Multianalytical investigation of wasters from the Tower 8/Porta di Nola refuse middens in Pompeii: Sr–Nd isotopic, chemical, petrographic, and mineralogical analyses

A total of nine representative pottery fragments belonging to ferruginous, carbonate, and thin‐walled wares were recovered in refuse middens outside the fortification wall of Pompeii and subjected to a program of multianalytical operations (thermal ionization mass spectrometry, X‐ray fluorescence, X‐ray powder diffraction analysis, scanning electron microscopy–energy‐dispersive spectrometry techniques, optical microscopy). The fragments bear manufacturing defects, indicating their local production in workshops located somewhere at Pompeii. These three groups display a similar volcanic coarse component that exhibits distinct chemical compositions. The volcanic component consists of alkali feldspar, clinopyroxene, plagioclase, minor garnet, and rock fragments (with primarily plagioclase and leucite), pointing to an origin from the Somma‐Vesuvius. The fingerprint of the Sr–Nd isotopes of carbonate ware suggests an affinity with high‐CaO clays from Rufoli, a subdivision of Salerno. Sr–Nd isotopes also suggest that clays from the Sorrentine Peninsula were used: A clay mixture of different argillified pyroclastic materials was employed for the low‐CaO ferruginous ware, whereas the low‐CaO thin‐walled ware was manufactured with a marine varicolored clay. The distribution of materials likely occurred by sea via the port at Salernum and Surrentum. The combination of different types of complementary data obtained through this program of analysis illustrates the importance of combining both quantitative petrographic and chemical characterization in the evaluation of archaeological pottery.


| INTRODUCTION
A variety of research projects carried out at the ancient Roman town of Pompeii in recent years has substantially enlarged the body of evidence at our disposal regarding the local manufacture of pottery (Cavassa, 2009;Cavassa et al., 2013Cavassa et al., , 2015Ellis et al., 2015;Grifa et al., 2021a;Peña & McCallum, 2009a, 2009bToniolo & Osanna, 2020).
This evidence includes the remains of pottery workshops that contain production fixtures (clay mixing basins, mounting pits for potter's wheels, kilns), potter's tools, and unused raw materials, as well as vessels and vessel fragments recovered either at these establishments or at other locales in or near the town that exhibit manufacturing defects, indicating that they are wasters discarded in the course of the manufacturing process. This evidence has the potential to provide new insights into the different wares and forms manufactured at Pompeii, as well as the geography, organization, technology, and chronology of Pompeiian pot- The program of analysis was aimed at determining the following: the general nature of the raw materials employed for the manufacture of the vessels; the likely locations of the sources of these raw materials; and the paste preparation and firing practices employed in the manufacturing process.
More generally, the determination of the specific compositional characteristics of pottery manufactured at Pompeii will facilitate the recognition of pottery of Pompeiian origin recovered both at Pompeii and at other archaeological sites.

