The ornamental stones of the Roman thermal baths of Teate Marrucinorum (Chieti, Italy): autoptic, geochemical and minero-petrographic multi-analytical characterisations

doi:10.21203/rs.3.rs-7979889/v1

Download PDF

Research Article

The ornamental stones of the Roman thermal baths of Teate Marrucinorum (Chieti, Italy): autoptic, geochemical and minero-petrographic multi-analytical characterisations

https://doi.org/10.21203/rs.3.rs-7979889/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Ancient Romans exploited aesthetic natural stones from many sites across their Empire around the Mediterranean, transporting them for thousands of kilometres, to decorate their buildings. Petrographically, these metamorphic, sedimentary and igneous rocks display considerable differences, ranging from simple white stones to vividly coloured lithotypes. The source Region of these coloured stones is typically reconstructed from autoptic (visual and comparative examination of macroscopic or hand-sample scale) determinations, also tacking advantage from the personal expertise of some specialists of the sector.

Here, the ornamental stones of the 2nd-century A.D. thermal baths of Teate Marrucinorum (Chieti, Abruzzo region, Italy) are considered and investigated by different and complementary methods. According to their autoptic features or mesoscopic textures, the initial 56 samples were divided into three categories: polychrome stones, grey-striped and white crystalline marbles. These rocks were analysed via bulk autoptic, mineralogical (XRPD) and geochemical (XRF) methods; also, representative thin sections were used for transmission optical microscope (TOM) petrographic and micro-Raman determinations. The δ¹⁸O and δ¹³C isotopic signatures were also characterised for white and grey-striped marbles.

The complementary and multi-analytical approach unveils that the grey-striped marble is Greco Scritto, the white marbles come from Carrara (Marmor Lunense) and Marmara Island (Marmor Proconnesium) sources, whilst the four polychrome stones correspond to Pavonazzetto Antico (Marmor Phrygium), Cipollino Verde (Marmor Carystium), Portasanta (Marmor Chium) and Breccia di Settebasi (Marmor Scyreticum). The coupling of qualitative observations with quantitative measurements further constrains the provenance and features of aesthetic rocks employed in the ancient town by the Romans.

Roman baths

polychrome stones

white marbles

autoptic

XRPD and XRF analyses

Rocks and sediments have been exploited for millennia to decorate artefacts and obtain building materials. During Roman times, huge amounts of lithologies were quarried and transported to build and decorate the external and internal surfaces of baths, theatres, villas, streets and monuments in general, as well as for statuary, either from surrounding areas or from many and faraway sites (Pensabene 1993, 2002). The importance of these activities and prices concerning ornamental stones is marked by the “Edictum de Pretiis Rerum Venalium” (Diocletian, 301 A.D.) (Gnoli 1988). The Romans distinguished Marmora, i.e. stones susceptible to being polished (mirror-like surface), from lapides, which could not be polished. These raw materials include i) tenacious sedimentary limestones, conglomerates, travertines and alabasters, ii) metamorphic lithotypes (like breccias, marbles) and iii) igneous rocks (lavas, lithified pyroclasts, granitoids stones, etc.) (Lazzarini 2007; Taelman et al. 2023). Petrographically, marbles are metamorphic rocks subdivided into pure and impure marbles if the carbonate minerals are > 95 or > 50 plus < 95 vol.%, respectively (Fettes and Desmons 2007). In archaeology, marbles are instead frequently defined following the Latin mining and thus unrelated to mineralogical and petrographic criteria; consequently, marmora are lithotypes characterised by high aesthetic, cutting and polishing properties, often divided into white or stricto sensu (s.s.) versus coloured or lato sensu (l.s.) (European Commission - Extra-Expo project 2015). Here, Roman white and polychrome marbles, along with a third grey-striped marble category, sampled from the ancient Teate Marrucinorum, the modern city of Chieti, (Abruzzo, Italy) (Figs. 1a,b), are investigated through a multi-analytical approach. The identification and provenance of the polychromatic samples were first established across mesoscopic textural and structural attributes by i) analysing polished rock slabs through specific autoptic expertise or ii) using comparative criteria with well-known and previously catalogued samples and/or classical treatises (Corsi 1845; Gnoli 1988; Borghini 1989). By contrast, the provenances of marble s.s. samples require additional petrographic and stable δ¹³C‰ and δ¹⁸O‰ isotopic determinations, in line with the updated protocol for white marble and grey varieties (Lazzarini 2007; Antonelli et al. 2009a; Yavuz et al. 2011; Antonelli and Lazzarini 2013, 2015). In parallel, all three marble categories were also distinguished according to their microscopic textures and structures determined via TOM (transmission optical microscope) observations on representative thin sections. These results and conclusions were complemented with quantitative mineralogical (XRPD and Raman) and geochemical (XRF-EDS) determinations to gain further characteristics of these rocks. This study enriches the characterisation portfolio of Greco scritto, white marbles and some polychrome lithologies present in the Teate Marrucinorum thermae and other archaeological sites (Agostini and Rossi 2012). In general, it provides an extended and complementary quantitative approach to analyse Roman ornamental polychrome stones and to constrain their geochemical, mineralogical and petrographic attributes.

Historical and archaeological context. The Roman Teate Marrucinorum, the former city of Chieti, developed on an italic settlement (Fig. 1a) favoured by the geographical strategic context dominating the Aterno-Pescara valley and the Tiburtina Valeria Roman road (Agostini 2018). The Teate Marrucinorum settlement thrived upon the hill of the periadriatic belt of Abruzzo, made of Plio-Pleistocene permeable sandy association, overlaying poorly permeable clays (ISPRA 2012); this geological association favoured the emergence of some natural springs nodal for ancient urban sites. In fact, the archaeological remains of Teate Marrucinorum include many important monuments, such as the Roman theatre and the Civitella’s amphitheatre, the sacred area of the temples and thermal baths complex with the cisterns; the latter site was accurately decorated with aesthetic stony panels, pilasters and mosaic (Agostini et al. 2002). These thermal baths were built between the 1st and 2nd century AD on the eastern slope of the hill (Fig. 1b) and associated with an extensive and nine-vaulted cistern system fed by waters tapped from a close spring (Fig. 1c-d). Unfortunately, this part of the ancient city underwent landslides and erosion that led to the collapse of the eastern part of the complex (Agostini 2018).

Nowadays, the thermal baths and cisterns are constituted of a few rooms accessible via a staircase from a transverse corridor (vestibulum) that led, through a colonnaded entrance, to a large room, paved with mosaic decoration (the probable apodyterium) and connected by an aisle with the praefurnium to produce heat (Fig. 1c-d) Three chambers used for hot baths, a tepidarium and two calidaria, are in the southern part, while the room for cold baths (frigidarium) equipped with semicircular tubs is near the apodyterium, which was partially ruined after the landslide between the 2nd and 4th century AD (Adinolfi et al. 2019). The large rooms were decorated with polychrome marble slabs and mosaics of great value, large carved columns and refined sculptures, demonstrating the importance of the Teate Marrucinorum.

