The article presents flint material from two Final Palaeolithic assemblages of the Arch Backed Piece technocomplex (ABP) from Nowy Młyn (part of the city of Skarżysko-Kamienna, Skarżysko County). The Nowy Młyn flint concentrations, designated Cypel II/89 and II/90,belong to the Rydno site complex on the Kamienna River, which consists of multiple remnants of Final Palaeolithic and Mesolithic campsites near hematite outcrops and mines (Fig. 1, 2). The flint artefacts from both assemblages were deposited without a stratigraphic sequence in sandy alluvial sediments of the river terrace. Within the area of the flint concentrations, there were pits with sandy infill coloured by a mineral admixture of hematite. The pits are considered either the remains of dwelling structures of human groups camping in the area or living areas from which the processed hematite sunk gravitationally into the soil level. Post-depositional disturbances in the form of windthrows have been observed within the concentrations (Fig. 3, 4). Remains of a hearth have been identified in the centre of flint concentration II/90; it was manifested as a cluster of several hundred small fragments of burnt animal bones, regarded as culinary remains burnt in a fire. They yielded the radiocarbon date of 11 390 +/-60 BP (Poz-18066), linking the concentration to the period from the mid-Allerød (GI-1c1) to the Gerzensee Oscillation (GI-1b).

Both concentrations are remains of residential camps, as indicated by the wide variety of tools and their large share (22.6% in assemblage II/89 and 18% in II/90) in the material composition (Tables 1, 2). The results of the refitting method have been used in the analysis of the flint material (Table 3).

Approx. 82% artefacts from flint concentration II/90 are burnt or over-heated, probably as a result of natural fires. Among the identified flint raw materials, mined and erratic chocolate flint is dominant; Cretaceous erratic flint as well as Turonian and Rauracian flints are also present. Moreover, 8 specimens made of obsidian, which in this context is an exotic raw material, have been found.

Cores include single-platform blade specimens from unprepared erratic concretions, single-platform blade cores with a prepared, sometimes faceted, striking platform, usually from mined chocolate flint nodules, as well as exploited changed-orientation flake cores, likewise mainly from mined raw material (Fig. 6, 7:1.3.5). Most of the cores, blades and flakes show signs of reduction with a soft hammerstone. In the blade and flake samples, the ratio of plain to faceted butts is approx. 48% to approx. 20%. Endscrapers are made from flakes and are usually very short specimens with an arched front (Fig. 8:1–10, 9, Table 4). Burins, formed on flakes, are dominant among the tools, dihedral and truncation burins in particular, with multiple burins constituting 45% of all the burin specimens (Fig. 8:12–14, 10–13, Table 5). The tools in question were often rejuvenated, which is documented by the number of burin scars and by refits with burin spalls.

Moreover, they bear multiple marks of secondary retouching as well as damages on the edges seen with a naked eye (Tables 6, 7). Out of the burin spalls, 38% are secondary pieces. Perforators and borers are made out of ordinary flakes (Fig. 14:1–5) and burin spalls (Fig. 14:6–11, 15). The tips of the products of these two groups were formed in different ways. The only backed piece possibly associated with the assemblage is a slender arched point in the type of small Federmesser points (Fig. 14:12, 16).

In flint concentration II/89, approx. 83% of the specimens were also burnt or over-heated. The main raw materials were mined and erratic chocolate flint and erratic Cretaceous flint. Moreover, one tool was made of obsidian, and a few artefacts were made of sandstone. Cores from mined chocolate flint, single-platform or with changed orientation, show evidence of intensive, sometimes circular, blade reduction with a soft hammerstone. Preparation is limited to striking platforms, which bear negatives of one or several removals or sometimes negatives of faceting. There are also cores from erratic concretions, exploited without any preparation (Fig. 17–20). Blades with plain butts (approx. 36%) are almost as numerous as specimens with faceted butts (approx. 34%); in the flake sample, plain butts (approx. 41%) far outweigh faceted ones (approx. 21%). In the endscraper group, flake specimens are dominant, with a few made from blades (Fig. 9, 21:1–13, Table 4). The dominant tools are burins, mainly flake ones but with a share of blade specimens; truncation burins are the most numerous (Fig. 13, 21:14–18, 22, 23, Table 5). There are fewer multiple specimens (20% of the items identified) than in concentration II/90, and there is a lower average number of burin spalls detached per burin (Tables 6, 7). Primary specimens (70%) predominate among the burin spalls. In the process of shaping burins, the procedure of splitting blanks was used. Perforators/groovers/borers are divided into those made from flakes and those formed on thick blades from split blanks or on bulky burin spalls from nucleiform burins (Fig. 15, 24:1–9). Of the 14 fragments of narrow backed blades, two have preserved points. Pieces with a straight back dominate over specimens with an arched back (Fig. 16, 24:11–25). The average width of the backed pieces, excluding a dimensionally different specimen from a massive blade with two backs (one cortical), is 0.81 cm.

