Samland belts are a form present during the oldest phase of Dollkeim-Kovrovo Culture of the Early Roman Period (H. Jankuhn 1933a; 1933b, p. 220–225; 1950, p. 55–56; W. Nowakowski 1996, p. 48–49, pl. 107). The belt was a broad strap with a solid clasp or buckle with an openwork attachment at one. At the other there usually was a rectangular openwork mount, and, along the edges of the belt, a row of smaller mounts with a curving body and narrow, profiled or trapeze-shaped terminals. Some belt sets also include tags for suspending an implement (Pl. 28). All these belt mounts were in bronze. The range of occurrence of Samland belt sets is limited to the territory occupied by Dollkeim-Kovrovo Culture – almost all these finds were recorded in Samland Peninsula (Pl. 1; cf. H. Jankuhn 1933a, fig. 15; 1950, fig. 3). The majority evidently belong in the mature phase B2, but some may date back even to phase B2a, a and a small number of specimens originate from phase B2/C1–C1a. One of the youngest belt sets, discovered at Tûlenino, raj. Polessk (Тюленино, рай. Полесск, former Viehof, Kr. Labiau), in grave 170 (Pl. 34c.d), was accompanied by two A V, series 1 fibulae (Pl. 35a; A. Bezzenberger 1895, p. 257; H. Jankuhn 1933a, pl. III:4; 1933b, p. 225). During phases B2/C1–C1a and C in Samland Peninsula we find belt mounts which because of their material, construction and ornamentation cannot be classified to the group of Samland mounts but which nevertheless are in a similar openwork style. The group includes e.g., a mount from the former Tengen, Kr. Heiligenbeil, grave 39 (R. Klebs 1877, p. 61, pl. II:11; BA 1880, pl. 12:486), and an openwork mount discovered at Grač’evka, raj. Zelenogradsk (Грачьевка, рай. Зелено¬градск, former Craam, Kr. Fischhausen; W. Gaerte 1929, p. 232, fig. 174d; H. Jankuhn 1933a, p. 197). Another group of openwork belt mounts are noted during phases B2/C1–C1a and C also outside the territory occupied by Dollkeim-Kovrovo Culture (W. Gaerte 1929, fig. 173b, 174a–c; H. Jankuhn 1933a, p. 197–199, fig. 17–19; J. Okulicz 1976, p. 196–197, fig. 8) cannot be included in the group of Samland belt mounts owing to their material, construction or ornamentation; neither can they be recognised conclusively as their derivative form. Particularly absorbing is the question of the origin of Samland belt sets because in Samland Peninsula we find no forms with an earlier chronology from which these mounts could be derived. It was accepted therefore that they had taken shape in consequence of outside influence. Oscar Almgren (1913, p. 269) suggested that they represent the effect of inspiration reaching from the Danube area. He noticed the similarity of Samland belt mounts to belts which were in use within the Roman Empire starting from the late Augustine period until mid-2nd century in Noricum and Pannonia, and by the Quadi and the Marcomanni (J. Garbsch 1965, p. 79–114, 164–234, maps 1, 2, 14–16; J. Okulicz 1976, p. 191–192). This hypothesis was upheld by Herbert Jankuhn (1933a) – author of an excellent monograph on Samland belt sets – and by most researchers concerned with this group of issues (J. Okulicz 1976, p. 191–198; R. Madyda 1977, p. 366–367; R. Madyda-Legutko 1984, p. 107). An argument in favour of adopting Almgren’s hypothesis is evidence on exchange between SE Baltic coast with southern Europe during the Early Roman Period which is documented by archaeological material, eg, inflow of Norician-Pannonian A.236–238 fibulae (J. Garbsch 1965, p. 71, 224, 227–228, 231–232, map 1), traces of this exchange may be found also in the written sources (Tac. Germ. 45; cf. J. Okulicz 1976). Some researchers despite accepting Almgren’s concept have proposed also other direction of contact which could have contributed to the emergence of Samland belt sets (R. Wo¬łągiewicz 1976; U. Kobylińska 1981; 1986). A common element shared by Samland belt mounts and Norician-Pannonian belts is the presence of: openwork attachment end, rectangular openwork mounts, and small mounts with a curving