We first collected as many different types of stones as possible; they are single rectangular-shaped pieces and only a fraction of them has a precise original location (Tab. S1), i.e. the specimens from the tepidarium, apodyterium and frigidarium are respectively 1, 12 and 28, while those from unknown sites are 15 (Fig. 1d). These 56 rock fragments were sampled and divided into 3 main categories: 6 white marbles, 5 grey-striped marbles and 45 polychrome stones (Tab. S1). The specimens of the latter category were visually compared with the most similar and known ornamental catalogued samples according to an expert-based judgement; moreover, available images in several databases were considered from references or by comparing samples to images of polychrome stones previously identified. Due to their mesoscopic heterogeneity, each sample was first photographed and then divided into two halves: i) one for further hand-specimen textural observations and Raman determinations and ii) the other for producing homogeneous powders for XRPD and XRF-EDS analyses, as well as representative thin sections using as small as possible specimen, on the order of 5 x 5 x 1 cm (Figs. 2a,b).

The microscopic textural (number, abundance, colour, size, shape and boundary of mineralogical phases) and structural (arrangement of phases in space) features were analysed using a polarised transmission optical microscope (TOM) Zeiss Axioskop 50 (Dept. INGEO, University “G. d’Annunzio”). The microscope is equipped with a Qimaging MicroPublisher 3.3 RTV digital camera linked to a computer with image analysis software Image-Pro Plus v.6.0 (Media Cybernetic Inc.) (Potere et al. 2023; Radica et al. 2024). Although the microscopic features are usually investigated only for white marble samples (marble s.s.), here they are adopted and the Maximum Grain Size (MGS) was measured for every stone, thus expanding the number of quantitative parameters and implementing the existing databases.

Bulk rock geochemistry of powdered samples, i.e. major and minor oxides, was quantified with a PANalytical Epsilon 3-XL X-ray fluorescence energy-dispersive spectrometer (XRF-EDS) (University of Milano-Bicocca). The X-ray source is a Rh anode operating at variable kV (4 to 50) and µA (1 to 3000) conditions. Loss on ignition (LOI) was determined by heating the samples at 1050°C for 5 hours. The collected data were preliminary analysed with the Malvern Panalytical Epsilon 3 software platform, using the Omnian-standardless model, which allows qualitative and quantitative chemical analysis of unknown materials without the construction of calibration curves. The quantitative analysis was then repeated in 6 different instrumental conditions, using the Panalytical WROXI® – synthetic, high-quality Certified Reference Materials for calibration.

Stable isotope ratios of oxygen and carbon were assessed using a Gasbench II preparation system coupled with a ThermoFinnigan Five Plus mass spectrometer operating in continuous flow mode, following the protocol described by McCrea (1950). The analysed powders were reacted with pure phosphoric acid at 70° C. Isotopic ratios are reported as δ¹³C‰ and δ¹⁸O‰ values relative to the Vienna-Pee Dee Belemnite (V-PDB) standard. These ratios were compared with the isotopic published dataset of white marbles and Greco scritto varieties (Antonelli et al. 2009a-b; Yavuz et al. 2011; Antonelli and Lazzarini 2015; Perna et al. 2023).

X-ray powder diffraction (XRPD) was carried out with an X’Pert PRO PANalytical diffractometer in Bragg-Brentano geometry (University of Milano-Bicocca). Each pattern was recorded in the 3°-71° angular range of 2θ, with a step size of 0.02° and a counting time of 2 s/step; the Cu Kα1 = 1.5406 Å incident radiation was produced with 40 kV and 40 mA. Data elaboration was made with X’Pert Highscore v.2.1 software (Malvern Panalytical, Malvern, UK), using the ICDD PDF2-2004 database, following the typical procedure of search-match reported in several other studies (Iezzi et al. 2003; Della Ventura et al. 2005). A semi-quantitative evaluation of the abundance of crystalline phases (wt.%) was assessed using the Reference Intensity Ratio (RIR) method (Hubbard and Snyder 1998; Chipera and Bish 2013), through which the intensity scaling factor of each mineralogical phase (I) was compared with a “virtual” corundum crystalline phase (I_cor.) not necessarily present in the XRPD patterns. Then, the semi-quantitative content (wt.%) of each crystalline phase is pursued by the ratio “I/I_cor” (Galderisi et al. 2022).

The XRPD results were complemented by unpolarised Raman spectroscopy measurements carried out on a few µm² portions of the specimens, allowing the textural identification of the most prominent carbonate and silicate minerals. Spectra were collected in the frequency range of 30-1550 cm^− 1 using a coherent sapphire single-frequency 488 nm laser as excitation source, and an iHR 320 Horiba spectrometer coupled with a Sincerity UV–Vis CCD camera (Institut für Glas und Keramik, Erlangen). The black crosses shown in the photos in Figs. 2a,b represent the Raman spots. For each sample, between 8 and 12 different spots were investigated with an OptoSigma PAL-50-L NA 0.42 objective. Spectra were background-subtracted with a linear function and normalised to the total area. CaCO₃ (Sigma Aldrich) was measured as well under the same conditions and used as a reference.

Autoptic determinations. The polychrome stones present different fabrics (textural and structural characteristics) (Fig. 2a). PNORN-1 and PNORN-2A have a saccharoidal texture, whilst the remaining stones have a brecciated to slightly deformed appearance. PNORN-1 has large purple veins that run across the white areas; PNORN-2A presents cm-sized parallel to slightly curved bands with greenish to whitish tones (Fig. 2a). PNORN-X1 shows a brecciated aspect made of centimetric white to dark red clasts in a pinkish matrix. PNORN-2C-1 and PNONR-2C-2 samples show similar characteristics, with many large, closely spaced white clasts (from a few to many cm) in a dark red to black matrix (Fig. 2a). PNORN-2C-3 samples also show similar characteristics, plus yellowish to dark red clasts. Compared to all the other rock samples, PNORN-4 has much darker hues and the white clasts have limited size compared to the other polychrome samples (Fig. 2a). Autoptic analysis indicates that stones of the thermal baths of Teate Marrucinorum correspond to 4 well known lithotypes: Pavonazzetto antico (Marmor Phrygium, PNORN-1), Cipollino Verde (Marmor Carystium, PNORN-2A), Portasanta (Marmor Chium, PNORN-X1) and Breccia di Settebasi (Marmor Scyreticum, PNORN-2C-1, PNORN-2C-2, PNORN-2C-3 and PNORN-4) (Tab. S2 and Fig. 2a). These outcomes corroborate with previous determinations reported in Agostini et al. (2002). The grey striped marbles (PNORN-3 series, Tab. S2) exhibit a white background with numerous dark grey to light grey, thin, convoluted veins, typical of the Greco scritto marble varieties (Fig. 2b). Instead, the third category of white marbles cannot be determined only by their mesoscopic appearance. PNORN-2B, PNORN-X2 and PNORN-X2-c exhibit a white tone, tending towards pale grey or bluish tones, frequently with graphitic parallel foliation or crossed by linear grey veins, whereas PNORN-X2-a, PNORN-X2-b and PNORN-X2-d samples have a warmer white tone, often with ivory or milky hues; the former three samples have calcite crystals larger than the latter three ones (Fig. 2b).

