5th ‐ Century BC Himera and the Campanian Connection: Petrographic and Archaeological Studies on Western Greek Amphorae from Poseidonia and Elea Unearthed in the Necropolis of Himera

: Within the frame of an in ‐ depth study of the corpus of about 560 western Greek transport amphorae (6th – 5th century BC) yielded from excavations at the necropolis of the Dorian ‐ Chalcidian colony of Himera in North ‐ western Sicily, one of the most interesting issues consists in the determination of their provenance. Based on archaeological considerations, nearly 100 items have been attributed to southern Campania, specifically to Poseidonia and Elea. The present paper proposes a detailed combined archaeological ‐ archaeometric investigation of 16 samples discovered at Himera and one at Jerba (Tunisia), of presumed Campanian provenance, compared with 4 local reference samples from Poseidonia and 6 samples of western Greek amphorae found at Pithekoussai and Elea, attributed to Poseidonia by previous archaeometric analysis. All samples have been submitted to a macroscopic fabric examination according to the standard methods of FACEM (Fabrics of the Central Mediterranean) and to petrographic investigation (polarised light microscopy) and digital image analyses of microstructures. Our study points to a Campanian provenance of the investigated amphorae and their distinction in a large group from Poseidonia and a small group from Elea. The identification of a numerous assemblage of 5th century BC Poseidonian transport vessels at Himera substantially underlines an earlier hypothesis about its ‘Campanian connection’ and allows for the reconstruction of an important Tyrrhenian commercial axis.


Introduction and Archaeological Background
The intensive, archaeological excavations of the eastern and western necropolis of Himera (in detail, see Section 2.1) have yielded c. 3100 transport amphorae reused in enchytrismos burials. Among this exceptionally large ensemble, c. 560 western Greek amphorae form the largest selection of this class ever unearthed in the central Mediterranean area. The term 'western Greek' indicates transport vessels developed by numerous Greek apoikiai of Magna Graecia, Sicily, the Ionian-Adriatic area and southern France. The earliest items of this family, dating to the first half of the 6th century BC, derive from a common archetype-Sourisseau's form 1α ( Figure A1a)-produced along the southern and eastern shores of Calabria, namely at Sibari, as well as in the region of Corkyra/Bouthrotum [1][2][3][4][5][6].
During the latter part of the 6th and early 5th century BC and contemporaneously to the latest examples of this shape, several colonies launch the production of Sourisseau's forms 2-3 ( Figure  A1b,c), which form a sort of morphological koiné conjoining many of the Greek colonies. For stillunknown reasons, from the second quarter of the 5th century BC onwards at the latest, the typological development of the western Greek amphorae shows at least three different tendencies: while the Ionian-Adriatic series are characterised by the appearance of 'Corinthian B' vessels of Sourisseau's form 5 ( Figure A1d), southern Calabrian workshops produce Gassner's rim shape 6 of almond-shaped profile ( Figure A1e) and central-Tyrrhenian sites introduce Gassner's rim shapes 4 and 7 (in detail, see Section 2.2).
Thanks to the selection of in-depth studies referenced above, the typo-chronological evolution of the local amphorae series of specific areas like Poseidonia (Roman Paestum), Elea (Roman Velia), or southern Calabria has already been outlined and their ceramic fabrics have been published within the frame of the FACEM (Fabrics of the Central Mediterranean) database. By contrast, we have to lament the still extremely fragmentary state of research in amphora production and circulation in the milieu of Sicilian colonies during the 6th and 5th-century BC. The systematic analysis of the ample corpus of western Greek amphorae found in the necropolis of Himera will offer a significant contribution to this issue and, more generally, break new ground in view of a diachronic serialization of well-characterised fabric groups attributed to clearly defined production areas or sites. The still ongoing study, which combines archaeological with archaeometric methods, will allow for a better understanding of economic 6th-5th-century BC interactions in north-western Sicily and beyond. Thus, the present paper constitutes a first step in this direction and anticipates some data on the circulation, in 5th-century BC at Himera, of transport amphorae from southern Campania ( Figure  A2).
