Campana plaques from Ossaia - La Tufa (Cortona, Arezzo): from

Per. Mineral. (2002), 71, SPECIAL ISSUE: Archaeometry and Cultural Heritage, 145-166 http://go.to/permin

PERIODICO di MINERALOGIAestablished in 1930

An International Journal of

MINERALOGY, CRYSTALLOGRAPHY, GEOCHEMISTRY,ORE DEPOSITS, PETROLOGY, VOLCANOLOGY

and applied topics on Environment, Archaeometry and Cultural Heritage

Campana plaques from Ossaia - La Tufa (Cortona, Arezzo):from archaeological hypotheses to archaeometric results

MAURIZIO GUALTIERI1*, GIOVANNA ROSSINI2 and BEATRICE MORONI3

I Dipartimento di Studi Storico-artistici, Universita di Perugia - Via Armonica 3,06123 Perugia, Italy2 Dipartimento di Scienze dell'Antichita, Universita di Roma "La Sapienza" Piazzale AldoMoro 5, 00185 Roma, Italy

3 Dipartimento di Scienze della Terra, Universita di Perugia, Piazza Universita, 06100 Perugia, Italy

ABSTRACT. - The Roman villa at Ossaia - LaTufa, near Cortona, was built during the lateRepublican period and monumentalized in theAugustan period. In the second half of 1st cent. ADsome parts of the villa were radically transformedand some of the rooms employed as productiveworkshops. The on-going excavations of the villa,have yielded a number of fragments of Campanaplaques. The iconographic type attested by theplaques is that of an Siren standing on top of anacanthus tuft turned upside down, from whichluxuriant flowered tendrils develop. Two verydifferent levels of stylistical refinement have beenrecognized in the terracotta fragments: the firstshows the typical features of the best productions ofLate-Republican/Early-Imperial age, the second is areproduction in poorer style of the sameiconographic type. The aim of this work was tocharacterize these different productions in order toestablish any connections with the sequence ofrearrangements undergone by the monumentalcomplex, and to clarify some aspects of productionand circulation of Campana plaques.

The samples were characterized by means ofstereoscopic and optical microscopy, and analyzedby scanning electron microscopy. At the same time,samples of urban production from Palatine Museumwere characterized using the same methodologies,and then compared to the samples from Ossaia - LaTufa. The results point to the existence of twodifferent groups of production, local and urban, ofarchitectural terracottas in the Ossaia - La Tufavilla, corresponding to the different levels ofstylistical refinement. Local production, in poorer

* Corresponding author, E-mail: [email protected]

style, is characterized by the sole presence of asedimentary fraction which has been attributed tolocal geological formations found in the Cortonaarea. Urban production, in high style, ischaracterized by the presence of a volcaniccomponent which has been attributed to pyroclasticunits of potassium alkaline magmatic rocks found inLatium and Campania. The archaeometric resultshave been interpreted assuming importation of theproducts in high style from Rome, in the period inwhich the villa probably passed into imperialproperty, and local production of the pieces inpoorer tyle, in a subsequent period of declineoccurred during the Flavian age.

RIASSUNTO. - La villa romana di Ossaia - LaTufa, costruita durante il periodo tardo repubblicanoe monumentalizzata tra la meta del I secolo a.c. el' eta augustea, subi una modifica radicale nellaseconda meta del I secolo d.C con la trasformazionedi alcune stanze in impianti produttivi. Durante 10scavo archeologico, tuttora in corso, sono statiriportati alla luce vari frammenti di lastre Campana.II tipo iconografico attestato dalle lastre e quello diuna Sirena stante su un calice di acanto rovesciato dacui si sviluppano lussureggianti racemi floreali. Taletipo iconografico, perc, ricorre a Ossaia in dueversioni sensibilmente differenti. La prima mostra icaratteri tipici della migliore produzione del periodotardo repubblicano - proto imperiale, la secondaconsiste invece in una riproduzione in stile deterioredello stesso tipo iconografico. Lo scopo di questolavoro e stato quello di caratterizzare le due diverseproduzioni al fine di definire la connessione con lasequenza di cambiamenti intervenuti nel complessomonumentale, e di contribuire al chiarimento di

146 M. GUALTIERI, G. ROSSINI and B. MORONI

alcuni aspetti della produzione e della circolazionedelle lastre Campana.

I campioni sana stati esaminati al lostereomicroscopio e al microscopio attica, eanalizzati in microscopia elettronica a scansione.Allo stesso tempo, alcuni campioni di produzioneurbana provenienti dal Museo Palatino sana staticaratterizzati mediante le stesse metodologieanalitiche per essere poi comparati can i campiani diOssaia - La Tufa. I risultati evidenziano l' esistenzadi due diversi gruppi di produzione, locale e urbana,nell' ambito delle terrecotte architettoniche. Laproduzione locale, in stile deteriore, e caratterizzatadall' esclusiva presenza di una frazione sedimentariariconducibile alle formazioni geologiche affiorantinei dintorni di Cortona. La produzione urbana, inbello stile, e caratterizzata dalla presenza di unacomponente vulcanica che e stata attribuita a unitapiroclastiche di serie alcaline potassiche affioranti inLazio e in Campania. I risultati archeometrici sanastati interpretati assumendo l'importazione deiprodotti di stile pili elevato da Roma, probabilmentenel periodo in cui la villa divenne di proprietaimperiale, e la produzione locale dei prodotti in stiledeteriore in un successivo periodo di trasformazioneintervenuto in eta flavia.

KEY WORDS: Terracotta, Campana plaques,petrography, mineral chemistry, provenance.

1. THE VILLA AT OSSAIAICORTONA: A CASE

STUDY IN THE «ARCHAEOLOGY OF PRODUCTION»

The excavation of the Roman villa at Ossaia(Cortona,AR), undertaken in 1992 jointly bythe University of Perugia and the University ofAlberta (Canada), has uncovered the mainresidential area of one of the earliest villas sofar documented in Northern Etruria . Thanks tothe systematic and large scale nature of the ongoing exploration, the data retrieved providesmuch information on the long history of thecomplex, between ca. 100 B.C. and the LateAntique (mid-5th c. A.D.)., its architecturallay-out and material culture (Fracchia andGualtieri 1996). In the light of the most recentexcavation data, it is also clear that the laterepublican villa was built on the site of a preexisting vicus (rural hamlet) dating back to the5th century B.C., which may have had the dualpurpose of a small nucleated settlement for the

rural population and the centre of a fairlysizable estate. Thus, the late Republican villa,in spite of the monumental architecture andimpressive pars urbana, shown by its centralresidential nucleus (an elongated complex onthe main terrace with a frontage of over 100metres overlooking the lower Valdichiana - fig.1), was certainly part of a larger productivecontext, developed over a span of a fewcenturies (Fracchia and Gualtieri 2001).

Thanks to the detailed documentationprovided by the stamped tiles, alreadydiscussed elsewhere from the viewpoint oftheir production and distribution (Gualtieri2000), we can better outline the majorconstruction phases of the early complex (1st c.B.C.- 1st c. A.D.) and, at the same time,provide reasonable hypotheses about changesin the ownership of the estate in which the villabelonged. The first owner of the villa complex,to judge from a large number of stamps datedto the beginning of the 1st c. B.C., must havebelonged to a local family of some means,bearing the Latin version of an Etruscan namewell attested in the territory between Arretiumand Cortona, ANILIUS or AVILlUS(Zamarchi Grassi 2001). Around the middle ofthe century, perhaps as a result ofintermarriage, the estate passed into the handsof the VIBII (Pansae?) a well known family ofancient Etruscan descent from Perugia who isalso known to have had connections withCortona. Another exceptional group of stamps,on both brick and tiles, with CAESARUM (tobe interpreted as Cai and Lucii Caesarum, onthe basis of similar documentation availablefrom a well known estate of Agrippa aroundVibo Valentia in Southern Italy), provides anunquestionable piece of evidence for thetransferring of the Ossaia villa into the Imperialfiscus by the later first century B.C. This phaseof the villa in which the figlinae belonged tothe family of Augustus, also providesimpressive documentation of architecturalrefinement (especially the well preserved areaaround the atrium) and interior decoration (theexceptional group of black and white mosaicslaid by musivarii in close contact with urban

Campana plaques from Ossaia - La Tufa (Cortona, Arezzo) ...

