Between Scania and Samland : from studies of stylistic links in the Baltic Basin during the Early Migration Period Bitner-Wroblewska, Anna Fornvännen 1991:4, 225-241 http://kulturarvsdata.se/raa/fornvannen/html/1991_225 Ingår i: samla.raa.se Between Scania and Samland. From Studies of Stylistic Links in the Baltic Basin during the Early Migration Period

By Anna Bitner-Wröblewska

Bitner-Wröblewska, A. 1991. Between Scania and Samland. From Studies of Stylistic I.inks in the Baltic Basin during the Early Migration Period. Fornvännen 86, Stockholm.

The question of the contacts between Seandinavia and lhe south-eastern Baltic zone in lhe Early Migration Period has so far received scant attention in lhe literature. The present paper seeks to demonstrate the striking proximity of stylistic affmities between these areas. A case in point is the impact of the Sösdala style on the formation of the West Balt står ornamentation. In turn, the Baltic står and spade-footed fibulae decorated in this style furnished inspiration for the shaping of lhe foot in certain cruciform fibulae. This traffic in stylistic ideas, often immediatdy modified to suit local tastes and traditions is particular­ ly evident in the case of crossbow brooches with a long, narrow foot. Other interesting observations are prompt ed by the sludy of drinking horn mountings in the case of which the existing chronological gap between the source area in Gotland and lhe secondary centre in the Baltic lands did not interfere with ihe survival of stylistic traditions.

Anna Bitner-Wröblewska, Panstwowe Muzeum Archeologiczne (State Archaeological Museum), Dluga 52-Arsenal, 00-950 Warszawa, Poland.

Foreword. Remarks on chronology Balt lands—in the period between the second Before embarking on the analysis proper, the half of 4th to around mid-5th c. (Godlowski term "Early Migration Period" used in the 1970, idem 1974, idem 1980). Unlike in other title needs to be explained; even more neces­ areas, the relatively consistent horizon of arte­ sary is an attempt to synchronize different facts from the Balt lands predudes the intro­ systems of chronology (Fig. 1). The period of duction of further subdivisions of phase D. At transition between 375—the convcntional most, it is possible to set apart forms having a starting point of the Migration Period—and slightly låter chronological position, yet still 450, roughly coincident with the appearance predating materials typical of the Late Migra­ of the horizon of ranked burial cemeteries in tion Period—phase E (Bitner-Wröblewska western Europé functions in the literature as 19916). stadium I of this civilisation or "phase D" of Tischler's phase D was used by J. Reichstein the Migration Period, otherwise called the in his analysis of north and north-western, Early Migration Period (Schlette 1975, pp. European cruciform brooches (Reichstein 181-183). The term phase D was introduced 1975). Separation of his material into older, by O. Tischler (Tischler & Kemke 1902, pp. younger and låter forms prompted him to 10-13) to describe phenomena occurring in introduce a further subdivision of phase D. certain areas of central Europé—induding the U. Lund Hansen in her discussion of trade

16-91H644 Fornvännen 86 (1991) 226 A. Bitner-Wröblewska centres in Denmark during the Roman and The turbulent devdopments taking place Migration Periods also uses the term "older between the Baltic and Black seas seem to Migration Period", coincident with phase D have had little impact on the Western Balt (Lund Hansen 1988, pp. 161-162, Fig. 3). In territories which are widdy recognized by general, particularly with regard to eastern scholars as the most stable settlement area, regions, Scandinavian archaeology utilizes B. unique in Central Europé in its continued Nerman's period VI (1935) thought to have development from the Roman Period well lasted until c. mid-6th c. according to the into the Early Middle Ages (Godlowski 1989, latest chronological definitions (Nerman p. 38). The process of formation of the West 1969, idem 1975), and thus can be synchro- Balt culture circle in the territory extending nized with Central European phase D and from the Samland Peninsula and the Mazur- with the beginning of phase E. Currently, ian Lakeland to western and central Lithuania there are suggestions to redefine the chro­ was complete in the late 1st c. AD (phase B2a) nology of the Migration Period on the basis of (Okulicz 1989, pp. 69, 85-86; Nowakowski new evidence from Gotland (Näsman 1984 a, 1989a, p. 145). Significant in this process was p. 71). In Seandinavia, the system of division the role played by the strong economic and of the Migration Period is seriously affected cultural centre in the Samland Peninsula by the question of periodisation of specific which maintained far-flung commercial and ornamental styles. One tendency is to take the cultural links, and functioned as intermediary evolutionary approach (Bakka 1973, idem in the dissemination of interregional civilisa- 1977), another, to note a partial overlap of tional and stylistic novelties in the eastern Bal­ individual motifs, often coincident in a single tic zone. Samland's special position was bol- specimen (Näsman 1984 a, pp. 60-62, 70-71, stered by its rich amber deposits which had Fig. 1). even led to the establishment of close, possibly direct, contacts with the Roman Empire. The Introduction existence of such contacts is confirmed by a Starting from the younger phase of the Late mentioned in Pliny (Naturalis Historia XXXVII, Roman Period and primarily throughout the 3 /11/, 45) of a state Roman expedition for Migration Period Barbarian territory wit- amber late in Nero's reign, which reached the nessed the departure of populations headed Baltic sea, and probably Samland as well (Ko- south for the limes. Tribes long settled along lendo 1981). the limes as well as those from the distant Following a period of decline during the north (the Nordic countries and what is now younger phase of the Late Roman Period the Poland) were drawn towards the disintegrat- Samland centre once again begins to flourish ing Empire by the promise of easy plunder. during the Migration Period. Certainly, the The appearance of Huns on the Black Sea, correct view on the considcrable settlement their progress into central Europé and the stability in the Balt lands quoted in this paper establishment of a poly-ethnic state with a should neverthdess not overshadow the evi­ centre in the Carpathian Basin added to the dence of cultural changes at work in the re­ intensity of migrations in this part of the Bar- gion during the period in question. A certain baricum. One example of these displacements westward expansion of the Samland popula­ may be the Rujrii, thought to have previously tion (identified by some with the Aestii, known been settled on the southern coast of the Bal­ to Jordanes (Getica, 36) and Cassiodorus (Va­ tic, who in Attila's time put in an appearance riae, V,2)) to the lower course of the Vistula, along the middle Danube as a satelite people areas settled earlier by the Widbark Culture of the Huns (Schmidt 1934, p. 119). Popula­ people and their successors, associated with tion shifts in the opposite direction are also Jordanes's Vidivarii (Jordanes, Getica 96) pre- recorded, such as the return of the Heruli to cipitated a demographic crisis in the Peninsu­ their homeland, mentioned by Procopius of la in late phase D and early phase E, evidenced Cesarea (Bell. Goth. II, 15, 1-2). by a large number of abandoned burial