Moreover, it represents the first archaeometric case study of a significant selection of amphorae of supposed Poseidonian production. A selection of production indicators and reference pottery from the area of Poseidonia has been investigated for comparison, along with additional samples of amphorae ascribed to Poseidonia and unearthed in the nearby Greek settlements of Elea and Pithekoussai (the modern Island of Ischia in the Bay of Naples). These comparator materials provide significant diagnostic elements and evidence to support the hypothesis on the connection between Poseidonia and Himera.

The Necropolis of Himera and their Funerary Contexts
The western necropolis of Himera has been investigated by systematic archaeological excavations during the years 2008-2010 by the Soprintendenza BB.CC.AA. (Beni Culturali e Ambientali) di Palermo during the construction of a new railway track between Palermo and Messina. More than 9400 graves of different types and funeral rites have been unearthed, dating within the late 7thcentury BC and 409 BC when the Greek colony was destroyed and definitively abandoned.
Together with the eastern necropolis explored during the previous decades, Himera's Archaicand Classical-period cemeteries [7] represent one of the most interesting and complete samples for the study of the funerary sphere in the western Greek milieu. Moreover, the excavations have yielded an enormous patrimony of finds helpful for many other aspects of research related to the colonial world, as in the case of transport amphorae under discussion herein.
The identification of a relevant number of amphorae from Poseidonia and Elea amongst the containers used for enchytrismos burials significantly enriches our view of imported transport vessels. At the same time, it offers a substantial contribution to the issue of Himera's relations with the Greek colonies of southern Italy. In fact, several archaeological materials have been recognised recently from both the urban area and the necropolis which testify to the presence of objects imported from Magna Graecia, bearing witness of cultural and economic interaction [8]. In this sense, a treasury of eleven stateri from Poseidonia stands out as one of the most important discoveries. The coins have been found in a young femaleʹs burial unearthed in the western necropolis. Without any doubt, the find appears to be symptomatic of interesting economic exchange between the two Tyrrhenian poleis [9].
More generally, the amphorae from Poseidonia and Elea appear to be distributed quite uniformly in the burial grounds of the Dorian-Chalcidian colony and do not show specific concentrations, which might indicate a special destination within the necropolis' topography. Out of the 16 amphorae discovered at Himera discussed in the present paper, only five items have been associated with a grave inventory useful for chronological indications (see also Table 1): 1. W6807-grave inventory represented by a black-glaze lekythos dating to the early 5th century BC. 2. W7253-grave inventory composed of five objects, including a black-glaze, globular lekythos with reserved resting surface and inferior part of the body, dating around the mid-5th century BC, and a guttus with raised handle and decoration like the lekythos of a type starting during the first half of the 5th century BC ( Figure A3). 3. W5683-grave inventory represented by a globular juglet, with raised handle, with comparisons in several items from Himera dating to the late-5th century BC. 4. W5660-grave inventory composed of two objects, a palmette lekythos decorated by rays on its shoulder and a small, black-glaze pitcher dating to the last decades of the 5th century BC. 5. W8082-grave inventory composed of three objects: a miniaturistic trefoil jug with reserved inferior part of the body, a miniaturistic skyphos with dots below the lip and a two-handled bowl with black-glaze handles, all dating around the mid-5th century BC (Figure 1a-c).