The assemblage in question also includes a sandstone tool with intense marks of battering on the tapered top (Fig. 26).

The inventories of flint concentrations II/89 and II/90 are typologically and technologically similar. Assemblage II/89 is distinguished by the predominance of blades among blanks and a greater share of blade tools, as well as a much lower degree of processing and exploitation of burins. Some typological fluidity has been noted between the burins and perforators, and a temporal convergence of their formation in the blank modification sequences, probably resulting from the fact that both types of tools were intended for working analogous hard organic raw materials, has been observed.

Romuald Schild isolated the assemblages from Nowy Młyn Cypel as the so-called Kamienna variant, postulating their older chronology than the one of the classic ABP assemblages and indicating their role in a possible local process of azilianisation – the adaptation of Magdalenian communities to the environmental changes of the late glacial period and the formation of communities with arch backed pieces. The features distinguishing the assemblages from Nowy Młyn Cypel from other ABP inventories are the technique of core reduction with a soft hammerstone, greater focus on blade production and the share of repetitive faceting of core striking platforms during exploitation. The older origin of the assemblages may also be evidenced by the predominance of burins in the tool group. On the other hand, the separation of Lacan burins in the burin group has been considered questionable.

In order to reconstruct the spatial organisation of the two campsites, an artefact distribution analysis, the results of refitting of the flint products and the ring and sector method proposed in 1989 by Dick Stapert were used. This last method is based on the model of functioning of an open-air camp of hunting communities around a central hearth, which was developed by Lewis Binford (Fig. 27). Measuring the distance of artefacts from the hearth makes it possible to identify zones of flint processing and economic activity, and either put forward a hypothesis of the presence of a dwelling/tent in a camp, if a barrier effect is found based on a bimodal distribution of artefacts in the rings (bimodal histograms of artefact frequency), or confirm the open-air nature of a camp with an unimodal distribution.

On the plans of both flint concentrations, rings were drawn every 0.5 m – in concentration II/90 around the central hearth (Fig. 28) and in concentration II/89 around the geometric centre (Fig. 38) – and their area was divided into 8 sectors.

For flint concentration II/90, the distance of cores, individual types of tools and burin spalls from the hearth was measured (Fig. 29), and so was the distance from the hearth of these categories of artefacts in total in individual sectors (Fig. 31). The centrifugal effect and size sorting process, characteristic of hunting camps, were confirmed. The discarded cores accumulated outside the flint concentration in a backward toss zone, while the smaller burin spalls remained closer to the campfire, in a drop zone. A clear asymmetry of the concentration and an accumulation of material on one, northern and western, side of the hearth have been observed. This has been expressed in a division of the flint concentration into a rich half and a poor half (Fig. 30). A synthetic picture of the distribution of cores, tools and burin spalls, and separetely of debitage, is presented in Fig. 32.

The numbers of chips per metre (Fig. 33) indicate the places of direct core reduction and tool shaping. The chips were concentrated to the north of the hearth, were slightly less numerous to the west of it and spread in the south-easterly direction, but did not generally exceed the distance of 1.5 m from the hearth (the eastern part of the flint concentration, with a large windthrow pit, was considered to a limited extent). Cores and elements of reduction refits were deposited primarily in the northern part of the flint concentration (Fig. 34), burins and burin spalls to the north and west of the hearth (Fig. 35). These zones were identified as the main places of core exploitation, manufacture of and working with burins. Endscrapers, on the other hand, were found primarily in the poor half of the flint concentration, to the east and south of the hearth, as were the sparse truncated pieces (Fig. 36). A rest area was probably situated in the southern part of the encampment, which was devoid of larger amounts of flint waste.

