This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright
Author's personal copy
Journal of Archaeological Science 40 (2013) 810e826
Contents lists available at SciVerse ScienceDirect
Journal of Archaeological Science journal homepage: http://www.elsevier.com/locate/jas
Research paper
A Late Roman ceramic production from Pompeii Celestino Grifa a, *, Alberto De Bonis b, Alessio Langella a, Mariano Mercurio a, Gianluca Soricelli c, Vincenzo Morra b a
Dipartimento di Scienze per la Biologia, la Geologia e l’Ambiente, Università del Sannio, Via dei Mulini 59/A, 82100 Benevento, Italy Dipartimento di Scienze della Terra, Università Federico II, Via Mezzocannone 8, 80134 Napoli, Italy c Dipartimento di Scienze Umane, Storiche e Sociali, Università del Molise, Via Mazzini 8, 86170 Isernia, Italy b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 12 September 2011 Received in revised form 27 August 2012 Accepted 28 August 2012
The Via Lepanto site is one of the best examples showing how the Vesuvian region was partially reconstructed and earlier re-occupied after Vesuvius’s eruption in the year 79 AD. The large amount of ceramic finds illustrates the typology in use in this area during the IV and V century AD. Analyses were focused on table and cooking ware productions. Archaeometric data were obtained using chemical and minero-petrographical methods (OM, XRD, XRF and SEM). Grain size measurements using Image Analyses on thin sections and a geochemical comparison with clayey deposits outcropping in the Campania region permitted the identification of the raw materials used for these pottery productions. XRD and SEM completed the data set, establishing the protocols used for pottery production in the Pompeii area during Late Roman period. The Via Lepanto site was part of an exchange network of markets with a periodic frequency, where locally produced and imported pottery was sold, indicating a flourishing network of exchanges spanning short, medium and long distances. Ó 2012 Elsevier Ltd. All rights reserved.
Keywords: Pompeii Pottery Clayey deposits Volcanic temper
1. Introduction The so-called Agro Nocerino-Sarnese, in the southern part of the Campanian Plain, had been an optimum for human settlements since pre-historic times, due to its fertile soils and propitious climate (Soricelli, 2001; Marturano et al., 2009). However, the scenario commonly proposed for this land immediately after the 79 AD eruption, which destroyed Pompeii, Ercolano and Stabia, is that of a bare and abandoned land. In fact, several authors hypothesise an abrupt interruption of all human activities immediately after the eruption and only a late and sporadic reoccupation of the area, up until III century AD (Soricelli, 2002). However, such reconstructions suffer a lack of interest on the part of archaeologists for the post-79 AD levels which consequently, were only summarily investigated or simply removed, in order to reach the pre-eruption layers. According to ancient sources (Suet. Tit. 8,4; Cass. Dio LXVI.24, 3e 4 in Gazzetti, 1976), the reconstruction process immediately after the disaster, which focused on agricultural infrastructures, public buildings and road networks, was managed by two imperial officers, the curatores restituendae Campaniae, and was funded by the imperial treasury (Soricelli, 2001, 2002). Several sites and settlements, in particular along the NuceriaePompei road (Fig. 1) * Corresponding author. Tel.: þ39 (0)824 363649; fax: þ39 (0)824 323623. E-mail address:
[email protected] (C. Grifa). 0305-4403/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jas.2012.08.043
(De Carolis and Soricelli, 2005), confirm a reoccupation of the area, such as the Porta-Vesuvio necropolis and the Moregine complex. The Via Lepanto site is one of the best examples of a reoccupation of the Vesuvius area after the 79 AD eruption (Fig. 1). It dates back to the first half of II century AD (as shown by the African Red Slip Ware, forms Hayes 7A, 8A). The site is located on the Nuceriae Pompei road, at approximately 1 km from the SE walls of the ancient city of Pompeii. It was abandoned after the 472 AD eruption, as shown by the “Pollena” pyroclastic deposits that covered the whole settlement (De Carolis and Soricelli, 2005). Some ceramic dumps, dating back to the late IVeearly V century AD by the African Red Slip Ware forms Hayes 59, 60, 61, 73 and 91, brought to light a large amount of pottery, among which samples of table and cooking ware. These were collected and analysed from a mineralogical and petrographical point of view. This research study aims at defining the technological features of some selected ceramic productions that were widespread in the Via Lepanto site, by means of an exhaustive mineralogical, petrographical and chemical characterisation, focussing on the identification of the clays and tempers used, as well as the firing technologies. Moreover, the whole data set aims at providing a useful archaeometric database on the ceramic productions from this area. An attempt of hypothesis on the provenance of the shards was also carried out in order to confirm the presence of active ceramic workshops in Pompeii and the surrounding area during the Late
Author's personal copy
C. Grifa et al. / Journal of Archaeological Science 40 (2013) 810e826
811
Fig. 1. Simplified geological map of Campania region (modified from Bonardi et al., 2009). Main roads (Shepherd, 1911) and clay sampling sites are also reported.
