Navigation – Plan du site

AccueilNuméros48-1Multiproxy approach based on imag...

Multiproxy approach based on image analysis, petrography, mineralogy and micropaleontology to compare carved limestone fragments, application to the study of a broken stone recumbent (Beatrix d’Avesnes, Valenciennes)

Approche multi-proxy basée sur l’analyse d’images, la pétrographie, la minéralogie et la micropaléontologie pour comparer des fragments de calcaires taillés ; application à l’étude d’un gisant brisé (Béatrix d’Avesnes, Valenciennes)
Éric Armynot du Châtelet, Michel Dubois, Silvia Gardin, Ludovic Nys, Vincent Hadot, Renaud Toullec, Sandra Ventalon et Philippe Recourt
p. 17-30

Résumés

Il est parfois difficile de démontrer une origine commune pour des objets de pierre brisés trouvés lors de fouilles archéologiques et conservés dans les collections des musées. Cette étude fournit une approche multi-proxy pour une investigation pétrographique. Deux éléments séparés d’un gisant (tête et corps) présentés au musée de Valenciennes (France) ont servi d’exemple pour utiliser des outils de minéralogie (composition en glauconite et quartz) et de micropaléontologie (foraminifères et coccolithes) associés à un traitement statistique, afin d’en tester la nature commune. La glauconite, la teneur en quartz, les foraminifères et les coccolithes ont été décrits et quantifiés dans les deux pièces du gisant. Les deux éléments sont constitués d’un calcaire crayeux du Turonien (Crétacé supérieur). Sur la base de la combinaison des résultats des proxys analysés aucun argument ne permet d’attester avec certitude que la tête n’appartient pas au corps du gisant.

Haut de page

Texte intégral

The Nikon Eclipse Ni-E automated microscope was acquired thanks to Région des Hauts-de-France council project Visionn’AIRR [grant number iM4]. We would like to thanks Yvan Coquinot (C2RMF) for his help during the discussion about the possible origin of the stone. The authors are grateful to Cesare Papazzoni with his help for determining the taxonomic and stratigraphic position of larger foraminifera. We then thank David Decroocq for looking for and collecting ancient bibliographical sources in several libraries. At last, the authors would particularly like to thank the two reviewers who, through their reading and suggestion work, have considerably improved the readability of this work.

Introduction

1During archeological excavations, statues and sculptures made of rock are often discovered and dug out as broken fragments. They can be sometimes simply restored by joining fragments, but in some cases, the absence of certain parts makes it impossible to reconstitute the whole artifact. Another approach consists in comparing the rocky material they are made of. Similarities in rock nature, general aspect, relative sizes, cutting methods offer good indications but can be insufficient to conclude about a possible common origin of two archeological fragmented artifacts. Amongst petrographic tools, thin sections are particularly pertinent preparations as they allow identification of the mineralogical composition and the fabric characteristics (Riederer, 2004).

2For sedimentary material, this characterization includes in particular petrography and rock texture, mineralogy, micropaleontology based on optical microscopy, and Scanning Electron Microscopy associated with chemical EDS analysis (Riederer, 2004). Recent studies used carbon and oxygen stable isotope signatures, in conjunction with fine petrographic analysis to decipher the provenance of protohistoric sculptural limestones (Fournier et al., 2023). The constituents provide information on the geological formation from which the rock originated, sometimes the outcrop itself (Cassar, 2010). Often, these archeological studies involve multiscale observations from a broad macroscopic characterization of the object especially its aspect and possible surface alteration, up to finer observations of the fabric trough microscopic description of the inorganic and bioclastic components (Goodman et al., 2009 ; Cassar, 2010 ; Wilkinson et al., 2010). For example, on a study of ceramic wares, Riederer (Riederer, 2004) demonstrates the interest of the mineralogical composition, occurrence of fossil fragments, grain size characteristics to sort, classify and identify the origin of ceramics. In this specific, well-studied case of pottery, in addition to mineralogy, microfossils could be used to trace the origin of the archeological material; hence diatoms and pollens are commonly used (Alhonen & Väkeväinen, 1981 ; Gibson, 1983 ; Matiskainen & Alhonen, 1984 ; Magid & Krzywinski, 1988). Petrographic and micropaleontological tools therefore exist. Amongst the disciplines of archeometry, the sourcing of flint stones has been extensively commented (e.g., Fernandes & Raynal, 2006), which made it possible to confirm the usefulness of the tools (cf Delvigne, 2019 and Delvigne, 2020 and cited reference). For statuary and pottery, unfortunately, these tools are often used individually.

3A question was addressed by Le Musée de Valenciennes (Valenciennes city Museum, France) whose collections include two broken recumbent statues found during archeological excavations. One of them was a headoff body, the second was a single head reclaimed by the museum, without any contextual information. The recumbent historical and carving characteristics (Le Bouck, 1650 ; Oursel, 1978 ; Hardy & Beaussart, 1980 ; Beaussart et al., 1987) are consistent with a common origin of the head and the body of the recumbent. At first glance, the two fragments present similarities, particularly the nature of the rocky material. Unfortunately, these fragments could not be tightly associated together, ca. 3 cm of the broken neck miss. As a consequence, some extra arguments are needed to assess the correspondence of the two pieces, providing an opportunity to test a multiproxy method.

4The strength of the present study is to combine several techniques (petrography, mineralogy, paleontology of micro- and nanno-fossils) to characterize the rock. The main question is: does this multiproxy approach provide robust quantified arguments for assessing the link between the head and the body of the recumbent statue?

Material and methods

5In 1973 two recumbent statues were exhumed from the embankments of Valenciennes city (France) in the Notre-Dame-de-Beaumont convent: one male and one woman (Museum Inventory number: M.V. 78.2.A and 78.1.A), the latter lacking its head (Figure 1a). In 2014, a woman head (Figure 1b) of unknown origin was brought to the museum by a private individual from Valenciennes (Hardy & Beaussart, 1980). Both woman’s broken pieces are the objects of this study.

The recumbent artifact

6The body of the statue (height = 136 cm; width = 48 cm; thickness = 29 cm) is a Gothic recumbent (Figure 1a); a lady dressed in a skirt and overcoat closed by buttons, covered with a large cord mantel, tight at the throat and slightly raised from the right hip. The drapery is folded in long waves. At her feet, 2 puppies or 2 small lions sit, too damaged to be determined with certainty (Figure 1c). Some traces of polychromy are visible in the hollow areas. The head (height = 26 cm; width = 21 cm; thickness = 19 cm) is covered with a gimp closed to the chin by a towel. The statue fragments initially attributed to Jeanne du Roeulx, also called Jeanne d’Écaussinnes (Le Bouck, 1650), are now considered as the recumbent of Béatrice d’Avesnes mother of Emperor Henry VII of Luxembourg (Nys, 2023). The head and body, collected independently, were hypothetically brought together by convergence between the sculptural invoices (Hardy & Beaussart, 1980), but suffer from the absence of the neck for a good match.