Poseidonia's and Elea's 5th-Century BC Amphorae Production through Himera's perspective
Exceptions being made for the earliest mid-8th to late-6th-century BC Pithecusanian amphorae of types A-B derived from Phoenician-Levantine prototypes [2,3] and based on the archaeological evidence currently available, Greek colonies located in the southern Tyrrhenian coastal region commence to produce transport amphorae not earlier than the late 6th or early 5th century BC. Indepth studies of stratified fragments from the Austrian excavations conducted at Elea, supported by systematic archaeometric research, have allowed for the reconstruction of a reliable, typo-chronological sequence of the 5th-century BC rim shapes of Poseidonia and Elea itself (for all typological references, in the following see Gassner [1]; Gassner and Trapichler [10]; Gassner et al. [11]; for the fabrics, see www.facem.at). Furthermore, the analysis of materials yielded by recent investigations at Piazza Nicola Amore has given evidence for the circulation of 5th-century BC amphorae produced at Poseidonia and Elea in the area of the Gulf of Naples and the existence of a local series [12] (for the fabrics, see www.facem.at).
At present, the earliest Campanian western Greek amphorae of Sourisseau's form 2 ( Figure 2a) have been fabricated at Poseidonia and date to the late 6th century BC. The repertoire of the first or second third of the 5th-century BC is characterised by Gassner's rim shapes 3-4 (Figure 2b,c) of thickened profile with a ridge at their bottom. Derived from rim type 4, in the course of the second half of the 5th century BC appears Gassner's shape 7 (Figure 2d) distinguished by slightly everted rims with internally concave profile, underlined by a ridge and combined with bulbous necks [11].
At Elea, amphorae production does not start earlier than the second quarter of the 5th century BC, following, however, the typological evolution outlined above for the series of Poseidonia [11]. By consequence, the 5th-century BC amphorae productions of this coastal, southern Tyrrhenian area, refer very clearly to a common morphological koiné.
Within the scope of provenance studies applied to the c. 560 western Greek amphorae unearthed in the necropolis of Himera, ceramic samples have been chipped from almost all of the vessels. By using binocular microscopy and digital photos of freshly broken surfaces (for this method, see http://facem.at/project/about.php#photography), items have been compared with reference samples of archaeological fabrics which have been already attributed [13]. This procedure has led to the identification of a large selection of transport vessels (N 97), presumably imported from Poseidonia, and a very small group of six items, hypothetically originating from Elea (previously, see Bechtold [14]). Representative samples of both archaeological sub-fabrics and types (Table 1) from both assemblages have been selected for petrographic analysis (see Section 4.1). In the following, we will discuss the results of this combined, archaeological-archaeometric study of 16 fragmentary western Greek amphorae attributed to Poseidonia (12 items) and Elea (4 items); one additional sample found at Jerba (Tunisia) was analysed in order to confirm the distribution of the class also in the Carthage's sphere of influence [14,15].
The amphorae selected for petrography (Table 1) document Poseidonia's complete 5th-century BC typological repertoire, which will be outlined briefly in the following. The chronologies conferred upon the amphorae are primarily based on morphological considerations and comparisons, combined with the dating of associated grave inventories, if present.
About 30 items out of the whole group from Poseidonia (N 97) refer to the earliest morphological stage of this class, classified as Sourisseau's forms 2-3/Gassner's rim shape 3 and dating in between the very late 6th and the first third of the 5th century BC. The presumably earliest items show an almost globular body (Figure 3b), while the bulk of this group is characterised by heart-shaped bodies ( Figure 3e) and shallow, cylindrical pegs close to Gassner's variant 1a. The majority of necks are cylindrical (Figure 3a,b), whereas others present short, truncated cones. Unfortunately, only three amphorae of this group are provided by grave inventories, which suggest a burial of the respective ceramic assemblage within the first quarter of the 5th century BC (Table 1). Around 30 amphorae can be referred to Gassner's rim shape 4. The vessels of this group present more or less bulbous necks, rather ovoid-shaped bodies and cylindrical pegs close to Gassner's variant 1b or type 3 (Figure 4a-c). Clearly incurving handles seem to be typical at least for some of the later items of this series, as is also suggested by a parallel from Lipari [16] (Figure 8, catalogue 27).
By contrast to the large Poseidonian assemblage dating to the first and second third of the 5th century BC, Himera's necropolis have yielded only six amphorae presumably produced at Elea ( Figure 5). W382 (Figure 5d-f) seems to represent a typical expression of the morphological evolution of the second third of the 5th-century BC series.