The elements of all types of artefact refits and the lines connecting them are arranged in a crescent, surrounding the hearth and open to the south-east (Fig. 37). The barrier visible in the frequency histograms in the form of a second maximum in the number of cores, tools and burin spalls at a distance of 2.5 m from the hearth, as well as the bimodal distribution of debitage in sectors with accumulation at the same distance (Fig. 29, 31, 32), allow hypothesising that a household functioned inside a dwelling of about 5 m in diameter, with a hearth in the middle. The entrance to the feature, despite the lack of a clear door dump, was most probably located in the south-eastern part of the concentration, or in its eastern part, where a relatively large amount of waste lay in an arrangement disturbed by a windthrow. A similar residential encampment of the Federmesser culture community, with comparably distributed artefacts and refit lines, which included a circular dwelling, was reconstructed for flint concentration 10 from Rekem, Limburg Province (Belgium). The resulting image of camp II/90 is consistent with Lewis Binford’s model (cf. Fig. 27:a), subsequently developed by Dick Stapert.

Flint concentration II/89, located at the edge of the sand quarry, has not been preserved in its entirety; however, the hematite feature had not been disturbed (Fig. 38). The spatial distribution of burnt and unburnt flints indicates that the latter accumulated inside the feature, which probably confirms its half-dugout nature (Fig. 40). The distance of the cores, tools and burin spalls (Fig. 41) from the geometric centre of the flint concentration and the total of these artefacts in each sector (Fig. 43, 44) showed a clear centrifugal effect in the case of the cores, which were all located on the periphery of the concentration. In contrast, no accumulation of small debris (e.g. burin spalls) was observed in the centre of the flint concentration, providing evidence for the absence of a hearth in the centre of the hematite feature. The asymmetry of the concentration was less marked than in the case of flint concentration II/90. The rich half was located in sectors 4–7, in the southern and western parts of the flint concentration (Fig. 42). Chips were concentrated inside the hematite feature, most abundantly in its western part and outside, near its southern edge (Fig. 45). Tools and cores were also accumulated near the same spots (Fig. 43); thus, two main zones of economic activity were reconstructed in these areas: one outside the feature, in sector 4, and one in the western part of the feature, in sector 6 (Fig. 44:a). Thanks to the refitting method, core exploitation (Fig. 46) and the manufacture of burins (Fig. 47) were confirmed in the southern cluster. The possibility of an external hearth at this location was suggested. In the western cluster, flaking and burin manufacture are less documented by refits, but the accumulation of backed piece fragments in the area allows an assumption that this was also a place for the preparation/re-arming of hunting weapons (Fig. 48:c). Most often such activities were carried out near campfires, which suggests that an additional hearth may have been located there. Numerous cores and burins were found to the north of the hematite feature; at the same time, there was a complete absence of burin spalls and very little debitage and waste. Elements refitted to the cores and burins from this part of the flint concentration were located inside the feature or in the southern cluster. The northern zone was therefore considered to be a waste disposal area and an external dump. A long reduction sequence of one core has also been attested there (Fig. 46, 47). A rest area may have been located in the north-eastern part of the feature, in sectors 2 and 3.

For flint concentration II/89, there are no clear indications allowing the reconstruction of a dwelling/tent. The bimodal nature of the histograms in sectors 4 and 6 results from the accumulation of artefacts in the discussed activity zones unrelated to the centre of the flint concentration, and in sector 8 – from the slightly distant external dump (Fig. 43). The distance of the rings with the accumulation of cores from those with the most numerous tools is also different, which may indicate a freer spread of artefacts (Fig. 41:g.h). The additionally analysed distribution of debitage products (Fig. 50) showed a clearly unimodal pattern in all sectors, with the exception of sector 5, where it is very uniform and exhibits characteristics of a door dump or a pathway causing artefact scatter. This, together with the accumulation of the most abundant debitage near the periphery of the hematite feature of probably half-dugout character, seem to indicate the presence of at least some protective covers on the north and east sides, while on the south side the feature was wide open. This image would be somewhat reminiscent of the reconstructed asymmetrical camps of Magdalenian communities, with a hearth and an economic activity zone in front of the entrance to semi-open shelters at, for example, Monruz, Canton de Neuchâtel (Switzerland) and would correspond in part to the encampment model created by André Leroi-Gourhan (1972).

The differences in the organisation of the encampments in the flint concentrations analysed are perhaps due to their functioning in different seasons – a colder and a warmer one. The hypothesis of the functioning of a closed dwelling in camp II/90 in a colder season is also supported by a much more intense exploitation of the tools, mainly burins. An attempt at reconstructing both campsites is presented in Fig. 51.