Sampling

7Initial observations of the two artifacts (body and head) and sampling were made in the premises of the Valenciennes City Museum. Only about 1 cm3 of rock material was sampled with small hammer and chisel from the body along the neck, at the back of the statue for esthetic and preservation reasons (sample latter called BODY) (Figure 1d and e). Two fragments equivalent in terms of the size were taken at the base of the head (samples called HEAD1 and HEAD2) (Figure 1f).

8These three samples were then prepared as uncovered thin sections and further analyzed using petrographical, mineralogical and micropaleontological tools.

Figure 1: Pictures of the recumbent.
Figure 1 : Photos du gisant.

Figure 1: Pictures of the recumbent. Figure 1 : Photos du gisant.

a) General view with a positioning attempt of the head and the body; the neck is missing; b) Detailed view of the head with details of the gimp; c) Detailed view on the mutilated puppies or lions at the feet of the lady; d, e, f) detailed view of the positioning of the sampling, one from the neckline of the body (sample BODY) and two from the back of the base of the head (HEAD 1 and HEAD 2).
a) Vue générale avec une tentative de positionnement de la tête et du corps ; le cou est manquant ; b) Vue détaillée de la tête avec détails de la guimpe ; c) Vue de détail sur les chiots ou les lions mutilés aux pieds du gisant ; d, e, f) vue de détail du positionnement des prélèvements, un sur l’encolure du corps (échantillon BODY) et deux sur l’arrière de la base de la tête (échantillons HEAD 1 et HEAD 2).

Petrographic observations and measurements

9The petrographic study (carried out on an Olympus BX60 polarizing optical microscope equipped with Spot Flex camera) was carried out to provide better understanding of mineralogical, diagenetic and paleontological features of the rock from the artifacts. To improve treatment possibilities through image analysis, computer and statistics, high resolution – plane-polarized light and cross-polarized light – images were acquired using a Nikon Eclipse Ni-E automated microscope and a Nikon Camera DS-Ri2. Used objective is TU PlanFluorEPI-P 10X NA 0.3. Mosaic and vertical stacking data (multiplane images) were acquired, merged and aligned with the NIS-AR software.

10The mineral identification was made with the classical petrographic observation and checked by means of Raman microspectroscopy. Raman spectra were compared to RRUFF database included in the software LabSpec (Lafuente et al., 2015). The Raman microspectrometer is a Horiba Jobin Yvon, HR 800 UV equipped with a laser Quantum, Torus 532 nm and 50 mW (green light) and spectra were recorded with an integration time of 3 to 30 s.

Micropaleontological content analysis

11Foraminifera we observed by using the polarizing optical microscope and the high-resolution images in all the three samples thin sections (BODY, HEAD1 and HEAD2). Only larger foraminifera (> 500 µm) are determined at a genus level. Too reworked to be determined at species neither genus level, the smallest foraminiferal specimens (50 to 150 µm) were quantified according to their test morphologies (uniserial, biserial, triloculine, quinqueloculine, planispiral and trochospiral). Uncertain forms are arranged in an “indeterminate” class. Single or two chambers ambiguous specimens, of ~20-30 µm are not considered. These abundant small foraminifera were counted following a grid of 1000 pixels by 1000 pixels (0.5329 mm²) to then be able to carry out statistical analyzes to discuss intra-sample and inter-sample variability.

12Calcareous nannofossils were studied using both optical and scanning electron microscope (SEM) in the two samples: HEAD1 and BODY. For the optical microscope, simple smear-slides from the raw material were processed according to standard preparation technique. For the SEM analysis a very small, fresh chip of rock was mounted on a carbon stub and analysed with a Flexsem1000 II Hitachi scanning electron microscope without metal or carbon coating.

Quartz quantification by chemical mapping and imaging

13As quartz is the main detrital mineral in carbonaceous environments (eg. Desoignies, 1967) we selected it for characterizing facies variation. After carbon coating, the recumbent thin sections were then mapped using a SEM equipped with an energy dispersive X-ray spectrometer (EDS). The EDS was used to conduct both qualitative and quantitative elemental analyses to identify the mineral particles composing the samples. Silicon mapping has been particularly studied because it makes it easy to deduce occurrences of quartz. Quartz was selected because of its high abundance and simple chemical composition of silica which allows it to be identified from a Si map. The extraction procedure of quartz from EDS observation of elemental distribution is illustrated in Figure 2a and b. Image processing was performed using free software Fiji (Schindelin et al., 2012). Image from silicon distribution is selected. After binarizing and thresholding of grey levels, eroding dilating for removing isolated pixel and filling holes (b), particles were automatically counted with Image-J tools. The Image-J macro is provided as appendix A1. Hence 3 list of quartz area (i.e. size) were produced, one of each sample.

Glauconite quantification by optical observation and imaging

14Because glauconite is a mineral already used as proxy (i.e., sea level variation in Hesselbo & Huggett, 2001) and is frequent in the geological formation of the region (Tribovillard et al., 2023), it was selected as proxy for characterizing our rocks. The content of glauconite is quantified by image analysis from the three samples (BODY, HEAD1 and HEAD2). Glauconite was selected because of its very characteristic hue in microscopy which is almost unique in sedimentary petrography, from bottle green to deep green in non-polarized light. To this purpose, the non-polarized light thin sections images were subdivided in 1000-pixels by 1000-pixels squares, that represent each exactly 0.5329 mm². The division into the 1000 × 1000 pixels squares allows a statistical comparison on the distribution of the density of minerals. As the samples are not square neither rectangle, only the 1000 × 1000 pixels squares completely filled by a rock zone are analyzed (165 images for the BODY slide, 114 for HEAD1, 22 for HEAD2). The too thin areas of the thin section that may introduce the strongest bias through the image analysis process are also avoided as much as possible (HEAD2 was sufficiently big to preserve a minimum number of square). The extraction procedure of glauconite is illustrated in Figure 2c, d and e. After 3 to 6 color deconvolutions (depending on the color range and then adapted to each thin section), green pixels were extracted from the image (d). Image is then binarized (e). Pixels were grouped after erosion, dilatation and filling holes: erosion and dilatation are used to remove isolated pixels (β) and filling holes is used to fill up minerals whose color variation did not allow direct extraction (α). Glauconites are then counted and surface was evaluated. Note that minerals situated on the edges of images are not considered for the calculation (δ). Image-J codes are provided on request to the corresponding author.

Figure 2: Image analysis methods for extracting the quartz from the SEM-EDS image (a and b) and the glauconite from thin sections (c, d and e).
Figure 2 : Méthodes d’analyses d’images pour extraire le quartz de l’image MEB-EDS (a et b) et la glauconite des lames minces (c, d et e).

Figure 2: Image analysis methods for extracting the quartz from the SEM-EDS image (a and b) and the glauconite from thin sections (c, d and e). Figure 2 : Méthodes d’analyses d’images pour extraire le quartz de l’image MEB-EDS (a et b) et la glauconite des lames minces (c, d et e).

Sample BODY is taken as an example. See text for details of the image processing used.
L’exemple est tiré de l’échantillon BODY. Se référer au texte pour le détail des traitements d’image employés.