body and profiled terminals (cf. H. Jankuhn 1933a; J. Garbsch 1965). Even so, the method of securing them was entirely different (cf. pl. 2a and pl. 28a4). Other differences are visible also in the construction and ornamentation of individual elements. This suggests that Samland belt sets must have taken form also under impact other than Danubian, but the latter, in the light of current research is irrefutable. A part of Samland belts were secured with R. Madyda-Legutko type 3 belt clasps (1990a; pl. 24c.i). Presumably, the people of Dollkeim-Ko¬v¬rovo Culture had borrowed the idea of construction of the clasp with a long and curved grip, joint with a cross-wise strip of bronze sheet, from the people of Wielbark Culture, and had added an openwork attachment end (H. Jankuhn 1933a, p. 186; R. Madyda-Legutko 1990a). This explanation finds support in evident contacts of Samland Peninsula and zone A of Wielbark Culture (R. Wołągiewicz 1981, p. 84) during phase B2 (W. Nowakowski 1989, p. 145–151; 1996, p. 98–99). The second group of Samland belt sets was secured using belt buckles type C10–12 – with a fixed tongue extending to the attachment end and a flexible frame (Pl. 3g, 4c, 30s). Belt fastenings with a similar construction but lacking an openwork attachment end occurred in number on territory of Przeworsk, Bogaczewo and Wielbark Cultures during phases A2/A3–B2a (cf. R. Madyda-Legutko 1987a, p. 16–18; W. Nowakowski 1995, p. 24–26; 2002, p. 139–143; P. Szymański 2005, p. 30–31; P. Iwanicki, A. Juga-Szymańska 2007, p. 46–53; T. Dąbrowska 2008, p. 38). Presumably, the people of Dollkeim-Kovrovo Culture had borrowed the idea of the described construction of belt fastening from the people of Wielbark Culture and had added an openwork attachment end (cf. H. Jankuhn 1933a, p. 178–182; R. Wołągiewicz 1976; U. Koby¬lińska 1981, p. 142–143; 1986, p. 380, 382–383). It is not possible to prove the existence of a chronological difference between Samland belts secured with a buckle and belts secured with a clasp. Both groups occur in greatest number during the mature phase B2. Samland belt mounts can be distinguished into forms with a modestly decorated attachment end, divided at mid-length with a strip of bronze sheet which extended into a grip or pin – type 3a belt clasps and type C10 belt buckles (Pl. 20b, 30s), and forms with a richly decorated attachment end filled with fine openwork – type 3b belt clasps and type C11 belt buckles (Pl. 23g, 24i). No chronological difference can be demonstrated between these two groups and the more elaborately decorated specimens cannot be recognised as later than the plain ones. Samland belt sets differed visibly from Norician-Pannonian belts in their ornamentation and only a part of Samland designs, as rectangular, stepped, cruciform, semicircular and vase-like openwork (Pl. 3g, 7a, 16j, 23g) find their counterparts in the Danubian provinces (Pl. 2d–f; H. Jankuhn 1933a, p. 193; J. Okulicz 1976, p. 194–196; U. Kobyliń¬ska 1981, p. 148–149; 1986, p. 385–388). Other types of ornament, as pairs of stamped rings and iron inlay, represent the effect of local fashion. Only a small number of finds of Samland belt mounts co-occurred in graves with weapons. Most were discovered next to a rich assortment of ornaments (see the Catalogue), and consequently, Samland belt sets should be interpreted as women’s dress accessories. The women inhabitants of the Samland Peninsula presumably were members of the elite. They fastened their clothing with fibulae (usually two to four), most typically, type A.60–62 and A.42, more seldom, with other forms – classified to type A.II/IV 2.1 of J. Andrzejowski (1994), A.109, A.IV, series 3, and A V, series 7 and 8. They adorned their head-dress – cap/kerchief – with a round disc fibula of repoussé bronze sheet, their neck – with a showy necklace of assorted glass and bronze beads. Usually, they also wore a pair of bracelets with a narrow hoop and profiled terminals (type IA; E. Blume 1912) and occasionally, also finger-rings.