Finally, at least nine amphorae can be attributed to Gassner's late Classical rim shape 7 ( Figure  4f-h) dating to the last two decades of life of the Greek colony. The excavation photo of W9240 ( Figure  4g) shows an elongated, ovoid-shaped body, tapering towards a rather massive knob-peg of rounded profile of Gassner's type 3.    [14,15]).
Sample M239/1 is a waster, found in the town of Poseidonia, ascribed to a batch of deformed fictile fingers, of possible votive function, welded in a block of concotto (clay hardened by firing), likely a kiln floor or a collapsed kiln vault [17]. Based on the discovery context it is possible to date the object to the first decades of the 3rd century BC. This was found in a dump relevant to an ergasterion located in the northern part of the southern urban sanctuary, used for the production of ceramics and figurines for those who frequented the sacred area [17].
Sample M226/1 is a wedge-shaped spacer found on the laying surface of the 14/68 tomb from the Licinella, a Lucanian funerary area located south of the ancient city. It is a peculiar type of spacer used to support large vessels, architectural elements or decorative slabs and appliques during firing [18]; a chronology between the end of the 5th and the beginning of the 3rd century BC is conceivable. The item was found along with other production indicators in a limited area of the burial ground, giving evidence of the existence of an artisan workshop aimed at the production of objects used in funerary circles [17,19].
Sample M231/2 is a ribbon handle of a kitchen ware (4th-3rd century BC) while M231/3 is a rim of a storage amphora (5th-4th century BC). They were collected among the sporadic materials in the area of the Laghetto necropolis, located to the north-east, outside the city walls. These finds are unusual as grave goods of this burial ground and could be the trace of rituals undertaken outside the burials.

Amphorae Found in Elea and Pithekoussai
Because of the rural nature of the plain of Paestum we can suppose that wine and oil were produced there and then exported in amphorae. However, at the current state of research we do not possess a detailed study of the local amphorae-production. The best overview is given by M. Cipriani and A. Pontrandolfo [20], but, as mentioned by the authors themselves, this list does not distinguish imports from local products. Of importance are also the four amphorae found in the sanctuary of Hera (Foce del Sele) in Poseidonia [21] and an amphora of form Sourisseau 4 from the necropolis of Santa Venera ( [22] and Figure 37 therein). A first archaeometric overview on pottery production in Paestum was given by A. De Bonis [23].
The identification of the local production of Poseidonia/Paestum thus relies on a series of analyses conducted by R. Sauer in the 1990s on amphorae found as imports at Elea/Velia, which allowed us to define the characteristics of this production by their comparison to local clay resources collected in the territory of Paestum [24]. Most of the samples stem from the mudbrick-houses of the first half of the 5th century BC, situated under the so-called Insula II, built in the Augustan period [1] (pp. 113-119, and Table 7 therein). Among these materials, imports from the neighbouring town of Poseidonia/Paestum are frequent [1] (p. 133, and Figure 62 therein), so that we have a reliable quantity of finds. To the production of Paestum belong the FACEM fabrics PAE-A-1, PAE-A-2, PAE-A-3, and PAE-A-4. Samples for PAE-A-1 were taken from the material of the Scarico Gosetti from Ischia, previously published by N. Di Sandro [25].
Generally, from a macroscopic point of view, the fabrics of amphorae produced at Poseidonia are orange to brown and display a clearly visible residue of carbonate as well as quartz and mica. The samples selected for this contribution (Table 2 and Figure A4) range from amphorae of the form Sourisseau 3 and 4 (M 37/26, M37/28, M10/22, M10/4, 5th century BC; M6/26, 5th-4th century BC) to amphorae of the form MGS IV (M10/5, late 4th and early 3rd century BC) [26]. They continue to be used at least until the beginning of the 2nd century BC (details will be provided in the volume in preparation by Gassner et al. [27]).