LAG0TRASIMENO

Fig. I - Sketch map of the Cortona area with location of the villa at Ossaia - La Tufa.

147


workshops - Gualtieri 2001). Interestingly, thestudy (in the course of publication) of the largequantities of Italian terra sigillata associatedwith the Augustan-julio Claudian periodshows, aside to unquestionable Arretineproducts, identifiable also by some of the bestknown stamps from those workshops, anumber of locally made wares, especially forthe later productions of this class of fine wares(personal communication by Dr.J.W. Hayes;also Kolonicki 2001).

A noticeable change in the second half ofthe first century A.D. is indicated not only bythe nature of the pottery assemblages but alsoby a fourth group of brick-stamps dated to thelast decades of the 1st c. A.D., with A.GELLI.POTNI, probably documenting a new transferof ownership, into private hands (a freedmantied to the family of the Gelli, attested in theterritory between Ossaia and Cortona byfunerary inscriptions of the early imperialperiod) , which might be connected with theworks of re-structuring documented in someparts of the central residential nucleus. Here,the rooms to the south-east of the large atriumshow radical transformations into 'workingareas' by the addition of a system of canalswhich cut through the previous mosaicpavements. The evidence so far available forthe later 1st and 2nd centuries A.D., particularlythe vast assemblage of locally produced 'Spellotype' amphorae (work in progress by M.McCallum), is quite consistent with a generalpicture of decline of importance of the onceluxurious pars urbana of the complex which ismatched by a phenomenon of enhancedagricultural (and other) production.

M.G.

2. CAMPANA PLAQUES: PREVIOUS STUDIES AND

RELEVANCE OF AN ARCHAEOMETRIC APPROACH

The definition Campana plaques indicates aclass of architectural terracottas, characterizedby an ornamental syntax much different fromthat of traditional Etrusco-Italic terracottas,

with figured and narrative motifs prevailingover geometric and flowered ones (typical ofthe latter).

Such reliefs were employed in publicbuildings, appointed to sacred or civicfunctions, and particularly prestigious privatebuildings, and they were more exclusively usedthan contemporary antefixes. This genre seemsto flourish in Roman urban context, wheremost refined testimonies turn out to be directemanation of the propaganda message adoptedby the central power. Nevertheless, aprogressively growing diffusion of such aproduction on a relatively wider scale, can benoticed both in Rome and in the suburbs,presumably in relation with phenomena oflocal evergetism and, even more withmanifestations of private taste. As figuredmaterials, Campana plaques can be associatedin mythological/narrative cycles, and they openup meaningful perspectives on systems ofvalues and forms of programmatic selfrepresentation, both in public and privatecontext.

The mechanisms and logic of diffusion ofthis class of artefacts has been a long lastingmatter of debate (Anselmino, 1981; Tortorella,1981b; Strazzulla, 1987a; Strazzulla, 1987b;for a general survey, see Torelli, 1983). Therole of Rome as propelling centre of models forarchitectural terracottas in general, startingfrom 2nd cent. BC (before Campana plaquesappeared), is unanimously acknowledged.Propagation of finished products can beallowed as well as diffusion of moulds or evenof plain pasteboard models, worked out againon demand in local workshops, or in thefiglinae of rustic villas for internal use only.The issue is still debated, as specific matter.

When we are faced by specimenscharacterized by an extremely low quality andautonomous decorative patterns as well asmotifs (though inspired by Roman models), weare undoubtedly led to assume a non-urbanproduction (Anselmino, 1981). Moreover,finding of remains of kilns - an exceptionalevent - with moulds and waste materials offersa decisive corroboration to the hypothesis of

Campana plaquesfrom Ossaia - La Tufa (Cortona, Arezza) ... 149

local production, as in the case of Saint-Justworkshop (Laubenheimer et al., 1989) and inthat of Bassano del Grappa workshop(Strazzulla Rusconi, 1984).

An archaeometric approach can be extremelyuseful, or even a determining factor, in specificcases. That is when the only opportunity fordefining the productive profile of differentworkshops, the transmission of models fromone workshop to another and, eventually, thediffusion of finished products, is based onconjectures related to considerations ofhistorical and/or stylistic nature.

Until now, this class of materials has beenanalysed in conformity with many differentperspectives. Description and classificatoryplanning out of 19 th century collectioncatalogues, culminating with publication of theCorpus by H. von Rohden and H. Winnefeld(1911), was followed by new approaches to thesubject (particularly during the last thirty yearsof 20th century). Attention focused on differentspecific problems, concerning mechanisms ofplaque production systems (Tortorella, 1981a,b), typologie-functional aspect (Calderone,1975) or, eventually, iconographic-stylistic andiconologic ones (particularly interesting areBorbein, 1968; Strazzulla, 1982-1983;Strazzulla, 1990). Interpretative hypothesesoften complete the editing of new catalogues oruniversity collections, and also excavationreports (most valid: Carettoni, 1973; Rizzo,1976-1977; Di Mino, 1981; Manca Di Mores,1982-1983; Strazzulla, 1982-1983; StrazzullaRusconi, 1984; Ciffarelli, 1988; Laubenheimeret al., 1989; Dupre and Revilla, 1991;Caravale, 1993; Strazzulla, 1990. As formuseum catalogues: Mielsch, 1971; Hedinger,1987; Perry, 1997). Nevertheless, studies onthis kind of production have been only rarelymatched by archaeometric analysis ofconstituent materials. On the other hand,archaeometric investigations on Campanaplaques refer only to a few pieces and aremarginal to more detailed studies on differenttypologies of architectural terracottas. Moreprecisely, archaeometric tests were carried outin the following four cases. The first one is that

of a single Campana plaque from themagnificent building of the Fondo Tuzet inAquileia (Strazzulla, 1982-1983, 1987a), whichhas been studied during a research concerningarchitectural terracottas of Roman Venetia.Then, two Campana plaques coming from theworkshop in Via Gallia in Rome, wereexamined in the context of a research onantefixes from the Antiqu.arium Comunale(Anselmino, 1977). In another case,fluorescence analysis was applied to a numberof amphora fragments and three Campanaplaques seemingly produced in the sameworkshop at Saint-Just (Laubenheimer et al.,1989). Finally, a test was carried out in order tospecify the production technique ofarchitectural terracottas from the Tarragonaterritory, which seems to have implied theapplication of archaeometric analysis to asingle Campana plaque coming from the region(Ramos Sainz et al., 1990; Vigil de la Villa etal., 1994), but the results of such analysis havenot been published.

In the case of Aquileia, test results provemeaningful, suggesting the hypothesis offinished products imported on the demand of ahighly prestigious client, presumably theImperial family. And there is an evidentcontrast with an apparently non-urbanprovenance of architectural terracottas fromRoman Venetia of a different type (and withthe hypothesis of a local production by urbancraftsmen of the same terracottas from theFondo Turet - as postulated by the author ofthe Corpus - as well).

