Fig 1.
Map of France indicating the archaeological sites from which artefacts were sampled for adhesive residues for this study (Base map made with Natural Earth.
Free vector map data @naturalearthdata.com).
Table 1.
List of sites/samples and relative chronology per archaeological site, with primary excavation references.
Detailed information on the archaeological sites (municipality and lieu-dits) and samples can be found in Table 1.
Fig 2.
Photographs (and drawings) of some of the artefacts from which residues were sampled, as divided into function groups.
a) Rim fragment with black surface treatment of a dolium from Camp d’Attila (TK8270), b) Rim fragment with black surface treatment of a vase from Camp d’Attila (TK8256), c) Rim fragment with visible adhesive residues around the ridge and a perforation hole (see white dash line) from Champ Pertaille (TK8240), d) Two rim fragments with perforation holes and black residues from La Warde (TK8229B), e) Vase foot entirely made of a dark brown matter which replaces the broken base from Les Auges (MR2600), f) Lump (broken into two pieces) with possible perforation hole from Camp d’Attila (TK8249), g) Lump from Les Robogniers (MR2606), h) Reconstructed vase fragments with black decorative pattern from Les Auges (TK8277), i) Base fragment with heterogeneous residues adhering to the interior surface from La Cheppe (TK8221). Photos by T. Koch, drawings by I. Turé and M. Saurel.
Fig 3.
Chromatograms of adhesive residues.
a) Vessel content (sample TK8221) with evidence for a complex composition typical of the presence of Pinaceae resin, birch tar and beeswax. b) Decorative pattern on a ceramic vase (sample TK8277) indicating beeswax. c) Adhesive lump only composed of birch tar (sample TK8247). DX = Diacids (X = carbon n), AX = Alcohols (X = carbon n), Cx = Alkanes (x = carbon n), CX:Y = Fatty acids (X = carbon n, Y = n of unsaturations), IS = Internal standard. Drawings in a-b by M. Saurel, photo in c by T. Koch.
Fig 4.
Plot of the overall results of the chemical identifications per adhesive function/group.
Fig 5.
PCA plots showing the difference in adhesive function groups based on the four birch tar production markers.
a) Plot of the archaeological birch tar samples from multiple sites. b) Plot of the archaeological birch tar samples from Camp d’Attila. The arrows and numbers indicate the loadings of the variables, 1 = 3-oxoallobetulane, 2 = 28-oxoallobetul-2-ene, 3 = allobetul-2-ene, 4 = allobetulin.
Table 2.
Results (p-values) from six Tukey-Kramer tests.
Tests compare the 3 adhesive function groups (repair, lump and surface treatment) in terms of the birch tar production markers (allobetulin, 3-oxoallobetulane, allobetul-2-ene, and 28-oxoallobetul-2-ene) used to differentiate the tars in the PCA (Fig 5A).