Polarised light microscopy (PLM) was carried out in order to investigate the petrographic composition and microstructural features of the ceramic samples. Thin section observation was performed with an OPTIKA V-600 POL (OPTO-LAB Instruments, Concordia sulla Secchia, Italy) microscope coupled with a ZEISS AXIOCAM 105 COLOR and ZEN 2.3 Lite software (Oberkochen, Germany) for acquiring the pictures. A total of nine specimens of amphorae discovered at Himera (8) and Jerba (1) were selected because of their homogeneous petrographic and microstructural features for modal, particle-size, and shape analyses, performed on collages of micrographs (40× magnification).
Modal analysis quantitatively assessed the percentage of matrix, voids, and non-plastic inclusions via a point-counting procedure [28] performed with the 3.2.1 version of the Leica Q Win software (Wetzlar, Germany) and a minimum of 2000 points per sample. This measurement can be considered representative as the maximum uncertainty is around 2.8% already at 1500 points [28,29]. The inclusions amount of other samples was estimated via comparative charts [30].
Grain size and shape measurements were performed using the 1.52a version of the public domain Java image processing software ImageJ (National Institutes of Health, Bethesda, MD) and statistically processed with the 3.4.4 version of the R Development Team software (R Foundation for Statistical Computing, Vienna, Austria) to obtain distribution density histograms [31,32]. The batch feature in Photoshop (CC 2015 version; Adobe Inc., San Jose, CA, USA) was used to automatically apply thresholds and make binary images. Thresholding was carefully inspected before the measurement was performed. Grain size distribution (GSD) was measured considering the minimum Feret (mF) value in millimetres for calculating the Krumbein phi scale (ϕmF = −log2 (mF)). This parameter in sedimentology is based on the effective diameter determining the grain size of a class obtained by sieving [33]. The circularity value (C = 4π (A/P 2 ), where A = area, P = perimeter) was considered as a shape descriptor [34,35].

Petrographic Features of Amphorae Discovered at Himera and Jerba
The thin section analyses of the amphorae discovered at Himera and Jerba (sample M149/54) showed that most samples are characterised by similar petrographic and microstructural characteristics. The matrix is frequently birefringent or shows a weak optical activity; an inactive matrix was observed in five samples (M179/141, M179/144, M179/154, M179/210, M149/54).
Common ware samples (M231/2, M231/3) basically show the same petrographic features of production indicators, except for a slightly more active matrix (Figure 8e  Porosity is low. The inclusions are moderately sorted and bimodally distributed and are represented by quartz and muscovite in the fine fraction. Coarse grains are mainly siliciclastic and are composed of quartz, sandstone, and feldspars (alkali feldspar and plagioclase). Carbonate fragments are also common and are often represented by decomposed grains (Figure 9e-g,i) and bioclasts (M10/4, M10/22; Figure 9g). Occasional clinopyroxene was detected in M6/26, M10/4, and M10/22 (Figure 9f,h,i).

Provenance and Technology
The petrographic analysis of the amphorae discovered at Himera and Jerba This petrographic composition is also typical of coarse wares produced in the territory of Poseidonia/Paestum [23] and is fully compatible with the geological features of the area, where sources of siliciclastic and carbonate rocks crop out along with minor volcanic layers [42,43].
The locally specific petrographic features mentioned above are also evident in the production indicators and common ware from Poseidonia used for comparison. Actually, local reference materials of common ware are very similar from a petrographic point of view to these coarse-grained production indicators and, in turn, are also homogeneous to each other from a chemical point of view [23]. Based on fabric analysis, M. Trapichler [44] grouped these common wares in the existing FACEM fabric PAE-C-2 (ribbon handle, M231/2), established for Poseidonia's coarse ware [10], and in the new fabric PAE-REG-C-4 (storage amphora, M231/3), which is very similar to or is a variation of PAE-C-2 with a coarser grain size [44].
