid)ken )s a+ ors+g m)ly)n3 .c5 5agyarors+g n).raj+i t)rk).e . ' =agyar 7ir6lyi vall6s0 ;s k:9oktat6sRgyi minist;rium. Budapest, *2#".

"@*

S+er)m megye, tn)+eti t)rk).e3 Budapest, *2##0*2*#.

#orontl vrmegye tn)+eti t)rk).e . Budapest, *2##0*2*#.

#orontl vrmegye, !Rldad,katas+teri "eos+tsanak tn)+eti v+lata3 * *@@ ###. Budapest, *2##0*2*#.

#emes vrmegye, !Rldad,katas+teri "eos+tsanak tn)+eti v+lata3 * *@@ ###. Budapest, *2##0*2*#.

Bcs2Bodrog vrmegye, !Rldad,katas+teri "eos+tsanak tn)+eti v+lata . * *@@ ###. Budapest, *2##0*2*#.

=rass,2S+Rr)ny vrmegye, !Rldad,katas+teri "eos+tsanak tn)+eti v+lata . * *@@ ###. Budapest, *2##0*2*#.

*5r[noy, I-p (ensit) de la .o.ulation Ser"o2Groate dans le Banat, a.ras la statistiNue o!!icielle Hongroise de /0/D .Cufr;nay, +aris, undated, but before *2*25.

Atanasi*, Pro5( A( ( Garte Cthnogra.hiNue du Banat de #emesvar3 -.ras la Garte Cthnogra.hiNue de la Hongrie orientale, .ar $l Balogh, et les statistiNues o!!icielles3 +aris, *2*2.

%e,eki, GrQ5 PK,: 5agyarors+g n).raj+i t)rk).e a n).srs)g ala.jn FCthno 2 gra.hische =arte [ngarns au! der :rundlage der BevRlkerungsdichte 5. Budapest, *2*2.

"@"

Mreytag0Bernds =arte von RumLnien3 5it den neuen :ren+en .=aHstab, * * ### ###5. Mreytag Bernd (mb$, Dien und Beip9ig .between *2*2 and *2"*5.

%eodoresc*, Genera, C( 7i !er*Yi*, Pro5( /asi,e( Rom_nia cu noile 4rontiere admise de Gongresul de la $aris Si numirele Rom_ne/ti ale tutoror TocalitZilor .la scara * * ### ###5. BucureFti, *2"#

Popa1B*rc6, Pro5( I(: Harta !i+icZ a Romniei , scara *I4##.###. 'teliere -ocec %o. -.'., BucureFti, around *2"#

- 5agyar"irodalom s+)tdara"olsaa trianoni ")kes+er+d)s s+erint . =agyar 7ir. 7:9ponti statis9tikai hivatal, =agyar 7ir. llami T;rk;p;s9et nyom6sa. Budapest, around *2"#.

$pek, Pro5(

r( R*do,5( =arte der deutschen Siedlungen :ro>2RumLniens .

7ulturamt des Perbands der Ceutschen in (roH0)umgnien. $ermannstadt, .-ibiu, ,agys9eben5 *2"<.

$ch;a,-, Hans( .a5 Das Deutschtum des Banats auf (rund der rumgnischen, sRdslawischen und ungarischen Polks9ghlungen aus den Gahren *2"# b9w. *2"*. .b5 Das Deutschtum der Batschka auf (rund der sRdslawischen und ungarischen Polks9ghlungen aus den Gahren *2"# b9w. *2"*. Beilagen 9ur +ublikation ]$andw:rterbuch des (ren90 und 'uslandsdeutschtums\. Petersen, $chee,, R*th, $ch;a,- 8Hrsg(@: Perlag Merdinand $irt, Breslau, *2<<.

Pa*, Pro5( /irgi,: Harta !i+icZ, .olitiZ Si etnogra!icZ a Banatului . * <## ###, TimiFoara, *2<E.

"@<

^_`a, bcdefg b(: reeymng3 e xqxliloxbf3 bmd3 mfmemg }mwfgyme m flge . pmlop * E## ###, p~l |unovrsn koyuxp nxm tp, *2<2.

BKtky, Asig-ondF 'og*to;ic>, 'Kro,yF [s %e,eki PK, GrQ5 . Crd)lynek )s 5agyarors+g keleti r)s+einek n).ra+i t)rk).e /0/D2"en3 * "##.### m;ret n;praj9i t;rk;pe alpj6n. -9erkes9tette ;s kiadja a =agyar -tatis9tikai T6rsas6g 'llamtuom6nyi Int;9ete, Budapest, *2@#.

Crd)lynek )s 5agyarors+g keleti r)s+einek n).ra+i t)rk).e a+ 07D2/0703 vi n).s+mllsok ala.jn3 - hivatalos magyar, csehs+lovk, romn )s jugos+lv n).s+mllsok adatai al.jn . -9erkes9tette ;s kiadja a =agyar -tatis9tikai T6rsas6g 'llamtudom6nyi Int;9ete, Budapest, *2@#.

'ra,,ert, =i,5ried 8Hrsg(@ .a5 8olkstumskarte von 9ugoslawien auf (rund der amtlichen Tghlungsergebnisse Rber die =uttersprache *2<*, mit Trennung der -Rdslawischen P:lker mit $ilfe der 7onfessionsstatistik und :rtlicher 'ngaben. .* "##,###. @# sheets5, +ublikationsstelle, Dien, *2@#0*. .b5 8olkstumskarte von )um:nien auf (rund der &rgebnisse der amtlichen rumgnischen Tghlung Rber die Polks9ugeh:rigkeit von *2<# .)ecensamintul (eneral al +opulatiei )omaniei *2<#, Pol. II, Bukarest *2<25 .* "##,###. @@ sheets5 +ublikationsstelle, Dien, *2@#0*. .c5 8olkstumskarte von ;ngarn< auf (rund der amtlichen Tghlungsergebnisse von *2<# ..* "##,###. "< sheets5. +ublikationsstelle, Dien, *2@*.

Her9ert, GUnter. (ie deutschen Siedlungen in den :e"ieten Banat, Bergland und Sie"en"Urgen nach der Bestandsau!nahme vom 7 Kovem"er /01D . 7artenbeilage 9u K(ie deutschen Siedlungen in RumLnien nach der Bestandsau!nahme vom 7 Kovem"er *2@#L3 Bearbeitet im Inst. MRr -tatistik und Bev:lkerungspolitik der deutschen Polksgruppe in )umgnien. $ermannstadt .-ibiu, ,agys9eben5, *2@*.

- viss+at)rt Bcska )s Baranya nem+etis)gi )s "irtok.olitikai t)rk).e3 '9 jjvid;ki *2@* agr6r0hivatal 6tvetel;vel;s a birtokpolitikai k;rd;sek tanulm6nyo96s6val megbi9ott minis9teri t6rak:9i bi9otts6g. bjvid;k, @ =ay

"@@

Romnia, kR+iga+gatsi t)rk)., Budapest, *2@*0@@.

RumLnien3 Ver altunsstand Akto"er /01/ . .* *.###.###5. Bearbeitet und herausgegeben von der +ublikationsstelle Dien, Dien *2@".

5agyarors+g, kR+iga+gatsi t)rk)., Budapest, *2@".

Weltkarte /B/DD DDD, Blatt Buda.est . Dar >ffice *2<3. -traHen, ,amen und (ren9en berichtigt, I? *2@". (ren9berichtigung nach angaben des 'uswgrtigen 'mtes vom *E.@.*2@<.