TOM. TOM observations reflect the first qualitative autoptic observations (Tab. S2). Calcite crystals are by far the most abundant mineral in all samples. PNORN-1 (Pavonazzetto antico) shows a heteroblastic texture with embayed crystals up to an MGS of 2.7 mm, plus polysynthetic twinning with some bendings, mosaic microstructure and recrystallised fractures; tiny and sparsely distributed opaque to black/brown/reddish unidentifiable (by TOM) crystals also occur (Fig. S1). Calcite crystals in Cipollino verde (PNORN-2A) vary from hetero- to homeoblastic, with an MGS of 2.1 mm, a lineated microstructure and embayed crystal boundaries. Along the lineations, grey to green sheet silicates are located, visible also at the mesoscopic scale (Fig. 2a). Additionally, some quartz crystals and again opaque black zones are present. All the other polychrome stones are brecciated, with micritic to cryptocrystalline calcitic clasts preventing observations of their crystal boundaries (Tab. S2 and Fig. S1). Although the PNORN-X1 (Portasanta) is not a metamorphic rock, its sedimentary clasts in veins show MGS of 1.6 mm. Breccia di Settebasi samples (PNORN-2C-1, PNORN-2C-2, PNORN-2C-3 and PNORN-4) display marble angular clasts having calcite crystals of small dimension (~ 0.06 to 0.2 mm). These clasts show some zones in which there are a few larger calcite crystals; according to Lazzarini (2007), the MGS for these samples can consider the recrystallisation in these portions; following this approach, the MGS are 0.8, 0.7, 1.7 and 2.4 mm for PNORN-2C-1, PNORN-2C-2, PNORN-2C-3 and PNORN-4, respectively (Tab. S2 and Fig. S1). In addition, pre-kinematic larger calcite porphyroclasts are visible. These clasts are well separated by dark veins hosting unspecified reddish to black tiny crystals.

Greco scritto sample PNORN-3 has a heteroblastic texture with embayed to sutured calcite crystals, polysynthetic twinning, a mosaic microstructure and a MGS of 2.7 mm (Tab. S2 and Fig. S1); it shows few microcrystalline veins or bands of oriented impurities/opaque minerals, often associated with the grey carbonaceous veins visible to the naked eye (Fig. 2b). PNORN-2B, PNORN-X2 and PNORN-X2-c white marble samples have heteroblastic texture. PNORN-2B has a mortar microstructure with embayed to sutured calcite boundaries and a MGS of 3.7 mm (Tab. S2 and Fig. S1). PNORN-X2 and PNORN-X2-c have similar characteristics (mortar microstructure, embayed to sutured crystal boundaries); MGS is 2.6 mm and 2.9 mm, respectively (Tab. S2 and Fig. S1). PNORN-X2-a has a texture from homo- to heteroblastic, mosaic microstructure and curved to straight crystal boundaries, with calcite crystals having a maximum size of 0.8 mm (Tab. S2 and Fig. S1). PNORN-X2-b and PNORN-X2-d are both homeoblastic, characterised by a mosaic microstructure with both straight and curved crystal boundaries (Tab. S2 and Fig. S1). PNORN-X2-b tends to be polygonal, with triple joints and its MGS is 0.7 mm, whereas in PNORN-X2-d triple joints are rare, with the maximum size of calcite of 0.6 mm (Tab. S2 and Fig. S1).

XRF-EDS. Bulk oxide compositions of all the samples are plotted in Fig. 3a and compiled in Tab. S3. The sum of CaO and LOI (Loss On Ignition) is invariably higher than 90 wt.% (CaO: 52.1 to 57.4 wt.%), while the other oxides sum up between 0.6 and 8.3 wt.%. CaO + LOI is the lowest and the highest for the Cipollino verde PNORN-2A and the Greco scritto PNORN-3-b, respectively (Fig. 3a). Polychrome stones show higher values of SiO₂ + Al₂O₃ + Fe₂O₃ + MgO (7.7 wt.% for Cipollino, 3.2 wt.% for Pavonazzetto, PNORN-1; PNORN-2A; 3.1 wt.% for Portasanta, PNORN-X1; from 1.8 wt.% to 5.8 wt.% for Breccia di Settebasi, PNORN-2C-1, PNORN-2C-2, PNORN-2C-3 and PNORN-4) compared to Greco scritto (0.6–1.5 wt.%) and white marble varieties (0.8-1 wt.%) (Fig. 3a). Bulk compositions of white marbles and Greco scritto marbles are similar to each other, with PNORN-3-d having the relatively highest MgO and lowest CaO contents (Fig. 3a). The highest amount of SiO₂ is reached by Cipollino verde (PNORN-2A, SiO₂: 4.8 wt.%) and in general, is higher for polychrome samples (from 0.4 wt. % to 1.5 wt.%) and lower for Greco scritto and white marbles (Fig. 3a). Al₂O₃ is relatively rich in Pavonazzetto and Cipollino (PNORN-1 and PNORN-2A); the highest Fe₂O₃ and MgO amounts are in Breccia di Settebasi PNORN-4 and PNORN-2C-3, respectively (Fig. 3a and Tab. S3). SnO₂ and MnO are invariably below the detection limit (LOD ~ 0.05 wt.%), except for Cipollino verde (MnO: 0.10 wt.%); TiO₂ is appreciable only for Cipollino verde, Pavonazzetto and PNORN-4 Breccia di Settebasi. Again, K₂O and SrO are both detectable in Cipollino, while K₂O and SrO are individually present in Portasanta and PNORN-3-b Greco scritto, respectively (Tab. S3).

Stable isotopes of oxygen and carbon. The stable C and O isotope ratios of white marble samples are reported in Tab. S3 and plotted in Fig. 4, in two plots corresponding to fine-grained (MGS ≤ 2.00 mm) and medium- to coarse-grained (MGS > 2.00 mm) white marbles. Isotopic δ¹⁸O and δ¹³C ratios of medium-grained samples are respectively − 1.54‰ and 2.18‰ for PNORN-2B, -2.03‰ and 2.75‰ for PNORN-X2, -2.31‰ and 3.50‰ for PNORN-X2-c (Fig. 4). Marble varieties having MGS ≤ 2.00 mm have stable isotope values of -1.83‰ and 2.25‰ (PNORN-X2-a), -2.19‰ and 2.34‰ (PNORN-X2-b), -1.95‰ and 2.24‰ (PNORN-X2-d) for δ¹⁸O and δ¹³C, respectively (Fig. 4). Greco scritto specimens (PNORN-3 series) have δ¹⁸O ranging from − 3.03‰ to -2.52‰ (PNORN-3-c and PNORN-3-d, respectively), while minimum and maximum values for δ¹³C are 2.26‰ (PNORN-3-d) and 4.22‰ (PNORN-3-b) (Fig. 4).