In the case of the workshop in Via Gallia, thestudy of two pieces of Campana plaquesindicates a characteristic fabric of urbanproduction, as also shown by the analysis of aconsiderable number of urban antefixes.Compared to the Roman situation, outcomes ofthe analysis on a sample of an antefix from theSettefinestre villa, in the territory of Grosseto,together with macroscopic tests carried out onclay of all architectural terracottas from thevilla, meaningfully reveal that all of them arelocally produced, including Campana typeplaques (Celuzza 1985). Different origins of


materials employed seem to correspond toalready verified characteristics of the villa,which was owned by Roman elite members andentrepreneurs but, in this case, not by membersof the Imperial family. This clearly shows aproductive character of the complex from itsvery beginning - second quarter of 2nd cent.B.C. - first quarter of 1st cent. A.D. (Celuzza1985).

Indeed the Saint Just analyses, indicatesubstantial degree of homogenety between thedifferent classes of artifacts produced at the site(Campana plaques and amphoras).

Though very selective, the examples ofarchaeometric analyses that have just beendescribed, show how productive such aresearch can be. These cases suggest thenecessity of extending archaeometric studies tothe same general contexts, in order to provehypotheses already formulated, as well as tonew contexts and, eventually, to the class ofmaterials on the whole.

3. ARCHAEOLOGICAL HYPOTHESES

The iconographic type testified by fragmentsof Campana plaques found at Cortona,represents an Siren standing on top of anacanthus tuft turned upside down, from whichluxuriant flowered tendrils develop.

Another variant of the same pattern ispresent on known reliefs (von Rohden andWinnefeld, 1911) coming from anotherproperty of the Imperial family (the fact is notnegligible), the Horti Sallustiani in Rome (fig.2). Although fragmentary, the mentionedspecimens from Rome are still more completethan those from Ossaia.

Yet, it must be pointed out that twoconsiderably different versions of the sameiconographic type have been traced at Ossaia.Execution levels are qualitatively different, andthis immediately allows to think of twodifferent productions. The former (fig. 3a) ismarked by relief carried out in the style of the

Fig. 2 - Iconographic type of fragmentary Campana plaque from the Horti Sallustiani (after von Rhoden and Winnefled, 1911).

a)

b)

Campana plaques from Ossaia - La Tufa (Cortona, Arezzo) ... 151

Fig. 3 - Campana plaques from the villa in Ossaia - La Tufa. Graphic reconstruction of the two different versions, in high(a) and poorer style (b), of iconographic type. Testified parts of fragments are in black, graphic integrations are in grey.


best products belonging to the LateRepublican/Early-Imperial period, quick andsmooth, with delicate chiaroscuro. As forflowered tendrils, the modelling is instead thin,clearly outlined, from time to time heavilyengraved, so that it looks like embossed metal.Motifs are finely stylized, though maintainingnaturalistic quality. The latter (fig. 3b) shows apoorer style «reproduction» of the sameiconographic type. Modelling appears nowextremely stiff and dry, with figures outlined ina most awkward schematic way,«geornetricized» .

What we have just described, leads us tothink that stylistically superior examples hadbeen imported from the centre of power,presumably during the first living phase of thevilla, between Late-Republican period (when itwas built) and Augustan age (when it becameproperty of the Imperial family). Alsoextremely meaningful is the fact that thecrowning border, decorated with small archesand palmettes, recurring with similar varietieson Aufsatzplatten, can only be punctuallycompared with the «Nike killing the bull»plaque (Coarelli, 1981; Tortorella, 1981 a),coming from the podium (Augustan phase) ofthe temple B in the sacred area of LargoArgentina. The Nike relief was clearly reused

when Agrippa started the restoration ofbuildings in the area, and it could date to thereworkings of the time of Pompey or Caesar.Thus, mould of such crowning border (one ofthe few elements allowing us to distinguish theproduction of different workshops) is testifiedin Rome since that period, and it is applied to apiece which was commissioned by one of themagni viri of the Late-Republican age.

According to the conclusions just drawn, itcan be inferred that, in the case of the Ossaia La Tufa villa, low quality plaques wereproduced by local workshops, on the occasionof the mentioned restructuring of parts of thecomplex during the later first century A.D.,when a number of rooms were turned intoproductive areas.

The study of palmette antefixes, with frondsturning to the inside (Pensabene and Di Mino,1983), only two fragments of which werefound in the villa (fig. 4), goes in the samedirection. Diffusion of the basic decorativepattern is connected to its usage in thebuildings of the Augustan period, where pieceswith such palmettes developing from anacanthus tuft were employed. The type found atOssaia - La Tufa, though only testified by afew fragments, is comparable to few knownpieces, dated to the second half of 1st cent.

Fig. 4 -Graphic relief of antefixes from the villa at Ossaia La Tufa.

Campana plaques from Ossaia - La Tufa (Cortona, Arezzo) '" 153

B.C., and coming from areas that, in two casesat least, were related to restoration works madeby Pompey and Augustus (fig. 5). The contextof utilization adds to the significance of thecomparisons just drawn, in a way similar towhat we have already argued for the Campanaplaques. More precisely, the type couldcorrespond to that of antefixes traced within theAugustan complex of the Palatine (Strazzulla1990) and on the eastern slope of theCapitoline Hill (Pensabene and Di Mino,1983). It is noteworthy that Filippo Coarelli(1984) ascribes examples of Campana plaquescoming from the Capitoline Hill to theworkshop of Asinius Pollio, the sameworkshop that hypothetically produced thePalatine plaques as well. Otherwise, the type ofantefixes could be the same as that testified inthe sacred area of Largo Argentina (Anselmino,1977). Once again, the latest circumstance isworthy of particular attention: as we havealready explained, fragments of Campanaplaques with identical crowning border as thatof pieces from Cortona come from the samearea.

Of course, it is impossible to establish if alsoantefixes of higher quality were employed inthe villa of Ossaia-La Tufa (as for the plaques,the hypothesis is that they were imported fromRome). As a matter of fact, examples found areindeed comparable to plaques of the lowestquality, because of execution level andmacroscopic aspect of the impasto.

4. FORMULATION OF THE ARCHAEOMETRIC

RESEARCH

The hypothesis formulated according to thisfirst test of fragments can find confirmationthrough archaeometric study. First of all, thefact that the different execution levels seem tocorrespond, even at an absolutely superficialobservation, to two different types of claystrongly stimulates prosecution of such aresearch.

Study of materials coming from a specificsite begins with the creation of a typology for

whatever class of ceramics. This, in order toestablish connection between findspots and theclass on the whole, even when there is nointerest in a successive archaeometric test. Asfor Campana plaques tout-court, andconsequently pieces from Ossaia, a typologycan not be created. Better, there is no sense indrawing comparisons in conformity with amorphologic classification. Obviously,different functional types of architecturalterracottas - revetment plaques, piercedcresting plaques, Aufsatzplauen and simas -,according to different functions, don't at alltestify different productive groups. Onlyidentical crowning border and, logically,correspondence of moulds, allows recognitionof production of a workshop or connectionsbetween different ones. Extensively, theconcept of «typology» must not be applied tothe bearing, but to crowning borders,iconographic types and, eventually, to the styleof reliefs.

As we have already seen, two groups offragments of Campana plaques from Ossaia-LaTufa were traced on the basis of suchconsiderations.