%a,,iKn, "erenc: Bcs2Bodrog Vrmegye, =R+iga+gatsi t)rk). . Budapest, *2@<.

[ngarn, Ver altungskarte Fdie =arte "eruht au! dem Ver altungsstand vom /3@3 /017)3 (rundlage verwaltungsmgHiges >rtsver9eichnis )umpfungarns fRr das Gahr *2@@. Bearbeitet von der +ublikationsstelle Dien, *2@@.

Roth, "ran>: (onausch L"ische Seidlungsge"iete 2 Banat3 Ver altunsstand vor /01/. =aHstab * E##.###. ,euland, Dochenschrift der Conauschwaben. Conauschwgbische Perlagsgesellschaft, -al9burg ^ +annonia0Perlag, Mreilassing in Bayern, year

kesmmilmjg {cqxmjg kfmg, k ~ reeymng3 pmlop * <## ###, lopl, year

fczcyng gflg, kfcw =>?@ABC3 pmlop * A## ###. upmzp, \lnspoup\, *2E#.

fczcyng gflg, kfcwg DCEFCG3 pmlop * A## ###. upmzp, \lnspoup\, *2E#.

"@A

fczcyng gflg, kfcwg DG/ HIJGCBIK>3 pmlop * A## ###. upmzp, \lnspoup\, *2E#.

fczcyng

gflg,

kfcwg

LGA?M>?I?3

pmlop

* A##

###.

upmzp,

\lnspoup\, *2E#.

fczcyng

gflg,

kfcwg

NCBI

D>O/

pmlop

* A##

###.

upmzp,

\lnspoup\, *2E#.

fczcyng gflg, kfcw DPFCJIK>3 pmlop * A## ###. upmzp, \lnspoup\, *2E#.

#o.ogra!ska karta topografskim

* "##,###. CelimiIno dopunjeno *2A3. (od. Inscrtano i9 *2<2.god. I9danja Pojnogeografskog Instituta,

9nacima

Belgrade in *2E<, *2EE

Javiajna karta Iu%nog Banata .)a9mera * "## ###5. +reJoru!eno od +rosvetno pedagoJkog 9avoda, +an!evo. I9rada I reprodukcija (eokarta, Beograd, *2EA. (od.

Harta

4i+icZ

Si

-dministrativZ

IudeVului

-rad3

Tiparul

exetecutat

la

intreprinderea poligrafia \%risana\ >radea ,r. *##2 ^ *24<.

=reis #emesch, Banat FRumLnien)3 %opy acquired from Banater +ost, =Rnchen, no bibliographical details, probably *23#s.

IudeVele $atrei. BucureFti, *23

/asi,e, Genera, -a0or ing(

rago-irF/ictor, Co,( Ing( BaeaF Gheorghe,

Co,( Ing( !*re2an*F Gheorghe, Ep*ran: RomZnia, -tlas Rutier3 &ditura -port0Turism, without place, *23.

"@E

/o0nogeogra5skog Instit*ta: #o.ogra!ska karta

* "##,###. Crugo i9danje

stampano *234. (od. -adUaj prema stanju i9 *23A.god., Belgrade, *234

$,a;inski, I,ona: Ver altungsgliederung Ast2 und SUdosteuro.a . In )ordan, Peter 8Hrsg(@ -tlas Ast2 und SUdosteuro.a . sterreichisches >st0 und -Rdosteuropa0Institut, Dien, *232.

'ocsis, 'Kro,y: Re.artiVie de lim"i materne n #ransilvania F-rdeal) Crd)ly anyanyelvi t)rk).e, S.rachenkarte von Sei"en"Urgen, /08W . $;ttorony kft., feleles kiad8. 7.P., Budapest, *22#.

'ocsis, 'Kro,y: Cthnic 5a. o! Vojvodina . Institute of (eography, $ungarian 'cademy of -ciences, *22*.

$a*9erer, !ichae,: BevRlkerungsverteilung in den lLndlichen :emeinden RumLniens3 In )ordan, Peter 8Hrsg(@ -tlas Ast2 und SUdo7steuro.a . sterreichisches >st0 und -Rdosteuropa0Institut, Dien, *22#.

Crd)lyi t)rk). )s helys)gn)vtr , * A##.###. %artogr6phia Budapest, *22*.

h`iejkl, mnjn: ineog g cdx fgnmsg glgfm, nxm tp, *22*.

IudeVule -rad, #imiS, GaraS2Severin3 %ommune and settlement administrative borders. -cale and publication details not given. TimiFoara, around *22".

B[,a, -[rnok a,e>redes $>a9Q 5agyar Honv)ds)g Ve+)rkra, T2712@@@, Deva. * "##.###, =agyar $onv;ds;g 7artogr6iai 9em, Budapest, *22".

PaskQ,

-[rnok

a,e>redes

)Q>se5 *rad.

8int[>etparancsnok@: * "##.###, =agyar

5agyar

Honv)ds)g

Ve+)rkra,

T2712V@,

$onv;ds;g

7artogr6iai 9em, Budapest, *22".

"@4

PaskQ,

-[rnok

a,e>redes T2712

)Q>se5

8int[>etparancsnok@: * "##.###, =agyar

5agyar

Honv)ds)g

Ve+)rkra,

Sombor .

$onv;ds;g

7artogr6iai 9em, Budapest, *22".

PaskQ,

-[rnok

a,e>redes

)Q>se5

8int[>etparancsnok@:

5agyar

Honv)ds)g Ve+)rkra, T2712@, QrenRanin. * "##.###, =agyar $onv;ds;g 7artogr6iai 9em, Budapest, *22".

PaskQ,

-[rnok

a,e>redes

)Q>se5

8int[>etparancsnok@:

5agyar

Honv)ds)g Ve+)rkra, T2712V@, SSeged. * "##.###, =agyar $onv;ds;g 7artogr6iai 9em, Budapest, *22".

PaskQ,

-[rnok

a,e>redes

)Q>se5

8int[>etparancsnok@:

5agyar

Honv)ds)g Ve+)rkra, T2712V@, TimiUoara. * "##.###, =agyar $onv;ds;g 7artogr6iai 9em, Budapest, *22".

PaskQ,

-[rnok

a,e>redes

)Q>se5

8int[>etparancsnok@:

5agyar

Honv)ds)g Ve+)rkra, T2712V@, Subotica. * "##.###, =agyar $onv;ds;g 7artogr6iai 9em, Budapest, *22".

IudeVul #imiS, scara * "## ###. 'dministrative map, TimiFoara, *22"

!i,o0e+i:,

r !iros,a+F %odoro+ic, !arina: S-$ VojvodinaB -grarna :ustia

Kaseljenosti .o katastarckim o.;tinama3 I9dao (eografski Inst. \ Govan %vigiU\ -',N, Beograd, *22<.

'icoMe+, $aMa Cthnic Structure o! the $o.ulation o! Voivodina Fsettlements) according to the !irst results o! the /00/ census3 Institute of (eography, ,ovi -ad, *22<

$tepi:,

r( !i,o-irF I,i:,

r( )o+n: Cthnic Structure o! the $o.ulation o! the /00/ census .

Voivodina Fsettlements) according to the !irst results o! Institute of (eography, Nniversity of Belgrade, *22<.

"@3

!aps and Geography

ent, Borden: Gartogra.hyB #hematic 5a. (esign3 (eorgia -tate Nniversity, *22E.

"ischer, Ho,ger: =arten der rLumlichen Verteilung der KationalitLten in [ngarn3 F(arstellungsmRglichkeiten und :ren+en ihrer @nter.retation am Beis.iel von ungarischen KationalitLtenkarten des /03 und 9D3 Iahrhunderts)3 $amburg, *22*.