Roman period. This would provide a useful indicator of a retrieved socialeeconomic activity. Mineralogical, petrographical and geochemical analyses may also provide information on the provenance of the pottery. This is particularly true for coarse-grained common ware and amphorae productions in which inclusions can be easily correlated to the rocks cropping out in the surrounding area. The relationship between the minero-petrographical composition of the coarse-temper is particularly effective in volcanic contexts, as reported for some ceramic productions of the Circum-Mediterranean volcanic area, such as those from Sicily (Barone et al., 2010), the Aeolian islands (Williams et al., 2008) and Pantelleria (Grifa et al., 2005a). With regards to ceramic productions from the Bay of Naples, the occurrence of well sorted, rounded volcanic temper, mainly composed of sanidine, Ca-rich pyroxene, plagioclase, pumices and scoriae, is largely attested in both fabrics of common wares and amphorae (Grifa et al., 2005b, 2006, 2009a). To complete the characterization of the potsherds, clayey raw materials located close to the Pompeii area and along the Late Roman period main roads (e.g. Appia and Traiana roads) were also sampled and compared with the investigated ceramic products, from a mineralogical and geochemical point of view (De Bonis, 2011; De Bonis et al., 2012). 2. Volcanological remarks Located not far from the Southern-eastern part of the city of Naples and within one of the most densely inhabited areas of the
world, this typical stratovolcano has always conditioned human behaviour in this portion of the Campania region. Mount Somma represents the oldest volcano that concluded its activities with a caldera formation. Within this caldera, the Vesuvius volcano successively formed (Cioni et al., 1999). The volcanic history of the SommaeVesuvius complex is still debated. After a first phase, which started circa 300e400 ka (Brocchini et al., 2001), it was only after the Phlegraean Campanian Ignimbrite eruption that activity was recorded (39 ka; Fedele et al., 2008). It was upon this eruption that the present volcanic complex was formed (Di Vito et al., 1998). Since its last eruption in 1944 AD, the SommaeVesuvius has experienced a quiescent status. From a petrological point of view, the volcanic products belong to the HK series and they are typical leucite-bearing rocks (Conticelli et al., 2004). The first historical Plinian eruption, recorded in the archaeological layers and largely used as a stratigraphic marker for the Bronze Age, is the Pomici di Avellino eruption (4.365 ka; Santacroce et al., 2008) which deposited large volumes of pyroclastic falls and flows, causing the migration of the prehistorical population to the surrounding areas. However, the most important Plinian event, also from a historical point of view, is the 79 AD Pompei eruption (Lanphere et al., 2007) directly observed and carefully described by Plinius the Younger. Successively, the 472 AD Pollena eruption (Santacroce et al., 2008) also had a strong impact on the Vesuvius area. These latter two eruptions represent important events for the Via Lepanto site, as they mark the beginning and the end of occupancy.
