Barrows near Arnhem; excavations on the Warnsborn estate, 1947-’48.In 1947 and 1948 six barrows were excavated on the Warnsborn estate, to the west of Arnhem. Although seemingly initiated for purely scientific purposes, no report was ever published, only some very short notes. This paper gives an account of the organization, the procedures followed and a critical (re)interpretation of the findings, on the basis of the field drawings and the field journal. The conclusion differs in many respects from the excavators’ original interpretations. It may be viewed as a cautionary tale for those relying on records of earlier investigations of this kind. The research was initiated and supervised by prof. Van Giffen, but in fact executed rather independently by his experienced field technician and draughtsman, assisted by two students, who some years later were to be appointed university professors and as such shaped post-war archaeology in the Netherlands. It was the period in which the State Service (ROB, now RCE) was founded, which, headed by Van Giffen, ultimately must have been the responsible institution. The barrows to be excavated were an arbitrary selection from the numerous mounds in the shallow valleys of the Heelsum and Wolfheze brooks: four dating to the Bronze Age, along the valley floor, and two from the Beaker period, slightly apart. All six appear to be part of a long row, as found in many regions across the country, comprising19 barrows. One of the Beaker burials (IV) had a characteristic early Single Grave inventory and was surrounded by a narrow palisaded ditch. The reliability of the observation of a corpse silhouette must however be doubted. The other Beaker barrow (V) was dated only on the basis of its appearance and the fossil soil conditions. It only shows some features that are hard to interpret. The group of four Bronze Age burials offers us a glimpse of the changing burial customs among a small local community. They are characterized by the absence of cremations and the exclusive practice of inhumation in all 27 burials documented. All recorded bodies were supine, with only one exception, a slightly flexed burial. None of these were equipped with any imperishable grave goods. It is remarkable that this small community in the interaction zone of the southern and northern burial traditions had exclusively followed the northern practice of extended inhumation. In the absence of radiocarbon dates and artifacts, dating had to be based purely on circumstantial evidence; with Middle Bronze Age A as the result, with a possible extension into MBA-B, i.e. 1800-1400 cal BC. Two of the Bronze Age barrows (I and II) are relatively small, simple and one-phased, with a single, supine central inhumation, one of these accompanied by a subsequent child burial. The two others (III and the large, so-called ‘Meelworstenberg’) had started similarly, but had grown to larger dimensions, up to 15 and 18 m in diameter, by the addition of new construction phases linked with new central graves. The larger one incorporated a small and low sand dune already present. The barrows became more complex and significantly different also by the introduction of secondary burials, in the form of 6 and 15 shaft graves respectively, orientated tangentially all along the barrows’ margins, and the raising of surrounding post circles. In one case this circle was only partially preserved and documented, but it was of quite impressive dimensions around the larger barrow, and seemingly not connected with a central burial, but with the secondary burials only. In both cases we observe a fundamental shift from a function as an exclusive tomb for selected individuals from the local community, as customary in earlier times, to a communal cemetery for numerous members of the community. This change in burial custom may serve as an argument for the contemporaneity of the last phase of both barrows, the simple barrows representing the earlier stage only.