'eates, )($(: [nderstanding 5a.s . Bongman, $arlow, *23".

'raak !()( P Or-e,ing "(): Gartogra.hyB Visualisation o! S.atial (ata3 Bongman, $arlow, *22E.

'ra,,ert, =i,5ried: 5ethodische $ro"leme der VRlker2 und S.rachenkarten, dargestellt am Beis.iel von =arten U"er Ysterreich und SUdosteuro.a . Internationales Gahrbuch fRr 7artographie *, *2E*.

'rischan, A,e.anderB Bi"liogra.hieB Banater =arten im =riegsarchiv Wien F/W8W2/0DD)3 In $teinacker, Haro,d 8Hrsg(@: -Rdostdeutsches 'rchiv, PI. Band. ). >ldenbourg Perlag, =Rnchen, *2E<.

!enyen( E( 8Ed(@: Volks2 und S.rachenkarten 5itteleuro.as3 2 Ceutsches 'rchiv fRr Bandes0 und Polksforschung ". Berlin ,o.l. *2<3.

!itche,,,

B()(=(:

4Llschung im

und

WahrheitB Iahrhndert

die und

=artogra.hie ihre

des

=ar.athen"eckens

+ an+igsten

@nter.retation

angesichts der ungarisch2rumLischen :egenU"erstellung3 Nnpublished B' dissertation, N&', ,orwich, *22* 3

!onkho*se( " P =i,kinson, HR(: 5a.s and (iagrams, their Gom.ilation and Gonstruction3 =ethuen, Bondon, *243

"@2

!on-onier, !ark: Ho Bondon, *22E

to Tie

ith 5a.s3 Nniversity of %hicago, %hicago and

Ro9inson et a,(: Clements o! cartogra.hy, si<th edition3 Gohn Diley and -ons, ,1, %hichester, etc, *22A.

RQnai, AndrKs: #)rk).e+ett #Rrt)nelem . +Rski 7iad8 7ft., Budapest, *22<. "E"pp..

RumLnische Tandkarten!Llschungen3 4estgestellt vom Staats issen scha!tlichen @nstitut der [ngari schen Statistischen :esell scha!t. 'thenaeum, Budapest, *2@#.

RumLnische ReL*i'(

Cthnogra.hische

Tandkarten

und

ihre

=ritik .

-taats0

wissenschaftliches Institut. Budapest, *2@". Re5erred to in the te.t as

%e,eki, GrQ5 PK,, RQnai, AndrKs: The Cifferent Types of &thnic =ixture of +opulation. )oyal $ungarian %arographical Institute ^ 'thenaeum, Budapest, *2<4.

=i,kinson, H(R(: 5a.s and $olitics3 - revie o! 5acedonia . Biverpool, *2A*.

o! the Cthnogra.hic Gartogra.hy

=itt,

=erner

#hematische

=artogra.hie3

5ethoden

und

.ro"leme,

#enden+en und -u!ga"en . ". 'ufl., $annover, *24#.

=o,5, )ose5: Iahrhunderts

Raum"e ertung, Verkehr und Vekehrs.olitik im Banat des /83 im Tichte der Tandes"eschrei"ungen . In !*se*5Ur

Geschichte des 'reises 'arasch1$e;erin, Rechit>a, in /er9ind*ng -it de- Instit*t 5Ur dona*sch;V9ische Geschichte *nd Landesk*nde, %U9ingen, *nd der #ni+ersitVt Ba9es1Boyai, ',a*sen9*rg . =ulturraummittlere und untere (onau3 #raditionen und $ers.ektiven des Jusammenle"ens . )eFiOa, *22A.

"A#

=ood,

enis: #he $o er o! 5a.s3 )outledge, Bondon, *22"

Ethnic Geography
BK,int, $Kndor - Bnt tele.Ul)stRrt)net)"l . In ' $ungarol8giai Int;9et

Tud8m6nyos 7:9lem;nyei, *A. -96m, P. ;vfolyam, bjvid;k, *24<, jjnius.

Ba,ogh,

PK,

- n).!ajok 5agyarors+gon . ' =agyar 7ir6lyi vall6s0 ;s

k:9oktat6sRgyi minist;rium. Budapest, *2#".

Bodor,

r Anta,: ()lmagyarors+gi tele.\t)sek tRrt)nete )s hatsa a mai

kR+lla.otokra. -tephaneum ,yomda ).T., Budapest, *2*@.

Boh-ann, A,: BevRlkerung und KationalitLten in Sudosteuro.a . =enschen und (ren9en, Band ", 7:ln, *2E2.

B*ch-ann, 'Kro,y: - d)lmagyarors+gi tele.\t)sek tRrt)nete3 @B Banat3 Budapest, *2<E.

p*rqi:,

r $,o+odan, and 'icoMe+,

r $aMa:

(evelo.ment o! the

$o.ulation in Baranya . +eopleXs =useum, Beli =anastir 0 =useum of Pojvodina, ,ovi -ad, *22".

"ischer,

r( Ho,ger (eutsch2ungarische Be+iehugen in der :eogra.hie der -onderdruck aus "ischer r $>a9ad+Kry .$rsg.5

J ischenkriegs+eit .

#echnologie und Wissensch!tsaustausch + ischen [ngarn und (eutschlandB -s.ekte der historischen Be+iehungen in Katur issenscha!t und #echnik3 >ldenbourg Perlag, =Rnchen, *22A.

Grothe,

r( H*go et a,( 8Hrsg@(: :rothes kleines Hand Rrter"uch des :ren+2

und -usland2(eutschtums . Pg. ) >ldenbourg, =Rnchen und Berlin, *2<"

"A*

Hosch, E P $ee;ann, G(: -s.ekte ethnischer @dentitLt3 Crge"nisse des !orschungs.rojekts P(eutsche und 5agyaren als nationale 5inderheiten im (onauraum \. Buchreihe der -Rdostdeuschen $istorischen 7ommission, Band <A. Perlag ). >ldenbourg, =Rnchen, *22*.

'i:oMe+,

r( $aMa: #he -utonomous $rovince o! VojvodinaB the $ro"lems o!

Cthnic @dentity3 ,ovi -ad, *22A.

'i:oMe+, *22E.

r( $aMa: Ghanges o! Cthnic Structure o! the $o.ulation o! the

Yugoslav Banat during the 9Dth Gentury3 Nnpublished manuscript, ,ovi -ad,

'ocsis,

r( 'Kro,y

- Vajdasg nem+etis)gi k).)nek s++ )ve F/88/2/08/)3

TerRleti 7utat6sok, Cebrecen or Budapest, *23<.

'ocsis, 'Kro,yF 'ocsisn[, Hodosi Es>ter: 5agyarok a Hatrainkon t6l a =r.t2mendenc)"en3 Tank:nyvkiad8, Budapest, *22*.

'ocsis, 'Kro,y

Buda.est )s r)gi,ja etnika t)rs+erkee+et)nek talakulsa

F/8HD2/00D)3 In M:ldraj9i rteste ?BIII. vf. *22@. <0@. MR9et, pp. "220<"@.

'ocsis, 'Kro,y and 'ocsis1Hodosi, Es>ter: Hungarian 5inorities in the Gar.athian Basin, a Study in Cthnic :eogra.hy3 =atthias %orvinus +ublishing, Toronto0Buffalo, *22A.