Fornvännen 86 (1991) Between Scania and Samtand 227

Pericxl VU Style II ABC 600- Style 600

Stadium IV Phase Younger D3/E1 550-- Style I "•' Late 550 Pericxl VI:2 Style 1 Stadium ni Phase E Phase E Older Style I Early 500- 500 Style 1 X£_. Phase D3

450- _Nydi Phase D Phase D 450 v Nydam N Late stadium Pericxl VI: 1 Stadium 1 >Style Sösdala \ D2 Phase D style Sösdala x 400 400- Style Phase Phase D Phase C3 Nordic O/Dl 350 350-- Stamp's I Style Nerman Bakka 1973, Näsman Lund Hansen Reichstein Godfowski author's 1935 1977 1984a 1988 1975 1970, 1974 suggestion 1980 Fig. 1. Chronological scheme.

grounds. A similar phenomenon is recorded the Galindai who returned to their homeland on the western Lithuanian coast where the after migrating south with the Goths could be decline in the number of sites dated to phase an ideal explanation for this phenomenon, it D can possibly be linked to a partial inland still awaits fuller documentation (Nowakowski migration of the population. However, the 1989*, pp. 120-123, earlier literature ibid.). area by far the most affected, was the Mazur- None of the changes in the West Balt lands ian Lakeland, settled in the Roman Period by were overly rapid in appreciably altering the the Bogaczewo Culture (Nowakowski 1991), a cultural set-up in the area. In contrast, the component of the West Balt Culture Circle neighbouring territories of the Wielbark and identified with the Galindai mentioned by Przeworsk Cultures experienced at the turn of Ptolemy (Ptol. Geogr. III, 5,8). The Bo­ the 4th and Sth c. a speedy decline in the gaczewo Culture became extinct at the turn of intensity of settlement and a breakdown of the Late Roman and Early Migration Periods. cultural development (Golowski 1985, pp. In the second half of the Sth or early 6th c. it 112-125; idem 1989, pp. 26-32). Only in was succeeded by a new culture unit, full- Pomerania was this process less violent; there, fledged from its inception—the Olsztyn the relic settlement of populations linked to Group ("masurgermanische" Kultur) (Oku- the traditions of the Roman Period lan- licz 1973, pp. 476^191). The inventory of the guished until the 520's, still maintaining stable Olsztyn Group, which is unique for this part contacts with interregional Germanic culture of Europé, features a wealth of artefacts with during the Migration Period. characteristics common to the interregional The existence of such lively relations is evi­ Germanic culture, Gepid and Ostrogoth in dent in hoards of gold coins and ornaments particular. While a tempting hypothesis that found in Pomerania, having a wide distribu­ the Olsztyn Group was formed by a splinter of tion elsewhere in the south-western part of

Fornvännen 86 (1991) 228 A. Bitner-Wröblewska

Fig. 2. Gold coin finds from 395-518 (acc. to Knapke 1941, Godlowski 1980).

Fornvännen 86 (1991) Between Scania and Samland 229

the Baltic basin (Knapke 1941; Werner 1949; Fagerlie 1967; Godlowski 1980; Herschend 1980 a, idem 19806; Kyhlberg 1980). This phenomenon of circulation and thesauriz- ation of gold in the Sth and the first half of 6th c. probably reflects processes of integra­ tion of social and political systems as well as of beliefs and customs occurring in this part of Europé. The impressive number of finds of gold coins in Pomerania, Gotland, Öland and Bornholm, and, to a lesser degree, Jutland, the Danish islands, the Scandinavian Peninsu­ la and north-eastern Germany stånds in obvi- ous contrast to their absence in the north- eastern zone of the Baltic (Fig. 2). The divid­ Fig. 3. Archaeological sites dated to the second half of lhe Sth and early 6th c. in the region around the ing line is particularly marked in the second Vistula estuary (acc. to Godlowski 1989). West Bal­ half of the Sth and early 6th c, running tic Circle: a - burial grounds, b - treasure hoard. through the regions east of lower Vistula (Fig. Widbark Culture: c - finds of solidi, d - ornament 3), i.e. the zone directly adjoining the Balt hoards, e - settlement. Aestii and the presufnably Germanic Vidivarii. Material remains of their occupation fit into widdy different categories of archaeological (Åberg 1918, Norberg 1931, Salin 1904, Fors­ evidence: of the Vidivarii—hoards of solidi re­ sander 1937, Lund Hansen 1969) is a local main—and of the Aestii—cremation burial variation of the stamp ornamentation which grounds survive in an area with a record of was widespread in Eureipe in the Early Migra­ coin finds dated to as late as the first half of tion Period; it evolved from a younger Roman the Sth c. The dividing line, undoubtedly eth­ (Period V) Scandinavian tradition supple­ nic in nature, probably also reflects differ­ mented by stylistic influences from provincial ences in customs and religious beliefs (God­ Roman belt sets. Artefacts in Sösdala style lowski 1989, pp. 34-35). Admittedly pro­ appear in the last few decades of the 4th and cesses of integration affecting the Baltic re­ become extinct before the mid Sth c. (Geiss­ gion in the Migration Period seem to have linger 1961, pp. 175-180; Bakka 1973, pp. bypassed the Balt territories and the entire 60, 85; Näsman 1984 a, p. 70, Fig. 1); they eastern zone of the Baltic sea. It would be far occur widdy in Seandinavia, where they are from correct, however, to asume that the Balt concentrated in Scania, as well as sporadically lands were in those days on the periphery. On in Schleswig-Holstein (axe-shaped harness the contrary, they functioned in dose associ- pendant from Dahmker, Geisslinger 1961, p. atiem with other regions of the Baltic coast 175, Fig. 1:1, la) and Pomerania. Several and only the plane of these relations was wide- fibulae ornamented in this style are known ly different. from the area—Trzebiatowo, Swielino and Kielpin hoards, and from the sunken hut no. Sösdala style and its Balt variation 22 at Debczyno, site 3 (Godlowski 1980, pp. Balto-Scandinavian contacts can primarily be 68-70, earlier literature ibid.). seen in the sphere of shared stylistic motifs, Stamp ornamentation influenced by the with a simultaneous fréquent and immediate, Sösdala style is seen to flourish in the West remodelling of mutually intercepted "novel- Balt Circle (Fig. 4) in the form of "western ties" to suit local predilections. A good Balt står ornament" (Åberg 1919, pp. 46-50; exemplification of this phenomenon is the im­ Godlowski 1970, pp. 97-98; idem 1974, p. pact of the Sösdala style on the emergence of 90; Madyda 1977, p. 387), which obviously the West Balt står ornament. The former shows a number of local traits somewhat di-