Author's personal copy
812
C. Grifa et al. / Journal of Archaeological Science 40 (2013) 810e826
3. Materials 3.1. Pottery Some Late Roman potsherd dumps, sealed by Pollena products (472 AD), brought to light several thousands of ceramic fragments, mainly represented by common ware (table and cooking ware). Fifty-two fragments from these ceramic classes (Table 1) were sampled for minero-petrographical analyses. From an archaeological point of view, the pottery from Via Lepanto can be compared to similar products circulating in the region between the end of IV century AD and the beginning of the V century AD (De Carolis and Soricelli, 2005). Three different types of table were investigated: Red Painted Ware (RPW), Painted Ware (PW) and Burnished Ware (BW). The Red Table 1 Archaeological information of 52 common ware ceramic samples from Via Lepanto. The colour code form Munsell Soil Colour Chart. Sample
Class
Body
PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL
BW BW BW PW PW PW PW PW CW CW CW CW CW CW CW CW CW CW CW CW CW CW CW CW CW CW CW CW CW CW CW RPW RPW RPW RPW CW CW RPW RPW RPW RPW RPW RPW RPW* CW RPW* RPW RPW RPW RPW RPW CW
2.5YR 5/8 2.5YR 5/8 2.5YR 5/8 5YR 6/3 5YR 6/3 5YR 6/3 5YR 6/3 5YR 5/6 5YR 6/3 5YR 6/6 5YR 6/6
1 2 3* 4* 5* 6 7* 8 9* 10* 11 12* 13 14 15* 16 17 18 19* 20 21* 22 23 24 25 26* 27* 28 29 30 31 32 33 34* 35* 36 37 38* 39* 40 41 42 43 44 45 46 47 48 49 50 51 52
Zoning
Heart
Rim
Slip 2.5YR 4/8 2.5YR 4/8 2.5YR 4/8
faded faded faded faded
5YR 5YR 5YR 5YR
6/3 6/3 6/3 6/3
2.5YR 2.5YR 2.5YR 2.5YR
5/8 5/8 5/8 5/8
faded
5YR 6/3
2.5YR 5/8
faded
5YR 6/3
2.5YR 5/8
faded
5YR 6/3
2.5YR 5/8
sharp
5YR 6/3
2.5YR 5/8
faded
5YR 6/3
2.5YR 5/8
faded
5YR 6/3
2.5YR 5/8
faded
5YR 6/3
2.5YR 5/8
5YR 6/3 5YR 6/3 5YR 6/3 5YR 6/3 5YR 5/6 5YR 6/6
3.2. Clayey raw materials
5YR 6/6 5YR 6/6 5YR 6/6 2.5YR 5/8 2.5YR 5/8 2.5YR 5/8 sharp sharp
5YR 6/3 5YR 6/3
2.5YR 5/8 2.5YR 5/8
sharp
5YR 6/3
2.5YR 5/8
sharp
5YR 6/3
2.5YR 5/8
2.5YR 5/8 2.5YR 5/8 2.5YR 5/8 2.5YR 5/8 2.5YR 5/8 2.5YR 5/8 2.5YR 5/8 2.5YR 5/8 2.5YR 5/8 2.5YR 5/8 2.5YR 5/8 2.5YR 5/8 5YR 4/3
Painted Ware (17 samples), mainly represented by open shapes (Fig. 2a), can be easily identified by a light reddish slip that covers the entire internal surface and partially overlaps the external rim of the artefact. Macroscopic observations permit the recognition of two different groups of fragments. The first is characterised by a fine and very hard paste. The colour of the paint is red/red-orange (2.5YR 4/8). A second group of samples shows a coarser and softer paste. The thin paint is red in colour (2.5YR 5/6). In both groups, the ceramic body has a redebrown colour (2.5YR 5/8) but some sandwich structures can be observed (sample PVL 39 and 41, Table 1). Examples of RPW have been found in late IVeearly V century AD contexts from Somma Vesuviana (Aoyagi et al., 2007) and Mercato S. Severino (Fiorillo, 2003). These can be compared with pottery from other Southern Italy sites: Naples (Arthur, 1994), Pratola Serra (Alfano, 1992), Ordona (Annese, 2000) and Calle (Di Giuseppe, 1998). A second tableware ceramic class is represented by 5 samples of the Painted Ware (Whitehouse, 1966) showing a greyish fine paste (5YR 6/3) with a redebrown decoration applied by a paintbrush or a cloth. These vessels (Fig. 2b) can be compared with analogous products from other Campanian sites: Cuma, Miseno (Grifa et al., 2005b, 2009a), Benevento (Lupia, 1998) and Caudium (Perrone, 2005; De Bonis et al., 2010). The last three samples of tableware are represented by small closed shapes with a fine red paste (2.5YR 5/8) and with a shiny red cover (2.5YR 4/8) on the external surface: the Burnished Ware (inset of Fig. 2b). Well represented in Vesuvius area, this ware is common in Naples and in other Campanian sites (De Carolis and Soricelli, 2005). The Cooking Ware samples (CW, 27 samples) show a very coarse paste. In general, the matrix shows a red/brown colour (5YR 6/3e 5YR 6/6) but in several samples, sandwich zoning of the ceramic body can be observed. Main attested shapes (Fig. 2c) are the casseroles that can be found in other Campanian contexts, in particular in Benevento (Lupia, 1998), Caudium (De Bonis et al., 2010), Neapolis (Carsana, 1994) and the Phlegraean area (Soricelli, 2000). Saucepans and lids also occur. Only one sample (Fig. 2c) can be referred to the Pantellerian Ware, a cooking ware widespread in the Mediterranean area between IV and VI century (Grifa et al., 2005a).