'ocsis,

r( 'Kro,y: Gontri"utions to the Cthnic :eogra.hy o! the Hungarians

in Vojvodina3 Becture held at 4th $ungarian0%roatian (eographical -ymposium, =ljet Island, *<0*Ath -eptember, *22A.

'o+Kcs,

r( A,a0os: #erUlet )s n).ess)gB a S+er"2Horvt2S+lov)n llam

n).s+mllsnak nem+e+is)gi adatai3 =agyar -tatis9tikai -9emle, IP ;vf. Budapest, *2"E.

"A"

La0os,

Arday:

Iugo+lviai

magyar

nem+etis)g

hely+ete

F/08/)3

In

5edvetnc '9 &BT& ;s a9 =77& T6rsadalomelm;leti foly8irat6nak mell;klete, jelent;sek a hat6ron tjli magyar kisebbs;gekrel. Budapest, *233.

Nyigri, I-re 8ed(@: - viss+at)rt ()lvid)k nem+etis)gi k).e3 $al6s9 irodalmi ;s k:nyvkiad8v6llalat, Budapest, *2@*.

Petersen, $chee,, R*th, $ch;a,- 8Hrsg(@: Hand Rrter"uch des :ren+2 und -uslandsdeutschtums3 Perlag Merdinand $irt, Breslau, *2<<.

PogKny,B[,a 8Hrsg(@: (ie SiedlungsverhLltnisse des [ngartums im "eset+ten SUdungarn . +ester0Bloyd (esellschaft, Budapest, *2@*

RQnai, *2<2

r( AndrKs: Kem+etis)gi .ro"l)mk a =r.t2menenc)"en3 In M:lraj9i

7:9lem;nyek, B?PII k:tet, p.@E*0@4<. ' =agyar M:ldraj9i T6rsas6g, Budapest,

$a0ti, Eniko #he Hungarian Kational 5inority in Yugoslavia a!ter /01H, Historical Sketch3 'cta $istorica, Tomus B???I?, 'cta Nniversitatis -9egediensis de 'ttila G89sef ,ominatae.

$ch-idt,

Horst

ieter(

Cin

versch undenes

(or!

im

Banat3

BevRlkerungs"iologische [ntersuchungen der "Rhmer2deutschen :emeinde Tinden!eld3 'rmin Paas Perlag, Bangenau0Nlm, *22*.

so-e7an, Pro5(

r( La*rian: (ie #hei>e"eneB eine natUrliche :ren+e

+ ischen RumLnen und [ngarn, geogra.hische und geo.olitische Studie3 -ibiu^$ermannstadt, *2<2.

%opony-y, P,ace1na-e Registers

Brandter, r( Pa*,: Ver+eichnis der deutschen Siedlungen in [ngarn F/011)3

-onderdruck aus dem 'Polkskalender der Ceutschen aus Nngarn. $rsg. Bandmannschaft der Ceutschen aus Nngarn, =Rnchen, *2AE.

"A<

(icVionar de localitZVi din #ransilvania . 7riterion, %luj, *22".

Geh,, Hans P Bader, Leo: Werkstatt"ericht mit $ro"eartikeln aus dem &WRrter"uch der donausch L"ischen Bekleidungsge er"e'3 -nhangB =arten und Artsnamen . $rsg. Institut fRr donauschwgbische (eschichte und Bandeskunde, TRbingen, *22<.

Gya,y, !ihK,y: 5agyar iga+gatstRrt)neti helys)gn)vle<ikon F/?972/011 5. 7;9irat. Budapest, *24E

'nie>sa,

Ist+Kn: Crd)lyi !Rldraj+i nevei . In Bukinich Imre .ed.5 Crd)ly . '

magyar t:rt;nelmi t6rsulat. * 'ugus9tus, *2@#.

L*g,

/iktor:

(eutsche

Artsnamen des

in

[ngarn .

-elbstverlag

des

reichenberger

Tweigvereins

'llgemeinen

Ceutschen

-prachvereins,

)eichenberg .Ceutschb:hmen5, *2*4.

!arko+iq, !i,ica: :eogra!sko2@storijski @menik Kasela Vojvodine +a .eriod od /8H73 :odine do (anas3 .Ta Tiste :)ogra.hico HistoriNue des Tieu< Ha"it)s de Vojvodina ans la $)riode de.uis /8H7 jusNua nos Iours) . I9dane Pojvoanskos mu9eja, ,ovi -ad, *2EE.

!ati0e+ics, La0os - mai Iugos+lvia terUlet)re vonatko+, !Rldraj+i nevek a kR+).kori s+er"2magyar vonatko+s6 okiratok"an3 In ' $ungarol8giai Int;9et Tud8m6nyos 7:9lem;nyei, "". -96m, PII ;vfolyam, bjvid;k, *24A, m6rcius.

!ati0e+ics, La0os - mai Iugos+lvia Bcska helys)gei a+ els magyar !Rldraji le<ikon"an3 In ' $ungarol8giai 7:9lem;nyek, <@. -96m. ' =agyar ,yelv, Irodalom ;s $ungarol8giai Int;9ete Tmas9. bjvid;k, *242.

!itche,,, Br*ce )(=(: Banat and BacskaB Register o! $lace2Kames3 Recorded Romanian, Yugoslav F Ser"ian), Hungarian and :erman Kame24orms, /8HD 2 /0093 @ncor.orating -dministrative -llocations and Tatitude2Tongitude Go2 ordinates o! Settlement Gentroids . Nnpublished manuscript, *22A.

"A@

!itte,strass, Otto $eidelberg, *22".

Artsnamen"uch .Teil des $istorisch0Bandeskundlicher

'tlas von -iebenbRrgen5. 'rbeitskreis fRr -iebenbRrgische Bandeskunde,

Pena+in, O,ga: =oi%evo, =ocsityev,, egy k)tnyelv kiskR+s)g !Rldraj+i n)v k)s+lete3 In ' $ungarol8giai Int;9et Tud8m6nyos 7:9lem;nyei, "". -96m, PII ;vfolyam, bjvid;k, *24A, m6rcius.

Popp, *2@<.

r $e+er: (ie #o.onymie Sie"en"Ugens3 In \Sie"en"Urgen \, Bukarest,

Reg[nyi, A(F $cherer, A(: (onausch L"isches Artsnamen"uch !Ur die ehemals und teil eise noch deutsch "esiedelten Arte in [ngarn, Iugosla ien Fohne Slo enien) so ie West2RumLnien FBanat und Sathmar)3 $rsg. 'rbeitskreis donauschwgbischer Mamilienforscher .'7dMM, Carmstadt, *23#5.

$*ci*, Pro5( Corio,an( (icVionar istoric al TocalitZVilor din #ransilvania3 &ditura 'cademiei )epublicii -ocialiste )omVnia. BucureFti, no date.

=eid,ein, )ohann Jur :eschichte der Kamenmadjarisierungen . -onderdruck aus !ayer, %heodor 8Hrsg(@ SUdostdeutsches -rchiv 3Band . Perlag ). >ldenbourg, =Rnchen, *2E4.

History
Ear,y !odern Period
Berindei, !ihneaF /einstein, Gi,,es: TMem.ire Attoman et les .ays

Roumains, /H112/H1H . ditions de lX;cole des hautes ;tudes en -ciences -ociales 0 +aris, $arvard Nkrainian )esearch Institute, %ambridge. 1ear

"ene7an, Costin (ocumente 5edievale BZnZVene F/11D2/WH7)3 &ditura Macla, TimiFoara, *23*.

