A.J. Nijboer, P.A.J. Attema & G.J.M. van Oortmerssen

INDEX

1.    Introduction
2.    Location of P13
3.    The excavation of P13
4.    Report of the finds
5.    Salt production
6.    Epilogue
7.    Catalogue (see full report printed in Palaeohistoria 47/48, 2005/2006)
8.    Fabrics (including related pottery from catalogue)
9.    Acknowledgements
10.  Postscript
11.  References

1. Introduction

The fabrics of the ceramics presented here derive from our excavations of a Late Bronze Age saltern on the coast near Nettuno (Rome, Italy).

A preliminary publication of this site can be found under:

Table 1: ratio protohistorical impasto and Roman potsherds in trenches A-H from the 2001 and 2002 campaigns.

 
Origin and composition of the protohistorical deposit at P13 are interpreted by us as various dumps in the dune sand and considered to have been related to activities taking place during the Late Bronze Age. The sheer number and size of the protohistorical sherds, the old fractures, the good preservation of the ceramics (hardly any abrasion) and the fact that it was possible to refit a substantial number of the sherds recovered, indicates that in situ protohistorical features cannot have been far removed from the spot where they were dumped. The protohistorical assemblage is characterized by an association of sherds of large jars (pentole) and often burnt tufa chunks with worked sides. That we do not deal with settlement debris is clear from the very limited number of table ware (tazze for example) and the limited number of bones. As a result of a not yet understood post-depositional process, occasionally Roman sherds intruded in the deposits of protohistorical origin. The intrusion of a few Roman sherds may perhaps be attributed to the planting of shrubs.
The study of the stratigraphy at P13 has resulted in the identification of a number of superimposed layers containing potsherds dating mainly to the Late Bronze Age of which layer 3 was identified as an undisturbed, in situ layer of protohistorical date. However, no specific features were found that point to an activity in situ. The appearance of the layer is rather to be interpreted as a dump related to nearby activities and resulted from the preparation of salt and/or fish as will be suggested below. The salterns and possibly other activity areas were located in the vicinity of the dune which we have excavated and are probably totally eroded by the sea. There is evidence that P13 was one of many of such activity areas, as comparable deposits were found nearby.
Along the coast between Nettuno and Astura a number of other deposits of protohistorical origin are exposed by marine erosion. A number of these were, like P13, first mapped by Fabio Piccarreta in the 1970’s. In a recent thesis this coastal area was investigated in more detail by the Italian protohistorian L. Alessandri in 1999-2000 extending the research area to the whole coastal stretch between Ostia and Terracina and further south up to the Garigliano. Alessandri was able to give more accurate datings to the sherd material and made important observations that are helpful in the interpretation of the deposits (Alessandri, 2000-2001; forthcoming). Of the sites reported by Alessandri, the Groningen Institute of Archaeology investigated in 2002 a section known as Depuratore on the shore near Nettuno in more detail. Like P13, this section had thick layers of protohistorical pottery. This site as well appears to be related to industrial activities dating mainly to the Iron Age (Attema et al, 2003; Tol, 2005 internal report). The Depuratore site will be published separately. We will now turn to a discussion of the classification of the P13 ceramics.


4. REPORT OF THE FINDS

The catalogue of ceramics excavated at P13 is presented in vessel classes. In the classification a basic distinction is made between open and closed vessel. An open vessel has its maximum diameter at the lip. A closed vessel has its maximum diameter somewhere between the rim and base. Diagnostic forms concern rims, bases, cord decorations, lugs and handles. The typology lists the following aspects: Class, Type, Provenience (trench/spit/layer), Date and Type parallels. As far as possible the Italian equivalent of the vessel class is given as well. A typology was made since so many sherds were similar, especially those pertaining to (large) jars. Smal sherds were placed in the typology if possible. Reconstructed fragments of jars often demonstrated that the lip form within one vessel varied, for example, from convex, bevelled to flattened. Therefore, lip forms could not be a defining characteristic of a type of jar. Sherds that could not be attributed are presented as miscellaneous.
Examples of each class and type are published in the full report. Some examples can be found under one of the six Nettuno fabrics described.
 