Fomvännen 86 (1991) 230 A. Bitner-Wröblewska

Fig. 4. Sites with artefacts decorated in West Baltic står ornament (a). Selected sites with artefacts ornamented in the Sösdala style (b).

Fornvännen 86 (1991) Between Scania and Samland 231

verging from the Scandinavian standard. One Neman. Isolated finds are known from the of them is the choise of items decorated by Mazurian Lakeland and central Lithuania and stamping: in Seandinavia this category in­ from the zone of mixed Balto-Livonian settle­ cludes elements of ccremonial horse trap- ment at the mouth of the Daugava river, as pings, strap mounts and, less frequently, sheet well as from southern Finland. The West Balt silver brooches with a rectangular head and horizon of står ornamented artefacts is an rhomboid foot. (Cf. contents of Sösdala indicator of phase D (Godlowski 1970, pp. hoard, N. Mellby parish in the collection of 97-98) and synchronic with the Sösdala style. Lunds Universitets Historiska Museum, inv. no. LUHM 25570; Salin 1904, Figs. 105-106; Står- and spade-like footed brooches Lund Hansen 1969, Figs. 1-2.) Balt inven­ (Slem-, Schaufelfussfibeln) and variants tories feature exclusivdy elements of human of related cruciform brooches attire such as brooches and belt sets. Further- The står- and spade-footed brooches belong more, stamping is the sole decorative tech­ in the category of stamp ornamented arte­ nique in evidence (the belt mounting from facts. With a crossbow construction, type former Greibau, grave 272 is an exception, Almgren VI. 2, their foot features a semicir- having additional transverse notches Åberg cular plate with a serrated rim or row of per- 1919, p. 46) in a Balt environment while in forations or points along the edge. A recent Seandinavia, it co-exists with other, less com­ dassification identifies seven types of this mon techniques, such as niello, notching, and brooch depending on the shape of the foot silver inlay. Stylized animal imagery frequent­ (Bitner-Wröblewska 1991 a). Basically, it is a ly employed in the latter environment is not in chronological indicator of phase D, its par­ evidence in Balt lands of the time. ticular variants appearing in its different The above differences tend to be local vari­ stages, the latest being dated to early phase E. ations of two styles which are closely related in The står- or spade-footed brooch originated terms of stamp ornamentation motifs (Salin in Samland and surrounding areas whence it 1904, Figs. 358-361; Åberg 1919, Figs. spread to other Balt territories as well as to 43-44). West Balt står-, spade- or trapeze Finnish lands (Fig. 5). In the course of its footed crossbow fibulae, tongue-shaped strap progress along the eastern coast of the Baltic ends and ferrules of oval buckles with a thick- the motif underwent local modifications, giv­ ened bow display representations of stars, ing rise to local variations such as type III, multiconcentric circles and semicirdes as well almost exclusivdy characteristic of the west­ as triangles filled with a combination of ern Lithuanian coast and the region along the various decorative elements, typical of the lower Neman, or types VI and VII produced Sösdala horizon (Tischler 8c Kemke 1902, in areas of Finnish settlement, notably in Esto­ Tabs. IV, XI). Stamp ornamentation fre­ nia and south-western Finland. quently appears on silver sheet plaques used This group of brooches is also known from to inlay entire items or portions thereof, as in Öland and Gotland (Hackman 1905, p. 161, tongue-shaped belt fittings, fibulae feet, me- footnote 2; Åberg 1919, p. 42; Statens ... topes on the uppermost part of the fibula bow 1908, pp. 230-231, Fig. 75). Four specimens or on its head. represent type II (classical—Algutsrum parish; The horizon of objects ornamented with a Lundegård, Köpingsvik parish; Lenstad, stamped står motif is primarily in evidence in Torslunda parish; Gårdby parish); one type the Samland Peninsula where several score VII (with a spade-like foot—Övre Ålebäck, such finds (Fig. 4) were registered; from this Gårdby parish); and one is a local variant of area the tradition of stamping certain types of type IIIA (with a markedly extended fan-like fibulae and belt sets spread to other Balt foot—a stray find from Gotland). Stylistic lands. Numerous specimens are known from analysis of these brooches coupled with the the region to the south of Samland, as well as study of the territorial distribution of types the western Lithuanian coast and the lower (Fig. 5) offers insight into the sources of ori-