"AA

=eid,ein, )ohann

(ie ungarische 4lUchtlings"e egung in der #Urken+eit . SUdostdeutsches -rchiv, V@@3

-onderdruck aus $teinacker, Haro,d 8Hrsg(@ Band . Perlag ). >ldenbourg, =Rnchen, *2E@.

Nicoara, %oader: (ie rLumlichen Hori+onte der rumLnischen (or!+ivilisation von Sie"en"Urgen und Banat an!angs der Keu+eit F/W8D2/8DD)3 In !*se*5Ur Geschichte des 'reises 'arasch1$e;erin, Rechit>a, in /er9ind*ng -it de- Instit*t 5Ur dona*sch;V9ische Geschichte *nd Landesk*nde, %U9ingen, *nd der #ni+ersitVt Ba9es1Boyai, ',a*sen9*rg . =ulturraummittlere und untere (onau3 #raditionen und $ers.ektiven des Jusammenle"ens . )eFiOa, *22A.

Ha9s9*rg Period
Roth, E: (ie $lanmL>ig angelegen Siedlungen im (eutsch2"anater

5ilitLrgren+"e+irk, /?WH2/89/ . Buchreihe der -Rdostdeuschen $istorischen 7ommission, Band <<. Perlag ). >ldenbourg, =Rnchen, *233.

"e,dtVn>er,

Oskar: der

Iose!

@@3

[nd

die im

(onausch L"ische Batscherland /?81

-nsiedlung3 2 /?8? .

(okumentation

=olonisation

Conauschgbische 7ulturstiftung, =Rnchen, *22#.

Lot>, "riedrich: (ie ersten deutschen =olonisten der Batschka F4utak, /?D9, Keusat+, /?D0 5. -Rdostdeutsche Pierteljahresblgtter, =Rnchen, *2EE.

Lot>,

"riedrich: 2

(ie (ie

Herkun!t

der

deutschen

=olonisten

der

Stadt

#emesch"urg

Crstkolonisten,

/?/?/?/8 .

-Rdostdeutsche

Pierteljahresblgtter, =Rnchen, *2EE.

!i,,eker, "e,i. Versuch einer -nsiedlung von S.aniern im Banat .

Banater

BRcherei, Perlag der artistischen 'nstalt T & 7irchnerXs Ditwe, Drschat9, *2<4.

!i,,eker,

"e,i.

(ie Wei>kirchene :egend3 5itteilungen U"er !Un!+ehn

verschollene Artscha!tenund geschichtliche daten von el! (Rr!ern des Be+irkes3 Conauschwgbisches 'rchiv, DeiHkirchener Beitrgge <, -al9burg, *24E.

"AE

ip,ich, Hans: (ie staatsrechtliche Stellung des Banats im /83 Iahrhundert3 -Rdostdeutsche Pierteljahresblgtter, =Rnchen, *2E4.

Roth,

Erik: (ie $lanmL>ig angelegten Siedlungen

im (eutsch2Banater

5ilitLrgren+"e+irk, /?WH2/89/ . Buchreihe der -Rdostdeuschen $istorischen 7ommission, Band <<. ). >ldenbourg Perlag, =Rnchen, *233.

"int6, A*re,: Golonis_ule ha"s"urgul in Banat F/?/?2/?70)3 Mada, TimiFoara, *24*.

$>entk,Kray, )eno: S++ )v ()l25agyarors+g uja"" tRrt)net)"l F/??02tl na.jainkig)3 #ekintettel a @@@3 =roly )s 5ria #er)+ia =ora"eli el+m)nyekre3 Temesv6rott, *342.

"ata, !arta: (ie -us anderung aus WUrttem"erg ins Banat im /83 und /03 Iahrhundert3 In !*se*- 5Ur Geschichte des 'reises 'arasch1$e;erin, Rechit>a, in /er9ind*ng -it deInstit*t 5Ur dona*sch;V9ische Geschichte *nd Landesk*nde, %U9ingen, *nd der #ni+ersitVt Ba9es1 Boyai, ',a*sen9*rg . =ulturraummittlere und untere (onau3 #raditionen und $ers.ektiven des Jusammenle"ens . )eFiOa, *22A

Loric>, P[ter - "nti tRrt)net\rs tRrt)nete a @3 )s a 3 S++ad"an . In $unarol8gia 7:9lem;nyek, @A. -96m, ?II ;vfolyam, bjvid;k, *23#, december.

Inter 1 =ar
Ianc*, Gheorgge (ie .olitische Haltung der Banater Sch a"en im Iahre /0/0 . In !*se*- 5Ur Geschichte des 'reises 'arasch1$e;erin, Rechit>a, in /er9ind*ng -it de- Instit*t 5Ur dona*sch;V9ische Geschichte *nd Landesk*nde, ',a*sen9*rg . %U9ingen, *nd und der #ni+ersitVt (onau3 Ba9es1Boyai, #raditionen und =ulturraummittlere untere

$ers.ektiven des Jusammenle"ens . )eFiOa, *22A.

"A4

RQnai A(: -+ /0/829D3 )vi kR+).2eur,.ai terUletrende+)s kritikja !Rldraj+i s+em.ont",l3 llamtudom6nyi Int;9et, Budapest, *2@<.

$ch-idt1RRs,er,

Andrea:

RumLnen

nach

dem

Crsten

WeltkriegB

die

:ren++iehung in der (o"drudscha und im Banat und die 4olgero"leme . +eter Bang, &uropgische Perlag der Dissenschaften, Mrankfurt a. =., Berlin, Bern, etc. *22@.

Boia, E*gqne( RomaniaMs di.lomatic Relations $eriod, /0/0 2 /01/3

ith Yugoslavia in the @nter ar

Cso,noky, )eno: #eleki $l :r,! . In M:lraj9i 7:9lem;nyek, B?PII k:tet, p."@20 "E4. ' =agyar M:ldraj9i T6rsas6g, Budapest, *2<2.

B*rks, Richard /. # o #eleki letters. In Gournal of %entral &uropean 'ffairs, Pol. 4, ,o. *. 'pril, *2@4.

GK,, Ist+Kn: Iugos+lvia a magyar tRrt)nettud,mny"an, /09D2/01D . In ' $ungarol8giai Int;yet Tudom6nyos 7:9lem;nyei, v. vfolyam, *@. -96m. ,ovi -ad ^ bjvid;k, *24<.

23L3 1 23J4
Beer, )(: 4lucht aus dem ser"ischen Banat und aus dem Batschka . 'rticle in (ie (onausch a"en, /011 2/0W13 -Rdosteuropa0Bibliographie, Bd.A, *2EE0*24#, teil ", entry E@*.

(okumentation der Vertrei"ung der (eutschen aus Ast25itteleuro.a3 Band @@@B (as Schicksal der (eutschen in RumLnien3 Bundesministerium fRr Pertriebene, MlRchtlinge und 7riegsgeschgdigte. Bernard (raefe, Berlin, *2A4.

(okumentation der Vertrei"ung der (eutschen aus Ast25itteleuro.a3 Band @VB (as Schicksal der (eutschen in Iugosla ien3 Bundesministerium fRr Pertriebene, MlRchtlinge und 7riegsgeschgdigte. Bernard (raefe, Berlin, *2A4.

"A3

Gos>tony, Peter: (er =am.! um Buda.est, /0111H3 Perlag -chnell und -teiner 0 =Rnchen, TRrich, *2E@.

Hasse,9,att, =erner 8Hrsg@: (ie Tage` Banat2Ser"ien, RumLnien ,. In Kation und Staat, (eutsche Jeitschri!t !Ur das euro.Lische KationalitLten.ro"lem . Dilhelm BraumRller, Dien/ =gr9, *2@<.

