The village Kozłówko, commune Kozłowo, distr. Nidzica, voiv. warmińsko-mazurskie (former Klein Koslau, Kr. Neidenburg) lies some 20 km to the south-west of the town Nidzica, on a small elevation on the left bank of the Szkotówka, right-bank tributary of the Nida R. During the 18th century relics of an ancient cemetery were identified near this locality. In volume II of his work Versuch einer wirtschaftlichen Naturgeschichte von dem Königreich Ost- und Westpreußen, F. Bock recorded the presence of fifteen small circular pavements which were to be found in the vicinity of the village Klein Koslau. In 1890 the Prussia Museum at Königsberg added to its collections a grave assemblage unearthed during farming on one of the small hills close to the village. In the autumn of 1891 Kozłówko was visited by J. Heydeck, vice-chairman of Altertumsgesellschaft Prussia. The discovery of further artefacts during this visit and survey of the site led to a decision to excavate. Archaeological investigation covering a fragment of the southern slope of the elevation identified eighty or so flat graves and a barrow. The material from Kozłówko passed to Prussia Museum where it remained until the evacuation at the end of the second world war. Its present location is unknown. Perhaps this material was not taken out of the city at all – this is suggested by the discovery made in 2000 in Kaliningrad, of a buckle originating from grave XXXVII. Heydeck published the results from the fieldwork in a brief report, providing a general description of the site and listing 39 grave inventories, illustrating twelve of them. This brief report is supplemented to some extent by information found in later works – E. Blume and R. Schindler, and in the private files of German archaeologists. Most of the flat graves were unurned cremations; only four appear to have contained cinerary urns (graves XVI, XXX, XXXV). The grave pits were basin-like, with a diameter of up to 1.5 m and a depth in the range of 35–45 cm. Individual burials had an irregular distribution within the cemetery, with larger and smaller spaces between them. A number of grave outlines were in contact and, in a few cases, later burials intruded on older grave pits. At the time of Heydeck’s investigation most of the graves were in good condition, only a few of them, found at the highest level, next to the northern margin of the cemetery, had suffered damaged from ploughing. A few graves were under circular pavements of fieldstones. The grave goods were discovered in 40 graves. The largest group of metal finds (17) were bronze brooches. Fourteen originated from grave assemblages (A, VII, IX, X, XVI, XVII, XXI, XXII, XXIII, XXV, XXVII, XXXI, XXXIV, XXXIX) one was a stray find. Arguably the earliest form is the fibula type A 41 discovered in grave A; it is a form characteristic mainly for the area of the Odra and the Vistula basin during phase B2/C1–C1a. Tendril brooches, the largest group, were noted in eight graves, a ninth was a surface find lacking context. Specimens close to types A161–162 were discovered in four graves (X, XVI, XXV, XXXIV). Specimens recorded in graves XVI (Fig. 4/XVI:1), XXV and XXXIV, with an arched or a knee-shaped bow, belong in phases C1b–C2, and a fibula from grave X, with a flat ribbon bow, is datable to phases C3–D. The remaining five tendril brooches are specimens with the end of the foot hammered flat into a catch, types A172 and A168/172. Four of these come from graves (VII, IX, XXXI, XXXIV), one was a stray find. Type A172 brooches are characteristic mainly for deposits attributable to so-called Gothic cultures. They are dated starting from phase C2 until the early segment of the Migration Period although they are regarded by some researchers as forms restricted to phase C3. Nevertheless, fibulae A172 may probably be recognised as a form diagnostic for the final phase of Wielbark Culture occupation corresponding to phases C3–D. Grave XXII contained a fibula type A185 with “an openwork” foot. Its bow was decorated with cross-wise incisions to imitate coiled wire seen Almgren 1 series VI group brooches. Similar forms were discovered e.g., at Niedanowo, graves 539 and 