XRPD. The XRPD patterns of the analysed stones are stacked in Figs. S2a,b, along with recognised minerals; the single XRPD patterns are reported in Fig. S4 to Fig. S21. The background is invariably flat and the Bragg reflections are extremely sharp and intense, indicating that all phases are highly crystalline and amorphous material is absent (Fig. S4 to Fig. S21). Calcite invariably shows the most intense Bragg peaks (Figs. S2a,b). The abundance of phases in wt.% per sample is tabulated in Tab. S4; the RIR quantification indicates that calcite ranges from 83.5 wt.% (PNORN-2A) to 99.0 wt.% (PNORN-X2-a and PNORN-X2-c) (Fig. 3b), whereas dolomite is relatively significant only for PNORN-X1 (Portasanta) and Breccia di Settebasi PNORN-2C-3, the two samples with the highest amounts of MgO (Figs. 3a,b). However, in other Breccia di Settebasi samples, dolomite is absent (Tab. S4). In Greco scritto, dolomite varies from 1 to 2 wt.%, while it is lacking in all white marble varieties (Tab. S4). The amount of quartz is low but present in all samples; it attains 0.5 wt.% in Pavonazzetto PNORN-1 and in Breccia di Settebasi samples, except for PNORN-2C-3, in which it arrives at 3 wt.% (Tab. S4). In Portasanta and Cipollino verde, the quartz content is 2.5 wt.% and 3.0 wt.%, respectively, while it is constant at 1 wt.% for white (less in PNORN-X2, qtz: 0.5 wt.%) and Greco scritto marbles (Tab. S4). Alkali-feldspars are also low but detectable for PNORN-1 (Tab. S4). Cipollino verde is also rich in muscovite and clinochlore in line with the highest SiO₂; similarly, PNORN-2C-3 and PNORN-4 samples have 3.0 wt.% clinochlore, while PNORN-4 hosts talc (Tab. S3 and Fig. 3a). Illite is 3.0 wt.% in Greco scritto PNORN-3 and PNORN-3-d; it is also detectable in medium-grained marbles PNRON-2B and PNORN-X2; it attains 4.0 wt.% in PNORN-X2-d fine-grained white marble (Tab. S4). Finally, hematite is detected in the four Breccia di Settebasi, while rutile is present only in PNORN-X2-b.

Raman spectroscopy. The micro-Raman data (Tab. S4 and Fig. S3) complement the XRPD results and enable the correlation of textural features with crystal–chemical attributes in the polychrome stones and the Greco scritto sample PNORN-3. The individual Raman spectra are displayed in Fig. S22 to Fig. S28. As expected, calcite displays the most intense Raman vibration centred at ∼1090 cm^-1 (black spectra in Fig. S3), but some other faint vibration modes are also present (red spectra in Fig. S3). In agreement with XRPD results, the Pavonazzetto specimen PNORN-1 exhibits the presence of alkali-feldspars, whereas the spectra of Cipollino and Portasanta (PNORN-2A and PNORN-X1, respectively) reveal the presence of quartz (Fig. S3). Unexpectedly, aragonite was detectable exclusively by Raman in PNORN-4 (Breccia di Settebasi specimen), which is also the only sample in which Fe oxides/hydroxides (attributable to hematite) could be identified (Fig. S3). Consistent with the XRPD data, in Greco scritto PNORN-3, dolomite was detected. Sheet-silicates are not determined by Raman, possibly due to their fine grain sizes and/or high localisation in domains.

White and Greco scritto marbles. The isotopic comparisons of white and grey-striped marbles with those present in reference databases are reported in Fig. 4, along with mineralogical (Figs. 3b, Fig. S2, S3) and especially petrographic (Tab. S2 and Fig. S1) parameters. They indicate that the white marbles analysed here have two different provenances: Carrara (Marmor Lunense) (PNORN-X2-a, PNORN-X2-b, PNORN-X2-d) and Marmara Island (Marmor Proconnesium) (PNORN-2B, PNORN-X2, PNORN-X2-c), corroborating the previous outcomes of Agostini et al. (2002). The provenance of the Greco scritto was instead undetermined in Agostini et al. (2002); our new isotopic results firmly hypothesise from the Asia Minor field (Fig. 4), considering provenance areas reported in Antonelli et al. (2009a), Yavuz et al. (2011) and the Murecine samples of Perna et al. (2023). In Fig. 4 our Greco scritto samples (PNORN-3, PNORN-3-a, PNORN-3-b, PNORN-3-c, PNORN-3-d) plus one from Agostini et al. (2002) plot close to the isotopic field recorded for marble from the Microasiatic quarry of Hasançavuşlar (near Ephesus, Turkey); furthermore, also our new TOM determinations match with an Ephesian origin for the Greco scritto samples of Teate Marrucinorum. Importantly, these data coupled with those reported in Antonelli et al. (2009b) and those recently published by Perna et al. (2023) suggest a possible enlargement of the Hasançavuşlar Greco scritto field towards higher δ¹⁸O and δ¹³C ratios (Fig. 4).

Polychrome stones. The polychrome stones are typically recognised only via autoptic determinations. Here, we compare and complement this aspect (Fig. 2) with geochemical (Fig. 3a and Tab. S3), mineralogical (Figs. 3b, Fig.S2, S3 and Tab. S4) and petrographic (Fig. S1 and Tab. S2) quantitative parameters. The Pavonazzetto polychrome PNORN-1 specimen is a marble s.s. (Tab. S2 and Fig. S1) due to its calcite > 95 wt.% with only minor alkali-feldspars and quartz (Figs. 3b, S2, S3 and Tab. S4); its heteroblastic mosaic texture is made of calcite grains with embayed contours and a MGS of 2.7 mm (Tab. S2, Fig. S1). These features are corroborated by high CaO and LOI, coupled with SiO₂ and Al₂O₃ contents close to 1 wt.% (Fig. 3a, Tab. S3). The opposite situation is presented by the impure (calcite < 95%) Cipollino verde PNORN-2A marble, the poorest in calcite (~ 84 wt.%) and the richest for the other remaining four minerals, i.e. dolomite, quartz, muscovite and clinochlore (Figs. 3b, S2, S3 and Tab. S4); coherently, the amount of SiO₂ + Al₂O₃ + Fe₂O₃ + MgO is the highest (Fig. 3a, Tab. S3). This stone has calcite grains invariably with embayed boundaries, displaying a (prevalently) heteroblastic mosaic texture; it is the unique rock with a foliated fabric and attains an MGS close to 2.1 mm (Tab. S2, Fig. S1). The Portasanta PNORN-X1 has the largest content of dolomite (5.5 wt.%) and a low, but significant amount of quartz (Figs. 3b, S2, S3 and Tab. S4); this paragenesis corroborates the relatively high contents of both SiO₂ and MgO > 1 wt.% (Fig. 2a, Tab. S3). Petrographically, the PNORN-X1 is a calcitic tectonic breccia, characterised by a brecciated texture of micritic/cryptocrystalline calcite clasts, with a (sedimentary, see above) MGS up to 1.6 mm in its veins (Tab. S2, Fig. S1). The four Breccia di Settebasi (PNORN-2C-1, PNORN-2C-2, PNORN-2C-3 and PNORN-4) samples are the same stone, although they appear different in hand-specimen at least for an unexpert eye (Fig. 2a); in fact, they represent the different facies of the same lithotype. They have the same brecciated and textural attributes of the marbles’ clasts; nonetheless, their MGS measured ranging from 0.7 mm up to 2.4 mm (Tab. S2, Fig. S1). These similarities are further agreed by their mineral contents, since quartz and hematite occur in all these stones; in addition, PNORN-2C-3 and PNORN-4 host also clinochlore, plus only dolomite in PNORN-2C-3 and talc for PNORN-4 (Figs. 3b, S2 and Tab. S4). These slight mineralogical differences are also reflected by SiO₂, Al₂O₃, Fe₂O₃ and MgO variations. These low but detectable geochemical, mineralogical and MGS differences are related to heterogeneities of the same original lithotype.