| THE ARCHAEOLOGICAL CONTEXT AND THE POTTERY ANALYZED
In 1978, an archaeological team from the Università Statale di Milano under the direction of C. Chiaramonte Trerè excavated a set of three large refuse deposits consisting of mixed material that had been dumped against the outer face of Pompeii's fortification wall in the north-central part of its circuit, between Tower 8 to the west and the Porta di Nola (Nola Gate) to the east (Chiaramonte Trerè, 1986) (Figure 1a,b). The Milan team subsequently published a catalog of the pottery, ceramic lamps, and vessel glass recovered in these features (Romanazzi & Volontè, 1986). During (Peña, 2020). In the course of this study, the PALHIP team recognized a modest number of pottery fragments that bore more or less prominent manufacturing defects (vessels marked by distortion and/or cracking, with ceramic bodies that display conspicuous reduction or discoloration, advanced vitrification, and/or bloating), and on the basis of this evidence, it inferred that these deposits had received small amounts of refuse originating at a pottery workshop (or, less probably, two or more workshops) likely located somewhere in the unexcavated part of the town, presumably at no great distance from this segment of the fortification wall. The available evidence indicates that this workshop was active for some undefinable period of time during the final quarter of the firstcentury BCE and/or the first half of the first-century CE. The PALHIP team evaluated the ceramic body of the fragments of waster pottery and fragments of pottery from these deposits that did not bear obvious manufacturing defects that were judged likely to be products of the same workshop. A macroscopic description of Group 1: the use of an unmodified (or largely unmodified?) "ferruginous" clay derived from the argillification of a parent material of volcanic origin; Group 2: the addition of volcanic sand temper to a carbonate clay of probable marine origin; Group 3: the use of a "ferruginous" clay derived from the argillification of a parent material of volcanic origin, possibly a fine fraction of the clay employed for the manufacture of Group 1.
The PALHIP team and the archaeometric research group based in the Dipartimento di Scienze della Terra, dell'Ambiente e delle Risorse, University of Naples Federico II, subsequently initiated a collaborative program of analysis involving these materials. This represents an extension of the Naples group's ongoing program of research involving the mineralogical and chemical characterization of archaeological pottery and relevant geologic materials from the broader Campania region. The Parco Archeologico di Pompei approved a program of analysis that involved the characterization of a limited number of specimens, reflecting the fact that the application of the suite of analytical procedures proposed envisaged the destruction of a small, though not insignificant amount (ca. 1-2 g) of each specimen analyzed. The PALHIP team accordingly selected a set of specimens for analysis that included three examples of each of the three groups that it had identified-designating these as PomT8PN1-9-with the assumption that this might permit-if only in a very minimal way-the recognition of a general group composition for  | 715 each of the three groups and perhaps also the identification of any compositional outliers. For all three groups, the three specimens selected were rim or base/knob fragments that clearly belonged to different vessels. For Groups 1 and 3, the set of specimens included a specimen that was unambiguously a waster, a second specimen that was thought likely to be a waster, and a third specimen that, while highly similar to one of the first two specimens in form, dimensions, and technique, and thought perhaps to belong to the same kiln load in which that other specimen had been fired, had a ceramic body that displayed more regular firing (rather than advanced vitrification and reduction), as this facilitated the characterization of the ceramic body and was thought likely to permit the acquisition of more complete and indicative information regarding texture, mineralogical composition, and firing temperature. The set of materials displaying manufacturing defects that was available for Group 2 was somewhat more limited than it was for the other two groups, and the three specimens selected were all specimens that were judged to be a probable or possible waster that did not display either advanced vitrification or reduction.
The three pottery groups, their manufacturing defects, and related profile, identified by the PALHIP team, were referred here as follows ( Vesuvius and 20 km to the southeast of the modern city of Naples ( Figure 1c).
To identify the raw materials used for the production of pottery at Pompeii, we focused our attention on the outcrops closest to the city and those that, despite being located at some distance, are more widespread and, in some cases, have been employed in historically recent times for the traditional production of ceramics.
Ethnographic studies of traditional potters indicate that in the vast majority of cases, these craftsmen employ clay and tempering material obtained from sources located within no more than ca. 7 km of the locus of manufacture (Arnold, 2006;Kelly et al., 2011). This practice is substantially a function of the wide availability of potting  .
Deposits of material suitable for use as a temper, in contrast, would have been widely available in the environs of Pompeii in the form of volcanic sand that had weathered out of SVVC formations. These rocks vary from potassic to ultrapotassic (e.g., Conticelli et al., 2004), with the degree of alkalinity and thus silica undersaturation increasing with time from (1) weakly silica-undersaturated (potassic series or KS) pre-caldera products (>8.9 ka), through (2) mildly silica-undersaturated (high-potassic series or HKS) syn-caldera products (from 8.9 ka to 79 CE), eventually to (3) strongly silica-undersaturated (HKS) post-caldera products (younger than the 79 CE eruption). KS rocks range in composition from K-trachybasalts to trachytes. Rocks of the mildly undersaturated HKS are phonotephrites, tephriphonolites, and phonolites, whereas those belonging to the highly undersaturated HKS range from leucite tephrite to leucite phonolite. The sands derived from these formations accordingly consist of grains of leucite, clinopyroxene, alkali feldspar, plagioclase, biotite, and olivine, with accessory garnet (Joron et al., 1987), along with volcanic lithics and leucite-bearing scoriae. Although volcanic sand of this kind would have been present in some larger or smaller amount in virtually any depositional basin, either at or in the environs of Pompeii, it probably would have proved convenient for Pompeii potters to obtain tempering material in quantity by collecting beach sand at some location along the shore somewhere in the immediate vicinity of the town. To the north of the mouth of the Sarno River, deposits of beach sand are composed almost exclusively of volcanic material derived from SVVC formations, whereas at the mouth of the river and along the coast to the south, these consist of a mixture of material originating in SVVC formations and carbonate rock fragments derived from the Monti Lattari (Garzanti et al., 2002;Morra et al., 2013).
It is widely accepted by scholars of the Roman world that the transport of cargoe by water would have been substantially less costly than their transport overland (e.g., Greene, 1986). The fact that Pompeii was situated on the coast and functioned as a port, thus, raises the prospect that the potters who worked there obtained their raw materials from sources located beyond the immediate environs of the town, especially in cases in which a source was GUARINO ET AL.   (Cavassa et al., , 2015, provided new and remarkable insights into the ceramic production cycle in the town. The outcomes of this study indicate that this production involved the use of marine clay with a composition strikingly similar to that of the clay from the Rufoli di Ogliara and Montecorvino Rovella sources (Grifa et al., 2021a(Grifa et al., , 2021b.
Worth noting in this connection is that small deposits of marine clay closer to Pompeii than those from Rufoli di Ogliara and Montecorvino Rovella exist at several locations in the Sorrentine Peninsula.
We can consider these sediments as the fourth clay source. They occur in the form of olistoliths and olistostromes containing beds of varicolored clays that belong to Castelvetere wedge-top arenaceous deposits (upper Tortonian-lower Messinian; Vitale & Ciarcia, 2018), which in the Sorrentine Peninsula are expressed locally also as "disturbed" clay-rich deposits that occur in reworked and weathered zones affected by landslide activity (Cesarano et al., 2018).
The fifth and last of the nonlocal clay sources worth considering is an extensive set of deposits of moderately calcareous marine clay containing substantial amounts of volcanic mineral grains and rock fragments that belong to the Phlegraean Fields that occur on the north slope of Monte Epomeo on the Island of Ischia, ca. 50 km to the west of Pompeii (Cava di Leccie unit; upper Pleistocene; Barra et al., 1992;De Bonis et al., 2013. In historically recent times, clay from these deposits was employed by potters at Casamicciola, a town on the north shore of Ischia, for the manufacture of architectural ceramics and pottery (Olcese, 2011). Documentary evidence indicates that this clay was transported to Naples for the manufacture of ceramics during the 18th and 19th centuries (Buchner, 1994), and compositional studies suggest that it was also transported there and to other locales in the GBNR during the Greek and Roman periods for the manufacture of pottery (e.g., De Greco et al., 2014;Grifa et al., 2009Grifa et al., , 2016Grifa et al., , 2019Munzi et al., 2014). The sources of this material, which would have been suited for the manufacture of tablewares and storage vessels, are situated no more than ca. 2 km from the coast and would have lain within no more than ca. 3-5 km of the Roman port at Aenaria, meaning that in Roman times, it could have been distributed by sea to distant coastal locations at a moderate cost.
These five raw materials are listed in Table 1, which include, in the order described above, Sant'Agnello clay (SO1) and Agerola clay (AGE1) as argillified pyroclastic deposits, Rufoli di Ogliara (RUF1) as a high-CaO marine clay ascribed to Apennine wedge-top basin successions of the Altavilla Group, a varicolored clay type from Piano di Sorrento (PDS1), and an Ischia Monte Epomeo North clay specimen (IS6).