549, and Pruszcz Gdański 6 (former Prauster Dreieck); their dating is confined to phases C2–D. Gave XXXI, datable to phases C3–D, contained a brooch with a lozengic foot, similar to type A175, known as Elbefibel. These forms are characteristic mainly for northern Europe where they are dated mostly to phases C2 and C3. In Wielbark Culture this fibula form is rare, is recorded at e.g., Bzowo, Malbork-Wielbark, grave 1334 [133?], Stary Targ, grave 241, and is dated mainly to phases C3–D. In grave XXVII there were two ‘caterpillar’ brooches (Raupenfibel), variant 1b, variety A, series II of M. Tuszyńska. This form is recorded mostly on the southern Baltic littoral, most often, in Wielbark Culture and in West Balt cultures, where it is dated mainly to phases C3–D. From grave XVII comes a fibula with a large half-round disc on the head, so-called Spiralplattenfibeln. This find from Kozłówko is related to type Soest-Ardey. The variant is observed mainly on the territory of Wielbark Culture and in the region between the Elbe and the Rhine; it is dated to the second half of the 4th and onset of the 5th century. Grave XXI contained a fragment of a brooch with a round disc on the foot, design ZG 47 of A. Kokowski. This form, quite popular in the Wielbark Culture, is dated to phase C3–D. Another group of finds from the cemetery at Kozłówko are ten bronze buckles – a single stray surface find and specimens originating from seven graves (VII, X, XIX, XX, XXI, XXVIII, XXXVII). Two specimens, type AC13–17, were discovered in grave XXVIII; graves VII and, probably, grave XX, contained type AD11 forms. Type AD30, dated to phases B2/C1–C2, is represented by the stray buckle discovered on the surface of the cemetery type AE4, from phases C1b–C2, is represented by a large specimen discovered in grave XXXVII. Grave XXI contained a buckle type AH11. A similar specimen, presumably, type AH14, was found in grave XX. A type AH16 or AH27 buckle was discovered in grave X. The three last finds are dated to phase C3–D. Fragments of thick-walled glass vessels, including types E 230 and E 237, were discovered in at least seventeen graves, not only in richly furnished burials but also assemblages with a very modest inventory. On most occasions they occurred in company of wheel-made vessels. This situation was recorded in no less than nine graves (IV–VII, IX, XI, XII, XXXI, XL). Hand-built pottery, mainly bowls from groups VI, XIV and XVIII of R. Wołągiewicz, were observed in twenty-seven features. Most of these vessels survived in fragmented form, their number in individual assemblages usually ranged from one to three, but in nine assemblages (VIII–X, XII, XVI, XXI, XXII, XXXI, XXIV) there were more of these pieces – up to no less than eight. These graves represented the best furnished burials in the cemetery. Finally, there were personal ornaments and dress accessories: a shield-headed bracelet (grave A), a finger-ring (grave XXXIV), bucket pendants (grave XXII), a bell pendant (grave XXXI), wire and bead pendants (graves VI, VIII, XXXI, XXXIV), glass beads (graves II, VI, VIII, X, XV, XVI, XVIII, XIX, XXVI, XXXI, XXXIV, XXXVII), and ornamental clay globular objects (grave XXXI). There was also a smaller group of items of everyday use: needles or awls (graves XXXVI, XXXIX, XXXVII), bone combs (graves VI, XXIX), an ear pick (grave XIX), spindle-whorls (graves VI, X, XXXI, XXXIV, XXXVIII, a stray find), fire-making tools (grave XXXII, a stray find) and metal casket mounts (grave XXXVII). The cemetery at Kozłówko was established during phase B2/C1–C1a, as is evidenced by the inventory of grave A. Very likely, graves XXX and XXXVI containing black cinerary urns also belong in this phase. The remainder of the reliably dated graves are younger and may be distinguished into two groups. Group I would include four graves (XVI, XXV, XXXVII, XXXIX) datable by the presence in their inventory of A161–162 brooches and a buckle with an omega-shaped frame to phases C1b–C2. Group II is larger and includes nine features (VI, VII, X, XVII, XX, XXI, XXVII, XXXIV, XXXI). The presence in them of brooches type A172, A VI series 2, glass bowls E 230 and E 237, and of buckles