When applying criteria normally used fortests of ceramic classes (Cuomo Di Caprio,1985; Olcese, 1993a, b; Olcese Hiener, 1993;Olcese, 1996; Olcese and Picon, 1995;Murialdo et al., 1998; Olcese, 2000) to suchclass of material - which would be betterassimilated to classes of products of lessrefined impasto, such as bricks and containers-, we can confirm that groups seem to differ asregards to the two productive phases. Theformer, called «supply phase», shows thatimpastos are different, while conclusionsemerging from the latter, called «conversionphase» (Giannichedda, 2000), allow us to pointout differences involving the making of themould, which was nevertheless used in bothcases. Anyway, while mould of one group wasvery accurately made, we can postulate that theother one was coarsely modelled, very likelyon the basis of an approximate casting of thebest examples, as shown by narrowing of themoulded relief.


(after Strazzalla 1990)

(after Anselmino 1977)

(after Pensabene and Di Mino 1983)

Fig. 5 - Urban reference models of antefixes of the palmettes type from Palatine, Largo Argentina and Capitol.


The first necessary step we have to take incase we want to determine the origin of thepieces, is that of comparing compositionalcharacteristics of materials we are testing withthose of a «reference group» (Olcese Hiener,1993).

In our case, we assumed that some of thereliefs were made in Rome, and this is the areawhere we must look for reference material. Asmain collections of Campana plaques weretemporarily inaccessible at the time ofsampling (those of the Roman NationalMuseum and those of the CapitolineMuseums), the one «reference group» we couldcreate is that of plaques from the PalatineMuseum. Actually, they represent a productionof a most certain origin, as they were attributedto the Augustan building called Portico of theDanaids and set in a precise historical context,between 36-26 BC (Carettoni, 1973).

G.R.

5. SAMPLING AND METHODS OF ANALYSIS

It seemed sufficient to examine two samplesof pieces of refined workmanship and twosamples of those remaining, because ofexiguous number and small dimension offragments from Cortona, which were allclassified at macroscopic level with the aid of alens. Samples of urban production were chosenaccording to the attempt of collecting arepresentative number of cases, «assaying»products made with impastos coloured withscarcely different nuances and bearing reliefsboth iconographically and stylisticallyheterogeneous, which cannot be surely ascribedto a particular cultural context, or at least to anidentical context of employment, though theybelong to the same architectural complex. Inthis respect, samples were chosen representingmytological-narrative motifs (samples P5, P7,P8 and PIO; tab. 1), well in accordance withAugustan suggestions, along withiconographies referring to Egyptian cult(samples P6 and P9; tab. I), not properlyadequate symbols for semi-public rooms at the

time of employment (the age immediately afterthe battle of Actium).

A list of the samples with inventory numbersand macroscopic characteristics is reported intable 1.

The samples, consisting of small chips ofmaterial, were observed first at the stereoscopicmicroscope and then cut and prepared to obtainpolished thin sections. The thin sections wereobserved using a polarizing microscope toattain textural characterization of the fabrics,and then analyzed by scanning electronmicroscopy coupled with EDS microanalysis(SEM-EDS) to characterize significant phasespresent within the fabrics. SEM-EDSmicroanalysis was performed at 15KV beamcurrent for a 60" count time.

G.R.;B.M.

6. MACROSCOPIC CHARACTERISTICS

Samples in different styles from Ossaia - LaTufa show very distinguishing macroscopicfeatures. In fact, the samples in high style (LTIand LT2) are pink, show irregular fracturingand are characterized by the presence ofmagmatic inclusions consisting of pyroclasticfragments, biotite and, in one case, pyroxene.The samples in low style (LT3 and LT4) have apeculiar reddish colour, show earthy fracturingand are characterized by the presence ofabundant silicoclastic inclusions consisting ofquartz, lithic fragments and, in one case,feldspar. They do not contain any kind ofmagmatic inclusion. The different mode offracturing in the two groups of samples couldbe the consequence of different amounts ofclay fraction within the pastes, namely tohigher amounts of clay in the earthy irregularlyfractured samples (LT3 and LT4) and loweramounts in the irregularly fractured samples(LTI and LT2).

The samples from Palatino Museum are allcharacterized by a noticeable presence ofmagmatic inclusions, mainly pyroclasticfragments but also pyroxene in one of thesamples. The pyroclastic fragments are often

TABLE 1

Summary ofmacroscopic characteristics of the samples.

Ut0\

* reported following the criteria by Olcese (l993a).§ defined by comparison with the Munsell Soil Color Chart.Abbreviations: Bi =biotite; Fld =feldspar; Px =pyroxene; Pyr =pyroclastic fragment; Qz = quartz; LF = lithoclastic fragment; OM = organicmatter traces

Label Inventory # Colour§ Fracture Inclusions Notes---

LT1 99048 7.5YR 7/4: pink irregular Bi, Pyr, PxLT2 97149 5YR 7/3: pink irregular Bi, Pyr Oriented pores;

laminated appearanceLT3 99061 5YR 5/8: yellowish redLT4 97247 5YR 5/8: yellowish redP5 380037 7.5YR 7/4: pinkP6 379625 7.5YR 7/4: pinkP7 379651 7.5YR 8/4: pinkP8 380077 IOYR 8/4: very pale brownP9 380014 5YR 7/3: pinkPIO 379629 7.5YR 8/4: pink

earthy irregularearthy irregular

flakyirregularirregular

rough irregularindentedirregular

Qz,LFQz, Fld, LF

Qz, Pyr, OMQz, Pyr, Px

PyrQz, Pyr, OM

PyrPyr, OM Laminated appearance

~oc:>tj

.~o:;0o~

~P'::l0-to

~o;;<lo~


millimetric or sub-millimetric in size. Quartz ispresent in three of the samples, with smallerdimensions with respect to the pyroclasticfragments. Three of the samples are alsocharacterized by the presence of blackishmoulds which have been attributed to organicmatter residues. The colour is pink in all thesamples but one, and the grain size of the paste,deducible by the mode of fracturing, isrelatively finer in sample P5, and relativelycoarser in samples P8 and P9 with respect tothe rest of the samples.

Two of the samples - one from Ossaia - LaTufa and one from Palatino Museum - arecharacterized by the presence of oriented poresand/or show a laminated appearance in thefractured surfaces. These characteristics can beboth related to the manufacturing process ofarchitectural terracotta, in which the claymatrix is stretched and then pressed within amould to obtain the decorative relief (CuomoDi Caprio, 1985).

7. PETROGRAPHY AND MINERAL CHEMISTRY

The petrofabric characteristics of the samplesare described in table 2, and showed in figure6. The matrix is pale brown to ochre andorange in the samples from Ossaia - La Tufa inhigh and low style, respectively, and palebrown to ochre in the samples from PalatinoMuseum. Total porosity is 3 and 5% in thesamples from Ossaia - La Tufa, and 2 and 5%in the samples from Palatino Museum, whereasthe mean size of the pores is about 300~m inthe formers, and 100 to 700~m in the latters.

The minerals and phases recognized at themicroscope are quartz, feldspar, clinopyroxene,phyllosilicates and pyroclastic fragments.Quartz is subhedral to anhedral, sub-spherical,angular to sub-angular in shape (afterKrumbein and Sloss, 1979). In the samples inlow style from Ossaia - La Tufa (samples LT3and LT4) it frequently forms poly crystallineaggregates (fig. 6c, d). The quartz abundance is2 and 7% in the samples in high style, 20 and25% in the samples poorer in style, and 1 to 3%

in the samples from Palatino Museum. The sizeof crystals is from 20 to 210~m in the samplesin high style, from 10 to 300~m in the samplesin poorer style, both from Ossaia - La Tufa,and from 10 to 400~m in the samples fromPalatino Museum.