Fomvännen 86 (1991) 232 A. Bitner-Wröblewska

Fornvännen 86 (1991) Between Scania and Samland 233 gin and routes of transmission of the pattern tant from the source of stylistic inspiration (Bitner-Wröblewska 1991a, 19916). Thus, (Fig. 6). The direction of its transmission re­ type II specimens from Öland appear to be mains an open question. It may well be that a direct imports from the source in the Samland land route existed, linking southern Finland Peninsula or perhaps the centre in western with Norrland and the western Norwegian Lithuania. Similarly, a damaged specimen coast. Spade-footed cruciform brooches from presumably belonging to type VII may have Norway showing dose affmities to F"innish been impeirted from an area of Finnish settle­ type VII seen to confirm this line of argument ment. A type III specimen from Gotland is an (Moora 1938, Tab. IX, 11; Hackman 1905, interesting example of external influences Fig. 123, Tabs. 3: 6; 4: 1). Alternatively, the from the lower Neman area or the western Baltic islands with participation of Scania Lithuanian coast incorporated into local tra­ could have been instrumental as a sui generis ditions. contact zone, which is evidenced by the pres­ While the fashion for crossbow fibulae with ence of forms typical of beith the western and a står- or spade-like foot did not spread in the the eastern Baltic basin, in the archaeological Scandinavian environment, this brooch in- material from Gotland, Öland and Beirnbolm. spired the shaping of the foot of certain cruci­ form specimens (Fig. 6), namely, the spade- Crossbow brooches with a long, narrow foot footed types S0ndre Gammelsrod, Ådland, (Armbrustfibeln mit Nadelscheide oder Sagland and Feering, as well as the so-called mit kurzem Nadelhalter) isolated forms (Einzelformen)—several fibulae Such a category of artefacts linking the east­ defying dassification (Reichstein 1975, pp. ern Baltic zone with Gotland, Öland and 37-39, 44, Tabs. 21: 1, 3-4; 22: 5; 32: 3; 41: Bornholm are crossbow broeiches with a full 7, 9; 107: 3, 7). The study of various morpho­ catch-plate, type A. VI, 2 and a long, narrow logical features demonstrates that the spade­ foot (Näsman 19846, map 12 a). They fall into like foot appears emly in late forms of cruci­ two groups: specimens with the foot visibly form brooches. Reichstein dates all the extended beyond the length of a short catch- quoted types of cruciform fibulae to his phase plate (of type: Nerman 1935, Fig. 367;

D3, the youngest specimens even as late as Tischler & Kemke 1902, Tab. V, 6), and those

D3/E| (cf. Fig. 1). In absolute chronology this featuring a foot flush with the catch-plate (of corresponds to the second half of the Sth and type: Nerman 1935, Fig. 368; Tischler & the early 6th cc, and the youngest Anglei- Kemke 1902, Tab. V, 21). A slightly different Saxon variants may be as late as the mid 6th c. local group is represented by fibulae from (Reichstein 1975, pp. 70, 107-109). Stylistic Bornholm, featuring a knob on the head pro- similarities between står- eir spade-footed jecting beyond the spring (type Klindt-Jensen brooches and Scandinavian cruciform speci­ 1957, Fig. 87:11). A number of specimens mens were observed in the past (inter alios also show varying degrees of similarity te> Rau- Åberg 1919, pp. 31-44) occasionally prompt- penfibeln (brooches with a Caterpillar bow) ing erroneeius conclusions on the formation (Tuszyriska 1988, pp. 177-187). of Balt fibulae under the impact of Scandi­ A recent study of crossbow fibulae viewed navian forms (Hackman 1905, pp. 159-162). from within the two varieties quoted above The clearly older chronological position of (the short and the long catch-plate) reveals a Balt and Finnish brooches obviously contra- number of interesting regularities in terms of elic ts such an assumption. both spatial and chronological distribution Finds of spade-footed cruciform brooches (Bitner-Wröblewska 19916, references ibi­ are concentrated mainly in southern and dem). The first variant occurs in cemcentra- south-western Norway, a region relatively dis­ tions primarily in Samland and Vistula Leman

Fig. 5. Distribution of siar-and spade-footed fibulae: a-type II, A-type III, c-type VII, d- types I, I V-VI.

Fornvännen 86 (1991) 234 A. Bitner-Wröblewska

Fig. 6. Distribution of crossbow fibulae with a står- or spade-like foot (a) and of cruciform specimens with a spade-like foot (b - acc. to Reichstein 1975).

Fornvännen 86 (1991) Between Scania and Samland 235

Fig. 7. Distribution of crossbow fibulae with a long, narrow foot: 1 a -with a short catch-plate (variant 1), b - with a long catch-plate (variant 2), 2 - Bomholm type.

and the Mazurian Lakeland (Fig. 7). A distinct 2. There, variant 1 is, with one exception, concentration is manifest in western Lithua­ lacking. nia and Gotland, the latter being much small­ Chronological relations obtaining between er. Finally, a number of single, dispersed the variants 1 and 2 of crossbow fibulae when specimens are known from outside the area mapped provide a slightly different spatial dis­ mentioned. In contrast, the spatial distribu­ tribution of the respective finds. Samland and tion of variant 2 shows other areas of concen­ Mazurian specimens of both variants are the tration (Fig. 7), i.e. western and central Lith­ oldest and most chronologically compact. uania, Öland and Gotland; the Mazurian They occur in assemblages with artefacts from Lakeland and the area around the Vistula Le- phase D and can thus be synchronized with man contain fewer finds. Northern Estonia the Sösdala stamp ornamentation horizon. and south-west Finland report finds of variant This would confine their chronology to the