2.5YR 2.5YR 2.5YR 2.5YR
4/8 5/6 5/6 4/8
2.5YR 2.5YR 2.5YR 2.5YR 2.5YR 2.5YR 2.5YR
4/8 4/8 4/8 4/8 4/8 4/8 5/6
2.5YR 2.5YR 2.5YR 2.5YR 2.5YR 2.5YR
4/8 4/8 4/8 4/8 5/6 5/6
BW ¼ Burnished Ware; PW ¼ Painted Ware; CW ¼ Cooking Ware; RPW ¼ Red Painted Ware, * Samples studied by image and modal analyses.
Ten samples of clayey raw materials (Table 2, Fig.1) were collected in order to compare their mineralogical and geochemical composition with that of the ceramic finds. Samples were chosen according to their proximity to both possible ancient sites of production and the main Roman roads of the Campania region (Fig. 1). They are mainly represented by basinal sediments, subordinate alluvial deposits and strongly weathered pyroclastic soils (Table 2). Basinal sediments mainly outcrop along the Apennine chain located in the eastern part of the Campania Region and are ubiquitously ascribed to different units such as the Sicilide Unit (samples BS1, BS2, GS1) or the Fortore Unit (sample MLV1), which all date from the Upper Cretaceous to the Lower Miocene (Table 2; Fig. 1) (Bonardi et al., 2009). Sample MLV1 was collected in the area surrounding the mud volcano known as the Bolle della Malvizza in the Miscano river valley, very close to the Traiana Roman Road (Fig. 1), which crossed the Apennine chain to the Adriatic coast. The GS1 sample crops out close to the town of Gioia Sannitica (Caserta province). Other basinal samples (SQ1 and TRE1), ascribed to Caiazzo Sandstones (Bonardi et al., 2009), come from the Northern Campania area, which is well-known for the presence of several Roman settlements. The SQ1 sample was collected near the town of Caiazzo (Caserta province). The TRE1 sample comes from the village of Treglia (Caserta province), a Roman settlement (Trebula),
Author's personal copy
C. Grifa et al. / Journal of Archaeological Science 40 (2013) 810e826
813
Fig. 2. The variegated repertoire of the common ware ceramic from Via Lepanto; a) the Red Paint Ware (RPW), bowls and dishes; b) the Painted Ware (PW), bowls and dishes, in the inset the shape representative of the Burnished Ware (BW); c) The Cooking Ware (CW), pans, pots, saucepan and lids, in the inset a typical saucepan of Pantellerian Ware.
where an ancient ceramic workshop is reported to have existed (Livadie, 1994). The PMV2 alluvial clayey sediment (Piana di Monte Verna, Caserta province) of the Volturno River plain is a holocenic clay Table 2 Clayey deposits description. Sample
Locality
Geological origin
Description
BS1
Bisaccia (AV)
Basinal sediment
BS2 GS1
Bisaccia (AV) Gioia Sannitica (CE)
Basinal sediment Basinal sediment
MLV1
Montecalvo Irpino (AV) Castel Campagnano (CE) Pontelatone (CE)
Basinal sediment
Piana di Monte Verna (CE) Velina di Castelnuovo Cilento (SA) Cascano di Sessa Aurunca (CE) Sant’Agnello (NA)
Alluvial sediment
Blue-greenish clayey silt Reddish clayey silt Reworked brownyellowish sediment Greyish clayey sediment Reworked brownish silty sediment Reworked brownish silty sediment Yellowish clayey silt
Alluvial sediment
Yellowish sandy silt
Pyroclastic deposit
Reddish clayey silt
Pyroclastic deposit
Brownish sandy silt
SQ1 TRE1 PMV2 VEL1 CSC1 SO1
Basinal sediment Basinal sediment
collected two meters below the ground level, in which bricks of Roman age were also found. VEL1 is an Alento River sediment (PleistoceneeHolocene), sampled in a former clay quarry in the area surrounding Velia, the ancient Greek colony of Elea. Other clayey raw materials derive from strongly weathered pyroclastic deposits that were widespread in the volcanic areas of the region. Such materials were most likely exploited in the past for pottery productions, especially in the Bay of Naples area, and are still used for the manufacturing of bricks for wood-burning ovens or cooking wares, following ancient techniques and taking advantage of their refractory properties. One sample (SO1) comes from a brownish lahar-like deposit of the Sorrento Peninsula, deriving from the activity of SommaeVesuvius. The other sample (CSC1) comes from the pyroclastic soils of the Roccamonfina volcano (De Bonis, 2011).