Class I     STORAGE JARS/PENTOLE
Diameter lip is larger than 30 cm.

Class VII       BASES

Class VIII     CORD DECORATIONS

Class IX         LUGS

Class X           HANDLES



The bulk of the excavated sherds consists of large jars, which are difficult to date exactly because they are of generic type (cf. the storage jars and jars in: Alessandri, 2000-2001; Mandolesi, 1999: p. 174; Cassano et al., 1978; Damiani et al., 1998: Tomb 2). The excavated storage jars/pentole as such can be dated to the Late Bronze Age as well as to the Early Iron Age. Most of the storage jars are coarsely made, which accounts for the varieties of lips within the types. The lip of a jar could vary from convex to flattened, to bevelled. The majority of the jar types are smaller versions of the storage jar types I.2, I.3 and I.4 and as such can neither be dated exactly.
The few drinking vessels (bowls and cups) excavated at the site have some type parallels from nearby Casale Nuovo. Bowl type III.1 and cup type V.1 are for instance also known from Casale Nuovo context US 141 and context US 11 (Angle et al., 1992; 1993). Type III.1 has type parallels from Bronzo recente to Bronzo finale while type V.1 can be dated to the early stages of the final Bronze Age.
In 2002 one worn sherd of depurated clay with slip decoration was recovered at P13 (catalogue entry 126): P13F5.3/1 under bowl miscellaneous. It might be a fragment of an Italo-Mycenean bowl due to the context in which it was found and the lack of exact parallels from Mycenean contexts (Mountjoy, 1999; Bettelli, 2002). Close type parallels of this sherd have not been found though as a bowl it is somewhat comparable to Italo-Mycenean and grey-ware bowls found at Broglio di Trebisacce (Bettelli, 2002: p. 203, no. 29; Belardelli, 1994; Vagnetti & Panichelli, 1994). The Italo-Mycenean sherd from P13 could well be a local product. In 2002 Prof. Vagnetti and Dr. Bettelli were so kind to inspect the sherd. They could not assign it on account of its poor diagnostics. The fabric contains some inclusions that can also be found in the vicinity of the site P13. The shallow horizontal groove on the outside just underneath the lip is atypical for Mycenean and Italo-Mycenean bowls but is one of the features of the impasto bowls type III.5 (see catalogue entry 126 in Nijboer et al 2006).

Considering the coarse character of the vast majority of the ceramics recovered, two sherds stand out on account of their careful manufacture: the above mentioned P13F5.3/1, i.e. the fragment of a bowl that we have classified as Italo-Mycenean (under bowl miscellaneous; catalogue entry 126) and the fine black cup P13D.S5.L2/11-14 (under Cups type V.1; catalogue entry 137). Both sherds have a fabric that contains hardly any visible inclusions and need therefore be classified as fine pastes. As such they are an anomaly within the repertoire excavated at P13 and can be considered as remnants of a fine ware tradition that is more apparent in other Italian regions. Italo-Mycenean ceramics were made in southern Italy and possibly in the lower Po Valley at Frattesina during the Late Bronze Age (Jones et al., 2002). No production centre of fine wares during this period has so far been discerned in southern Lazio though the excavators of Casale Nuovo have speculated on the existence of such a production centre at the site (Angle and Zarattini, 1987; Angle et al., 1992; Angle et al., 1993; Loney, 2000; Mater, 2005: pp. 96-97).
The uncommon fabric of the fine black cup P13D.S5.L2/11-14 (under Cups type V.1; catalogue entry 137) resembles somewhat the account of a few sherds recovered at Torre Mordillo in southern Italy. These three sherds are listed as ceramica tornita a pasta grigia con superficie nera and are dated to the final phases of the Bronzo recente period (Vagnetti, 2001: pp. 329-330). It is however unlikely that P13D.S5.L2/11-14 is formed on a wheel.