Fornvännen 86 (1991) 236 A. Bitner-Wröblewska mid Sth c. Both variants of crossbow brooches by the construction of rung brooches (Arm- from other Balt and Finnish lands occur with­ brustsprossenfibeln) dated to phase E (Åberg in a much wider chronological framework. 1919, Figs. 181-182). Specimens from Got­ The youngest of them are found in assem­ land and Öland tend to have pronounced blages dated to phase E (and in absolute chro­ elongation and occasionally, tapering of the nology cover the second half of the Sth c). It foot (Nerman 1935, Figs. 371, 375). Another is noteworthy that while the chronological po­ local feature there is the presence of a double sitions of the two variants partially overlap, cord and of stamp ornament, also present on variant 2 clearly persists in use longer. As far the variant 1 of crossbow fibulae (Nerman as brooches from Öland, Gotland and Born­ 1935, Figs. 374-375). Some of these charac­ holm are concemed, it is more difficult to teristics, such as stamp ornament or pro­ draw the line between the chronologies of the nounced elongation of the foot, occur spo- variants. They appear to be contemporaneous radically in Balt and Finnish material—clear and all fit within Period VI: 2, which started, evidence of Scandinavian references and pos­ according to the latest evidence around the sibly of regular contacts between production mid Sth c. centres. The demonstrably earlier chronological po­ sition of brooches from Balt territories sug­ Drinking horn mountings gests that the stylistic initiative had sprung This category of artefacts testifies to more from this area, namely Samland, very rapidly than just the enduring nature of contacts be­ spreading ovcr the eastern Baltic zone, also tween production centres. A case in point may reaching the Islands of Gotland, Öland and be drinking horn terminals of type D.S and Bornholm, often, as has already been said, rim fittings of type k.6, in the dassification by modified to suit local tastes and tradition. J. Andrzejowski (1991, ibid. bibliography). (Isolated specimens also reached Scania from The former have two plates and an interven- the Baltic islands—e.g. finds in XIX c. collec­ ing bi-conical thickened junction; the latter tion of antiquities at Lunds Universitets His­ are made of wide richly embossed band fit­ toriska Museum, inv. no. LUHM 3614, tings. Both occur in two concentrations—one LUHM 12771.) The variant from Bornholm in Gotland, the other in the Balt lands (Fig. 8). could be a case in point. Morphological analy­ While type D.S has a clearly earlier dating, all sis of variant 1 reveals striking stylistic similar­ of the mountings dated from the Late Roman ities between Scandinavian and Balt speci­ Period (period V), most of them to its earlier mens: the proportions are identical or very phase (inter alia Skällhoms, gr. 3; Havor, gr. similar, although local features are also in evi­ 95, 223 a, 2236). With few exceptions (Lub- dence, such as stamp ornamentation on the lino, gr. 26; Gibaiciai), Balt specimens cannot bow and foot of Scandinavian brooches be dated prior to phase D and E (Period VI (Tischler & Kemke 1902, Tab. V, 6, 8-10; and part of VII). The appearance of drinking Nerman, 1935, Figs. 367, 373). Specimens horns (with mountings) in the eastern Baltic belonging to variant 2 from Balt environ- zone should doubtless be linked to influences ments show greater diversity, e.g.: Samland from Gotland which in the local environment and Mazurian brooches are markedly more were transformed and enriched. This is best slender and lighter than their Lithuanian seen in the case of upper rim mountings. counterparts (Åberg 1919, Figs. 52, 57-59). Specimens found in Geitland are exclusivdy In addition, certain of the latter feature large made from bronze and rather modestly deco­ trapeze-shaped metopes at their top, occa­ rated (Almgren & Nerman 1923, pp. 96-99, sionally in combination with a transverse pro­ nos. 297, 304-305, 324). Balt mountings on jection at the base of the bow, at the peiint the other hand, particularly specimens from where it becomes the foot (Åberg 1919, Figs. Lithuania, were partly or wheilly made of sil­ 61-63)—stylistic features clearly indicative of ver and richly ornamented (Kazakevicius a låter chronological position as is confirmed 1987). The custom of drinking from such ves-

Fomvännen 86 (1991) Between Scania and Samland 237

Fig. 8. Distribution of drinking horn mountings (acc. to Andrzejowski 1991): a terminal mounting lype D.S; b - upper rim type k.6.

sels reached the Balt lands from Gotland. The drinking horns. Both these finds are dated to chronological gap between the source area in the earlier phase of the Late Roman Period. Gotland and the secondary centre of finds of This sheds new light on the nature of inter- drinking horn mountings in the Balt lands did change as well as on the capacity for storage not affect the survival of stylistic traditions. of foreign models and their latter assimilation This gap is visible in particular in the case of in the local environment. band upper rim mountings, and filled by items conforming to the style but entirely din- Conclusions stinct in term of function, the so-called this Studies of links between the West Balt terri­ from Babieta and Mojtyny (Hollack & Peiser tories with other lands of the Baltic sea cannot 1904, Tab. VII, 416; Peiser 1921, Fig. 1), give too much attention to the question of manufaclured from embossed decorated Balto-Scandinavian contacts. To be sure, B. sheet metal with ornamentation meitifs similar Nerman explored the subject in the past to those present on upper rim mountings of (1929), but, being primarily interested in ma-