4. Analytical techniques The mineralogical and textural properties of the pottery fragments were investigated on thin sections with a Leitz Laborlux 12POL polarising microscope. On selected representative samples of each ceramic class, accurate measurements of the non-plastic inclusions were carried out with an image analysis software (Leica Q-Win). Point-to-point modal analyses were carried out on
Author's personal copy
RPW
CW
qz/fd* ¼ quartz and feldspar microcrystalline; cpx ¼ clinopyroxene, san ¼ sanidine, pl ¼ plagioclase, bt ¼ biotite, garn ¼ garnet, pum ¼ pumices, sco ¼ scoriae, lc-sco ¼ leucite-bearing scoriae, obs ¼ obsidians, cc ¼ calcite, por ¼ pores, mat ¼ matrix, tot incl ¼ total inclusions; volc ¼ volcanics. Statistical parameters of f size and logit (SF). median; s ¼ Standard Deviation; mean.
0.535 0.573 0.594 0.341 0.675 0.507 0.517 0.731 0.594 0.693 0.649 0.333 0.854 0.656 0.349 0.236 0.378 0.380 1.094 0.834 1.001 0.915 0.900 0.883 0.975 1.050 1.046 1.096 0.983 0.905 1.116 0.906 0.903 1.006 0.953 1.053
s median
0.344 0.488 0.415 0.223 0.542 0.507 0.441 0.616 0.446 0.542 0.541 0.203 0.624 0.539 0.282 0.097 0.201 0.227 3.887 4.387 4.160 4.374 3.655 3.601 3.337 3.303 3.485 2.986 3.176 3.892 4.786 4.188 4.328 4.519 3.945 4.140 0.608 0.501 0.746 0.674 0.735 0.698 0.847 0.848 0.668 0.946 0.783 0.887 0.557 0.664 0.776 0.740 0.928 1.120 3.926 4.383 4.255 4.461 3.752 3.664 3.317 3.365 3.550 2.998 3.239 4.028 4.800 4.250 4.466 4.605 4.107 4.454 15.99 1.93 6.00 2.86 10.95 17.75 20.02 24.35 23.65 16.19 8.64 19.86 16.14 10.60 17.35 12.03 24.51 23.00 1.11 2.28 2.40 2.06 1.67 1.02 1.67 1.79 1.24 1.28 1.37 1.45 1.39 1.34 1.19 1.43 1.31 1.53 44.62 27.44 27.20 30.75 35.65 46.01 35.62 32.83 42.74 41.12 39.16 38.29 36.50 36.40 41.76 36.38 40.63 36.80 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 49.75 62.43 65.20 63.25 59.46 46.75 59.50 58.67 52.88 52.75 53.59 55.71 50.75 48.60 49.90 52.10 53.25 56.20 5.63 10.13 7.60 6.00 4.89 7.24 4.88 8.50 4.38 6.13 7.25 6.00 12.75 15.00 8.34 11.52 6.12 7.00 0.00 0.00 0.00 0.00 0.22 0.63 1.00 0.50 0.13 1.33 1.42 0.00 0.00 0.00 0.00 0.00 0.00 0.20 4.13 0.00 0.20 0.10 0.00 1.00 0.38 0.00 4.75 1.88 0.00 0.43 3.42 1.80 1.72 3.11 4.13 5.40 0.00 0.00 0.20 0.00 0.00 0.00 0.00 0.00 0.63 0.00 0.00 0.00 0.00 0.00 0.00 0.00 1.13 0.00 3.38 0.89 1.20 0.54 3.23 1.62 2.88 4.17 1.13 2.75 0.00 7.71 4.29 2.20 3.34 2.87 5.24 8.40 1.63 0.18 0.00 0.00 0.22 1.25 0.00 0.00 0.50 0.75 0.00 0.00 2.50 0.00 4.67 1.12 5.50 3.00 0.00 0.00 0.00 0.00 0.00 0.50 0.00 0.17 0.00 0.18 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.30 0.37 2.13 0.67 0.63 1.00 0.00 2.86 1.60 2.40 2.64 0.60 0.13 0.80 1.00 0.20 0.80 1.22 2.64 2.00 3.00 4.50 4.00 4.50 4.45 2.43 2.35 1.20 1.98 2.78 3.75 1.80 2.23 0.36 1.40 0.67 4.04 8.13 10.13 12.17 10.38 3.75 4.19 5.86 1.50 2.20 2.76 1.43 3.00 2.80 3.62 0.30 2.20 0.33 0.52 2.88 1.50 2.67 1.63 1.38 0.00 0.57 0.48 0.80 0.24 0.12 1.63 0.80 28.63 25.51 21.20 27.89 24.48 27.63 14.60 7.98 18.96 23.60 29.10 18.43 20.36 25.80 24.41 24.35 16.12 13.60 3 4 5 7 9 10 12 15 19 21 26 27 34 35 38 39 44 46 PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL PVL BW PW
logit (SF)
mean
s median
f volc mat/inc tot inc tot mat por cc obs lc-sco sco pum garn bt pl san