Hasse,9,att, =erner 8Hrsg@: (ie Tage` RumLnien, [ngarn3 Keue Crge"nisse der deutschen [msiedlung3 Titeratur"erichte 2 aus Jeitschri!ten und Jeitungen P-u! dem Weg ins Reich \. In Kation und Staat, (eutsche Jeitschri!t !Ur das euro.Lische nationalitLten.ro"lem . Dilhelm BraumRller, Dien/ 'pril0=ai, *2@@.

Ho,>trVger, Hans (ie Cvakuierung der (eutschen aus Kordsie"en"Urgen, Sathmar und dem rumLnischen Banat a" Se.tem"er /011 im Tichte des =riegstage"uches der Herresgru..e SUdukraine . In \Cer Nngarndeutsche $eimat Teitung der Ceutschen as Nngarn\, Molgen *4 Gg. .*24<5 nr. "*, "", "<, "@/ *3. Gg. ,r *,".

)anko, $epp: Weg und Cnde der deutschen Volksgru..e in Iugosla ien . Beopold -tocker Perlag, (ra9 0 -tuttgart, *23".

L*-ans, /a,dis O(: HimmlerMs -u<illiaries3 #he Volksdeutsche 5itelstelle and the :erman Kational 5inorities o! Curo.e, /077 2 /01H3 The Nniversity of %arolina +ress, %hapel $ill and Bondon, *22<.

!at*ska, !Krton: Retaliation . +Rski 7iad8 7ft., Budapest, *22A.

!erk, As*>sa

- "ukovinai S+)kelyek Bcsk"a tele.Ul\t)se a+ egyh+\

!orrsok tUkr)"en F/01/2/011)3 7Rl:nlenyomat a Cun6ninnen 0 Tis96ninnen c. 7:tetbel .-9erk. B6rth G6nos5. 7ecskem;t, *22A.

Reinerth, 'ar, !(: Jur :eschichte der (eutschen in RumLnien /077 "is /01H3 'rbeitsgemeinschaft fRr -Rdostdeutsche Polks0 und $eimatforschung\ . >rt Catum

"A2

RQnai,

r( AndrKs: Crd)ly tjai )s a+ 6j hatr . In M:lraj9i 7:9lem;nyek, B?PIII

k:tet, p."<20"A#. ' =agyar M:ldraj9i T6rsas6g, Budapest, *2@#.

$a0ti, Eniko

S+)kely tele.\t)s )s nem+etis)g.olitika a Bcsk"an /01/ .

'kad;mia 7iad8, Budapest, *23@.

$a0ti, Eniko: ()lvid)k /01/2/0113 #he Southern Region, /01/ 2 /0113 #he Southern Slav $olicy o! Hungarian :overnments . 7ossuth 7:nyvkiad8, Budapest, *234.

$cherer,

Anton

8Hrsg(@:

[n"ekannte

SS2:eheim"erichte

U"er

die

Cvakuierung der SUdostdeutschen im Akto"er und im Kovem"er /011 so ie U"er die .olitische Tage in RumLnien , [ngarn, der Slo akei, im Ser"ischen Banat und im P[na"hLngigen Staat =roatienP3 (ra9 -elbstverlag, *22#.

%o-asic,

inko: Kationality $ro"lems and $artisan Yugoslavia , in Gournal of

%entral &uropean 'ffairs, Pol. E, ,umber ". Guly, *2@E.

/R,k,, Ekkehard: (eutsche und [ngarn im West2Banat F/01/2/011 5. Ceutsche in Nngarn. -Rdost0&uropa -tudien, Band @A.

In

/R,k,, Ekkehard: (er West2Banat /01/2/0113 (ie deutsche, die ungarische und andere Volksgru..en . -tudia $ungarica. -chriften des ungarischen Instituts =Rnchen. (lassl^P:lkl^$:sch .$rsg5. Trofenik, =Rnchen, *22*.

=*escht, ). *2E2.

Iugosla ien und das (ritte Reich . -eewald Perlag, -tuttgart,

Aer0a+ic, /(: #he losses o! Yugoslav .o.ulation in the Second World War . (eographical +apers 3, Nniversity of Tagreb, *22*.

Post 1 23J4
"E#

/aso+ic, !i,orad Kajnovije Kaseljavanje Grnogoraca u Kekim Backim Selima . ,ovi -ad, *2A2.

%ri5*noski,

r I+an " . A .osleratnom neseljavanju stanovni;tva i+ KR

5akedonije u tri "anatska naselja 2 Ia"uka, =aaaevo i :logonj . =atica -rpka, +osebna I9danja, ,ovi -ad, *2AE.

N*ri:,

r /,adi-ir Kajnovije naseljavanje Baake kolonistima i+ Hrvatske3

KaseljaB Baaki :ra%ac, =ljaji%evo, ono.lja, Stani;i% @ Riica3 =atica -rpka, +osebna I9danja, ,ovi -ad, *2E#.

"reiho55er, Sklaven in BZrZgan3 ,o further details known.

!arineasa, /iore, r /ighi, BZrZgan . Including maps

anie,: Ruslii MH/B 4ragmente din de.ortarea n TocalitZVi din care s2au !Zcut de.ortZri Fcele

su"linate) and TocalitZVi din BZrZgan . 'sociaOia Mo*tilor CeportaOi Wn BVrVgan. &ditura =arineasa, TimiFoara, *22@.

!isce,,any
A Pesti Hir,ap 'j6nd;ka t)ved)sei3 Budapest, *2<*. @ga+sgot 5agyarors+gnakE #rianon kegyetlen

Ga*<, Ada,9ert 'ar,: (as Woj odina2$ro"lem . -onderdruck aus Cer Conau0 )aum Teitschrift des Morschungsinstitutes fRr Mragen des Conauraumes, * Gahrgang, <^< $eft, -al9burg, *2AE.

Ho-onnay, E,e-er Wilsons $rin+i.ien in der SUdungarn24rage . Cokumente 9um Teitgeschehen, $eft ". +annonia0Perlag, Mreilassing in Bayern, *2A2.

5agyar 4igyel3 ' =agyarok Pil6gs9evets;gnek Taj;koytat8ja. IP ;vfolyam, Budapest, *22A.

"E*

#he Hungarian C<ce.tion, -pecial Issue. Balkan Dar )eport, >ctober^,ovember *2@@.

$ogromB Jeitschri!t !Ur "edrohte VRlker3 Iugosla ien +er!Lllt . (esellschaft fRr bedrohte P:lker. (:ttingen, Bern, Dien. -ept.^ >kt. *22#

%he

ona*sch;a9en

Genera, %e.ts
Anna9ring, !atthias: :eschichte der (onausch a"en, Band 93

Volksgeschichte der (onausch a"en in Iugosla ien . Perlag \-Rdost0-timmen\, -tuttgart0=:hringen, *2AA.

Anna9ring,

!atthias:

:eschichte

der

(onausch a"en,

Band

73

Volksgeschichte der (onausch a"en in RumLnien . Perlag \-Rdost0-timmen\, ,euhausen^M. bei -tuttgart0=:hringen, *2AE.

(ie (onausch a"en3 (eutsche Siedlung in SUdosteuro.aB 'usstellungskatalog. $erausgegeben vom Inneneministerium Baden0DRrttemburg. Gan Thorbecke Perlag, -igmaringen, ". 'uflage, *232. (ie (onausch a"en3 Institut fRr donauschwgbische (eschichte und

Bandeskunde .$rsg.5, TRbingen, *22<.