Statistical processing

15The comparison between the three samples from the recumbent is based on the proxy enumeration distinguished on the thin section observation. Occupied surface area and number of grains per unit area of both dominant minerals (unaltered glauconite and quartz) and density of foraminifera are considered and compared between all data sets using non-parametric tests, Kruskall-Wallis and the Dunn test. For the Dunn test, a Bonferroni adjustment method is applied. The calculation and representation were done using the R software (R Core Team, 2017) and the packages tidyverse (Wickham, 2017), ggpubr (Kassambara, 2020a) and rstatix (Kassambara, 2020b). R codes are provided on request with the data.

Results

Analysis of the recumbent samples

Diagnosis of the samples and stratigraphical origin

16Macroscopic examination of the recumbent fragments and the three samples BODY, HEAD1 and HEAD2 reveals that the rock is a white-yellowish limestone including numerous black points, identified as glauconite.

17Microscopically, the texture is described as wackestone-packstone (Dunham, 1962): this texture of limestone is at the boundary between a grain supported sediment whom intergranular space is filled by micritic matrix (packstone) and a mud supported sediment with non-touching and floating allochems (wackestone). In the 3 recumbent samples, the numerous foraminiferal chambers and fragments are encased in and filled by a fine matrix, entirely composed by coccoliths (Figure 3f, 3g). There is no pervasive overgrowth or dissolution while fragmentation is frequent. Fragments of urchin spines and isolated pentagonal echinoderm plates are scattered within the thin section (Figure 3c). Many fragments of small calcite shells that may belong to ostracods are also observed. Added to this highly fragmented fauna are brachiopods and bivalve shells (Figure 3d-3), and fish teeth (Figure 3d and e).

18The three thin sections HEAD1, HEAD2 and BODY examination revealed planktonic and benthic foraminifera (subclasses of Spirillinana, Miliolana, Nodosaria and Rotaliana) (e.g., Figure 3d-2 and 3e-2). A single large foraminifera is observed in the HEAD1 sample with a relatively flattened morphology (Figure 3a). At first it could be determined as a Nummulitidae from Operculina genous originating from Eocene. However, while remaining very wary due to the very large size (2.34 mm), the organization of lamellar layer may attribute it to a polymorphinid or even a robertinid that may originate from the Cretaceous.

19HEAD1 and BODY displayed rich, abundant and almost identical calcareous nannofossil assemblages moderately preserved. The occurrence of common coccoliths “key species” such as Quadrum gartneri, Eiffellithus eximius, Gartnerago obliquum, Kamptnerius magnificus, Eprolithus moratus, E. floralis , Lucianorhabdus maleformis (Figure 3f and 3g) helped to determine an age not older than middle Turonian, according to the most used biozonation schemes (Sissingh, 1977 ; Perch-Nielsen, 1985 ; Burnett, 1998).

20Sparsely distributed in the carbonate samples, quartz and glauconite were observed in addition to clearly determined phosphate grains and uncommon flakes of muscovite. All are confirmed with Raman analysis. The glauconite is particularly noticeable in the HEAD2 sample with an unaltered form (“bottle green” coloration, Figure 3b-1) or including iron oxides giving brown to black colors linked to the alteration degree. The very fine angular quartz crystals (< 100 µm) (Figure 3b-2) are easily identified from the SEM-EDS maps as the mineral is associated to silicium whom SEM-EDS spectrum is characteristic. Dark nodules were identified as manganese oxides (Figure 3b-3) using EDS analysis. A trigonal dolomite partially crystallized is observed in sample BODY.

Figure 3: Pictures from the thin sections.
Figure 3 : Photos des lames minces.

Figure 3: Pictures from the thin sections. Figure 3 : Photos des lames minces.

a) an occurrence of the large foraminifera polymorphinid or even a robertinid from HEAD1 sample; b1) glauconite grains occurrence from HEAD2 with a carbonate matrix; b2) quartz grains; b3) and dark nodules; c1) urchin plate; d1) tooth, d2) biserial micro-foraminifera, d3) bivalve bioclast; e1) tooth, e2) trochispiral planktonic micro-foraminifera; f) Optical micrographs of key calcareous nannofossil species: 1) Eprolithus moratus; 2) E. floralis; 3) Quadrum gartneri; 4) Eiffellithus eximius; 5) Lucianorhabdus maleformis; 6) Ahmuellerella octoradiata; 7) Kamptnerius magnificus; 8) Gartnerago obliquus; g) SEM micrographs of key calcareous nannofossil species: 1) Coccosphere of Watznaueria sp; 2) Eprolithus moratus; 3) E. eximius, distal view; 4) Gartnerago obliquuum; 5) Eiffellithus eximius, proximal view; 6) Tranolithus orionatus.
a) occurrence du grand foraminifère polymorphinide ou robertinide (échantillon HEAD1) ; b1) occurrence de grains de glauconite avec une matrice carbonatée (HEAD2) ; b2) grains de quartz ; b3) nodules sombres ; c1) plaque d’oursin ; d1) dent, d2) fragment de micro-foraminifère, d3) fragment de bivalve ; e1) dent, e2) micro-foraminifère planctonique trochispiralé ; f) Micrographies optiques des principales espèces de nannofossiles calcaires : 1) Eprolithus moratus ; 2) E. floralis ; 3) Quadrum gartneri ; 4) Eiffellithus eximius ; 5) Lucianorhabdus maleformis ; 6) Ahmuellerella octoradiata ; 7) Kamptnerius magnificus ; 8) Gartnerago obliquus ; g) Micrographies MEB des principales espèces de nannofossiles calcaires : 1) Coccosphère de Watznaueria sp ; 2) Eprolithus moratus ; 3) E. eximius, vue distale ; 4) Gartnerago obliquuum ; 5) Eiffellithus eximius, vue proximale ; 6) Tranolithus orionatus.

Comparative study of the three samples

21Glauconite content – The glauconite is studied through the surface it occupies in thin section and the number of grains. Both these values are compared in between the 3 thin sections. Data will be available on request.

22The proportion of surface occupied by fresh glauconite reach up to 7.2% in HEAD2 thin section (Figure 4a). A mean of 2.96% of the surface of HEAD2 is covered by fresh glauconite (Figure 4a) (Table 1). There are significant differences between the three considered thin sections (Kruskal-Wallis test for proportion of surface occupied by glauconite is significant with p-value = 3.74 × 10-10). The comparison by pairs show that BODY and HEAD1 cannot be considered as significantly different (Dunn test with p-value = 7.87 × 10-1, Table 2), whereas HEAD2 is considered as significantly different from BODY and HEAD1 (Dunn test with p-value < 0.05).

23The mean number of glauconite grains ranges from 3.8 to 14.8 per mm² and exceptionally up to 50 in HEAD2. Kruskal-Wallis test is significant with p-value = 3.04 × 10-26, there are significant difference between the three samples. The comparison by pairs shows that the number of glauconite per mm² in BODY could be considered as different from both HEAD1 and HEAD2 (Dunn test with p-value < 0.05) (Table 2), whereas HEAD1 and HEAD2 cannot be considered as different (Dunn test with p-value = 6.12 × 10²).