Feldspar is mostly K-feldspar but a fewplagioclase is also present in some of thesamples from Palatino Museum. Unfortunately,it was not possible to define exactly thecomposition of all the feldspars due to thesmall size of some of the grains, typicallyranging from 10 to 60~m. The amount offeldspar is a bit higher in the samples fromOssaia - La Tufa than in those from PalatinoMuseum.

Clinopyroxenes were found only in samplesLT1 and P6. In sample LT1 the clinopyroxeneis loose within the matrix, whereas in sampleP6 it is embedded in a pyroclastic fragment(fig. 6a, f). The shape is subhedral andfragmented, the mean size is 350~m in thesample from Ossaia - La Tufa and 500~m inthe sample from Palatino Museum, thecomposition is salitic in both cases (tab. 3).

Phyllosilicates, present in all the sampleswith similar amounts, have been mostlyascribed to white mica and biotite, but in somecases the small size of the grains (ranging from20 to 100~m) and the vivid colour of matrixmade it difficult to discriminate and to analyzethese minerals. For this reason the term«phyllosilicate» was preferred to the term«mica» to describe the typology of minerals asa whole. A large biotite crystal embedded in apyroclastic fragment is present in one of thesamples from Palatino Museum (sample P7;fig. 6g). The composition of this mineral isreported in table 3.

Pyroclastic fragments are present in all thesamples except in those in low style fromOssaia - La Tufa (samples LT3 and LT4). Theamount of pyroclastic fragments is 2 and 3% inthe samples from Ossaia - La Tufa, 3 and 7% inthe samples from Palatino Museum, whereasthe size of the fragments is from 1OO~m to1.3mm in the samples from Ossaia - La Tufa,and from 100~m to 2.7mm in the samples from

VI00

TABLE 2

Summary of the petrography of the samples front Ossaia - La Tufa (LT) and Palatino Museum (P).

MATRIX GRAINS

Label Colour* p** Psize# Abund. So§ Sizerange Sizemean Pyr Pyrrange Qz QZrange Fld Fldrange Phyl Phylrange CPX CPXmean(%) (1-.1111) /Shape (/-Il11) (I-Im) (%) (I-Im) (%) (I-Im) (%) (I-Im) (%) (I-Im) (%) (I-Im)

I~of poresac:::

LTI pale brown 3 350 few; large MIL 30-350 150 2 100-400 7 40-210 2 30-50 <2 30-80 <1 350;>r-l

LT2 ochre 3 250 many; small M/H 20-1300 50 3 100-1300 2 20-70 <1 20-30 <3 40-60 mLT3 orange 5 300 large; irregular L 10-300 120 20 30-300 1 50-60 <2 20-60

,::::

0LT4 orange 5 300 large; rounded L 10-250 120 25 10-250 1 30-40 <2 20-50 :::0P5 ochre 5 700 rounded M/H 30-1100 50 5 170-1100 2 50-400 <1 30-40 <2 20-40 0

VJ

P6 ochre 3 250 closed; irregular M 30-2000 200 7 170-2700 2 30-150 <1 30-40 2 40-100 <1 500VJ

~P7 ochre 3 400 fissures H 10-1800 50 5 400-2000 1 20-70 <1 10-30 <2 40-60 ~

;:l

P8 pale brown 3 200 many; small M/H 10-350 50 3 100-350 2 10-90 <1 20-30 <2 20-40 0-t:.O

P9 pale brown 5 200 many; small M/H 10-2500 200 5 100-2500 3 20-150 <1 20-30 <2 20-40~

PIO ochre 2 100 small; closed M/H 20-650 50 5 200-650 2 20-50 <1 30-40 <2 20-40 0

""0* parallel nicol view

,3;

** total porosity~ mean pore size§ Sorting: L =low; M =medium; H =highData of total porosity and relative abundanceof grains have been obtained by visual estimate (Shvetsov diagrams) at 40x and IOOx magnification.Mean pore size has been estimated by direct measurementof the mean diameter of pores.Abbreviations:Pyr =pyroclastic fragments; Qz =quartz; Fld =feldspar; Phyl =phyllosilicates;CPX =clinopyroxene

Campana plaques from Ossaia - La 159

Fig. 6 - Petrofabric characteristics of the samples: LTI (a), LT2 (b), LT3 (c), LT4 (d), PI (e), P2 (f), P3 (g), P4 (h), P5 (i)and P6 (1). Optical photomicrographs, parallel nicols, 40x magnification.

C\o

TABLE 3

Representative chemical analyses ofpyroclastic fragments (pf), pyroxenes (px), spinels (sp) and biotite (bi) from Ossaia - La Tufa(LT) and Palatino Museum (P). I~

0

LT1-pfl LT1-pf2 LT1-px P6-pfl P6-pf2 P6-px P6-sp P7-pfl P7-pf2 P7-bi P9-pfl P9-pf2 P10-pfl P10-pf2 l~....,m

Si02 54.30 51.55 49.21 53.09 51.81 47.76 3.70 50.74 56.93 40.19 50.72 54.98 56.83 58.04 FTi02 0.62 1.04 1.09 12.06 1.05 0.63 3.10 1.50 0.26 0.42 0.39 0A120 3 17.64 18.23 5.49 21.49 14.72 5.60 2.90 14.95 22.02 20.05 18.04 20.39 21.90 3.71 i'CI

0

Cr203 0.32 0.43 ~

~FeOtot 5.17 10.37 6.21 10.91 4.77 6.44 79.06 10.77 4.25 9.28 10.80 4.39 4.44 5.29~

MnO 1.16 ::l0.-

MgO 3.37 3.45 12.72 2.25 2.21 13.12 4.31 3.55 19.18 6.91 3.60 3.39 13.23 ~

CaO 16.12 9.02 25.42 6.77 25.50 25.56 1.12 12.62 5.51 6.21 14.89 6.69 19.33 z0

Na20 1.83 1.88 3.31 1.61 4.09 2.31 0.37 4.50 ::00

K20 1.56 5.49 1.12 0.98 3.94 3.02 8.20 3.52 1.13 1.82 ~

'" Point analyses performed in different positions within the phases. Data representative of groups of analyses for each phase.

Campana plaques from Ossaia La Tufa (Cortona, Arezzo) ... 161

Palatino Museum. The pyroclastic fragmentsare generally rounded in shape andcharacterized by remarkable vesicular porositywhich is particularly evident in samples P5 andP6. Small spinel grains (tab. 3) have beenfound within one of the pyroclastic fragmentsin sample P6.

The chemical composition of pyroclasticfragments is mildly alkaline with KINaapproximating unity, two different KlCa ratiosand corresponding Fe amounts, highaluminium and low titanium (tab. 3). It iscomparable to that of rocks belonging to thePotassium Series (Peccerillo and Manetti,1985).

Comparison between the samples in high andlow style from Ossaia - La Tufa revealssignificant differences between them. In fact,the samples in low styIe do not containmagmatic phases, are very enriched in quartzand show lower sorting of grains. On the otherhand, samples from Palatino Museum show awide range of fabric characteristics as for thesize and amount of phases, resulting in mediumto high sorting of grains within the matrix.

8. DISCUSSION OF ARCHAEOMETRIC RESULTS

The results point to the existence of twodifferent kinds of paste in the terracotta fromOssaia - La Tufa, corresponding to differentdegrees of stylistic refinement. In fact, thesamples in high style (LTI and LT2) arecharacterized by the presence of a magmaticcomponent mainly consisting of pyroclasticfragments, whereas the samples in low style(L T3 and LT4) do not show any kind ofmagmatic phase and are characterized by agreat abundance of quartz. These observationspoint to different provenance of the materialsemployed in the preparation of the differentkinds of paste.