| ANALYTICAL TECHNIQUES
The nine specimens of pottery, and abovementioned raw materials, were subjected to a battery of analyses selected with a view to elucidate the research questions indicated above. The set of operations to which each of the specimens was subjected is indicated in Table 1.

| OM
Basic features of the ceramic body, including texture, color, and birefringence of the matrix, and the type, condition, abundance, and sorting of inclusions were evaluated in thin section employing an OPTIKA petrographic microscope equipped with a Zeiss Axiom 105 color camera running ZEN 2.2 (blue edition) software. The abundance, size range, and angularity of inclusions were estimated by reference to comparator charts (Terry & Chilingar, 1955).

| SEM
A fresh fracture surface of the specimen was examined by SEM to evaluate the degree of sintering undergone by the ceramic body with a view to assist in the estimation of maximum firing temperature (Maniatis & Tite, 1981). for the major elements and 5%-10% for trace elements (Cucciniello et al., 2017(Cucciniello et al., , 2018. The standards employed are reported in Table S1.

| XRPD
Loss on ignition (LOI) was determined by pre-drying 1 g of powdered sample material overnight at 110°C and then heating this to 1000°C.

| TIMS
The Sr-Nd isotopic composition of representative pottery and clay samples was determined via TIMS. The high compositional homogeneity of the groups allowed us to analyze one sample of CW (PomT8PN7) and TWW (PomT8PN5); from the FW group, two samples were selected (PomT8PN3 and PomT8PN4). Samples PDS1, AGE1, and RUF1 were selected from the clay deposits described in Section 3.  (Fabbri et al., 1994;Maggetti, 2001). The statistical treatment included data obtained from the analysis of several high-CaO and low-CaO clays from the Campania region to find the best correspondence between the pottery and the type of clay used, considering the geological origin; some of these clays were also subjected to the analysis of the fine fraction obtained by means of a rigorous refining process (data from De Bonis et al., 2013 and other unpublished data).
The variance results are given in Table S1, whereas the geologic details of main Campanian clays are presented in Table S2.

| Mineralogical and microstructural analyses
A semiquantitative summary of the results obtained by XRPD for all nine specimens is presented in Table 4. Figure 8 (Table 4).

| XRF analysis
The data obtained by XRF for all nine specimens are presented in       and lower in the CW specimens (128-142 ppm). F I G U R E 6 (a) Classification of clinopyroxenes from the FW, CW, and TWW samples according to Morimoto (1988). (b) Compositional variation is shown in the CaO (wt%) versus MgO (wt%) diagram. Main fields from Campania volcanic districts are reported for comparison (Melluso et al., 2012, 2014, andreferences therein, andProf. Melluso, unpublished data  The results are presented in Table 5 and shown in Figure 11. Rovella (samples labeled RUF and MCR; Figure 12a).