In general close type parallels of the pottery excavated at P13 are found at Ficana (Malmgren, 2001) and at Torrionaccio Period 1, 2, 3 and 4 (Cassano et al., 1978). This site in the Agro Tarquinese was recently dated by Pacciarelli to Bronzo Recente – Bronzo Finale 3, in absolute years according to him from 1325/1300 – 950/925 BC (Pacciarelli, 2000: pp. 68, 95-107). The complete lack of incised decorations at P13 also on the thousands of wall fragments not catalogued can have a chronological significance since these decorations are more common in the Bronzo finale than in the Bronzo recente period.
At this stage of research we date P13 on account of the pottery excavated to the Late Bronze Age. The characteristics of the sherds (not worn and large fragments that could partly be reconstructed; see for example cup 5.1; catalogue entry 137) and of the context (high concentration of sherds in a relatively small area) indicates that the deposit at P13 is formed in a relatively short period. Two radiocarbon samples from animal bones found at P13 in the lower levels of trench F (spits 9 and 10) date the context to the 14th - 11th century BC.
Interpretation of the calibration of the 14C dates in combination with the associated pottery makes an absolute date for P13 around the 12th century BC likely, which implies a context that can be assigned to the early stages of the Italian Bronzo finale period.


5. SALT PRODUCTION

A direct parallel for P13 is provided by a specific protohistorical context in a district called Le Saline on the coast of South Etruria at a distance of ca. 9 km from the ancient site of Tarquinia. The context is described as an organic stratum with an extension of ca. 10 m “contenente esclusivamente resti di grandi contenitori di impasto rosso-bruno”. The large containers are described as consisting “in gran parte da olle e doli di medie e grandi proporzioni, realizzate probabilmente in loco con un caratteristico impasto di colorazione rosso-bruno, a volte decorati con cordoni plastici sul corpo o sotto l’orlo” (Mandolesi, 1999: p. 174). Like at P13, table ware is almost absent in the artefact assemblage of this context and there is no mention of settlement debris in the form of bones, seeds or building remains. The site is part of a series of exposures of protohistorical strata in an area of about 60 ha, which are generically dated to the Villanovan period (Mandolesi, 1999: p. 174). These strata are found both on the coast and more inland. A second context furnishing a direct parallel comes from Scolo dei Prati in the same area. In località Fontanile delle Serpi, during cleaning of the canal, a stratum became visible that contained again “quasi esclusivamente resti pertinenti a grande contenitori assegnabili tipologicamente alla Tarda Età del Bronzo e, al massimo più tardi, alla Prima Età del Ferro”. Also at this site hardly any tableware was recovered and there is no mention of associated settlement debris (Mandolesi, 1996: p. 56). Scolo dei Prati may represent “il primo approccio verso la costa di una rinnnovata comunità tarquiniese che andava lentamente radicandosi, tra Bronzo Finale e Primo Ferro, sul territorio di pertinenza” and is seen as part of a series of minor sites with specific functions for the control and exploitation of resources (exploitation of the lagoonal environment for salt and fish). Towards the end of the Bronze Age, Mandolesi postulates also resident sites in the coastal area towards the inland “a ridosso del littorale”. Mandolesi interprets the coastal sites in close relationship with the developments at Tarquinia itself: “Significativo è il fatto che il processo di appropriazione della costa avviene in concomitanza con il primo intenso sviluppo dell’insediamento proto-urbano della Civita” (Mandolesi, 1996: p. 57). On account of the present evidence Mandolesi postulates the growth of a complex settlement system in the early Iron Age with Tarquinia at its centre.