Table 1: Glauconite grains characteristics in each thin section.
Tableau 1 : Caractéristiques des grains de glauconite dans chaque lame mince.

Glauconite characteristics BODY HEAD1 HEAD2
Number of analysed cell 148 133 22
Analysed surface (μm2) 78968669 70965088 11738586
Number of counted glauconite 301 642 174
% Surface covered by glauconite 0.54 0.58 2.96
Mean glauconite surface (μm2) 1409 642 1995
Median glauconite surface (μm2) 1065 169 406

Table 2: Results of Dunn test for multiple comparisons between the samples.
Tableau 2 : Résultats du test de Dunn pour des comparaisons multiples entre les 3 lames minces.

Thin section 1 Thin section 2 n1 n2 Statistic p p.adj p.adj.signif
Surface occupied BODY HEAD1 148 133 1.12 2.62E-01 7.87E-01 ns
Surface occupied BODY HEAD2 148 22 6.57 5.09E-11 1.53E-10 ****
Surface occupied HEAD1 HEAD2 133 22 5.94 2.87E-09 8.60E-09 ****
Number of grains BODY HEAD1 148 133 9.55 1.28E-21 3.83E-21 ****
Number of grains BODY HEAD2 148 22 7.33 2.30E-13 6.89E-13 ****
Number of grains HEAD1 HEAD2 133 22 2.32 2.04E-02 6.12E-02 ns

Top part by considering the surface occupied by glauconite. Bottom part by considering the number of glauconite grains. The differences are considered significant for p-values < 0.05.
Dans la partie haute du tableau en considérant les glauconites par la surface occupée sur chaque lame mince et dans la partie basse en considérant le nombre de grain. Les différences sont considérées comme significatives pour des valeurs de probabilité < 0,05.

24If we just consider the number of glauconite, the head and body were sculpted from rocks with different facies. However, by observing these same glauconites but this time, according to their surface, the significant difference is observed within the same piece of recumbent. As a consequence, glauconite surface is not an argument to differentiate or to associate the head from the body, the variations observed are probably due to subtle variations in glauconite abundance within the same facies.

Figure 4: Box and whisker plot for measurement on glauconite grains, quartz grains and foraminiferal individuals.
Figure 4 : Boîtes à moustaches calculées à partir des mesures sur les grains de glauconite, les grains de quartz et les individus de foraminifères.

Figure 4: Box and whisker plot for measurement on glauconite grains, quartz grains and foraminiferal individuals. Figure 4 : Boîtes à moustaches calculées à partir des mesures sur les grains de glauconite, les grains de quartz et les individus de foraminifères.

The box presents the interquartile Q1-Q3 with the median as intermediate line. Whiskers present minimum and maximum values. For each of the three thin section from the recumbent : (a) proportion of surface occupied by glauconite, (b) Number of glauconites per mm², (c) area of the quartz grains, (d) density of foraminifera per mm². For more visibility log scale are used when necessary.
Le rectangle représente l’interquartile Q1-Q3 avec la médiane comme ligne intermédiaire. Les moustaches présentent des valeurs minimales et maximales. Pour chacun des trois échantillons : (a) proportion de surface occupée par la glauconite, (b) nombre de glauconites par mm², (c) proportion de surface de quartz, (d) densité de foraminifères. Pour plus de lisibilité, une échelle logarithmique est utilisée lorsque nécessaire.

25Quartz content – The number of quartz grains per mm² ranges from 16 to 21. Their average areas range from 215 to 394 μm² (Figure 4c). Some outliers reach more than 3000 µm² in HEAD2 sample. There are no significant differences between the samples (Kruskal-Wallis test with p-value = 0.288). As a second preliminary conclusion, no difference can be underlined in between the body and the head of the recumbent. Data are provided as supplementary material.

26Microforaminiferal content – 722 small (30 to 125 μm), 30 large (125 to 500 μm) and 1 very large foraminifera (length = 2.34 mm) were observed. The density of the foraminifera is not the same in the thin sections (Kruskal-Wallis test with p-value = 1.364.10-9) (Figure 4d). Data are provided as supplementary material. The comparison by pairs shows that HEAD1 foraminifera density is significantly different than BODY and HEAD2 (Dunn test with p-value < 0.05, Table 3). BODY and HEAD2 foraminiferal density cannot be considered as different (Dunn test with p-value = 0.126) (Table 3). As a third preliminary conclusion, with this third proxy we cannot concluded to any evidence of similarity neither dissimilarity in between the head and the body of the recumbent. As for glauconite, it is probably subtle variations in facies that are at the origin of the differences in foraminifera density between the head and the body.

Table 3: Results of Dunn test for multiple comparisons between the densities of foraminifera.
Tableau 3 : Résultats du test de Dunn pour des comparaisons multiples entre les densités de foraminifères.

Thin section 1 Thin section 2 n1 n2 statistic p p.adj p.adj.signif
BODY HEAD1 163 133 5.42 5.87E-08 1.76E-07 ****
BODY HEAD1 163 22 -2.03 0.0416 0.126 ns
HEAD1 HEAD2 133 22 -4.76 0.00000193 0.00000579 ****

The differences are considered significant for p-values < 0.05.
Les différences sont considérées comme significatives pour des valeurs de probabilité < 0,05.

27Calcareous coccolith nannofossils – HEAD and BODY displayed rich, abundant and almost identical calcareous nannofossil assemblages. Thanks to these observations we cannot evidence dissimilarity between the head and the recumbent.

Discussion

Does the head belong to the body?

28In order to test the concordance between the head and the body of the statue, this study focuses on the petrographic content to consolidate the previous more traditional observations in archeology such as historical documents, costumes, and carving details. Microfossils and mineralogy analyses have the advantage of being based on elements that still exist while the engraving marks are worn away by time (Newman, 2005).

29By summarizing all the relationships between petrographic elements analyzed in this study, there are 5 proxies with equivocal conclusions (Figure 5). When considering glauconite number, head and body belongs to different rocks. When considering quartz surface and coccoliths, keeping in mind coccoliths community was not evaluated in HEAD2 samples, both head and body belongs to the same rock. When comparing HEAD1 and HEAD2, as both glauconite surface and foraminifera density are not significantly comparable, they cannot be considered.

30The fact that there are even dissimilarities between the two samples HEAD1 and HEAD2 underlines that facies variations are possible. Considering the surface of glauconites is representative of their size, this surface is proportional either to the dynamics of the sediment which will allow the formation and transport of glauconites, or to the chemistry of the sediments that contain the elements Al, Fe and Mg in different concentrations. This dynamic or this chemistry is therefore different between HEAD1 and HEAD2. The glauconites are distributed along a large range of size, they are not well sorted. It means that the sources is local with glauconites either reworked from close outcrops (the region is rich in glauconite facies from Jurassic and Cenomanian (Desoignies, 1967 ; Tribovillard et al., 2023) or that the glauconite is newly formed that is the case of Upper Cretaceous Boulonnais outcrops, 100 km to the west (Tribovillard et al., 2023). No previous studies focus on spatial variability of facies but thanks to our results we could affirm that just along few millimeters, large variations are possible.