Fomvännen 86 (1991) 238 A. Bitner-Wröblewska terials from phase E and, even to a greater connections discussed in the present paper extent, the Early Middle Ages, he largely ig- obviously do not exhaust the subject. The nored the interchanges between Scandinavian question of the influence of trapeze-footed and south-eastern lands which occurred in the brooches (Tischler & Kemke 1902, Tab. IV, Early Migration Period. That the problem is 12-16, 18-20, 22-24) on similar specimens indeed significant becomes evident when one from Gotland (Nerman 1935, Figs. 47-48, begins to assemble and analyse facts hereto- Textfig. 6) and Öland (Åberg 1923, Fig. 215) fore overlooked or scattered in the literature. still remains to be examined. The same is true It is then obvious that the question is essential of isolated finds of other Balt imports in Sean­ to the understanding of recent relations be­ dinavia—such as certain types of brooches, tween areas bordering on the Baltic. Of elements of belt sets, bracelets, etc. E.g. the course, phase D cannot be understood as a unique finds of fragments of rhomboid-foot- chronological watershed in Balto-Scandina- ed fibulae from Gotland (Nerman 1935, Text­ vian contacts. Many issues addressed by the fig. 7) and Scania (Strömberg 1961, Tab. present paper have a much larger chronologi­ 53: 3) which have no counterparts among the cal range (being rooted in phase D they persist local fibulae types, yet are closely akin to the in the subsequent phase). The emphasis on Baltic materials (Tischler & Kemke 1902, Tab. the Early Migration Period is intended to IV, 11). Similar sources need to be found for draw the attention of scholars to the earlier belt set components from Gotland (Nerman origin of certain phenomena as well as to their 1935, Figs. 484-485, Textfig. 186) and Scania scale in the period under discussion. (Strömberg 1961, Tab. 52:8). Oval buckles It becomes apparent that there is a striking with a thickened frame and a rectangular fer- stylistic similarity between, on the one hand, rule as well as tonjrue-shaped strap-ends re­ the south-east regions of the Baltic basin and, present a cultural and ethnic indicator of on the other, Gotland, Öland, Bornholm and Sambian costume in phase D, while the bow- Scania, well illustrated by the impact of the curved strap end from Scania also has close Sösdala style on the emergence of the West analogies with Samland material (Bitner- Balt står ornament. In turn, står or spade- Wröblewska 1989, pp. 165-175, eadem footed brooches decorated with this orna­ 19916). Analogies for bracelets with widen- ment inspired the shaping of the foot of some ing, thickened extremities discovered in Got­ cruciform fibulae. This exchange of stylistic land (Nerman 1935, Figs. 419-420) are to be novdties, often speedily modified to suit local found on the western Lithuanian coast (Åberg taste and habit, is particularly well exempli- 1919, Figs. 184-193). However, the author fied by crossbow brooches with a long, narrow feels that a complete study of all elements of foot. In this case the stylistic initiative can exchange would have unnecessarily swelled probably be traced to the Samland Peninsula, the volume of the paper without contributing the rapidly spreading motif finding its way to significantly to the conclusions. eastern reaches of the Baltic basin as well as to Similarity of stylistic ideas testifies to the Gotland, Öland and Bornholm. Furthermore, existence of a direct, lively interchange be­ it is also worth noting that individual produc­ tween West Balt workshops and their counter­ tion centres seem to have been in regular parts in Gotland, Öland and Scania. Such con­ touch (as confirmed by morphological study tacts between Seandinavia and often quite re- of the specimens). Finally, a study of drinking mote areas to the south during the Migration horn mountings offers an interesting insight Period are evidenced by inter alia the männer into the travel of motifs. The chronological of expansion of Animal Style I (Haseloff gap between the source area in Gotland and 1981, Fig. 512) or of the bird head frieze the secondary centre of finds of drinking horn motif on fibulae and buckles (Näsman 1984 a, mountings in the Balt lands did not affect the pp. 71-73). The spread of stylistic ideas and survival of stylistic traditions. of specimens was acederated by migrations. The cases of Balto-Scandinavian stylistic Certainly, itinerant artisans whose presence is