The nature of the site P13 and its associated ceramics do mark a semi-permanent settlement (relation to the sea; closely packed, substantial, well preserved sherds of large vessels; fragments fit and bases/rims could be reconstructed; a majority of crudely made storage jars; traces of burning etc.). The limited range of ceramic types of the vast majority of sherds collected and the complete lack of incised decorations also on the thousands of wall fragments not catalogued, indicate that most of the pottery was produced for a specific purpose. Some of the base fragments had particular colours, which have been labelled “salt colours” (Lane, 2001: p. 41; Crosby, 2001: pp. 410-412). Some containers from P13 display a range of pinks, whites, greys and lavender on the inner surface possibly caused by direct contact with brine (water saturated or nearly so with salt) under moderately high temperatures. Base fragment P13D.s5.l2/1 exhibits the following discolourations weak red (10R 4/2, 4/3, 5/3), pale yellow (2.5Y 8/2), black (2.5Y 2.5/1: Munsell 1994), pale red (10R 6/2), pale pink (5RP 8/2) to greyish red purple (5RP 4/2: Rock-color chart, Munsell, 1963) (fig. 8.). The explanation for the appearance of these colours on ceramics while producing salt is unclear though Matson obtained comparable colours while firing salt-containing clays (Matson, 1971). Nevertheless these characteristic colours do occur only occasionally on vessels, which have been used for the manufacture of salt. Crosby describes the salt making evidence from the Bourne-Morton Canal (Lincolnshire, England) from the late Iron Age to the pre- and early Roman period and states that five container sherds (from a total of 117) “exhibit pinkish or lavender ‘salt colours’ on the inner surface” (Crosby, 2001a: p. 293).
 

Fig. 8    "Salt colours" on base P13 D S5.L2/1 (overview and detail)

Another indication for the interpretation of P13 as a saltern is the composition of different types of ceramics recovered. “Briquetage” is taken to mean the ceramic equipment (containers, supports etc.) as well as the fragmented debris of hearths/ovens used in the processing of sea salt (Lane and Morris, 2001: p. 8; Crosby, 2001b). The composition of the briquetage from several salterns is marked by the predominance of large containers for the boiling of brine in order to produce salt crystals (cf. Lane and Morris, 2001: p. 252; Chowne et al., 2001; Daire, 1994). A comparable predominance of large containers is found at P13.

A reconstruction of P13 as a salt producing site would involve the processing and manufacture of salt from tidal sources. The tide would fill man made basins slightly towards the interior, which could be closed off from intruding seawater. Under influence of wind and sun the salt concentration of the seawater in the basin would increase due to evaporation of the water. Salt will crystallise once the water is saturated with salts. The process of salt winning often involved in prehistoric times artificial evaporation of water in containers above fires in order to produce salt crystals (Gouletquer et al., 1994: pp. 123-161). Whether this process took place at P13 is hypothetical since structural features have not been excavated though burnt fragments of tufa, which might have been used as supports for the containers, were found frequently at P13. The evidence strongly points to the exploitation of sea resources while using fire and large containers. Other options for activities around P13 could be the boiling of food in seawater for the necessary human salt consumption or the production of preserved fish and processed fish products.

Several fishponds have been recorded in Roman times along this stretch of coast (Higginbotham, 1997). No hard evidence such as fishbones was recovered but the exploitation of the fish resources around P13 seems likely also in prehistory. The importance of salt for both humans and animals can not be stressed enough but will not be covered in this article. An acclaimed introduction to the topic was published in 2002 (Kurlansky, 2002).

There are several other prehistoric sites along the west coast of Italy associated with salt production. Well known are the salt-beds at the river mouth of the Tiber which were connected through Rome to the Sabine and Umbrian hinterland by the Via Salaria, a road in use from the Bronze Age onwards (Giovannini, 1985; Coarelli, 1988a; 1988b). Waarsenburg put forward the hypothesis that a comparable situation may have existed along the Astura with Satricum as a fording place (Waarsenburg, 2001). Mandolesi assembled the possible evidence for the early use of the saline near Tarquinia, where comparable pottery as to the ceramics recovered at P13 was found during a survey (Mandolesi, 1999: pp. 174-176, 194-204). The coastal area near Pisa also functioned as a saltmaking centre since the Bronze Age (Pasquinucci and Menchelli, 2002; 1999).