Other samples analysed for comparison are the amphorae found in Pithekoussai and Elea ascribed to Poseidonia and which testifies the circulation of amphorae from this town within the Campanian area. Samples discovered in Elea are ascribed to the FACEM fabrics PAE-A-2 (M6/26, M10/5) and PAE-A-3 (M10/4, M10/22), which are very similar to each other [10]. According to R. Sauer, samples discovered in Elea can be grouped in his petrographic type RVA003 (sample M6/26) or are very close to it. This type is marked by the presence of quartzose grains (quartz and minor sandstone), occasionally of metamorphic origin, carbonate fragments (pseudomorphic and bioclasts), alkali feldspar (occasionally sericitised), sporadic plagioclase, and muscovite. Volcanic particles such as clinopyroxene occur in minor amounts along with rare garnet. Samples M37/26, M37/28 found in Pithekoussai belong to the petrographic type RVA003a, which is analogous to the former and only differs in its slightly lower carbonate content. These samples well reflect the typical mineralogical association of the area of Poseidonia [23] and share common features with other local samples of common ware and bricks. It is worth noting the occurrence of volcanic inclusions frequently observed in the investigated samples from Poseidonia. They are derived from the pyroclastic fall deposits related to Neapolitan volcanoes (Phlegraean Fields and Somma-Vesuvius), in particular, from pyroclastics ascribed to eruptions of Mt. Somma older than that of AD 79 [43,45]. A distinctive marker of the activity of this volcano is garnet [46][47][48], which was found in one sample collected in Pithekoussai (M37/26) and is common in samples of the Sauer's petrographic types RVA003/RVA003a from Poseidonia.
Bringing up the amphorae found at Himera again, we noticed four samples (M179/133, M179/135, M179/147, M179/281) which are siliciclastic-rich and do not show any occurrence of carbonate. These samples display a composition very similar to that of samples of the Sauer's petrographic type RVA001, typical of local transport amphorae from Elea [49]. Amphorae of this type are characterised by abundant siliciclastic inclusions mainly represented by quartz, sandstone, alkali feldspar and minor plagioclase. Feldspars often show a typical sericitisation as in the analysed samples M179/133, M179/135, where metamorphic lithics also occur in slightly higher amounts with respect to other amphorae. Sporadic clinopyroxene (M179/135, M179/281) and volcanic lithics (M179/135) can also be found [49,50]. These petrographic features are distinctive of Elea [24,49] and point to a possible provenance of samples M179/133, M179/135, M179/147, M179/281 from this settlement.
The analytical considerations made so far provide a good possibility for distinguishing the pottery of Poseidonia from that of Elea, especially considering carbonate grains that are virtually absent in the latter (Table 4).
On the other hand, a provenance of the investigated amphorae from the area of Himera should be excluded because of both archaeological reasons and fabric analysis [51], which did not find any similarity with local coarse ware from Himera. Archaeometric argumentations on pottery and raw materials from Himera analysed so far [52,53] do not seem to point out an affinity with the selection of analysed amphorae. Additional analytical techniques in progress on this selection (unpublished data) and on Himerian amphorae (see Bechtold et al. [54] for preliminary results) will provide further evidence to distinguish local productions from imports.
Microstructural analysis provided some technological clues. In particular, the negatively skewed grain size distribution along with the weak bimodality occasionally shown in the density plots ( Figure 7) pointed out the addition of temper [31,34]. The moderate sorting of particles, the shape, and their relatively small size (mostly fine sand with MGS not higher than c. 670 μm) indicate a good selection of the temper used for the mixture. Actually, tempering can improve the shock resistance of ceramics. An experimental study by De Bonis et al. [55] demonstrated that discontinuities formed at the interfaces between temper grains and clay matrix help dissipate mechanical energy, thus providing a good toughness to the ceramic body, which can tolerate shocks better with no damage incurred. Furthermore, a temper amount around 20%, similar to that observed for the analysed amphorae, associated to the quite low porosity would favour the waterproof characteristics, and is a fair compromise between water resistance and toughness, especially for transport amphorae stressed for their specific use [55].