31Similarly, the density of foraminifera is different between HEAD1 and HEAD2 (Figure 5). Once again, a lateral variation of the facies should be considered. With a single contribution of planktonic foraminifera the substrate could be homogeneous, but as soon as there is a mixture of planktonic and benthic foraminifera, the distribution pattern can reveal patches, even on the scale of a few cm (Armynot du Châtelet et al., 2017).

32The vast majority of quartz is detrital and depends on large-scale environmental conditions, which can explain why there is no difference if we admit the variability inside facies. In the same way, coccoliths being exclusively planktonic, we can assume a homogeneous deposition in the sediment. However, it should be noted that in our case the coccoliths were not analyzed in the HEAD2 sample and that no quantitative tests were carried out.

Figure 5: Summary of the relationships tested in between the three samples BODY, HEAD1 and HEAD2, throughout the five petrographical and micropaleontological proxies.
Figure 5 : Synthèse des relations testées entre les trois échantillons BODY, HEAD1 et HEAD2, à travers les cinq proxys pétrographiques et micropaléontologiques.

Figure 5: Summary of the relationships tested in between the three samples BODY, HEAD1 and HEAD2, throughout the five petrographical and micropaleontological proxies. Figure 5 : Synthèse des relations testées entre les trois échantillons BODY, HEAD1 et HEAD2, à travers les cinq proxys pétrographiques et micropaléontologiques.

Stratigraphic position

33The concordance of stratigraphic position is another argument to bring together the elements of the recumbent, on one hand by the nature of the fossils, but on the second hand, in relation to the sources of the rocks which could have supplied the stonemasons. In our study, the gathering of information from micropaleontological proxy, small foraminifera and coccoliths allows the determination of the stratigraphic position and permits to define the potential paleoenvironmental origin along carbonate platform. Classical depositional models such as Wilson standard facies belt (Wilson, 1975) are then used to evaluate the paleo-position of the sample. The fact that the rock comes from a continental shelf will unfortunately not help in the determination of a sliced facies as Quinn & Day (2007) could do. Indeed, these facies are often diffuse and less contrasted than lagoon and reef facies for example, which are characterized by short-scale lateral facies variations are characteristic.

34The observed coccoliths species allows to assign an age not older than mid-Turonian, at the end of the Cretaceous. Coccoliths are more commonly used in archeology than foraminifera for material origin reconstruction. They could be used from the Triassic to modern deposit stratigraphy (Gardin et al., 2012) and especially in late Cretaceous with the largest diversity (Brown et al., 1987). The major advantage is that even within a small amount of material they are numerous and easily identified. This is a particularly sensitive point when performing destructive analysis in precious archeological remains. In the present case, only a small volume of the recumbent was sampled to prevent significative damage to the statue. Many studies using coccoliths could be cited in several archeological fields such as their study in mortar to trace the origin of chalk (e.g., Falkenberg & Mutterlose, 2021), the origin of the chalky craft from the Amiens-Renancourt deposits (Deneuve et al., 2018), the origin of masonries for building (Lübke et al., 2018), the origin of the melting cast (clay) used before of bronze statues casting (Fiorentino, 1998), or the origin of material used for ceramics (Quinn et al., 1998 ; Burton et al., 2018) or artifact (e.g., Persico and Cogliati, 2014). Our study confirms the importance and efficiency of calcareous nannofossils for archeological remains studies, especially in the case of the larger classical stratigraphical foraminifera may not be used. In the present case, the sample size needed to obtain more stratigraphic fossils would be close to one dm3, which is obviously impossible.

Origin of the rock and sampling effort

35The question of the origin of the rock that composes the recumbent is asked. Many of the recumbents of the Notre-Dame-de-Beaumont convent were made of “Pierre bleue” (“Blue stone”), a blue-grey limestone of Paleozoic age exploited in many quarries of North of France and Hainaut regions in Belgium (Region of the family of Beatrix of Avesnes). This blue-grey limestone is macroscopically completely different (color, hardness) from the stone of the recumbent and of distinct age. The nomenclature of the stones used in building in North of France is vague, often with local names, or according to the appearance or facies (Dollé, 1924). The “Pierre d’Avesnes” (Avendersteen in Tolboom et al., 2009) covers numerous varieties of white stones from chalky formations. It was exploited for about 500 years (1400-1900) (Tolboom et al., 2009). The term more restrictively designates a rock mined in the Scheldt basin (Hainaut-Cambrésis-Artois areas) (Tolboom et al., 2009). A review of Cretaceous fine limestone quarry shows numerous good candidates for the origin of the stone. As previous works did not pay particular attention to the provenance of rocks used for recumbent sculpture, a development of the present study would be a full sampling as exhaustive as possible with the lingering question of the subtle variations in facies that have been highlighted.

36Such sampling would have a second interest, making it possible to quantify the degree of heterogeneity of the chalky facies at the scale of the outcrop. Indeed, even if the deposition process in carbonate platforms involves mechanisms on a regional scale, local variations are possible (Ramdani, 2023). Qualifying this variation would make it possible to quantify the possible variability during the analysis of microfacies for example glauconites and quartz to admit that the measurements fall within the error bar or are statistically different.

Mineralogical qualification and further developments

37The use of mineralogical qualification (from simple petrographic microscope to the use of heavy technology) is a traditional approach for tracing the origin of a rock especially clay minerals or metals (Hughes et al., 1998 ; Zaykov et al., 1999 ; Wisseman et al., 2002). A major contribution of our approach is the quantification of minerals on thin section surface using petrographic microscope and SEM-EDS mapping associated with image analysis. This approach was already used on small material at the size of an agglutinated foraminifera (Armynot du Châtelet et al., 2013 ; Armynot du Châtelet et al., 2014). For the thin section our methodology focused only on well colored and highly contrasted grains after petrographical mapping (glauconite) and chemically well evidenced after SEM-EDS mapping (quartz). Further development on less sharp minerals should be proposed in the near future for being adapted to any material origin. The best solution, would be the analysis of a complex image in which each image representing a chemical element is positioned. The technic is equivalent to the analysis of a multi-band image, as in remote sensing.

38Finally, it should be noted that to obtain a robust statistical basis, a large number of samples would be necessary, both on the head and on the body. However, as we are working on a museum piece, it has already been exceptional to be able to benefit from 3 small rock fragments. Portable XRF could be of great interest to distinguish anything other than quartz. Foraminifera in a carbonate matrix would remain undifferentiated. If more sample was available, petrographical and micro paleontological tools that were used could be completed by other petrophysical methods such as mercury porosity, isotopic, trace and major element measurements and other micro-fossils or minerals either specific or frequent if available. All these tools may also be developed depending on the mineralogical nature and the quantity of the available material.

Conclusions

39In conclusion, after reviewing the results from the different proxies used (glauconite, quartz, foraminifera, and coccolith), we found no significant variations among the three samples. The only differences noted are interpreted as subtle facies variations, evidenced by the significant differences between the characteristics of the two samples from the recumbent head.