In the samples in high style, the chemicaland mineralogical characteristics of volcanicclasts and, particularly, the salitic compositionof clinopyroxene are consistent with thevolcanic units of the Roman Comagmatic

Province. The Roman Comagmatic Province isone of the best known occurrences of potassicalkaline volcanism. It consists of a series ofvolcanic centres located along the Tyrrhenianborder of the Apennines, from South Tuscany,through Latium to the Naples area, withprevailing pyroclastic products and subordinatelava flows (Peccerillo and Manetti, 1985). Theage of volcanic activity is 0.8 (VulsiniMountains) to 0.03Ma (till present; SommaVesuvius) (Barberi and Innocenti, 1967;Nicoletti, 1969; Fornaseri, 1985). Based onpetrological characteristics, two main serieshave been distinguished within the RomanComagmatic Province: a High-potassiumSeries (HKS) and a Potassium Series (KS)(Appleton, 1972). The KS include slightlyundersaturated to saturated rocks (alkalibasalts,trachybasalts, latites and trachytes) with lowerK20 and K20/Na20 ratios than the HKS rockswhich, on turn, consist of stronglyundersaturated rocks (leucitites, leucitetephrites, leucite tephritic phonolites andleucite phonolites).

The chemical composition of pyroclasticfragments, particularly the K20lNa20 ratio,and the lack of leucite in the samples in highstyle from Ossaia - La Tufa is in betteraccordance with KS than with HKS, therefore aprovenance from KS can be postulated. Thishypothesis leads to restrict the area ofprovenance of raw materials to the Vulsini,Sabatini, Ernici, Roccamonfina, PhlegreanFields and Somma-Vesuvius districts, whereKS rocks outcrop alone or together with HKS.Moreover, the exclusive presence ofpyroclastic over lava fragments, and thechemical composition of these fragments pointto a provenance from the Phlegrean Fieldswhere pyroclastics of composition compatiblewith that of the analyzed samples largelypredominate over lavas (Peccerillo, personalcommunication).

In the samples in low style from Ossaia - LaTufa, the characteristics of quartz are consistentwith those of sediments belonging to theMacigno formation. Macigno is a quartzfeldspar-micaceous Oligo-Miocene turbiditic


complex with silty marls and clays and limymarly, calcarenitic interbedded layers(Jacobacci et al., 1970). The size, sorting andabundance of quartz, and the remarkablepresence of quartz aggregates with cataclasticstructure have been considered peculiarfeatures of Macigno sediments distinguishingthem from Marnoso Arenacea sediments(Cipriani, 1961; Cipriani and Malesani, 1963a,b). In particular, the presence of quartzpolycrystalline aggregates with cataclasticstructure has been put into relation with adifferent origin of quartz - from metamorphicand magmatic rocks, respectively - in Macignoand Marnoso Arenacea formations (Ciprianiand Malesani, 1963b). The roundness andcloudiness of the larger quartz grains is anevidence of transport within an alluvial basin,whereas the subhedral and anhedral smallgrains probably derive from mechanicalfragmentation of the polycrystalline aggregatesduring transport. This points to a provenance ofraw material from Cortona surroundings, wheresandy sediments of eluvial origin in riverlacustrine facies widely outcrop in the plain.

Samples from Palatino Museum, thoughquite different one from each other, are all tobe ascribed to the same volcanic Districts asthe samples in high style from Ossaia - LaTufa. In fact, they are all characterized by thepresence of pyroclastic fragments, and, in oneof the fragments (sample P6), saliticclinopyroxene and magnetite grains are alsopresent. This is in good accordance with a localprovenance, from Urban Roman context,originally postulated for this reference group ofsamples.

The mutual differences among the samplesfrom Palatino Museum are difficult to beinterpreted. They may be the consequence ofdifferences in the intrinsic composition of rawmaterial, that is different provenances of rawmaterial, or the consequence of differentpreparation of the paste from raw material, thatis different workshops and/or modes ofpreparation within the same workshop. In thefirst case pyroclastic grains were alreadypresent within the raw material and did not

suffer severe purification at the workshop. Inthe second case pyroclastic grains wereintentionally added to the matrix to obtain thepastes. Detailed bulk chemical analysis ofpastes and raw materials would be necessary totry to solve this question, however, by analogywith other similar occurrences (Cuomo DiCaprio, 1985), the second case is the mostlikely hypothesis.

Comparison between samples in high stylefrom Ossaia - La Tufa and samples fromPalatino Museum reveals some analogies andsignificant similarities. In particular, sampleLT 1 shows petrofabric analogies (similarsorting of grains within the matrix) to sampleP6, and geochemical affinities (similarcomposition of pyroclastic fragments) tosample P9, whereas sample LT2 shows clearsimilarities in sorting, abundance and sizerange of grains to samples P8 and P9. If a largequantity of data representative of contemporaryUrban productions from different attestedworkshops is available, these observations willbe helpful for a correct assessment ofprovenance.

B.M.

9. CONCLUDING REMARKS

Archaeometric characterization of samples ofCampana plaques from Ossaia - La Tufa pointsto the existence of two different groups ofproduction, local and urban, of architecturalterracottas, corresponding to very differentstylistic levels. Local production, in 'low' style,is characterized by the sole presence of asedimentary fraction which has been attributedto local geological formations outcropping inthe Cortona surroundings. Urban production, in'high' style, is characterized by the presence ofa volcanic component which has beenattributed to pyroclastic potassium alkalinemagmatic rocks outcropping in Latium andCampania. According to Cuomo Di Caprio(1985), importation of raw material was a farless common practice than importation ofproducts in antiquity, therefore urban


production is to be intended as importedmaterial coming from the centre of power. Inthe light of these points, the results areinterpreted as follows.

In the period in which the villa in Ossaia La Tufa probably passed into imperialproperty, some products of good workmanshipwere imported from Rome to Ossaia. Theseproducts are similar to the products from thePalatine manufactured on commission byAugustus, and show the same icononographictype present at Horti Sallustiani, anotherimperial property previously belonging toCaesar and, then, to Sallustius. It is possiblethat the restricted entourage gravitating aroundthe centre of power imported into the newproperty high craftmanship products with moreelaborate figurative projects, the same projectswhich were already present in public andprivate buildings in Rome belonging to thesame authority, for propaganda and selfrepresentation purposes. Similar conclusionswere obtained by Strazzulla (1987a) from theanalysis of a single terracotta sample from avilla in Aquileia, probably one of theresidences of emperor Tiberius. In a subsequentperiod of decline occurring in the Flavian age,local products of inferior quality appeared inthe decorative reliefs of the villa.

The presence of terracotta reliefs of verydifferent stylistical quality, and similararchaeological interpretation, has beendocumented in other occurrences such as thevilla of Voconius Pollio in Marino (Rizzo,1976-77). In this case a large quantity of finequality terracotta sherds, along with a smallquantity of bad quality pieces reproducing thesame iconographical elements in declining and'poorer' style, were found. The former wereinterpreted as products of urban manufacture,whereas the latter were considered the resultsof local reproduction. Considering thehistorical significance of these findings, this isof particular importance for an extension ofresearch to the whole class of architecturalterracottas.

The preliminary results of this investigationare so promising as to encourage future

development of research with the acquisition ofnew analytical data on a larger number ofsamples from Ossaia and from Romerepresentative of different typologies andvarieties within the same typological class.Petrographic investigation is to be implementedby quantitative fabric analysis of textures afterdigital treatment of thin section images. On theother hand, bulk quantitative chemical andmineralogical data by X-ray fluorescence andX-ray diffractometry are necessary to clarifythe relationships between the products and thesource areas of raw materials. This kind ofinvestigation is now in progress on a newcollection of twenty samples of CampanaPlaques from the Roman National Museum,now in deposit in the caveau of PalazzoMassimo in Rome. Moreover, sampling inOssaia is going to be extended to kiln discardsfound within the perimeter of the ancientproperty (Fracchia and Gualtieri, in press),whereas geological investigation of thepossible sites of provenance of raw materialshas already started.