Saltwinning along the Tyrrhenian Sea appears to have been a fairly common and necessary activity during the Late Bronze Age. Lane and Morris who published various salterns in the Fenland of Eastern England from prehistoric times till the Roman period, proposed a model of two contrasting phases of production and settlement, which they have labelled an “Opportunistic” phase and a “Permanent Settlement” phase (Lane and Morris, 2001: pp. 385-388). The “Opportunistic” phase is characterised by non-permanent, seasonal settlements in the vicinity of the salterns. Saltmaking was periodical and accompanied by transhumance, looking for suitable spring and summer grasslands. To quote Lane and Morris ‘The excavation of such an identified saltern site must reveal quantities of briquetage and a paucity of well-preserved animal bone and domestic pottery. During this phase, settlement sites unaccompanied by briquetage debris are not found within 2 km from the known saltern sites. Saltmaking consisted of exploration and seizing the opportunity to make salt when and where environmental conditions allowed’ (Lane and Morris, 2001: p. 385). The excavated evidence from P13 conforms remarkably well to this description.


6. EPILOGUE

In this web contribution we discussed an archaeological site that provides data on the exploitation of the coastal strip between Nettuno and Torre Astura in South Lazio during the Late Bronze Age. Our work elaborates on the inventory of archaeological sites made by Piccarreta in the 1970’s as part of his topographical studies for the Forma Italiae series (Piccarreta, 1977) and follows on the work done recently by Alessandri on the protohistory of the coastal strip between Ostia and Formia (Alessandri, 2001-2002; Attema et al., 2003).
Central to our studies was the excavation of P13, a single period site of which only a tip is preserved. The site’s stratigraphy has revealed that we deal with in situ deposits, an interpretation that is underscored by the well preserved and hardly abraded, sherds that were excavated, and by the considerable size of the fragments of those vessels that could be reconstructed. The majority of the potsherds belong to large vessels while a limited number of sherds is classified as tableware. The sherds are found in association with tufa chunks with often burnt sides. The deposits from which the potsherds derive do, however, not contain any features and the sherds and tufa chunks are therefore interpreted as refuse. A salt-working site is implied by the large size of the vessels that could be reconstructed, the ‘salt-colours’ on some of the sherds and the fact that the sherds were found stacked. Study of the existing literature shows that close packing of fragmented containers is indeed characteristic for such sites. The discolouring on some of the bases in combination with the many shattered vessels and the occurrence of tufa chunks that are interpreted by us as supports to hold the vessels during the evaporation process, moreover strongly point to a specific type of salt-making technique labelled briquetage. A date of P13 to the late Bronze Age (probably 12^th century BC) on basis of the pottery typology is confirmed by the ¹⁴C dates. Several comparable exposures in this area, mapped by Piccarreta, Alessandri and ourselves, indicate that P13 is certainly not an isolated case and that briquetage activities may have continued well into the Archaic period along this coastal strip.


7. CATALOGUE

One of the aims was to obtain a typology for the Late Bronze Age sherds recovered at P13 in order to process the numerous finds. This proved to be difficult on account of the variety of lip forms within one vessel fragment. We encountered some vessels with lips varying from flattened to convex to pointed etc. within one fragment. Thus the lip form could not be considered a determining characteristic for the typology therefore reducing its significance. The types are defined by rim form and open/closed vessel. The coarsely made storage jars and bowls and less the jars as type show considerable variety, which is illustrated in the catalogue. Evolution of shapes within a type is gradual and some fragments are clearly transition forms between specific types. The more carefully made vessels within the catalogue do exhibit much less variety in lip forms.
Single sherds that were not placed into the typology are listed under miscellaneous pentole (here translated with large jars), miscellaneous olle (jars) or miscellaneous scodelle (bowls). They have been classified as such because of their form, fabric or on account of their surface treatment. Quite a few of the sherds under miscellaneous storage jars and jars are made with more care than the generally coarse storage jars and jars that were placed in the typology.
Not all sherds assigned to a specific type have been incorporated into the catalogue. We have decided to include all the miscellaneous sherds because published contexts referring to approximately the 12th century BC are still rare in the Mediterranean. The number of sherds not assigned to a type is limited, though the present catalogue might give another impression. Less than 10% of the sherds could not be attributed to a specific type of vessel. The number of large jars at Piccarreta 13 is disproportionate for a regular settlement context (about 80-85% of the total number of sherds found at the site). This percentage can be compared with the assemblage of ceramics from the Late Bronze Age settlement at Torrionaccio. At Torrionaccio the bowls and jars are dominant in all features examined (158 bowls/ciotole (51%), 82 large jars/pentole (27%) and 16 storage jars/doli (5%) on a total of 308 (see tabel 3, Cassano et al., 1978: p. 187). Therefore we maintain one of our assessments in the preliminary publication (Attema et al, 2003) that the distribution of types at P13 is unusual and does not comply with a regular settlement context.
 