with a thickened frame, establishes the dating of this group fairly reliably as phases C3–D. The dating of late phase C2 or onset of phases C3–D is suggested for grave XXII by the presence in it of the fibula with the ‘openwork foot’ and of a glass bowl E 230. Left outside group I and group II, are seven assemblages furnished only with ceramic vessels and fragments of glasses which prevented their attribution to either of these groups. The exceedingly small number of graves attributable to the earliest phase of the cemetery and the domination of burials dating from a later period suggest that the investigation took in mainly the youngest fragment of the grave-site. What is striking is that a cemetery with such an large surface area contained so few features (on average, 3–4 graves per 100 m2); could it be that the boundaries of this site were identified incorrectly? The cemetery at Kozłówko belongs to a local micro-region of Wielbark Culture settlement defined as the Nidzica concentration which occupied the northern fringe of the present day province of Mazowsze, extended in the south to the Działdówka R., in the east, to the Nida R.; its northern boundary was defined by a series of lakes bordering on the moraine heights of Garb Lubawski. The western boundary is more elusive – the westernmost site attributed to the Nidzica concentration may have been recorded near the village Gródki, distr. Działdowo (former Grodtken, Kr. Soldau). Outside this relatively compact area we can probably attribute to the same group isolated sites identified on the left bank of the Nida R. – at Pokrzywnica Wielka, Szczepkowo Zalesie and Muszaki (former Muschaken, Kr. Neidenburg), all of them in distr. Nidzica. So far from this area we have a record on a dozen-odd sites of Wielbark Culture. Most were investigated before WW II and this affects the amount of data available at present. The beginning of Wielbark Culture settlement in this area started presumably during phase B2/C1. Numerous burials from this period were identified at nearly all the sites of the Nidzica concentration – at Kozłówko, Niedanowo 2, Turza Wielka, distr. Działdowo (former Groß Tauersee, Kr. Soldau), Litwinki, distr. Nidzica (former Littfinken, Kr. Neidenburg), Turza Mała, distr. Nidzica (former Klein Tauersee, Kr. Neidenburg), Wilamowo, distr. Nidzica (former Wilmsdorf, Kr. Neidenburg), and possibly, also Pielgrzymowo, distr. Nidzica (former Pilgramsdorf, Kr. Neidenburg) and Sławka Wielka, distr. Nidzica (former Groß Schlaffken, Kr. Neidenburg). During phase C1b–C2 most of the cemeteries already established earlier were still in use – as documented by substantial evidence from this period recorded at e.g., Kozłówko, Niedanowo 1 and 2, Litwinki and Sławka Wielka. Also noteworthy is the rich barrow burial no. I identified at Pielgrzymowo, which is placed mostly in late phase C2, or possibly at the onset of phases C3–D. At the close of the Roman Period and the beginning of the Migration Period the number of graves in some cemeteries visibly increase, e.g., at Niedanowo and Kozłówko. The presence of numerous objects foreign to the Wielbark tradition, e.g., glass vessels, suggest that the people inhabiting in northern Mazowsze took part in exchange with the outside world. The special role of the Nidzica concentration is confirmed by the grave inventories recorded at Kozłówko. The higher than average number of glass vessels and wheel-made wares discovered at this cemetery documents the existence of intensive exchange between the community of users of this grave-site with the Black Sea region. Individual finds (e.g., A175, A185 brooches), which may be referred to materials characteristic for western Pomerania, lands on the Elbe R. and Scandinavia, attest to connections with the south-western region of the Barbaricum. Both directions of exchange may be attributed to the existence at the close of Antiquity of a network of extensive contacts, in which – it seems – the local populations had a significant share. Connections with the eastern Baltic region are evidenced by the find of an iron pendant discovered in grave XXXI, a form distinctive above all for Dollkeim-Kovrovo Culture.