Optical microscopy observations of the investigated samples have already been reported in a previous paper (De Carolis et al., 2009) and this data highlighted a close relationship among textures and ceramic classes. A brief description of the main features of the group samples is reported below focussing mainly on the image and modal analyses data of selected samples. The results of a point-to-point modal analysis, statistical parameters of f size and logit (SF) are reported in Table 3. The BW samples (PVL 1, 2 and 3 e in bold numbers, are samples which underwent image and modal analysis) showed a fine-size grains of quartz, feldspar and volcanics scattered in a brown isotropic clay matrix (49.8%) and a 0.2 mm thick vitrified
cpx
5.1. Optical microscopy
qz/fd*
5. Results
Sample
ca. 3000 points for each sample, evaluating the abundance of paste constituents (grains, matrix and pores). Using a manual procedure, image analyses also permitted measuring some shape parameters of the inclusions, such as the minimum (Am) and maximum (AM) axis of single grains, assuming the grain inscribed in an ellipse. The two axes were used to calculate the Krumbein f size (f ¼ log2 AM; Krumbein and Sloss, 1963) and the Am/AM axis ratio that was assumed, in this study, as a shape factor (SF). As SF values range from nearly 1.0 for circular grains to nearly 0.0 for elongated grains, a logit transformation, Logit (SF) ¼ Ln(SF/(1 SF)), was applied (Prakongkep et al., 2010). Optical conditions for particles measurements varied as a function of the inclusion grain size. Image resolution was 1280 1024 pixel, for finer grains captured with 40 magnification (1 mm ¼ 480 pixel) and for coarser grains captured using 25 magnification (1 mm ¼ 320 pixel). All images were captured in parallel-polarized light in order to avoid extinction problems. Bulk chemical compositions of the shards and the reference clay samples were evaluated by XRF (Philips PW 1400): ten major elements (Si, Ti, Al, Fe, Mn, Mg, Ca, Na, K, P as oxide %) and 9 trace elements (Ni, Rb, Sr, Zr, Nb, Sc, V, Cr, Ba in ppm) were determined on pressed pellets. Analytical procedures were carried out according to Melluso et al. (2005). Volatile content (LOI) was determined measuring mass lost (1 g of powdered sample) heated at 1000 C. XRF analyses were not performed for sample PVL 28 and 41 due to a lack of sufficient material. A statistical multivariate approach was carried out on the XRF data set to better evaluate the chemical behaviour of the ceramic fragments in a multidimensional space and to verify homogeneity in the different data populations. Statistical treatment of the Grain-Size Distribution (hereafter GSD) and the XRF data was carried out using the R version 2.10.0 software (R Development Team, 2005). Two analytical approaches were united in order to evaluate the firing temperatures of the ceramic shards. X-ray diffraction (XRD) permitted the identification of the mineralogical composition of the potsherds, which depends on the base-clay and possible submicroscopic phases related to the firing dynamics (Philips PW 1730/3710 diffractometer, CuKa radiation 40 kV, 30 mA, curved graphite monochromator, scanning interval 3e80 , step size ¼ 0.020 2q, counting time 5 s per step). Scanning Electron Microscopy (SEM Jeol JSM 5310) observations, carried out on goldcoated fresh-fractured fragments, provided information in terms of microstructures and the sintering degree of the clay matrix (Maniatis and Tite, 1981). Bulk mineralogy of clay-rich sediments was investigated by XRD on randomly oriented powders; clay minerals were identified on the