Archaeological and Historical Remarks
The identification, at Himera, of an exceptional high number of nearly hundred 5th-century BC transport vessels imported from Poseidonia confirms the close relationship between the two cities already inferred earlier [56]: among the c. 460 last quarter of the 6th-5th century BC western Greek amphorae recorded from the Himerian necropolis, the selection from Poseidonia totals around 22%, representing the second class in frequency after a huge group of southern Calabrian vessels. Amphorae from Poseidonia reach Himera regularly from the very late 6th-early 5th century to the end of the city in 409 BC, with a possible peak of influx during the middle decades of that century.
Recent provenance studies of amphorae from several southern-central Mediterranean sites have already given evidence for the circulation of apparently smaller quantities of 5th-century BC Poseidonian vessels in Tunisia (Carthage, Ghizène/Jerba, for one analysed sample M149/54), see Table 1), in the Sicilian Channel at Pantelleria, in the hinterland of Solunto at Pizzo Cannita [14], and possibly also at Lipari [1] (pp. 200-201).
Within the scope of the on-going project (see acknowledgement), the fabric analysis of further assemblages of western Greek amphorae yielded by the Grotta Vanella dump and the Mango sanctuary at Segesta and by the intensive excavations and field surveys at Entella has allowed for the identification of a handful of 5th-century BC amphorae from Poseidonia also at these western Sicilian inland-sites.
From a geographical point of view, the massive presence of the class under discussion at Himera fits nicely in the distribution pattern outlined above. Any differences which occur between our colony and the other sites are predominantly confined to quantitative aspects. However, even if on the basis of an admittedly very small number of analysed samples (16-19 items for each site), at least at the three Punic sites of Ghizène/Jerba [15], Carthage [57], and Cossyra [58] 5th-century BC, Poseidonian amphorae consistently represent the third group in frequency, after the Ionian-Adriatic and southern Calabrian series.
The new, quantified ceramic evidence from Himera justify, in any case, the reconstruction of a veritable trading axis, connecting the southern Campanian area to the westernmost Greek apoikia of Sicily's northern coast. However, the phenomenon of the massive documentation, in the Dorian-Chalcidian colony, of important quantities of storages vessels imported from Poseidonia, appears to be more complex and has to be embedded in the wider frame of the city's commercial and also its probable social interaction with the southern Etruria [59] (p. 220), as well as with the Campanian area [8,56] from the second half of the 7th century BC onwards. Concerning the latter, Cuma stands out as the most probable candidate, not only for the provenance of several architectural terracottas, but also for the transfer, to Himera, of technological experiences related to the construction of roofs and even of artisans during the late Archaic and early Classical period [8]. A special relationship probably with Poseidonia has been suspected on occasion of the discovery of 11 silver coins from Poseidonia in a mid-5th century BC grave unearthed in Himera's necropolis (see Section 2.1).
Unfortunately, almost nothing is known from the written sources about Poseidonia's 6th -5th century BC history [60] (p. 11). However, its agricultural vocation -derived from its geographic location in the fertile Sele River plain-has been sufficiently outlined. An important archaeological indicator is the interruption of the city's coinage (with a begin around 540/30 BC) from c. 490-475 BC [61]. This iatus is usually related to political changes and the probable introduction of a democratic regime. Contemporaneously, in 476 BC and as a result of the repression of its citizens, at Himera the despot Terone from Akragas repopulates the city by the abduction of Dorian people, perhaps in part from Poseidonia [9]. In the following, during the third quarter of the 5th century BC, Poseidonia becomes the most productive mint in Magna Graecia [61]. This evidence harmonises perfectly with the massive presence, in Himera's necropolis, of amphorae from Poseidonia dating to the second and third quarter of the 5th century, which underlines impressively the cities 'Campanian connection' during the decades following the battle of 480 BC and the re-population implemented by Terone. Finally, the flux of the class at Himera might have decreased towards the late 5th century BC, again in parallel with the reduction of Poseidonia's coinage and historically related to the increasing Lucanian pressure on the city [61].
