In June 1927, two artefacts – an iron shield boss and a fragment of a small clay bowl – were donated to the National Museum in Warsaw; both were found under unknown circumstances at Grzebsk, Mława County. The shield boss can now be found in the collection of the Polish Army Museum, where it was moved as a deposit of the National Museum before 1939, while the bowl appeared – quite unexpectedly – in the pottery storage of the Iron Age Department of the State Archaeological Museum (PMA) in Warsaw, where it was ‘discovered’ in 1988. It is not quite clear how it found its way to the PMA; what is known is that this must have happened no later than in 1980. According to notes on the catalogue cards of both artefacts, drawn up still in the National Museum, they were found in a grave “covered with a flat stone, with smaller stones around it”, together with “a clay idol, which crumbled after unearthing, an iron sword, and a couple of spurs”. The grave marks an otherwise unknown cemetery of the Przeworsk Culture. We do not have any details about its location other than it was (is?) probably situated on the grounds of the former estate in the village of Grzebsk. The catalogue cards and inventory book of the National Museum list the artefacts as donated by Damian Gniaz­dowski, however, a different name – Wacław Gniazdowski – can be found in the delivery book of the Museum. The latter is true, as we know that Damian took possession of the Grzebsk estate no earlier than in 1889 and no later than in 1892, then sold the manor farm in 1902 or 1903, and moved with his family to Łępice, Pułtusk County, where he died in January 1922. The grave would have been discovered between 1889/1892 and 1902/1903, thus Damian’s son Wacław, born in 1894, must have recounted the description of the grave that he heard from his father. The small bowl from Grzebsk (Fig. 1) is typical of Przeworsk Culture pottery from the Early Roman Period and corresponds to type VI/1 in the classic typology by Teresa Liana; its unpreserved base could have been convex or concave, possibly – although this would have been completely unique – flat. Similar bowls are common at cemeteries in northern and eastern Mazovia, for example, Niedanowo 2, Nidzica County, Modła 2, Mława County, or Kamieńczyk 2, Wyszków County. Their chronology at the three cemeteries falls within the horizon of phase B1 and the older stage of phase B2. The characteristic star-like ornament on the body connects the bowl from Grzebsk with a group of vessels considered – with reservations – as more or less distant imitations of ribbed Roman glass bowls. Our specimen can be regarded – after Morten Hegewisch – as a “creative plagiarism”. The shield boss (Fig. 2:a.b) belongs to conical forms corresponding to interregional types Bohnsack 8, Jahn 5, and Zieling I1a, typical of the end of the Late Pre-Roman Period and the beginning of the Roman Period. Its surface, especially on the flange, is heavily corroded. Nevertheless, there are visible remains of so-called fire patina, attesting that the object was at some point on a funeral pyre. Only one rivet with a slightly convex, circular head has been preserved, however, rivet holes indicate that the boss was originally attached to a shield with twelve regularly spaced rivets (Fig. 2:c). Such a large number of rivets indicates that the boss should be counted among the older conical forms of Late Pre-Roman shield bosses of the Przeworsk Culture corresponding to type Bochnak 15 and dated to phases A3 and A3/B1, i.e. the end of the 1st century BC and very beginning of the 1st century AD. This fits with dating of other north-Mazovian graves with shield bosses type Bochnak 15, e.g. from Lemany, Pułtusk County, Legionowo, Le­gionowo County, and possibly also from Niedanowo 1, Nidzica County and Łysa Góra at Gródki, Działdowo County. The small iron nail stuck in the head of the preserved rivet is an interesting element (Fig. 3). Similar to the rest of the artefact, it is covered with fire patina, which indicates its original, ancient provenance. It may indicate an unusual manner of repairing the shield, probably following damage it sustained in a fight. Such a solution, consisting of hammering in another rivet, or a nail as it may be, instead of replacing the damaged rivet, may indicated the ad hoc nature of the repair or lack of access to a specialised workshop. The location of the cemetery remains unknown. It was certainly situated within Damian Gniazdowski’s estate. It is probably what a primary school teacher from Grzebsk referred to in 1926 as a “pagan cemetery” on the grounds of the manor farm, already in the possession of the Rudowski family, where “pots with ashes” were being unearthed. It may be the site registered during field walking in 1998 within the limits of a large gravel pit in the northern part of the village of Grzebsk (Fig. 4, 5). Potsherds and damaged graves in the walls of the gravel pit were discovered there – the site was identified as a Przeworsk Culture cemetery from the “Roman Period”. During verification of the site in 2018, traces of graves in the gravel pit could no longer be observed, however, fragments of characteristic sepulchral pottery of the Przeworsk Culture from the Early Roman Period were found in the gravel pit itself and its immediate vicinity. More information about this site can only be obtained through archaeological excavations. However, we will probably never know whether the cemetery that yielded the artefacts described here and the cemetery discovered in 1998 are one and the same.