40We attempted to juxtapose proxies that have never been used together for this type of study before. Unfortunately, the samples we possess are too small for further analysis. The chalky facies presented a real challenge in obtaining a clear answer, and the cumulative evidence does not allow us to definitively conclude whether the head belongs to the body of the recumbent figure. Given the current state of evidence, we still consider the sculpture study and the historical study, which affirmed the concordance of the two elements, to be valid as we have no counterarguments. The use of these proxies should be tested in other facies.

Haut de page

Bibliographie

Alhonen, P., Väkeväinen, L., 1981. Diatom analytical studies of Early Comb ceramic vessels from Åland. Suomen Museo, 1981: 67‑78.

Armynot du Châtelet, E., Frontalini, F., Francescangeli, F., 2017. Significance of replicates: Environmental and paleoenvironmental studies on benthic foraminifera and testate amoebae. Micropaleontology, 63, (5): 257‑274.

Armynot du Châtelet, E., Frontalini, F., Guillot, F., 2014. A new methodological approach to evaluate the mass and volume distributions of agglutinated foraminiferal tests: a pilot investigation with Textularia earlandi. Micropaleontology, 60, (1): 5-12.

Armynot du Châtelet, E., Frontalini, F., Guillot, F., Recourt, P., Ventalon, S., 2013. Surface analysis of agglutinated benthic foraminifera through ESEM-EDS and Raman analyses: an expeditious approach for tracing mineral diversity. Marine Micropaleontology, 105: 18-29.

Beaussart, P., Dion, M.-P., Duée, R., 1987. Richesses des anciennes églises de Valenciennes. Musée des beaux-arts de Valenciennes 28 nov. 87-28 fév. 88. Musée des beaux-arts (Valenciennes, Nord), Valenciennes.

Brown, J. R., Gowen, R. J., McLusky, D. S., 1987. The effect of salmon on the benthos of a Scottish sea loch. Journal of Experimental Marine Biology and Ecology, 138: 143‑157.

Burnett, J. A., 1998. Upper Cretaceous. In P.R., B. (dir.). Calcareous nannofossil Biostratigraphy. Kluwer Academic Publishers, London.

Burton, M. M., Quinn, P. S., Tamberino, A., Levy, T. E., 2018. Ceramic composition at chalcolithic Shiqmim, northern Negev Desert, Israel: Investigating technology and provenance using thin section petrography, instrumental geochemistry and calcareous nannofossils. Levant, 50, (2): 237‑257.

Cassar, J., 2010. The use of limestone in a historic context – the experience of Malta. In Smith, B. J., Gomez-Heras, M., Viles, H. A., Cassar, J. (dir.). Limestone in the Built Environment: Present-Day Challenges for the Preservation of the Past. The Geological Society of London, London.

Delvigne, V., 2019. Barremian–Bedoulian flint humanly transported from the west bank of the Rhône to the Massif-Central Highlands–A diachronic perspective. C. R. Palevol, (18): 90‑112.

Delvigne, V., Fernandes, P., Tuffery, C., Raynal, J.-P., Klaric, L., 2020. Taphonomic methods and a database to establish the origin of sedimentary silicified rocks from the Middle-recent Gravettian open-air site of La Picardie (Indre-et-Loire, France). Journal of Archaeological Science, 32: 102369.

Deneuve, E., Paris, C., Jacquier, J., Lancelot, S., Lascour, V., Hérisson, D., Gardin, S., Brasseur, B., Godart, A., Fagnart, J.-P., Coudret, P., Vallin, L., Masson, B., White, R., Christensen, M., 2018. Objets en craie du gisement d’Amiens-Renancourt 1. De la fabrication à la mise en valeur. Rapport d’activité 2017 du Projet Collectif de Recherche. Ministère de la culture et de la communication.

Desoignies, J., 1967. Carte géologique (Geological map) au 1:50000, feuille de Valenciennes XXXVI-6.

Dollé, L., 1924. Etude sur les eaux souterraines de la région de Cambrai. Université de Lille, Faculté des Sciences de l’Université de Lille, Lille.

Dunham, R. J., 1962. Classification of Carbonate Rocks According to Depositional Textures. Memories of the American Association of Petroleum Geologists, Special Volumes, 1: 108‑121.

Falkenberg, J., Mutterlose, J., 2021. Calcareous nannofossils in medieval mortar and mortar-based materials: A powerful tool for provenance analysis. Archaeometry, 63, (1): 19-39.

Fernandes, P., Raynal, J.-P., 2006. Pétroarchéologie du silex : un retour aux sources. C. R. Palevol, (5): 829-837.

Fiorentino, A., 1998. The potential of nannofossil analysis applied to archaeological studies: The case of the Riace’s Bronzes. Journal of Nannoplankton Research, 20 : 101‑103.

Fournier, F., Ouass, A., Rochette, P., Bromblet, Ph., Léonide, Ph., Conesa, G., Marié, L., Boularand, S., D’Ovidio, A.-M., Vigié, B., Del Furia, L., Lemke, A.-L., Niaissa, T., Tendil, A., Fleury, J., 2023. Provenance of sculptural limestones in protohistoric Provence (SE France): Insights from carbonate rock petrography and stable isotope geochemistry. Journal of Archaeological Science: Reports, 48: 103922.

Gardin, S., Krystyn, L., Richoz, S., Bartolini, A., Galbrun, B., 2012. When and where the earliest coccolithophores? Lethaia, 45: 507‑523.

Gibson, A. M., 1983. Preliminary results of diatom analysis of clays and pre-historic pottery from the Millfield basin. Northern Archaeology, 4: 33‑44.

Goodman, B. N., Reinhardt, E. G., Dey, H. W., Boyce, J. I., Schwarcz, H. P., Sahoglu, V., Erkanal, H., Artzy, M., 2009. Multi-proxy geoarchaeological study redefines understanding of the paleocoastlines and ancient harbours of Liman Tepe (Iskele, Turkey). Terra Nova, 21: 97‑104.

Hardy, A., Beaussart, P., 1980. Peintures et sculptures gothiques du couvent des dames de Beaumont à Valenciennes. Revue du Nord, 62, (247): 903‑914.

Hesselbo, S. P., Huggett, J. M., 2001. Glaucony in Ocean-Margin Sequence Stratigraphy (Oligocene-Pliocene, Offshore New Jersey, U.S.A.; ODP Leg 174A). Journal of Sedimentary Research, 71, (4): 599‑607.

Hughes, R. E., Berres, T. E., Moore, D. M., Farnsworth, K. B., 1998. Revision of Hopewellian Trading Patterns in Midwestern North America Based on Mineralogical Sourcing. Geoarchaeology, 13, (7): 709‑729.

Kassambara, A., 2020a. ggpubr: “ggplot2” Based Publication Ready Plots. R package version 0.2.5.

Kassambara, A., 2020b. rstatix: Pipe-Friendly Framework for Basic Statistical Tests. R package version 0.6.0.