Contemporary ceramic producers at both Rufoli di Ogliara and
Montecorvino Rovella state that the clay that they extract from the Apennine wedge-top basin deposits that outcrop at these two locations has a tendency to crack during drying due to its high rate of shrinkage, and they accordingly mix it with coarse material of various kinds when preparing their paste to counteract this phenomenon (Peña & Kane, 2016). Pompeiian potters likely added volcanic sand temper to the paste employed for the manufacture of CW pottery for this same purpose.
As previously noted, the location of the outcrop of Apennine wedge-top basin clay at Rufoli di Ogliara raises the possibility that in Roman times, material from this source was distributed by sea via the port at Salernum to distant coastal settlements, such as Pompeii (see also Grifa et al., 2021a). The shortest route sailing distance for a coasting voyage from Salernum to Pompeii would have been ca. 36.5 nautical miles (67.6 km) long (http://www.andiamociavela.it/files/ assume that in many cases, a voyage of this kind could have been completed at an average speed of between 2 and 4 knots (3.7-7.4 km/h) (Casson, 1951;Pryor, 1989), we can estimate a one- for the garnets assayed that are typical for the SVVC (Scheibner et al., 2007 and Prof. Melluso, unpublished data). Both of these groups can be assigned to Mannoni's Fabric Group 1a (Mannoni, 1984;Peña & McCallum, 2009b).
The results of XRF confirm that the specimens in both groups were manufactured with a low-CaO clay. The TWW specimens are more homogenous than FW specimens, and they exhibit nonoverlapping ranges of values for most of the major and trace elements measured. This may suggest the use of two distinct low-CaO clayey raw materials. To identify the possible nature and geographic origin of these materials, we subjected the chemical data for these obtained by XRF to HCA. In this case, PCA showed that 95% of the cumulative variance of the sample population was explained with the first seven components (TiO 2 , Nb, K 2 O, Rb, MgO, Al 2 O 3 , Ni). The GUARINO ET AL.
| 729 resulting HCA dendrogram indicates the existence of a high level of chemical homogeneity among the three specimens belonging to each of the two groups, confirming the compositional integrity of these as production groups (Figure 12b).
The TWW (Group 3) specimens show a good affinity with a set of clays composed of specimens from several widely separated locales at a considerable distance from Pompeii, including three specimens of alluvial clay (CET2, PMV2, VEL1) and seven specimens of so-called varicolored clays of marine origin (BS1, BS2, SMV1, SCP1, CPR2, SQ1, PDS1) (see Table S2 for geologic details). Interestingly, the TWW specimens are clustered close to the PDS1 clay specimen. This clay, which was collected on a hilltop in the Piano di Sorrento area (Figure 1c), can be attributed to olistostromes and olistoliths composed of varicolored clays that belong to the Castelvetere wedge-top deposits (Vitale & Ciarcia, 2018). This association is amplified by the results of the program of isotopic analysis. Here, the TWW specimen analyzed (PomT8PN5) exhibited an Sr isotope ratio that was substantially higher than that obtained for the other pottery specimens analyzed ( Figure 11). The PDS1 clay specimen exhibited a more radiogenic Sr isotope ratio, and the theoretical mixing curve extending from the AQM2 volcanic sand to this clay ( Figure
T A B L E 5 New chemical analysis (XRF) of major oxides (wt%), trace elements (ppm), and LOI (   Indeed, isotopic analysis showed affinity between the FW group and the Sorrento clay, also suggesting a mixing with another volcanicderived clay. This would indicate that a mixing process, such as the one described above, was implemented also in the past to correct plasticity and improve processing. In fact, the composition of argillified pyroclastic deposits in the area exhibits a certain variability, and it is likely that in the area of S. Agnello, there may be a material with a composition more similar to that of AGE1. The robustness of the isotope method showed differences between the two analyzed FW specimens that indicate the proportions of the abovementioned mixture. source area for the purpose of acquiring a load of one or the other of these two materials, or to put in at one of these locations for this purpose in the course of a trip being made for some other purpose. As noted above, a set of four passages in the ancient literary sources supports the inference that Surrentum was a source of clay that was extensively employed for the manufacture of pottery by craftsmen who operated in this area during more or less the same period as that when the pottery that is the subject of this program of analysis was manufactured. The most interesting of these is an epigram in the Anthologia Graeca (11.27), probably composed at some point between the 30s and 50s CE, which indicates that Surrentum possessed a highly regarded potting clay that it terms trecheȋa (rough) and myrípnoe (sweet breathing), implying that this was employed for the manufacture of vessels for the drinking and/or the storage/packaging of wine. A passage in Pliny the Elder's Historia