8. FABRICS

8.1. Aims
Diagnostic impasto sherds from P13 were analysed to establish possible variations occurring in their fabrics. During the study of the P13 ceramics it was noted that there were various distinct groups of fabrics. As such it resembles the characteristics of the ceramics excavated at Casale Nuovo where also several pastes are reported leading to speculations regarding the unsystematic production of the pottery (Loney, 2000; Mater, 2005: pp. 96–97). Unfortunately the fabrics encountered at Casale Nuovo are not published in detail.
Fabric analysis may also give data on technological aspects and their relation to function and form of pottery. As such fabric analysis complements pottery typologies and in combination it may result in a detailed image of continuity and development in local pottery production. For two reasons fabric analysis of the pottery of P13 seemed useful. Firstly to establish whether the fabrics would reveal specific characteristics that could be related to their supposed function in the salt production process. Secondly to compare the fabrics to those of the nearby located protohistorical settlement of Satricum as published in Attema et al. (2001/2002). This was done in order to detect any differences and analogies that might exist between the Late Bronze Age fabrics found at P13 and the Iron Age fabrics from Satricum.

8.2. Method
The descriptions of the fabrics of P13 are based on both the macro- and microscopical study of fresh fractures. Fresh fractures were studied with a magnification between 6x and 30x and list measurable characteristics such as colour and hardness as well as quantity and type of inclusions, their sorting and particle size. In addition, of each fabric a minimum of two thin-sections is given to back up these descriptions with microscopical observations. It must be noted that generally speaking the thin-sections confirm the characteristics described on basis of the fresh cross-sections, be it that in thin-sections structurally a higher percentage of inclusions is detected than in fresh cross-sections. This is especially the case with fabrics having percentages above 20% inclusions. The number of diagnostic impasto sherds studied in this way was 235 on a total of 43,929 sherds. Of these 235 sherds or sherd assemblages, 210 (89%) could be grouped in one of six fabrics (see table 3). NETTUNO fabric 1 is represented best with 99 sherds (42%). Descriptions follow the method used for the fabrics of Satricum by which the characteristics are given in coded form (cf. Attema et al., 2001/2002: ‘Fabrics of architectural terracottas and pottery from Satricum’). An example of such a code in the present study is NETI/(II).AD*.ms-ps(2-4).a, variety of characteristics, whereby NET stands for the location code (Nettuno); the Roman numerals for the colour family, (in this case red and orange); A and D for the most dominant  inclusions; ms and ps for sorting characteristics of inclusions, with 2-4 denoting particle size and a referring to the total percentage of inclusions. Fabrics with comparable characteristics will, generally speaking, also have comparable codes. This, however, does not hold for fabrics with only few inclusions, as these lack the discriminating factors of dominance of certain inclusions, sorting and particle size. A full description of the six fabrics is given that were discerned in the diagnostic sherds of P13 is given in section 8.5, as well as black and white photographs of the cross-sections and thin-sections.  