Lafuente, B., Downs, R., Yang, H., Stone, N., 2015. The power of databases: the RRUFF project. Highlights in Mineralogical Crystallography. W. De Gruyter, Berlin, Germany.

Le Bouck, S., 1650. Sépultures et épitaphes qui se retrouvent en l’avant dicte Eglise de Beaumont. Histoire ecclésiastique de la ville et Comté de Valentienne. Valencienne.

Lübke, N., Mutterlose, J., Börste, N., Kaplan, U., 2018. A micropalaeontologically based provenance analysis of masonry and floor tiles from the medieval cathedral of Paderborn (northern Germany). Archaeometry, 60, (6): 1170‑1183.

Magid, A. A., Krzywinski, K., 1988. Extraction of pollen from pottery. Arkeologiske Skrifter, 4: 196‑204.

Matiskainen, H., Alhonen, P., 1984. Diatoms as indicators of provenance in Finnish sub-Neolithic pottery. Journal of Archaeological Science, 11: 147‑157.

Newman, R., 2005. An Overview of Current Scientific Research on Stone Sculpture. Scientific Examination of Art: Modern Techniques in Conservation and Analysis. National Research Council. Proceedings of the National Academy of Sciences, Washington, DC: 58-71.

Nys, L., 2023. Le gisant présumé de Beatrix d’Avesnes (février 1321), mère de l’empereur Henri VII de Luxembourg, au musée des Beaux-Arts de Valenciennes. Revue du Nord, 105: 331-353.

Oursel, H., 1978. Sculptures romanes et gothiques du Nord de la France. Catalogue dans le cadre de l’exposition consacrée aux sculptures romanes et gothiques du Nord de la France au Musée des Beaux-Arts de Lille. 72: 143‑145.

Perch-Nielsen, K., 1985. Mesozoic calcareous nannofossils. In Bolli, H. M., Saunders, J. B. ,. Perch-Nielsen, K. (dir.). Plankton Stratigraphy. Cambridge University Press, Cambridge.

Persico, D., Cogliati, C., 2014. Micropalaeontology in archaeology: reconstruction of provenance of a prehistoric artefact through microfossil and lithological analyses. Journal of Nannoplankton Research, 33, (1): 13‑17.

Quinn, P. S., Alaimo, R., Montana, G., 1998. Calcareous nannofossil analysis of ceramics and probable raw materials from an ancient punic kiln site on the island of Mozia, western Silicy. Journal of Nannoplankton Research, 20, (2): 85‑87.

Quinn, P. S., Day, P. M., 2007. Calcareous microfossils in Bronze Age Aegean ceramics: Illuminating technology and provenance. Archaeometry, 49: 775‑793.

R Core Team, 2017. R: A language and environment for statistical computing. R Foundation for Statistical Computing. Vienna, Austria.

Ramdani, A. I., 2023. Multi-scale geophysical characterization of microporous carbonate reservoirs utilizing machine learning techniques: An analog case study from an upper Jubaila formation outcrop, Saudi Arabia. Marine and Petroleum Geology, 152: 106234.

Riederer, J., 2004. Thin section microscopy applied to the study of archaeological ceramics. Hyperfine Interactions, 154: 143‑158.

Schindelin, J., Arganda-Carreras, I., Frise, E., Kaynig, V., Longair, M., Pietzsch, T., Preibisch, S., Rueden, C., Saalfeld, S., Schmid, B., Tinevez, J.-Y., White, D. J., Hartenstein, V., Eliceiri, K., Tomancak, P., Cardona, A., 2012. Fiji: an open-source platform for biological-image analysis. Nature methods, 9, (7): 676‑682.

Sissingh, W., 1977. Biostratigraphy of Cretaceous calcareous nannoplankton. Geologie en Mijnbouw, 65: 37‑65.

Tolboom, H., Dusar, M., Dubelaar, W., Dreesen, R., Elsen, J., Groessens, E., van der Star, C., 2009. Avendersteen.

Tribovillard, N., Bout-Roumazeilles, V., Abraham, R., Ventalon, S., Delattre, M., Baudin, F., 2023. The contrasting origins of glauconite in the shallow marine environment highlight this mineral as a marker of paleoenvironmental conditions. Comptes Rendus. Géoscience, 355, (S2): 213‑228.

Wickham, H., 2017. tidyverse, collection of package for data science.

Wilkinson, I. P., Tasker, A., Gouldwell, A., Williams, M., Edgeworth, M., Zalasiewicz, J., Christie, N., 2010. Micropalaeontology reveals the source of building materials for a defensive earthwork (English Civil War?) at Wallingford Castle, Oxfordshire. Journal of Micropalaeontology, 29: 87‑92.

Wilson, J. L., 1975. Carbonate Facies in Geologic History. New York.

Wisseman, S. U., Moore, D. M., Hughes, R. E., Hynes, M. R., Emerson, T. E., 2002. Mineralogical Approaches to Sourcing Pipes and Figurines from the Eastern Woodlands, U.S.A. Geoarchaeology, 17, (7): 689‑715.

Zaykov, V. V., Bushmankin, A. P., Yuminov, A. M., Zaykova, E. V., Zdanovich, G. B., Tairov, A. D., Herrington, R. J., 1999. Geoarchaeological research into the historical relics of the South Urals: problems, results, prospects. In Pollard, A. M. (dir.). Geoarchaeology: exploration, environments, resources. Geological Society, London.