Naturalis (31.60), probably written in the years immediately before 77 CE, states that Surrentum was renowned for the production of a type of vessel termed a calix, which is probably to be understood as a vessel for the drinking of wine. An epigram by Martial, probably composed at the end of the first or the very beginning of the F I G U R E 11 87 Sr/ 86 Sr and 143 Nd/ 144 Nd isotope ratios of PomT8PN7 (CW), PomT8PN5 (TWW), PomT8PN3 (FW), and PomT8PN4 (FW) from Pompeii, AGE1 clay (argillified pyroclastic deposit) and RUF1 and PDS1 clays (Apennine marine clays). Two mixing curves between RUF1 (orange line) and PDS1 (green line) clays mixed with AQM2 temper and a mixing line extending from SO1 and AGE1 clays (blue line) are elaborated. The light blue field represents the isotopic composition of Neapolitan volcanic products (data from Brown et al., 2014;D'Antonio et al., 2007D'Antonio et al., , 2013D'Antonio et al., , 2016Di Renzo et al., 2007, 2011. Symbols as in Figure 10. indicates that Surrentum manufactured vessels for both the storage/ packaging of the wine that it produced (presumably amphorae) and for the drinking of this wine (presumably calices).
Taken together, these passages indicate that the Surrentum area possessed sources of coarse or gritty clay that was employed throughout much of the first-century CE for the manufacture of amphorae and wine-drinking vessels, or perhaps sources of two distinct clays, given the different performance properties generally required for the manufacture of these two quite different kinds of vessels. In light of this observation and the ease with which it would have been possible to transport clay from the Surrentum area to Pompeii by sea, it seems possible that the FW group, which consists principally of heavy, utilitarian forms analogous to amphorae, was manufactured in a coarser paste produced using volcanic-derived Sant-Agnello clay, whereas the TWW group, which is composed mainly of drinking vessels, was manufactured in a finer paste produced using marine Piano di Sorrento clay.
Additional important information related to firing technology was obtained by means of XRPD and SEM analyses on the nine specimens included in the program of analysis and selected for one or more readily apparent production defects that point to irregular firing. We refer to equivalent firing temperature (EFT) as mineralogical and microstructural transformations occurring upon firing, depending not only on maximum firing temperature but also on time and redox conditions of the kiln atmosphere (De Bonis et al., 2017).
For the FW specimens (Group 1), the regular presence of hematite in samples PomT8PN2 and PomT8PN4 (Table 4) points to prevailing oxidizing firing conditions and EFT not lower than 750°C (Nodari et al., 2007), whereas remainders of illite/mica indicate that the EFT would not have exceeded 950°C . The presence of hematite and hercynite in the third specimen (PomT8PN3) points to both oxidizing and reducing firing conditions, whereas mullite indicates EFT higher than 1000°C . This is also consistent with the continuous vitrification (Table 4 and Figure 9a) of the matrix observed in this specimen (Maniatis & Tite, 1981).
F I G U R E 12 Hierarchical cluster analysis dendrogram resulting from multivariate statistical analyses of chemical data for high-CaO (a) and low-CaO (b) specimens (pottery and clays). Labels of samples ending with "cf" indicate the respective clay fraction of some selected specimens . Additional information on clay specimens is reported in Table S2. Symbols as in Figure 10 [Color figure can be viewed at wileyonlinelibrary.com] For the CW specimens (Group 2), the illite/mica in traces in one of these (PomT8PN8) indicates EFT closer to 950°C, whereas traces of analcime could be related to post-earthen weathering of pottery (Schwedt et al. 2006). The absence of illite/mica in the other two CW samples (Table 4) would suggest that firing exceeded 950°C. Moreover, the higher amounts of pyroxene detected in CW specimens contrast with the amount of the same mineral observed in the temper via OM, thus suggesting its occurrence as newly formed phases as well (e.g., Izzo et al., 2021). Newly formed pyroxene starts to form at about 850°C and increases its amount with the temperature. The continuous vitrification (Figure 9b) detected in the specimen in this group for which SEM analysis was performed (PomT8PN7) suggests a firing temperature higher than 1000°C.
However, the presence of newly formed gehlenite (Table 4) allowed for a more precise estimation, indicating that EFT would not have exceeded 1050°C .
For the TWW specimens (Group 3), the presence of hematite in all three indicates prevailing oxidizing firing conditions. Traces of illite/mica in two of these (PomT8PN5 and PomT8PN6) point to EFT not higher than 950°C, whereas in the other specimen, EFT could be slightly higher (Table 4). The extensive vitrification (Figure 9c) of the matrix in the specimen for which SEM analysis was performed (PomT8PN5) suggests a maximum firing temperature, somewhat lower than that attested for the specimens belonging to the other two groups observed at SEM, most likely in the range of 850-950°C (Maniatis & Tite, 1981).