B.M.;G.R.

ACKNOWLEDGMENTS

The excavations at Ossaia/Cortona wereundertaken in 1992 thanks to the support of theComune di Cortona and the SoprintendenzaArcheo1ogica della Toscana. We wish to thank Prof.M. Torelli, the Director of the Project for the ParcoArcheo1ogico at Cortona, for much helpful advice,and Prof. H. Fracchia, the Coordinator of theCanadian team at Ossaia-La Tufa, for informationon unpublished material. Dr. A. Bottini,Superintendent and Dr. P. Zamarchi Grassi,Inspector for the Cortona area, have helped theproject in many aspects and are to be thanked forgranting permission to take the terracotta samples.Archaeometric research on the Campana plaques ispart of a wider research project undertaken in thecontext of the Ossaia excavations and aimed at thestudy of pottery , amphora and brick production atthe site, on which much advice is owed to Dr. lW.Hayes (Oxford) and M. McCallum (State Universityof New York at Buffalo). The authors wish to thankDr. G. Olcese for useful suggestions/ anddiscussions, and Prof. G. Poli and S. Tortorella forcritical review of the manuscript.


Research was supported by C.N.R. (Comitato perla Scienza e la Tecnologia dei Beni Culturali,Rome), the Ministero per I'Universita e la RicercaScientifica (Rome), the University of Perugia andthe Social Sciences and Humanities ResearchCouncil of Canada (Ottawa).

REFERENCES

ANSELMINO L. (1977) - Terrecotte Architettonichedell 'Antiquarium comunale di Roma. -1- Antefisse.L' Antiquarium, Roma, 144 pp.

ANSELMINO L. (1981) - Le antefisse fitiili. In:Societa romana e produzione schiavistica. Vol. IJ.Merci, mercati e scambi nel Mediterraneo , A.Giardina and A. Schiavone (eds.), Laterza, Bari,209-218.

ApPLETON J.D. (1972) - Petrogenesis ofpotassiumrich lavas from the Roccamonfina Volcano,Roman Region, Italy. J. Petrol., 13,425-456.

BARBERI F. and INNOCENTI F. (1980) - VolcanismeNeogene et Quaternaire. 26 Congo Geol, Inter.,Paris, Guide al'excursion 122A, 99-104.

BORBEIN A.H. (1968) - Campanarelie]s,Typologische und stilkritische untersuchungen.Mitteilungen des Deutschen ArchaologischenInstituts (Rom. Abt.), Erg.-Heft 14, 216 pp.

CALDERONE A. (1975) - Sulle terracotte«Campana». Bollettino d' Arte del Ministero per iBeni Culturali e Ambientali, 60, 65-74.

CARAVALE A. (1993) - Lastre Campana di tipoarcaistico dallo scavo della Meta Su d ans .Bullettino della Commissione ArcheologicaComunale in Roma, 95, 71-82.

CARETTONI G. (1973) - Nuova serie di grandilastre [ittili «Campana». Bollettino d' Arte delMinistero per i Beni Culturali e Ambientali, 58,75-87.

CELUZZA M.G. (1985) - Tecnica e tipologia deirivestimenti [ittili. In: Settefinestre. Una villaschiavistica nell'Etruria romana. Vol. I. La villanelle sue p arti., A. Carandini (ed.), Panini,Modena, 91-99.

CIFFARELLI F.M. (1988) - Ventotene: lastre«Campana» dalla villa di Punta Eolo. DocumentaAlbana, 10, 11-16.

CIPRIANI C. (1961) - Ricerche sulle arenarie: III)la composizione mineralogica di una serie dirocce della formazione del macigno. Per. Mineral.,30,23-59.

CIPRIANI C. and MALESANI P.G. (1963a) - Ricerchesulle arenarie: VII) La composizione mineralogicadi una serie di rocce della formazione marnosoarenacea. Per. Mineral., 32, 303-342.

CIPRIANI C. and MALESANI P.G. (1963b) - Ricerche

sulle arenarie: VIII) Determinazioni microscopichesulle arenarie delle [ormazioni del macigno emarnoso-arenacea. Per. Mineral., 32, 343-385.

COARELLI F. (1981) - Topografia e sto ria. In«L'area sacra di Largo Argentina. 1», Comune diRoma, Roma, 9-49.

COARELLI F. (1984) - Roma sepolta. A. Curcio,Roma, 191 pp.

CUOMO DI CAPRIO N. (1985) - La ceramica inarcheologia: antiche tecniche di lavorazione emoderni metodi d'Ln.dagine . «L'Erma» diBretschneider, Roma, 365 pp.

DI MINO M.R. (1981) - Terrecotte architettonichedalla zona del monumento a Vittorio Emanuele. In«Archeo1ogia Laziale 4», Quaderni del Centro diStudio per l' Archeologia etrusco-italica 5, CNR,Roma, 119-125.

DUPRE X. and REVILLA V. (1991) - LastrasCampana en Tarraco (Hispania Citerior) y suterrito rio, Mitteilungen des DeutschenArchaologischen Instituts (Abt. Madrid), 32, 117140.

FORNASERI M. (1985) - Geochronology of volcanicrocks from Latium (Italy). Rend. Soc. It. Mineral.Petrol., 40, 73-106.

FRACCHIA H. and GUALTIERI M. (1996) - Theimperial villa at Ossaia (Arezzo, Italy).Preliminary data on the territory of RomanCortona. Echos du Monde Classique/ClassicalViews 15, 157-200.

FRACCHIA H. and GUALTIERI M. (2001) - Primi datisul territorio di Cortona in eta romana. In: Diecianni di archeologia a Cortona, C. Masseria (ed.),Giorgio Bretschneider, Roma, 183-215,216.

FRACCHIA H. and GUALTIERI M. (in press) - La villaimperiale di Ossaia ed il territorio cortonese ineta romana (Rome).

GIANNICHEDDA E. (2000) - S. V. Pro du.zione(archeologia della). In: Dizionario di Archeologia,R. Francovich and D. Manacorda (eds.), Laterza,Roma-Bari,231-236.

GUALTIERI M. (2000) - Figlinae, domi nobiles edapprovvigionamento di laterizi nell'Italia Centromeridionale: due casi di studio. In: La Briqueantique et medievale: production etcommercialisation d'un materiau, P. Boucheron,H. Brois and Y. Thebert (eds.), Collection del'Ecole Francaise de Rome, 272, 329-340.

GUALTIERI M. (2001) - La villa imperiale diOssaia/Cortona (AR): mosaici di eta tardorep ubblicana ed aug uste a, Atti dell' VIIIColloquio AISCOM (Firenze, febbraio 2001),Ravenna, 295-308.

HEDINGER B. (1987) - Die Campana-Reliefs del'Arch.dolog isch.en Sammlung del' UniversitiitZurich. Antike Kunst 30, 70-88.


JACOBACCI A., BERGOMI C., CENTAMORE E.,MALATESTA A., MALFERRARI N., MARTELLI G.,PANNUZZI L. and ZATTINI N. (1970) - Noteillustrative della Carta Geologica d'Italia: Foglio122, Perugia. Servo Geol. It., Roma.