',(,a*9e,

!an5red:

(eutsch2BRhmische

Siedlungen

im

=ar.athenraum .

Dissenschaftliche Beitrgge 9ur (eschichte und Bandeskunde >stmitteleuropas. $erder0Institut, =arburg ^ Bahn, *23@.

NVg,er,

r %ho-as: :eschichte der (eutschen au! dem :e"iete RumLniens3

Teilver:ffentlichung in =ar.thenrundschau3 =ronstLdter Wochenschri!t , Gahrang ??I?, 7ronstadt^Bra*ov, *< Ganuar, *22E, eines provisorishen, im 'uftrag des Cemokratischen Morum der Ceutschen in )umgnien, 'ufsat9es, der dem rumgnischen Nnterrichtsministerium abgegeben wurde.

"E"

$en>, )ose5: (ie (onausch a"en und ihre Kach"arn . 'rbeitsgemeinschaft donauschwgbischer Behrer. +annonia0Perlag, Mreilassing *2A2.

$che,,, G*nther 8Hrsg(@: (eutsche :eschiche im Asten Curo.as, Band A, P(as Tand an der (onauP3 -iedler Perlag, Berlin, *22A.

$t*-pp, 'ar,: Weiter anderung deutsche =olonisten aus [ngarn FBatschka) und dem Banat nach SUdru>land in den Iahren /8D1 2 /8/W3 -Rdostdeutsches 'chiv, ?I?. ^ ??. Band, ). >ldenbourg Perlag, =Rnchen, *24E^*244.2

%a55erner,

AntonF

$ch-idt,

)ose5F

$en>,

)ose5:

(ie

deutsche

Astsiedlung der Keu+eit und die (onausch a"en im .annonischen Becken . +annonia Perlag, Mreilassing^Bayern, *2E#.

%he

ona*sch;a9en o5 Ro-ania
)*,i*s A(: :eschichte der Banater Berglanddeutschen

Ba*-annn,

Volksgru..eB ein Beitrag +ur :eschichte des #emeser Banats3 sterreichische Bandsmannschaft, &ckart0-chriften, $eft *#2. Dien, *232.

Hro-adka,

Georg:

=leine

Ghronik

des

Banater

Berglands3

Perlag

-Rdostdeutsches 7ulturwerk, =Rnchen, *22<.

%he

ona*sch;a9en o5 D*gos,a+ia

=*scht, )(: (ie Woi odina und ihr (eutschtum, .olitische :eschichte, BevRlkerungsverhLltnisse, irtscha!tlische Tage . +ublikationsst9lle Dien, *2@#.

=*escht, )(: :eschichte der (eutschen in Iugosla ien, /071 2 /011 . Im -elbstverlag des Perfassers, 7ehl a. )h., *2EE.

%he

ona*sch;a9en o5 H*ngary
"E<

=a,>, Erich: (as (eutschtum in den /09D "ei [ngarn ge"lie"enen #eilen von Batschka und Banat3 )udolf =. )ohrer Perlag. BrRnn0=Rnchen0Dien, *2@<.

ona*sch;V9ishe =orks on the Banat

ip,ich, Hans: (as Banat 2 die $orta Arientalis dreimal +erstRrt3 Beitrgge 9ur (eschichte der Tschanad0Temeschwarer Ci:cese, *#<#0*2A#. Perlag %hrist Nnterwegs, =Rnchen, *2A".

ip,ich, Hans: (as Banat als deutsche =ulturlandscha!t3 -Rdostdeutsche $eimatblgtter, =Rnchen, *2AA.

Herrscha5t, Hans

(as BanatB ein deutsches Siedlungsge"iet im (onauraum3

:eshichte, Wirtscha!t und =ultur einer deutschen Volksgru..e . Perlag (ren9e und 'usland, Berlin, *2@#. 'lso

(as BanatB ein deutsches Siedlungsge"iet in SUdosteuro.a . Perbesserte und

erggn9te 'uflage. Pg. (ren9e und 'usland (mb$, Berlin, *2@".

!U,,er1G*tten9r*nn, Ada- (eutsche =ultur"ilder aus [ngarn3 Perlag von (eorg $einrich =eyer, Beip9ig, *32E.

]Hei-at9Ucher]
A*ssch*< der Ge-einde =oi,o;it> 8Hrsg(@ Heimat"uch der :emeinde Woilo it+. 'usschuH der (emeinde Doilowit9. >rt .*23*5.

BVr, )ohannF !U,,er, Phi,ipF Re*ter, Bert 8Hrsg(@: :eschichte einer deutschen :emeinde in der Batschka3 Si at+ /?8W 2 /011 . Im 'uftrage des $eimatausschusses. +annonia0Perlag, Mreilassing, *2E<. %ritique in $teinacker, Haro,d 8Hrsg(@: -Rdostdeutsches 'rchiv, PI. Band, ). >ldenbourg Perlag, =Rnchen, *2E<.

"E@

B,*-, 'onrad: Tie"ling, /?8W 2 /07W3 Cruck der $onterus0Buchdruckerei und Perlagsanstalt der evang. Bandeskirche N.B. in )umgniein. -ibiu ^ $ermannstadt, *2<E.

"asse,, HorstF $ch-idt, )ose5: -n (onau und #hei>B Banater Tese"uch3 Bandsmannschaft der Banater -chwaben. +lace 1ear.

Hei-atsortge-einscha5t Crinnern3 *22".

',einsanktpeter1%otina

8Hrsg(@:

=leinsankt.eter2#otina /8172/0073 Cin Heimat"uch +um Tesen, Schauen und

He<, Niko,a*s: Heimat"uch der drei Sch estergemeinden Sveti2Hu"ert, Gharlevil und Soltur im Banat F/??D2/09?)3 Plg. ,ikolaus $eH, -v. $ubert. ,ach *2@@.

Hornyatschek, )ose5< (aro a 2 =ranichstLtten, :eschichte vom -u!sieg und Kiedergang einer deutschen :emeinde im Banat . $eimatsortsgemeinschaft Carowa 0 7ranichstgtten, .$rsg.5 *22*.

',*gesher>, L( Et a,(: 5ercydor!B die :eschichte einer deutschen :emeinde im Banat. $eimatsortsgemeinschaft =ercydorf .$rsg,5, .>hne >rt5 *234.

'R,,er, Henrich: $anaevoB die #oten2 und Vermi>tenliste von der deutschen BevRlkerung der (onaustadt $antscho a im Banat . Nnpublished manuscript acquired from author, -indelfingen, *22A.

Preyer, )ohann N(: 5onogra.hie der kRniglichen 4reistadt #emesvr ^ 5onogra!ia Ara/ului li"er GrZiesc #imiSoara &ditura 'marcord, TimiFoara, *22A. .neue, 9weisprachige 'usgabe5.

Rieser, Hans1Heinrich: #emes ar3 :eogra.hische Beschrei"ung der Banater Hau.tstadt. -chriftenreihe des Instituts fRr donauschwgbische (eschichte und Bandeskunde. Gan Thorbecke Perlag, -igmaringen, *22".

"EA

$en>, )ose5: -.atin und die -.atiner3 4estschri!t +ur 9DD Iahr!eier der donausch L"ischen :ro>gemeinde -traubing, *2@2. -.atin FBatschka)3 'patiner 'usschuH -chwimmbach bei .$rsg.5 im rahmen der irchlichen $ilfsstelle, =Rnchen,

#r9an, "ran>: Iahrmarkt im BanatB das (or! rings um den -:ro>en BrunnenQ3 Crinnerungen an unsere Heimatgemeinde3 $rsg. $eimatortsgemeinde Gahrmarkt .Gahr >rt5.