Fornvännen 86 (1991) Between Scania and Samland 239

discemible in archaeological material from Forssander, J. E. 1937. Provinzialrömisches und the Sth and låter c, must have participated in germanisches. Stilstudien zu den schonischen the transmission of novd patterns, ornaments Funde von Sösdala und Sjörup. Meddelanden från Lunds universitets historiska museum, pp. and decorative motifs (Werner 1962, idem 11-100. 1970). Intermarriage could have been an­ Geisslinger, H. 1961. Fruhvölkerwanderungszeit- other factor (Näsman 19846, p. 122). Last but liches Zaumzeugzubehör von Dahmker, Kreis not least was the exchange between economic Herzogtum Lauenburg. Offa 17/18, pp. and cultural centres. One of the partners of 175-180. Samland could have been the centre in Godlowski, K. 1970, The Chronology of the Late Roman and Early Migration Periods in Central Scania, recently remarked by Ch. Fabech Europé. Zeszyty Naukowe Uniwersytetu Jagiellons- (1990), functioning alongside Öland, Gotland kiego. Prace Archelogiczne 11. Kraköw. and Bornholm, in the lively contacts develop­ — 1974. Chronologia okresu pöznorzymskiego i ing between the lands bordering on the Baltic wczesnego okresu wedrowek ludöw w Polsce pölnocno-wschodniej (Die Chronologie der jiin­ geren Kaiserzeit und der friihen Völkerwande­ rungszeit in Nord-Ost Polen). Rocznik Biatostocki Figs. 2-8 drawn by Monika Bajkowska XII, pp. 9-109. — 1980. Zur Frage der völkerwanderungszeitli- References chen Besiedlung in Pommern. Studien zu Sach­ Almgren, O. & Nerman, B. 1923. Die ältere Eisenzeit senforschung 2, pp. 63-106. Gotlands II. Stockholm. — 1985. Przemiany kutturowe i osadnicze w pofud- Andrzejowski, J. 1991. Okucia rogöw do picia z niowej i srodkowej Polsce w mlodszym okresie przedr- mlodszego okresu przedrzymskiego i okresu zymskim i w okresie rzymskim (Kultur- und Besied- wplywöw rzymskich w Europie srodkowej i pöl- lungsveränderungen in Siid- und Mittelpolen nocnej. Pröba klasyfikacji i analizy chronolo- während der jungeren vorrömischen Eisenzeit giczno-terytorialnej. (Drinking-horn Mountings und römischen Kaiserzeit). Wrodaw-Warsza- from Late Pre-Roman and Roman Periods wa-Kraköw-Gdarisk-Lodz. Found in Central and Northern Europé. An At­ — 1989. Ziemie polskie w okresie wedrowek ludöw. tempt of Classificalion and Analyse of Chronolo­ Problem pierwotnych siedzib Slowian (Poland in gy and Distribution.) Materialy Starozytne i Wczes­ the Migration Period). Barbaricum, pp. 12-63. nosredniowieczne VI, in press. Hackman, A. 1905. Die ältere Eisenzeit in Finland. Bakka, E. 1973. Goldbrakteaten in norwegischen Helsingfors. Grabfunden. Datierungsfragen. Fruhmittelalter­ Haseloff, G. 1981. Die germanische Tierornamen­ liche Studien 7, pp. 53-87. tik der Völkerwanderungszeit. Studien zu Salins — 1977. Stufengliederung der nordischen Völker­ Stil I. Vorgeschichtliche Forschungen 17. Berlin. wanderungszeit und Ankniipfungen an die kon­ Herschend, F. 1980 a. Tvä studier i öländska guld­ tinentale Chronologie. Archäologische Beiträge fynd. I. Det myntade guldet. Tor 18, 1978-1979 zur Chronologie der Völkerwanderungszeit. An­ (1980), pp. 33-194. tiquitas. Reihe 3, 20, pp. 57-60. Bonn. — 1980i. Två studier i öländska guldfynd. II. Det Bitner-Wröblewska, A. 1989. Elementy baltyjskie w omyntade guldet. Tori8, 1978-1979 (1980), pp. kulturze wielbarskiej (Baltische Elemente in der 195-294. Widbark-Kultur). Kultura wielbarska w mlodszym Hollack, E. 8c Peiser, F. 1904. Das Gräberfeld von okresie rzymskim, II, pp. 161-177. Lublin. Moythienen. Königsberg. — 1991 o. Zapinki z gwiazdzistq i lopatkowata Kazakevicius, V. 1987. Motifs of Animal Decorative nözka z poludniowo-wschodnich wybrzezy Bal- Pattern on Bindings of the 5th-6th Centuries tyku (Står- and spade-like footed Brooches from Drinking Horns from Plinkaigalis burial ground, south-eastern Baltic Zone). Wiadomosci Archeolo­ (Lithuania). Finskt Museum, pp. 45-63. giczne 51, in press. Klindt-Jensen, O. 1957. Bornholm i folkevandrings­ — 1991 6. The Southeastern Baltic Zone and Sean­ tiden. Köbenhavn. dinavia in the Early Migration Period. Knapke, W. 1941. Aurei- und Solidi-Vorkommen "Swiatowit" 37, in press. an der Siidkiiste der Ostsee. Acta Archaeologica Fabech, Ch. 1990. Sjörup—an Old Problem in a 12, pp. 79-118. New Light. Meddelanden från Lunds universitets Kolendo, J. 1981. A ta recherche de V ambre baltique. historiska museum 1989—1990. New Series 8, pp. L 'expedition d 'un chevalier romain sou Néron. War­ 101-119. szawa. Fagerlie, J. M. 1967. Late Roman and Byzantine Kyhlberg, O. 1980. Vikt och värde. Stockholm stud. Solidi Found in Sweden and Denmark. Numis­ arch. 1. Stockholm. matic Notes and Monographs 157. New York. Lund Hansen, U. 1969. Kvarmkäsefundet—en ana-