Comparisons of mineralogical, geochemical and MGS data from the literature. The differentiation of several ornamental Roman stones could be challenging to an inexperienced researcher. Thereby, to complement and further corroborate autoptic determinations, we compare the mineralogical, geochemical and MGS salient attributes of the lithotypes from the Roman thermal baths of Teate Marrucinorum (Chieti, Italy) (Figs. 5, 6 and 7). This approach could help integrate and enlarge the few existing databases to characterise these stones further. The mineralogy of white marbles and Greco scritto are mainly made of calcite (> 94 wt.%) plus minor amounts of dolomite (only Greco scritto, ≤ 2 wt.%) and quartz ± illite ± rutile, invariably ≤ 5 wt.% (Tab. S4); conversely, Pavonazzetto, Cipollino verde and Breccia di Settebasi polychrome stones host alkali-feldspar (Pavonazzetto), muscovite + clinochlore (≤ 13 wt.%), hematite ± talc clinochlore (≤ 6 wt.%), respectively, in addition to calcite, quartz and eventually dolomite (Tab. S4).

In Fig. 5, two coupled triangular diagrams and six binary plots of oxide ratios (SiO₂/Fe₂O₃, SiO₂/Al₂O₃, SiO₂/MgO, Al₂O₃/Fe₂O₃ and Al₂O₃/ MgO) are displayed, allowing a straightforward visualisation and discrimination of compositional clusters from isolated ones (Fig. 5). The polychrome Pavonazzetto sample (PNORN-1, black circle in Fig. 5) is invariably well separated in any of these eight plots, showing the highest value of Al₂O₃ between the polychrome samples. The Cipollino verde (PNORN-2A, red triangle in Fig. 5) is also perfectly separated from any other stones, except in the triangular SiO₂ vs Fe₂O₃ vs Al₂O₃ and in the SiO₂/Fe₂O₃ vs SiO₂/Al₂O₃ plots, where it overlaps with the Portasanta (PNORN-X1, dark red square in Fig. 5). This Portasanta (PNORN-X1, dark red square in Fig. 5) sample is the poorest in Fe₂O₃ content between polychrome ornamental stones analysed here; it is poorly discriminable from other Roman stones, except in the plots SiO₂/MgO vs SiO₂/Al₂O₃ and Al₂O₃/MgO vs SiO₂/Al₂O₃ (Fig. 5). The Breccia di Settebasi stones, i.e. PNORN-2C-1, PNORN-2C-2, PNORN-2C-3 and PNORN-4 (yellow diamond, blue triangle, pink hexagon and grey triangle, respectively, in Fig. 5) are all clustered and isolated from other samples in the SiO₂ vs Fe₂O₃ vs Al₂O₃ triangular plot, plus the SiO₂/MgO vs the SiO₂/Fe₂O₃, Al₂O₃/MgO vs SiO₂/Fe₂O₃ and SiO₂/Al₂O₃ vs SiO₂/Fe₂O₃ binary plots (Fig. 5). The other white varieties and Greco scritto stones have a primary geochemical distinction from the polychrome stones since they have invariably (very) lower amounts of Fe₂O₃ (both triangular plots of Fig. 5, Tab. S3 of the Online Resource 1); by contrast, they are wholly or poorly discriminable among themselves in both triangular and binary diagrams (Fig. 5).

Below, these previous data are considered together with geochemical plus MGS (Figs. 6 and 7); such comparisons are limited in the literature for polychrome stones and force us to consider also decorative marbles from modern excavations. In Fig. 8 and Tab. S5, the provenance sites, ancient name, petrographic type and possible quarries of the ornamental stones of the thermal baths from Teate Marrucinorum (Chieti, Italy) are resumed. The whole chemistry compositions of white marble are similar between the different varieties considered for the comparison, with only minor discrepancies for the MgO and CaO (Fig. 5); the same situation is also valid for the various Greco scritto rocks (Fig. 6). Again, the major oxides of polychrome Pavonazzetto, Cipollino verde and Portasanta stones analysed here overlap with those from literature; in contrast, the Breccia di Settebasi is more variable than the oxide ranges measured in this study (Fig. 6), proposing that an enlarged bulk geochemical characterisation of these ornamental stones could be valuable.