| CONCLUSIONS
This study involved the physical and compositional characterization of nine pottery specimens belonging to three different groups that were recovered in the excavation of a set of refuse middens deposited against the exterior of the fortification wall of Pompeii.
These specimens, apparently manufactured at some point during the last quarter of the first century BCE or the first half of the first century CE, bear manufacturing defects that together with the very high firing temperatures (up to 1200°C), indicate that they are wasters presumably originating at a nearby pottery workshop. They can thus inform us about various aspects of pottery production at Pompeii.
The results indicate that one of the three groups-carbonate ware (Group 2) consisting of jars and basins with a light-colored body, was manufactured using a high-CaO clay to which volcanic sand was added as temper. The composition of the clay is compatible with that attested for marine clay from an outcrop at Rufoli di Ogliara in the outskirts of Salerno, 28 km to the east-southeast of Pompeii, and may well be from this source.
The tempering material is volcanic sand deriving from a formation belonging to the Somma-Vesuvius Volcanic Complex that was likely beach sand collected along the shore of the Bay of Naples to the north of the mouth of the Sarno River, probably at no great distance from Pompeii.
The other two wares-ferruginous ware (Group 1)-consisting of utilitarian vessels with a coarse, ferruginous body-and thin-walled ware (Group 3)-consisting of thin-walled drinking vessels in a gritty, ferruginous body-were both manufactured with a low-CaO clay and contain coarse volcanic inclusions that might be either a natural component of this clay or added temper. The nature and specific provenance of the raw materials employed to manufacture these two groups remain uncertain and need to be explored further through These results are significant for the evidence that they provide regarding the sources of the raw materials that Pompeiian potters employed to manufacture the wares that they produced, as it suggests that for a significant portion of their output these craftsmen utilized, clays from sources lying beyond the immediate environs of the town that probably reached Pompeii by sea. To the extent that pottery production at Pompeii involved the use of clays from distant sources, transported by sea via the port at Salernum and Surrentum, that was perhaps distributed to multiple production centers around the greater Bay of Naples Region and possibly even beyond, the ambiguities introduced by this practice will complicate the recognition of ceramic compositional attributes-mineralogical, chemical, and isotopic-that can be regarded as diagnostic of an origin at Pompeii.