KOLODNICKI D. (2001) - Terra Sigillata potteryfrom Ossaia-La Tufa. An Archaeometric analysis,Unpublished M.A. Thesis, University of Alberta,Edmonton.

KRUMBEIN W.C. and SLOSS L.L. (1979) Stratigraphy and sedimentation. Freeman & C.,Amsterdam, 660 pp.

LAUBENHEIMER F., ODIOT T. and LECLERE H. (1989)- Sous Auguste, un atelier de potiers italianisanta Saint Just (Ardeche). In: Melanges P. LevequeII, Les Belles Lettres, Paris, 295-329.

MANCA DI MORES G. (1982-1983) - Terrecottearchitettoniche e problemi topograjici: contributiall'identificatione del Tempio di Quirino sui ColleQuirinale . Annali della Facolta di Lettere eFilosofia, Univ. Perugia, Sez. Studi Classici, 20,323-360.

MIELSCH H. (1971) - Romische Archirekturterrakotten und Wandmalereien im AkademischenKunstmuseum Bonn. Mann, Berlin, 48 pp.

MURIALDO G., OLCESE G., PALAZZI P. and PARODI L.(1998) - La ceramica comune in Liguria nel VI eVII secolo. In: Ceramica in Italia: VI-VII sec.: Attidel convegno in onore di John W. Hayes. Roma,11-13 maggio 1995, L. Sagui (ed.), All'Insegnadel Giglio, Firenze, 227-251.

NICOLETTI M. (1969) - Datazione argon potassiodi alcune vulcaniti delle regioni vulcaniche ciminae vicana. Per. Mineral., 38, 1-20.

OLCESE G. (1993a) - Archeologia e archeometriadei lateri zi bollati urbani: primi risultati eprospettive di ricerca. In: The inscribed economy:production and distribution in the Roman empirein the light of instrumentum domesticum. Theproceedings of a conference held at the AmericanAcademy in Rome on 10-11 January, 1992, W. V.Harris (ed.), University oj Michigan, Ann Arbor,121-128.

OLCESE G. (1993b) - Le ceramiche comuni diAlbintimilium. Indagine archeologica earcheometrica. Quad. del Dip. di Archeologia eStoria delle Arti dell'Universita di Siena, SezioneArcheol., 35, All'Insegna del Giglio, Firenze, 369 pp.

OLCESE G. (1996) - Ceramiche comuni di originetirrenica centro meridionale tra il II secolo a.C. eil I d.C: problemi aperti. L'evidenza dei reperti diAlbiruimilium. In: Les cerarniques communes deCampanie et de Narbonnaise (I« s. avo J.-C.-IIes.ap.J.-C.). La vaiselle de cuisine et de table. Actesdes Journees d'etude (Naples 1994), ColI. CJB, 14,421-445.

OLCESE G. (2000) - S. V. Archeometria. In«Dizionario di Archeologia», R. Francovich andD. Manacorda (eds.), Laterza, Roma-Bari, 24-29.

OLCESE HIENER G. (1993) - Il contributo delleanalisi di laboratorio allo studio e aliaclassificazione della ceramica in archeologia. In«Arch.eometria della ceramica. Problemi dimetodo. Atti 8 Simcer, Rimini 10-12 Novembre1992», Ed. Int. Centro Ceramico, Bologna, 35-53.

OLCESE G. and PICON M. (1995) - Ceramica inarcheologia e archeometria: qualche riflessionemetodologica sulle determinazioni di origine.Archeologia Medievale, 22, 429-432.

PECCERILLO A. and MANETTI P. (1985) - Thepotassium alkaline volcanism of Central-SouthernItaly: a review of the data relevant to petrogenesisand geodynamic significance. Trans. Geol. Soc. S.Afr., 88, 379-394.

PENSABENE P. and SANZI DI MINO M.R. (1983) Museo Narionale Romano. Le Terrecotte, III, 1 Antefisse. De Luca Editore, Roma, 375 p.

PERRY R. (1997) Die Campanareliejs. Philippvon Zabern, Mainz am Rhein, 71 pp.

RAMos SAINZ M.L., VIGIL DELA VILLA R. andGARCIA JIMENEZ R. (1990) - Empleo de latecnica de diffraccion por rayos X, en el estudio deterracotas arquitectonicas romanas de Ampurias yTarraco. Cuadernos de Preistoria y Arcqueologia.Universidad Autonoma de Madrid, 121-136.

RIZZO M. A. (1976-1977) - Su alcuni nuclei dilastre «Campana» di provenienza nota. Rivistadell'Istituto Nazionale di Archeologia e Storiadell' Arte, 23-24, 5-93.

STRAZZULLA M.J. (1982-1983) - Sistemi decorativiprivati in eta augustea: una villa Imperiale adAquileia? Annali della Facolta di Lettere eFilosofia, Univ. Perugia, Sez. Studi Classici, 20,463-487.

STRAZZULLA M.J. (1987a) - Aquileia e Roma:botteghe urbane e botteghe locali nella produzionedi terrecotte architettoniche. In «Aquileia e Roma,Atti della XVII settimana di Studi Aquileiesi, 24-29Aprile 1986», Antichita Altoadriatiche, 30,151-164.

STRAZZULLA M.l. (1987b) - Le terrecottearchitettoniche della Venetia Romana. Contributoallo studio della produrione jittile nella Cisalpina(II a.C.-II d.C.). «L'Erma» di Bretschneider,Roma, 462 pp.

STRAZZULLA M.J. (1990) - Il principato di Apollo.Mito e propaganda nelle lastre «Campana» daltempio di Apollo Palatino. «L'Erma» diBretschneider, Roma, 147 pp.

STRAZZULLA RUSCONI M.J. (1984) - Terrecottearchitettoniche del Museo Civico di Bassano delGrappa. Archeologia Veneta, 7, 167-190.

TORELLI M. (1983) - Edilizia pubblica in Italia


centrale tra guerra socia Ie ed eta augustea:ideologia e classi sociali. In «Les "Bourgeoisies"municipales italiennes aux /Ie et Ier siecles avoJ.-c. Centre Jean Berard. Institut francais deNaples 7-10 december 1981», Naples, 241-250.

TORTORELLA S. (1981 a) - Le lastre Campana.Problemi di produzione e di iconografia. In «L'artdecoratif a Rome a la fin de la Republique et audebut du principat. Table ronde organisee parI' Ecole francoise de Rome (Rome, 10-11 mai1979)>>, ColI. de l'Ecole franc. de Rome, 55, 61-100.

TORTORELLA S. (1981b) - Le lastre Campana. In«Societa romana e produzione schiavistica. Vol.II. Merci, mercati e scambi nel Mediterraneo», A.Giardina and A. Schiavone (eds), Laterza, Bari,219-235.

VIGIL DE LA VILLA R., GARCIA R., CALA V. andRAMOS M.L. (1994) - Estudio mineralogico yquimico de terracotas arquitectonicas de epocaromana de la Tarraconense. In La ciudad en elmundo romano. Actas XIV Congreso internacionalde arqueologia cldsica, Tarragona 5-11/9/1993,Comite Organitzador del XIV c.I.A.c., Tarragona,430-431.

VON ROHDEN H. and WINNEFELD H. (1911) Architektonische Riimis ch.e Tonreliefs del'Kaiserzeit. W. Spemann, Berlin-Stuttgart, 318pp.

ZAMARCHI GRASSI P. (2001) - Arezzo antica.Prolegomena per uno studio della citta e del suoagro. In: Etruschi nel tempo, Catalogo dellaMostra (Arezzo 2001), Firenze, 37-54.

Documents

Campana plaques from Ossaia - La Tufa (Cortona, Arezzo): from