By contrast, 5th-century BC transport vessels from Elea are extremely rare at Himera and currently absent at other central Mediterranean consumption sites [57], confirming Velia's supposed, scarce agricultural surplus production during the Classical period. The economic force of the Phocaean city is usually explained by its commercial activity based on trade with not-locally generated commodities. As a hypothesis, the supra-regional distribution of Poseidonia's 5th-century BC amphorae has been attributed to Elea [1], also in view of the absence, at Poseidonia, of a suitable harbour [60] (pp. [34][35]. Notwithstanding and by contrast to Poseidonia, Elea's 5th-century BC coinage appears to be very weak and characterised by a low circulation of proper money-as well as of the amphorae-outside the city's territory [61], which seems to contradict an intensive involvement in 'international' trade.
Hence, the issue of the commercial vectors responsible of the arrival at Himera of both amphorae from Poseidonia and architectural terracottas, presumably from Cuma, is still unclear. Unfortunately, no 5th-century BC ship-wreck carrying southern Campanian ceramic artefacts has thus far been found along the Tyrrhenian-Sicilian route and the question remains unanswered. It seems likely, however, that at least until the battle of Himera in 480 BC a special role at this regard has been invested by the allied colony of Rhegion [62].
The combined, archaeological-archaeometric research outlined in this paper highlights the extraordinary importance of provenance studies of non-local pottery from ancient consumption sites. In the present case of western Greek amphorae and with special regard to the late Archaic-early Classical rim shape 3 of Gassner's typology, produced by several Greek apoikiai (see Section 1), only the analysis of their ceramic fabrics and petrographic composition may lead to a correct geographic attribution.
Concerning Himera and based on the identification of the largest assemblage of 5th-century BC transport amphorae from Poseidonia ever documented outside the production region, we can now confidently sustain a special relationship between these two colonies. The phenomenon blends in with Himera's particularly close commercial and probably also social interaction with the coastal, southern-central Tyrrhenian region from its foundation period onwards. More in-depth studies in the material culture of both production and consumption sites need to be done in order to formulate hypotheses regarding possible traffic agents.  Acknowledgments: B.B. acknowledges the support of the Austrian Science Fund (FWF): P 30030-G25 for the project "Trade of western Greek amphorae (6th-5th century BC) from the perspective of Himera (western Sicily): a contribution to the identification of production centres, typological repertoires and distribution patterns". Furthermore, she thanks Ch. Michelini and A. Corretti (both Scuola Normale Superiore di Pisa) for the opportunity to sample an ample assemblage of western Greek amphorae from Entella and its hinterland. She is grateful to M. de Cesare, P. Cipolla (both Università di Palermo) and M. Quartaro (Pisa) for the possibility to sample smaller selections of western Greek amphorae yielded by the Grotta Vanella dump and the Mango sanctuary at Segesta. The issue of this research will be published in a joined contribution. Finally, she thanks F. Ferlito for his contribution to the ceramic fabric analysis and R. Lampl (Universität Wien) for the composition of Figures 2-5 and A1. A.D.B. acknowledges support from the Austrian Science Fund (FWF): M 1918-G25 for the project "Ceramic production in the plain of Paestum". All the authors thank the staff of the Parco Archeologico di Paestum and, in particular, the Director G. Zuchtriegel for providing the necessary authorization for sampling, the former Director M. Cipriani for technical support to the project. We also thank B. Huber for the drawings of Figure A4 and S. Gallagher for the English revision of the text. Finally, the authors wish to thank the Editors of Minerals, the Guest Editor of the Special Issue (G. Balassone) and the three anonymous reviewers for their suggestions that significantly improved this manuscript.

Conflicts of Interest:
The authors declare no conflict of interest.