8.3. Results
From the fabric descriptions given under the internal links below, it results that the vessels of P13 were hand made. As regards NETTUNO fabric 1,
NET I.=.ws*(1-4*).d, (large) FeMn nodules / (large) Augite / (hardness +) ,
this conclusion is based on the relatively large but few inclusions present in the clay, their uneven distribution, the random orientation of clay pellets and the specific patterns in shrinkage fissures. With NETTUNO fabric 2,
NET I.(ad*)eq.ps-vps(1-4).c, (large) FeMn nodules / (large) Augite,
it is especially the sorting, the variety in fissures and the variety in orientation of clay pellets that indicates that we deal with hand made pottery. In the case of NETTUNO fabric 3, 
NET I/(II).AD*.ms(-ps)(2-4).b, ((large) FeMn nodules), ((large) Augite, ((large) Flint,
it is the random orientation of fissures in different sizes, supported by the weak micaceousness of the clay pellets in all directions under crossed polarizors in thin-section that points to handmade vessels. This is assumed in spite of the preferential orientation of fissures noted in the thin-section that might indicate the use of a slow wheel. In the case of NETTUNO fabric 4,
NET I/(II).AD*.ms-ps(2-4).a, variety of characteristics,
it was impossible to establish whether the vessels of this fabric were modelled by hand and/or by slow wheel due to the high proportion of inclusions and the anhedral shape of the minerals that do not facilitate preferential orientation. It was observed, however, that the clay matrix of this fabric is hardly micaceous and that, where visible, the pellets are orientated randomly. Furthermore, it was observed that there is no difference in orientation between inclusions just beneath the surface and in the core that would indicate finishing by slow wheel. The many small and multidirectional fissures in the clay matrix moreover suggest the absence of preferential orientation. In combination these factors lead to the conclusion that also the vessels of this fabric were modelled by hand. Finally, in the clay matrices of NETTUNO fabric 5,
NET I.EH.vps(1-4).b, (large) FeMn / crumbling,
and NETTUNO fabric 6,
NET I.eh.vps(1-4).c, (large) FeMn nodules / (hardness+),
traces of insufficient blending, relatively large Ferro-Manganese nodules and fissures going in variable directions were noted suggesting that also the vessels of these fabrics were handmade.

Indications for surface treatment of P13 pottery were noted in a thin-section of NETTUNO fabric 2 (plate 2.4 above), which showed a zone with a changing clay colour and increase of small inclusions just below the surface at the interior side of the sherd. It may be that a fine tempered variety of the clay was used to finish the surface. This, we would, however, not call a veritable slip, even though the surface was treated with care. The observed difference could as well be caused by insuffiecient blending of the clay-pastes used. A comparable characteristic was observed in a thin-section of fabric NETTUNO fabric 3. Here a zone just beneath the surface was identified in which exclusively smaller inclusions were present, indicating that the object was finished with a clay containing different tempers. The transition between the two areas is characterized by a slight colour change and a shifting average size of inclusions. Surface treatment appears, however, not to be a structural characteristic of the sherds of P13.
In the fabric descriptions of both NETTUNO fabrics 1 and 2, it is mentioned that the clay used to form the vessels seems hardly processed. The insufficiently blended clays of fabrics NETTUNO fabrics 5 and 6 also testify to the poor attention paid to the preparation of the clay.
As to the issue of the preparation of the clay, we may turn to tables 4 and 5 in which a summary of the characteristics of the fabrics is given. On basis of these data we suggest that, while Ferro/Manganese and Augite may have been natural inclusions of the unprepared or hardly prepared clay, sand consisting of Quartz and Feldspars may have been added intentionally to improve its workability and/or firing characteristics. This is based on the observation that NETTUNO fabrics 1 to 4 show an increasing percentage from less than 5% for fabric NETTUNO fabric 1 to more than 20% for fabric NETTUNO fabric 4. This increase regards Quartz and Feldspars only, with Ferro-Manganese nodules and Augite remaining rather constant and even decreasing in fabrics with higher total percentages. Therefore it is assumed that Ferro-Manganese nodules and Augite were natural ingredients and that Quartz and Feldspars may have been added. Apparently the modelling of pottery was not hampered by the occasional to frequent presence of large inclusions, as table 4 shows, though during the preparation of the paste the largest inclusions will have been removed. The kind of inclusions found in all of the fabric samples that have been examinated, indicate that the ceramics deriving from P13 were produced locally.