=ei5ert, Ladis,a*s !ichae, Herkun!t der =olonisten und Cntstehung der 5undart von Heu!eld im West"anat . In $teinacker, Haro,d 8Hrsg(@: -Rdostdeutsches 'rchiv, PI. Band, ). >ldenbourg Perlag, =Rnchen, *2E<.

Ao,,ner, Anton Tie"ling "lie" "is heuteB (urch ge esene deutsche (Rr!er des Banats . In KCer ConauschwabeL, "4 'ugust, *22A.

"EE

!igration
"r*-kin G(: $o.ulation Ghanges in Curo.e since /070 (. 'llen Nnwin. Btd., Bondon, *2AI.

)ackson, !ar+in: Ghanges in Cthnic $o.ulations o! Southeastern Curo.eB Holocaust, 5igration and -ssimilation, /01D 2 /0?D . In ,ationalitgtenprobleme in -Rdosteuropa. .&d. -ch:nfeld, )oland.5. "A. Band (egenwartskunde -Rdosteuropas, herausgegeben >ldenbourg Perlag, =Rnchen, *234. Nntersuchungen 9ur vom -Rdost0Institut, ).

'ocsis,

r(

'Kro,y:

5igrci,s

!olyamatok

vajdasg"an

msodik

vilgha"or6 6tn F5igration $rocesses in Vojvodina FYugoslavia) a!ter World War @@)3 -eparatum ex 'M:ldraj9i rtesite Tom. ???IP, ,o. @ . Budapest, *23A.

'osinski, L(A(: 5igrations o! $o.u@ation in Cast2Gentral Curo.e !rom /0702 /0HH3 0 (eographica +olonica. "., *2E@.

'osinski, L(: Ghanges in thc Cthnic Structure o! Gentral Curo.e, /07D2/0WD3 The (eographical )eview A2. <., *2E2.

'osinski, L(A( red(r: $eo.le on the moveB Studies on @nternal 5igration . International (eographical Nnion, %ommn. on +opulation (eography. Bondon, *24A.

'osinski, L(A(: @nterregional migration in Cast2Gentral Curo.e . In +eople on the =ove. -tudies on Internal =igration. B.'. 7osinski and ) =ansell +rothero .eds.5. =ethuen and %o. Btd., Bondon, *24A.

'*,ischer, E( !(: #he (is.lacement o! $o.ulation in Curo.e3 International Babour >ffice, =ontreal, *2@<. B-& =ain %oll $B<A3* IE*.

'*,ischer, E( !(: Curo.e on the 5ove3 War and $o.ulation Ghanges, /0/?2 /01?. 0 %olumbia Nniversity +ress, ,ew 1ork, *2@3.

"E4

$checht-an, )( B. Curo.ean $o.ulation #rans!ers /0702/01H3 0 >xford Nniv. +ress, ,ew 1ork, *2@E.

$checht-ann, >ctober *2@E.

)oseph

B(:

#he

Climination

o!

:erman

5inorities

in

Southeastern Curo.e .. In Gournal of %entral &uropean 'ffairs, Pol. E, ,umber ".

$checht-ann, )oseph B(: Resettlement o! #rans!erred Volksdeutsche in :ermany . In Gournal of %entral &uropean 'ffairs, Pol. 4, ,umber <. >ctober *2@4.

$checht-an, )( B(: $ost ar $o.ulation #rans!ers in Curo.e3 /01H2/0HH . 0 Nniversity of +ennsylvania +ress. +hiladelphia, *2E".

"E3

Econo-ics, Hydrography
Banater Aeit*ng( =ochen9,att 5Ur %e-esch, Arad *nd Banater

Berg,and: =reis -radB C.idemie2:e!ahr vor"eugen3 Bei :urahont sind 9D von 97 Brunnen verunreinigt3. Temeswar, *#. Ganuar, *22E.

Beschrei"ung der Banater (omLne und des :ru"en"esit+es in BRhmen3 +robably official \-T&(\ +ublication, no year given, but around *2*#. B.% Tamarski %. Cittmarsch, Dien.

Eger, %ho-as: (as regionale Cnt icklungsge!Llle in Iugosla ien . -chriften der (esamthochschule +aderborn. Merdinand -ch:ningh/ +aderborn, =Rnchen, Dien, TRrich, *23#.

G,aser, *2<2.

r( La0os < -+ -l!Rld r)gi v\+raj+a )s a tele.Ulsek . In

M:lraj9i

7:9lem;nyek, B?PII k:tet, p. "240<#4. ' =agyar M:ldraj9i T6rsas6g, Budapest,

Iano2, Ioan: (ie jUngste Cnt icklung der rumLnischen (onaustLdte3 Tokale und regionale VerknU.!ungen3 In 'arasch1$e;erin, Rechit>a, in !*se*- 5Ur Geschichte des 'reises /er9ind*ng -it deInstit*t 5Ur

dona*sch;V9ische Geschichte *nd Landesk*nde, %U9ingen, *nd der #ni+ersitVt Ba9es1Boyai, ',a*sen9*rg . =ulturraummittlere und untere (onau3 #raditionen und $ers.ektiven des Jusammenle"ens . )eFiOa, *22A.

)ordan, $onia (ie kaiserliche Wirtscha!ts.olitik im Banat im /83 Iahrhundert3 Buchreihe der -Rdostdeutschen $istorischen 7ommission, Band *4. Perlag ). >ldenbourg, =Rnchen, *2E4.

O,ar, !artin: (as Banater2:e"irge3 Cine regional2touristische -nalyse . In !*se*- 5Ur Geschichte des 'reises 'arasch1$e;erin, Rechit>a, in /er9ind*ng -it de- Instit*t 5Ur dona*sch;V9ische Geschichte *nd Landesk*nde, ',a*sen9*rg . %U9ingen, *nd und der #ni+ersitVt (onau3 Ba9es1Boyai, #raditionen und =ulturraummittlere untere

$ers.ektiven des Jusammenle"ens . )eFiOa, *22A

"E2

%e-esy, /it[> Budapest, *2<2.

r( G>o>o: - temeskR+i v\+s+a"alyo+sok a V@@@ s++ad"an

In M:lraj9i 7:9lem;nyek, B?PII k:tet, p.@2@0A#2. ' =agyar M:ldraj9i T6rsas6g,

%*rnock,

r(

a+id #he Resita @ndustrial Gom.le<B $ers.ectives in Historical

:eogra.hy . >ffprint from (eoGournal "2.*, 3<0*#". 7luwer 'cademic +ublishers, Cordrcht, Boston, Bondon, *22<.

Principa, $o5t;are #sed

Operating $yste-s
5icroso!t, Windo s 0H 2 $an2Curo.ean Cdition, 5icroso!t Windo s K 13D Workstation3

O55ice $*ites
Word $er!ect Suite ?, 8 !or Windo s . Dord +erfect %orporation, *22E 5icroso!t A!!ice 0?3 =icrosoft %orporation, *22<023

Graphics $o5t;are
%orel Craw 4, 3. %orel %orporation *233024

GI$ $o5t;are
Ctnogra!, 73D Teleki Bas9l8 'laptv6ny .-ebek, Cr. B6s9l85, Budapest, *22". 5a.vie er 93/, 73D. (olden -oftware Inc., (olden, %olorado, *22"04. =apInfo @.A =apInfo %orporation, *223. A2

)$

"4#