Haut de page

Table des illustrations

Titre Figure 1: Pictures of the recumbent. Figure 1 : Photos du gisant.
Légende a) General view with a positioning attempt of the head and the body; the neck is missing; b) Detailed view of the head with details of the gimp; c) Detailed view on the mutilated puppies or lions at the feet of the lady; d, e, f) detailed view of the positioning of the sampling, one from the neckline of the body (sample BODY) and two from the back of the base of the head (HEAD 1 and HEAD 2). a) Vue générale avec une tentative de positionnement de la tête et du corps ; le cou est manquant ; b) Vue détaillée de la tête avec détails de la guimpe ; c) Vue de détail sur les chiots ou les lions mutilés aux pieds du gisant ; d, e, f) vue de détail du positionnement des prélèvements, un sur l’encolure du corps (échantillon BODY) et deux sur l’arrière de la base de la tête (échantillons HEAD 1 et HEAD 2).
URL http://journals.openedition.org/archeosciences/docannexe/image/12426/img-1.jpg
Fichier image/jpeg, 722k
Titre Figure 2: Image analysis methods for extracting the quartz from the SEM-EDS image (a and b) and the glauconite from thin sections (c, d and e). Figure 2 : Méthodes d’analyses d’images pour extraire le quartz de l’image MEB-EDS (a et b) et la glauconite des lames minces (c, d et e).
Légende Sample BODY is taken as an example. See text for details of the image processing used. L’exemple est tiré de l’échantillon BODY. Se référer au texte pour le détail des traitements d’image employés.
URL http://journals.openedition.org/archeosciences/docannexe/image/12426/img-2.jpg
Fichier image/jpeg, 745k
Titre Figure 3: Pictures from the thin sections. Figure 3 : Photos des lames minces.
Légende a) an occurrence of the large foraminifera polymorphinid or even a robertinid from HEAD1 sample; b1) glauconite grains occurrence from HEAD2 with a carbonate matrix; b2) quartz grains; b3) and dark nodules; c1) urchin plate; d1) tooth, d2) biserial micro-foraminifera, d3) bivalve bioclast; e1) tooth, e2) trochispiral planktonic micro-foraminifera; f) Optical micrographs of key calcareous nannofossil species: 1) Eprolithus moratus; 2) E. floralis; 3) Quadrum gartneri; 4) Eiffellithus eximius; 5) Lucianorhabdus maleformis; 6) Ahmuellerella octoradiata; 7) Kamptnerius magnificus; 8) Gartnerago obliquus; g) SEM micrographs of key calcareous nannofossil species: 1) Coccosphere of Watznaueria sp; 2) Eprolithus moratus; 3) E. eximius, distal view; 4) Gartnerago obliquuum; 5) Eiffellithus eximius, proximal view; 6) Tranolithus orionatus. a) occurrence du grand foraminifère polymorphinide ou robertinide (échantillon HEAD1) ; b1) occurrence de grains de glauconite avec une matrice carbonatée (HEAD2) ; b2) grains de quartz ; b3) nodules sombres ; c1) plaque d’oursin ; d1) dent, d2) fragment de micro-foraminifère, d3) fragment de bivalve ; e1) dent, e2) micro-foraminifère planctonique trochispiralé ; f) Micrographies optiques des principales espèces de nannofossiles calcaires : 1) Eprolithus moratus ; 2) E. floralis ; 3) Quadrum gartneri ; 4) Eiffellithus eximius ; 5) Lucianorhabdus maleformis ; 6) Ahmuellerella octoradiata ; 7) Kamptnerius magnificus ; 8) Gartnerago obliquus ; g) Micrographies MEB des principales espèces de nannofossiles calcaires : 1) Coccosphère de Watznaueria sp ; 2) Eprolithus moratus ; 3) E. eximius, vue distale ; 4) Gartnerago obliquuum ; 5) Eiffellithus eximius, vue proximale ; 6) Tranolithus orionatus.
URL http://journals.openedition.org/archeosciences/docannexe/image/12426/img-3.jpg
Fichier image/jpeg, 762k
Titre Figure 4: Box and whisker plot for measurement on glauconite grains, quartz grains and foraminiferal individuals. Figure 4 : Boîtes à moustaches calculées à partir des mesures sur les grains de glauconite, les grains de quartz et les individus de foraminifères.
Légende The box presents the interquartile Q1-Q3 with the median as intermediate line. Whiskers present minimum and maximum values. For each of the three thin section from the recumbent : (a) proportion of surface occupied by glauconite, (b) Number of glauconites per mm², (c) area of the quartz grains, (d) density of foraminifera per mm². For more visibility log scale are used when necessary. Le rectangle représente l’interquartile Q1-Q3 avec la médiane comme ligne intermédiaire. Les moustaches présentent des valeurs minimales et maximales. Pour chacun des trois échantillons : (a) proportion de surface occupée par la glauconite, (b) nombre de glauconites par mm², (c) proportion de surface de quartz, (d) densité de foraminifères. Pour plus de lisibilité, une échelle logarithmique est utilisée lorsque nécessaire.
URL http://journals.openedition.org/archeosciences/docannexe/image/12426/img-4.jpg
Fichier image/jpeg, 358k
Titre Figure 5: Summary of the relationships tested in between the three samples BODY, HEAD1 and HEAD2, throughout the five petrographical and micropaleontological proxies. Figure 5 : Synthèse des relations testées entre les trois échantillons BODY, HEAD1 et HEAD2, à travers les cinq proxys pétrographiques et micropaléontologiques.
URL http://journals.openedition.org/archeosciences/docannexe/image/12426/img-5.jpg
Fichier image/jpeg, 617k
Haut de page

Pour citer cet article

Référence papier

Éric Armynot du Châtelet, Michel Dubois, Silvia Gardin, Ludovic Nys, Vincent Hadot, Renaud Toullec, Sandra Ventalon et Philippe Recourt, « Multiproxy approach based on image analysis, petrography, mineralogy and micropaleontology to compare carved limestone fragments, application to the study of a broken stone recumbent (Beatrix d’Avesnes, Valenciennes) »ArcheoSciences, 48-1 | 2024, 17-30.

Référence électronique

Éric Armynot du Châtelet, Michel Dubois, Silvia Gardin, Ludovic Nys, Vincent Hadot, Renaud Toullec, Sandra Ventalon et Philippe Recourt, « Multiproxy approach based on image analysis, petrography, mineralogy and micropaleontology to compare carved limestone fragments, application to the study of a broken stone recumbent (Beatrix d’Avesnes, Valenciennes) »ArcheoSciences [En ligne], 48-1 | 2024, mis en ligne le 03 janvier 2026, consulté le 13 avril 2026. URL : http://journals.openedition.org/archeosciences/12426 ; DOI : https://doi.org/10.4000/12w35

Haut de page

Auteurs

Éric Armynot du Châtelet

Laboratoire d’Océanologie et de Géosciences, UMR 8187 LOG, U. Lille, CNRS, ULCO, IRD. Cité Scientifique, 59650 Villeneuve d’Ascq, France

Michel Dubois

Univ. Lille, Institut Mines-Télécom, Univ. Artois, Junia, ULR 4515 – LGCgE, Laboratoire de Génie Civil et géo-Environnement, 59000 Lille, France

Silvia Gardin

Centre de Recherche en Paléontologie – Paris (CR2P), Sorbonne Université/CNRS/MNHN. Campus Pierre et Marie Curie, case 104, 5 Place Jussieu, 75252 Paris, France

Ludovic Nys

Univ. Polytechnique Hauts-de-France (Valenciennes), CRISS, Centre de Recherche interdisciplinaire en Sciences de la Société, 59300 Valenciennes, France

Vincent Hadot

Musée des Beaux-Arts de Valenciennes et Service Archéologique Municipal, Boulevard Watteau, 59300 Valenciennes, France

Renaud Toullec

UniLaSalle, B2R – U2R 7511, Geosciences Department (GEOS), 19 rue Pierre Waguet, BP 30313, 60026 Beauvais CEDEX, France

Sandra Ventalon

Laboratoire d’Océanologie et de Géosciences, UMR 8187 LOG, U.Lille, CNRS, ULCO, IRD. Cité Scientifique, FR-59650 Villeneuve d’Ascq, France

Philippe Recourt

Laboratoire d’Océanologie et de Géosciences, UMR 8187 LOG, U. Lille, CNRS, ULCO, IRD. Cité Scientifique, 59650 Villeneuve d’Ascq, France

Haut de page

Droits d’auteur

CC-BY-NC-ND-4.0

Le texte seul est utilisable sous licence CC BY-NC-ND 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont susceptibles d’être soumis à des autorisations d’usage spécifiques.

Haut de page
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search