A similar comparison of the MGS values is presented in Fig. 7, mainly for white marbles since only a few works exist on polychrome samples (Arnoldi et al. 1999; Badouna et al. 2016, 2020; Bağci 2020; Carroll et al. 2008; Çelik and Sert 2020; Columbu et al. 2014; Lazzarini 2007). The MGS of our Pavonazzetto is just slightly larger than the average MGS of the white variety of the Docimium (Pavonazzetto) fine-grained marble, usually below 2 mm (Antonelli and Lazzarini 2015; Bağci 2020; Capedri and Venturelli 2004; Çelik and Sert 2020; Columbu et al. 2014). In addition, the MGS value of the PNORN-1 sample (Fig. 7) is well within the range of the Pavonazzetto stones (Al-Bashaireh 2022). Hence, the PNORN-1 sample corresponds to the Marmor phrigium extracted in the Afyon region in Turkey (Al-Bashaireh 2021; Attanasio et al. 2015) (Fig. 8, Tab. S5). The MGS of other Cipollino stone(s), even of modern time excavations, such as those from Badouna et al. (2016, 2020) or Apuan Cipollino from Arnoldi et al. (1999), are invariably smaller than that of PNORN-2A, being around 0.3 mm; conversely, the MGS of PNORN-2A is indeed in the range of impure marble corresponding to the Cipollino verde anciently labelled Marmor carystium from the Euboea region in Greece (Al-Bashaireh 2022; Lazzarini 2007, 2019) (Fig. 8, Tab. S5). The maximum grain-size in sedimentary veins (see above) of our PNORN-X1 Portasanta perfectly overlaps the measurements from literature by Lazzarini (2007) (Fig. 8). In addition, the presence of micritic clasts showing ooid shapes (Fig. S1) and the presence of quartz and dolomite (Figs. 3b, S2, S3) in both PNORN-X1 and the Portasanta of Lazzarini (2007) and Carroll et al. (2008) further support this conclusion, i.e. our Portasanta correspond to the tectonic breccia called Marmor chium of Romans, quarried in the Greek island of Chios (Gnoli 1988; Lazzarini 2007) (Fig. 8, Tab. S5). The equivalent of our PNORN-2C-1, PNORN-2C-2, PNORN-2C-3 and PNORN-4 Breccia di Settebasi samples have been primarily characterised by autoptic attributes (Taelman and Antonelli 2022); they show some analogies with the Italian Breccia Medicea, a metabreccia quarried at Serravezza and Stazzema villages, in the Apuan Alps (Lazzarini 2019; Taelmann et al. 2019). The MGS of our PNORN-2C-1, PNORN-2C-2, PNORN-2C-3 and PNORN-4 and those of the Breccia di Settebasi from literature have the same minimum and maximum values (Lazzarini 2007; Karambinis and Lazzarini 2015; Fig. 7). It can be thus concluded that the four PNORN-2C-1, PNORN-2C-2, PNORN-2C-3 and PNORN-4 stones from Chieti are the same metaconglomerate corresponding to the ancient Marmor scyreticum, also known as Breccia di Settebasi, from the Skyros Island of Sporades in Greece (Fig. 8, Tab. S5).

As previously discussed, the MGS values of the white marbles from Teate Marrucinorum have been compared with those present in the literature. In particular, Fig. 7 shows 7 of the chief white marbles used in antiquity sourced from the database in Antonelli and Lazzarini (2015). The coupling of petrographic/textural and isotopic signatures, as formerly treated, pointed out that the fine-grained decorative marbles is Marmor lunense from Carrara in Italy (PNORN-X2-a, PNORN-X2-b, PNORN-X2-d) and medium-grained specimens are Marmor proconnesium, excavated in Marmara Island, Turkey (PNORN-2B, PNORN-X2, PNORN-X2-c) (Fig. 8, Tab. S5). Finally, the analysed grey-stripped Greco scritto from Chieti (PNORN-3, PNORN-3-a, PNORN-3-b, PNORN-3-c and PNORN-3-d) show chemical compositions that are close both to that of Ephesus (Turkey) or Cap de Garde (Annaba, Algeria provenance) (Columbu et al. 2014; Gallala et al. 2017) and MGS values comparable between these two stones (Fig. 7) (Antonelli et al. 2009a, b; Yavuz et al. 2011; Antonelli and Lazzarini 2013; Columbu et al. 2014; Taelman et al. 2019; Perna et al. 2023); as previously assert, the more suitable provenance for these marbles s.s. is from Hasançavuslar in Turkey (Fig. 8, Tab. S5).

The ornamental stones decorating the thermal baths of Teate Marrucinorum are first divided into three polychrome, grey-striped and white marble categories (Fig. 2). The polychrome ones are then sub-grouped into four different stones corresponding to Pavonazzetto, Cipollino verde, Portasanta and Breccia di Settebasi, according to their autoptic discrimination, while the grey-striped type corresponds to the Greco scritto samples (Tab. S2). Instead, the white marbles require further geochemical (Figs. 3a, 4 Tab. S3) and petrographic (Fig. S1, Tab. S2) determinations. The two white marbles have isotopic (δ¹⁸O, δ¹³C) signatures (Fig. 4) plus MGS (Tab. S2) and microscopic features indicative of provenance from the ancient quarries of Ancient Prokonnesos in Turkey and Alpi Apuane in Italy (Tab. S5). In addition to these standard determinations, further mineralogical and geochemical analyses from this study and in previous studies provide further quantitative discriminations (Tab. S3,S4 and Figs. 5, 6 and 7). All these data represent possible new datasets to extend and support available datasets to discriminate rocks used by the Romans and depict provenance from different sites in the Mediterranean (Fig. 8).

The existing data on the thermal bath of Teate Marrucinorum indicate that the Pavonazzetto and Greco scritto were used to decorate two different parts of the apodyterium, while the same ambient of the cold tub of the frigidarium were covered with Cipollino verde, Proconnesian marble and Breccia di Settebasi, (PNORN-2A, PNORN-2B, PNORN-2C-1, 2C-2 and 2C-3); the different facies of the Breccia di Settebasi (PNORN-4) was used in the tepidarium (Fig. 1).

Supplementary Information. Tables from Tab. S1 to Tab S.5 are given in the Supplementary Material1.xlxs. Supplementary figures from Fig. S1 to Fig. S28 are in the Supplementary Material2.pdf

Acknowledgements. Most of this study was conducted during the Ph.D. of A. Casarin; it was funded by the “Fondi Ateneo of the University G. D'Annunzio” and PRIN (2017J277S9_003) project “Time Scales of Solidification in Magmas: Application to Volcanic Eruptions, Silicate Melts, Glasses, Glass-Ceramics” awarded to G. Iezzi and to the project “DPC-ReLUIS” awarded to G. Brando.

Competing interest. The authors declare no competing interests.

Data availability. The authors declare that the data supporting the findings of this study are available within the paper and its supplementary information files.

Author contributions.Conceptualisation: A. Casarin, G.I., F.A.; sampling: G.I., M.I.P., E.C., R.T., I.C., M.G.M., D.P., D.R.; methodology: A. Casarin, F.A., A. Cavallo, F.R., M.R.C.; data curation: A. Casarin, G.I., F.A., A. Cavallo, F.R., M.R.C., D.d.L.; writing-original draft preparation: A. Casarin, G.I., F.A.; writing-review and editing: A. Casarin, G.I., F.A., M.R.C., D.d.L., I.C., M.G.M., G.B., F.R., A. Cavallo, M.I.P., E.C., R.T., D.P., D.R.; resources: G.I., G.B., F.A., A. Cavallo, D.d.L.. All authors have read and agreed to this version of the manuscript.