Fomvännen 86 (1991) 240 A. Bitner-Wröblewska

lyse af Sösdalastilen og dens forudsaHninger. baltyjskich w polowie pienvszego tysiaclecia nas- Aarb0ger, pp. 63-102. zej ery (On the archaeological Identification of — 1988. Handdszentren der römischen Kaiserzeit the Balts in the mid of the First Millennium und Völkerwanderungszeit in Dänemark. In: A.C). Barbaricum, pp. 64-100. Trade and Exchange in Prehistory. Acta Archaeo­ Peiser, F. 1921. Die Trinkhorn-Ränder des Prussia- logica Lundensia 8:16, pp. 155-166. Museums. In: Festschrift Adalbert Bezzenberger, Madyda, R. 1977. Sprzaczki i okucia pasa na zie- pp. 114-120. Göttingen. miach polskich w okresie rzymskim (Bell Buckles Reichstein, J. 1975. Die kreuzförmige Fibel. Zur and Mountings on the Polish Territories in the Chronologie der späten Kaiserzeit und der Völker­ Roman Period). Materialy Starozytne i Wczesnos­ wanderungszeit in Skandinavien, auf dem Kontinent redniowieczne IV, pp. 351—412. und in England. Offa-Biicher 34. Neumunster. Moora, H. 1938. Die Eisenzeit in Lettland bis etwa Salin, B. 1904. Die Altgermanische Thierornamentik. 500 n. Ch. 2. Tartu. Stockholm, Neue Aufl. 1935. Nerman, B. 1929. Die Verbindungen zwischen Skan­ Schlette, F. 1975. Römische Kaiserzeit und Völker­ dinavien und dem Ostbalticum in der jungeren Ei­ wanderungszeit. In: Wege zur Datierung und senzeit. Slockholm. Chronologie der Urgeschichte, pp. 171-185. Ber­ — 1935. Die Völkerwanderungszeit Gotlands. Stock­ lin. holm. Schmidt, L. 1934. Geschichte der deutschen Stamme bis — 1969. Die Vendelzeit Gotlands. Tafeln. Slockholm. zum Anfang der Völkerwanderung. Die Ostger­ — 1975. (herausg. von A. Lundström). Die Vendel­ manen. Miinchen. zeit Gotlands. I, 1, Stockholm. Statens ... Statens Historiska Museum och K. Mynt­ Norberg, R. 1931. Moor- und Depotfunde aus dem kabinettet. Tillväxten under år 1908. Fcrmvånnen 5. Jahrhundert nach Chr. in Schonen. Acta Ar­ 3, 1908, pp. 201-311. chaeologica II, pp. 104-111. Strömberg, M. 1961. Untersuchungen zur jiin­ Nowakowski, W. 1989 a. Kultura wielbarska a za- geren Eisenzeit in Schonen. Völkerwanderungs- chodniobaltyjski krag kulturowy (Wielbark-Kul- zdt-Wikingerzeit. I, II. Acta Archaeologica Lun­ tur und Westbaltischer Kulturkreis). In: Kultura densia, 4:4. wielbarska w mlodszym okresie rzymskim, II, pp. Tischler, O. & Kemke, H. 1902. Ostpreussische Alter­ 143-159, Lublin. tiimer aus der Zeit der grossen Gräberfelder nach — 19896. Studia nad ceramika zachodniobaltyjska Christi Geburt. Königsberg. z okresu wedreVwek ludöw. Problem tzw. pu- Tuszyriska, M. 1988. O zapinkach z gasienicowatym charöw na puslych nözkach (Studies on the West kablakiem w obrebie kultury widbarskiej (Uber Balts Pottery from the Migration Period. Prob­ die raupenförmigen Biigdfibdn im Bereich der lem of the Hollow Pedestafs Beakers). Barbari- Widbark-Kultur). Kultura wielbarska w mlodszym cum, pp. 101-147. okresie rzymskim. I, pp. 177-187, Lublin. — 1991. Kultura bogaczewska na Pojezicrzu Ma- Werner, J. 1949. Zu den auf Öland und Gotland zurskim od schylku pöznego okresu przedrzyms- gefundenen byzantinischen Goldmiinzen. kiego do starszej fazy pöznego okresu wplywöw Fornvännen 44, pp. 257-286. rzymskich. Pröba analizy chronologiczno-kul- — 1962. Fernhandel und Naturalwirtschaft im turowej. (The Bogaczewo Culture at the Masuri- östlichen Merowingerreich nach archäologi­ an Lake Region from Late Pre-Roman Period to schen und numismatischen Zeugnissen. Bericht older Phase of Late Roman Period. An Attempt der Römisch-Germanischen Kommission 42, 1961 of chronological and cultural Analyse.) Materialy (1962), pp. 307-346. Starozytne i Wczesnosredniowieczne VII, in press. — 1970. Zur Verbreitung fruhgeschichtlicher Me- Näsman, U. 1984 a. Zwei Relieffibdn von der Insel tallarbeiten (Werkstatt-Wanderhanlwerk-Han- Öland. Praehistorische Zeitschrift 59, pp. 48-80. dd-Familienverbindungen). Antikvariskt Arkiv — 1984 b. Glas och handel i senromersk tid och folk­ 38 ( = Early Medieval Studies 1), pp. 65-81. vandringstid. En studie kring glas från Eketorp-ll, Åberg, N. 1918. De germanska sljärnornamenten Öland, Sverige. Aun 5. Uppsala. under 3- och 400-talet e. Kr. Antikvarisk Tidskrift Okulicz, J. 1973. Pradzieje ziem pruskich od pöznego för Sverige 21, 3, pp. 1-51. paleolitu do VII w. n. e. (History of the Ancient — 1919. Oslpreussen in der Völkerwanderungszeit. Prussian Territory). Wrodaw—Warszawa-Kra- Uppsala. köw-Gdarisk. — 1923. Kalmar läns förhistoria. Uppsala. — 1989. Pröba identyfikacji archeologicznej ludöw

Fornvännen 86 (1991) Between Scania and Samland 241

Sammanfattning

Uppsatsen behandlar under rubriken "Från utbyte av både varor och idéer. Sådana kem- Skåne till Samland" stilhistoriska förbindelser takter går lätt att spåra på det stilistiska områ­ över Östersjön under tidig folkvandringstid. det då det gäller ornerade föremål som t. ex. Redan då och långt tidigare hade det nätverk dräktspännen och bältesbeslag, även om de av handelsförbindelser börjat byggas upp, nya intrycken snabbt anpassats till vad lokal som under vikingatiden så påfallande band smak och tradition krävde. Som ett påtagligt samman stammar och stater i södra Östersjö­ exempel på denna typ av interaktion kan man området. Dessa förbindelser bröts inte ens se den skandinaviska Sösdalastilens inflytande under den oroliga folkvandringstiden då kri­ på den västbaltiska stjärnornamentiken, som giska horder var i rörelse för att inte missa gav ett återflöde genom den inspiration som chansen till lättfånget byte i det sammanfal­ baltiska fibulor, ornerade i denna stil, gav på lande romarriket. Tvärtom kan man konsta­ formgivningen av vissa typer av skandinaviska tera de då t.ex. påfallande stilistiska över­ korsformiga spännen. Uppmärksamhet har ensstämmelser inom keinstbantverket mellan å också ägnats den skandinaviska sedvänjan att ena sidan områdena i sydöst och å andra sidan dricka ur metallbeslagna horn. Detta bruk Gotland, Öland, Bornholm och Skåne. De et­ förmedlades till Baltikum från Gotland. På niska skillnaderna till trots synes kontakterna östbaltiskt område kom den stilistiska tradi­ mellan dåtidens skandinaver och balter haft tionen att fortleva trots att ett visst kronolo­ en fastare förankring än vad fallet varit under giskt glapp synes föreligga. Studien visar bur senare delen av 1900-talet. intresserad man i alla tider varit grannar emel­ Från sent 300-tal och genom 400-talet före­ lan att från motparten tillägna sig nya idéer kom inom södra Östersjöområdet ett livligt och följa modet.

17-91Hf>44 Fornvännen 86 (1991)