Adinolfi G, Agostini S, Carmagnola R, di Carlo M, Mancini MC, Menozzi O, Palumbo D, Petraccia V (2019) Le Terme Romane di Chieti: un caso esemplificativo. https://eurotech.udanet.it/wp-content/uploads/2020/07/POSTER_TermeRomaneChieti-scaled.jpg Accessed June 2024
Agostini S (2018) The stratified growth of Chieti from roman times to tomorrow: a new, geology-based conscience in city planning and renewal. Alp Mediterr Quat 31(2):195-205. https://doi.org/10.26382/AMQ.2018.13
Agostini S, Rossi MA (2012) La presenza di marmi e pietre policrome in siti archeologici dell’Abruzzo. Quaderni Di Archeologia d’Abruzzo: Notiziario Della Soprintendenza per i Beni Archeologici Dell’Abruzzo: 4, 2012, 87–93. https://doi.org/10.1400/250309
Agostini S, Mariottini M, Romano J, Rossi MA (2002) Polychrome stones from the Roman baths in Chieti (Abruzzo, Italy). In: Lazzarini L (ed) Interdisciplinary studies on ancient stone: ASMOSIA VI, Proceedings of the Sixth International Conference of ASMOSIA, Association for the Study of Marble and Other Stones in Antiquity, Venice, 15–18 June 2000, pp 73–78
Akbulut ZF (2022) Experimental investigation of the surface properties of accidental stained carbonate natural stones. Case Stud Constr Mater 17: e01626. https://doi.org/10.1016/J.CSCM.2022.E01626
Al-Bashaireh K (2021) Ancient white marble trade and its provenance determination. J Archaeol Sci Rep 35: 102777. https://doi.org/10.1016/J.JASREP.2020.102777
Al-Bashaireh K (2022) The white marbles and polychrome stones of the five-aisled basilica at Gadara (Umm Qais), Jordan: archaeometric characterization for provenance identification. Archaeol Anthropol Sci 14(2): 1–18. https://doi.org/10.1007/s12520-021-01470-6
Antonelli F, Lazzarini L (2013) White and coloured marbles of the roman town of Urbs salvia (Urbisaglia, Macerata, Marche, Italy). Oxf J Archaeol 32(3): 293–317. https://doi.org/10.1111/ojoa.12016
Antonelli F, Lazzarini L (2015) An updated petrographic and isotopic reference database for white marbles used in antiquity. Rendiconti Lincei 26(4): 399–413. https://doi.org/10.1007/S12210-015-0423-4/
Antonelli F, Nestola, F (2021). An innovative approach for provenancing ancient white marbles: the contribution of x-ray diffraction to disentangling the origins of Göktepe and Carrara marbles. Sci Rep 11(1): 22312. https://doi.org/10.1038/s41598-021-01800-7
Antonelli F, Lazzarini L, Cancelliere S, Dessandier D (2009a) Minero-Petrographic and Geochemical Characterization of ‘Greco Scritto’ marble from Cap De Garde, near Hippo Regius (Annaba, Algeria). Archaeometry, 51(3): 351–365. https://doi.org/10.1111/J.1475-4754.2008.00408.X
Antonelli F, Lazzarini L, Cancelliere S, Dessandier D (2009b) Volubilis (Meknes, Morocco): Archaeometric study of the white and coloured marbles imported in the Roman age. J Cult Herit 10(1): 116–123. https://doi.org/10.1016/J.CULHER.2008.04.006
Arnoldi C, Azzaro E, Barbieri M, Tucci P (1999) Petrographic and geochemical features of the “Cipollino Verde” marble from the Apuan Alps (northern Tuscany, Italy) and archaeometric implications. Period Mineral 68: 145–162
Attanasio D, Bruno M, Prochaska W, Yavuz AB (2015) Analysis and Discrimination of Phrygian and other Pavonazzetto-like marbles. In: Pensabene P, Gasparini E (ed) Interdisciplinary studies on ancient stone: ASMOSIA X, Proceedings of the Tenth International Conference of ASMOSIA, Association for the Study of Marble & Other Stones in Antiquity, Rome, 21-26 May 2012, pp 753-764
Badouna I, Koutsovitis P, Karkalis C, Laskaridis K, Koukouzas N, Tyrologou P, Patronis M, Papatrechas C, Petrounias P (2020) Petrological and Geochemical Properties of Greek Carbonate Stones, Associated with Their Physico-Mechanical and Aesthetic Characteristics. Minerals 10(6): 507. https://doi.org/10.3390/MIN10060507
Badouna I, Koutsovitis P, Laskaridis K, Patronis M, Papatrechas C (2016) Aesthetic characteristics of greek ornamental stones associated with mineral, geochemical and structural properties. Bull Geol Soc Greece 50(4): 1771–1780. https://doi.org/10.12681/bgsg.14104
Bağci M (2020) Mineralogical, petrographic, and geochemical characterization of colored İscehisarmarbles (Afyonkarahisar, W-Turkey). Turk J Earth Sci 29(6): 946–975. https://doi.org/10.3906/yer-2001-20
Beltrame C, Percic T, Lazzarini L (2019) The archaeometric identification of the marbles of the Roman shipwrecks of Capo Granitola (TP), Isola delle Correnti and Marzamemi I (SR). J Archaeol Sci: Rep 23: 953–967. https://doi.org/10.1016/J.JASREP.2018.11.021
Borghini G (ed.) (1989) Marmi antichi, De Luca Edizioni D’Arte, Rome, Italy
Capedri S, Venturelli G (2004) Accessory minerals as tracers in the provenancing of archaeological marbles, used in combination with isotopic and petrographic data. Archaeometry, 46(4): 517–536. https://doi.org/10.1111/j.1475-4754.2004.00171.x
Carroll M, Montana G, Randazzo L, Giarrusso R (2008) Recovering evidence for the use of marble and coloured limestone in the first century AD in excavations at the sanctuary of Venus in Pompeii. Fasti On Line Documents & Research 119: 281-311
Çelik MY, Sert M (2020) The importance of “Pavonazzetto marble” (Docimium-Phrygia/Iscehisar-Turkey) since ancient times and its properties as a global heritage stone resource. Environ Earth Sci 79(9): 1–18. https://doi.org/10.1007/S12665-020-08943-2/FIGURES/17
Chipera SJ, Bish DL (2013) Fitting Full X-Ray Diffraction Patterns for Quantitative Analysis: A Method for Readily Quantifying Crystalline and Disordered Phases. Adv Mater Phys Chem 03(01): 47–53. https://doi.org/10.4236/ampc.2013.31a007
Columbu S, Antonelli F, Lezzerini M, Miriello D, Adembri B, Blanco A (2014) Provenance of marbles used in the Heliocaminus Baths of Hadrian’s Villa (Tivoli, Italy). J Archaeol Sci 49(1): 332–342. https://doi.org/10.1016/J.JAS.2014.05.026
Corsi F (1845) Delle Pietre Antiche. Dalla Tipografia Salviucci, Rome, Italy
Della Ventura G, Iezzi G, Redhammer GJ, Hawthorne FC, Scaillet B, Novembre D (2005) Synthesis and crystal-chemistry of amphiboles along the magnesioriebeckite – magnesio-arfvedsonite series as a function of fO₂. Am Mineral 90: 1375-1383. https://doi.org/10.2138/am.2005.1682
European Commission - Extra-Expo project (2015) Marble, the divine stone. Its extraction, trade and use since the beginning of the humanity until the present
Fettes D, Desmons J (2007) Metamorphic Rocks: A Classification and Glossary of Terms. Cambridge University Press
Galderisi A, Iezzi G, Bianchini G, Paris E, de Brito J (2022) Petrography of construction and demolition waste (CDW) from Abruzzo region (Central Italy). Waste Manag 137: 61–71. https://doi.org/10.1016/J.WASMAN.2021.10.028
Gallala W, Younès A, Gaied ME, Hadjzobir S, Sparks K, Molli G (2017) A multi-analytical approach for determining the provenance of the marbles from Ruspina Roman baths (Monastir, Tunisia). Archaeol Anthropol Sci 9(6): 1275–1285. https://doi.org/10.1007/S12520-016-0323-2/FIGURES/9
Gnoli R (1988) Marmora romana. Edizioni dell’Elefante, Rome, Italy
Hubbard CR, Snyder RL (1988) RIR - Measurement and Use in Quantitative XRD. Powder Diffr 3(2): 74–77. https://doi.org/10.1017/S0885715600013257
Iezzi G, Della Ventura G, Cámara F, Pedrazzi G, Robert JL (2003) ^BNa–^BLi solid-solution in A-site-vacant amphiboles: synthesis and cation ordering along the ferri-clinoferroholmquistite–riebeckite join. Am Mineral 88(7): 955-961. https://doi.org/10.2138/am-2003-0702
ISPRA (2012) Carta Geologica d’Italia alla scala 1:50.000, Progetto CARG. Foglio n. 361 “Chieti”. Servizio Geologico d’Italia
Karambinis M, Lazzarini L (2015) The Roman marble quarries of Aliko Bay and of the islets of Rinia and Koulouri (Skyros, Greece). In: Pensabene P, Gasparini E (ed) Interdisciplinary studies on ancient stone: ASMOSIA X, Proceedings of the Tenth International Conference of ASMOSIA, Association for the Study of Marble & Other Stones in Antiquity, Rome, 21-26 May 2012, pp 791-804
Lazzarini L (2007) Poikiloi Lithoi, Versiculores Maculae: i marmi colorati della Grecia Antica: storia, uso, diffusione, cave, geologia, caratterizzazione scientifica, archeometria, deterioramento. Serra F (Ed)
Lazzarini L (2019) Ancient Mediterranean polychrome stones. In: Artioli G, Oberti R (ed) The Contribution of Mineralogy to Cultural Heritage. EMU Notes in Mineralogy, Vol. 20, pp. 367–392. Mineralogical Society of Great Britain and Ireland. https://doi.org/10.1180/EMU-notes.20.10
McCrea JM (1950) On the Isotopic Chemistry of Carbonates and a Paleotemperature Scale. J Chem Phys 18(6): 849–857. https://doi.org/10.1063/1.1747785
Pensabene P (1993) Marmi antichi. Problemi d’impiego, di restauro e d’identificazione. L'Erma di Bretschneider, Rome, Italy
Pensabene P (2002) Committenza edilizia a Ostia tra la fine del I e i primi decenni del III secolo. Lo studio dei marmi e della decorazione architettonica come strumento d’indagine. Mélanges de l’École Française de Rome. Antiquité, 114(1): 181–324. https://doi.org/10.3406/mefr.2002.10698
Perna S, Antonelli F, Lazzarini L (2023) Archaeometric analysis of the ‘greco scritto’ marble slabs from the Edificio dei Triclinii at Murecine (Pompeii, Italy). Archaeometry, 65(1): 1–16. https://doi.org/10.1111/arcm.12811
Potere D, Iezzi G, Scisciani V, Tangari AC, Nazzari M (2023) Provenance and deposition of a lithified volcanic-rich layer (VRL-5.5) at 5.5 Ma from Central Apennines (Italy). Sci Rep 13: 6880. https://doi.org/10.1038/s41598-023-33256-2
Radica F, Casarin A, Iezzi G, Bravo M, de Brito J, Galderisi A, Brando G, Nazzari M, Scarlato P (2024) Cement vs aggregates and textures of aggregates in a mortar: Comparative image analysis methods and analytical protocols. Constr Build Mater 453: 139033. https://doi.org/10.1016/J.CONBUILDMAT.2024.139033
Russell B (2013) Gazetteer of Stone Quarries in the Roman World. Version 1.0. Data set/Database: http://oxrep.classics.ox.ac.uk/databases/stone_quarries_database/
Taelman D, Antonelli F (2022) Provenance of the white and polychrome marbles used for the architecture and sculpture of roman Sentinum (Sassoferrato, Marche, Italy). Archaeometry 64(1): 1–24. https://doi.org/10.1111/arcm.12690
Taelman D, Delpino C, Antonelli F (2019) Marble decoration of the Roman theatre of Urvinum Mataurense (Urbino, Marche region, Italy): An archaeological and archaeometric multi-method provenance study. J Cult Herit 39: 238–250. https://doi.org/10.1016/j.culher.2019.03.009
Taelman D, Mei O, Antonelli F (2023) On the Use of White Marbles in Roman Forum Sempronii (Fossombrone, Marche, Italy): An Overview of the Archaeometric Data on Architectural Elements. Heritage 6(3): 2942–2955. https://doi.org/10.3390/HERITAGE6030156
Tarquini S, Isola I, Favalli M, Battistini A, Dotta G (2023) TINITALY, a digital elevation model of Italy with a 10 meters cell size (Version 1.1). Istituto Nazionale di Geofisica e Vulcanologia (INGV). https://doi.org/10.13127/tinitaly/1.1
Yavuz AB, Bruno M, Attanasio D (2011) An updated, multi-method database of Ephesos marbles, including White, Greco Scritto and Bigio varieties. Archaeometry 53(2): 215–240. https://doi.org/10.1111/j.1475-4754.2010.00542.x

No competing interests reported.

Download PDF

Editorial decision: Revision requested
22 Dec, 2025
Reviews received at journal
22 Dec, 2025
Reviews received at journal
15 Dec, 2025
Reviews received at journal
03 Dec, 2025
Reviewers agreed at journal
10 Nov, 2025
Reviewers agreed at journal
07 Nov, 2025
Reviewers agreed at journal
30 Oct, 2025
Reviewers invited by journal
30 Oct, 2025
Editor assigned by journal
30 Oct, 2025
Submission checks completed at journal
30 Oct, 2025
First submitted to journal
29 Oct, 2025

You are reading this latest preprint version

The ornamental stones of the Roman thermal baths of Teate Marrucinorum (Chieti, Italy): autoptic, geochemical and minero-petrographic multi-analytical characterisations

Status:

Version 1

Abstract

Figures

Introduction

Samples and Methods

Results

Discussion

Conclusions

Statements and Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1