Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://hbar.phys.msu.ru/gorm/dating/vesuvius_ama.pdf Äàòà èçìåíåíèÿ: Tue Jul 25 00:00:00 2006 Äàòà èíäåêñèðîâàíèÿ: Thu Feb 27 20:53:30 2014 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: prominence

Bull Volcanol (2004) 66:703­724 DOI 10.1007/s00445-004-0348-8

ORI G I NAL PA PER

Claudia Principe · Jean Claude Tanguy · Simone Arrighi · Anna Paiotti · Maxime Le Goff · Ugo Zoppi

Chronology of Vesuvius activity from A.D. 79 to 1631 based on archeomagnetism of lavas and historical sources
Received: 1 September 2002 / Accepted: 29 January 2004 / Published online: 25 May 2004 Springer-Verlag 2004

Abstract The activity of Vesuvius between A.D. 79 and 1631 has been investigated by means of precise archaeomagnetic dating of primary volcanic deposits and taking into account the stratigraphy of lavas and tephra, historical written accounts, archaeological evidence related to the developing urbanisation, and radiocarbon ages. We found that the historical records are highly useful in constraining the timing of the main events, even if the data are often too scarce and imprecise for ascertaining the details of all phases of activity, especially their magnitude and emplacement of all the deposit types. In addition, some eruptions that took place in the 9th and 10th centuries appear to be unnoticed by historians. The archaeomagnetic study involved 26 sites of different lavas and 2 pyroclastic deposits. It shows that within the 15 centuries which elapsed between A.D. 79 and 1631, the effusive activity of Vesuvius clustered in the relatively short period of time between A.D. 787 and 1139 and was followed by a 5-century-long repose period. During this time Vesuvius prepared itself for the violent explosive eruption of 1631. The huge lavas shaping the morphology of the coast occurred largely through parasitic vents located outside the Mount Somma caldera. One of these parasitic vents is located at low elevation, very close to
Editorial handling: J. McPhie C. Principe ()) · S. Arrighi Istituto di Geoscienze e Georisorse (IGG), CNR, Via G. Moruzzi 1, 56124 Pisa, Italy e-mail: principe@igg.cnr.it Tel.: +39-0503-152335 Fax: +39-050-3152360 J. C. Tanguy · M. L. Goff University of Paris 6 and Institut de Physique du Globe de Paris, 94107 St. Maur des Fosses Cedex, France A. Paiotti Via Pietro Nenni 25, 55042 Forte dei Marmi, Italy U. Zoppi ANSTO ­ environment, Private Mail Bag 1, 2233 Menai, NSW, Australia

the densely inhabited town of Torre Annunziata. Among the various investigated lavas, a number of which were previously attributed to the 1631 eruption, none is actually younger than the 12th century. Therefore it is definitively concluded that the destructive 1631 event was exclusively explosive. Keywords: Archaeomagnetism · Historical records · Vesuvius · A.D. 79­1631 · Lavas

Introduction
Vesuvius' activity from A.D. 79 to 1631 involved both late Roman activity, which occurred from A.D. 79 to 472, and medieval activity (Rolandi et al., 1998), which took place from A.D. 472 to 1631. This terminology is based on the rough temporal coincidence between the fall of the Roman Empire and the beginning of Middle Age (A.D. 476), which lasted until 1492. During this period, two small-scale plinian eruptions occurred, in A.D. 472 and 1631, respectively (Rosi & Santacroce, 1983; Rosi et al., 1993). In fact, there is little information regarding the volcanic activity of the Vesuvius between the plinian eruption of A.D. 79 and the nearly continuous activity that occurred after the small-scale 1631 plinian eruption (ArnÐ et al., 1987). This continuous activity was made up of lava effusions (Fig. 1), together with strombolian, violent strombolian, and subplinian episodes (Arrighi et al., 2001). Little data exist in the literature on the time of deposition and distribution of the lavas discharged during the 15-century-long period from A.D. 79 to 1631. Moreover, determination of the stratigraphic position of the so-called "medieval" lavas within the overall stratigraphy of the Vesuvius has been hindered by urban development along the Vesuvius' coast. In spite of this unclear stratigraphic position, many authors, even in recent literature, attributed many of these large lavas that shape the coast morphology to the 1631 eruption (Di Girolamo, 1970; Burri & Di Girolamo, 1975; Rolandi & Russo, 1989; Rolandi et al., 1993; Lirer et al., 2001). This


704 Fig. 1 Sampling map (topography from IGM-DTM data, courtesy of G.Vilardo, INGVVesuvian Observatory, Naples). Roman numerals: sampling sectors as discussed in the text. White dots: sampling sites (see Table 3 for the number of samples at each site). The distribution of lavas erupted between 1631and 1944 (simplified from the CNR, 1986, geological map), and the pyroclastic flow deposits of the 1631 eruption (from Rosi et al. 1993) are also shown

idea of large 1631 lavas originates from a relatively recent reconstruction of the 1631 eruption (Le Hon, 1865,1866), that has been uncritically adopted by almost all later authors. According to Rosi et al. (1993), the 1631 eruption did not produce any lava flow. This matter has important implications for determining the degree of volcanic risk in the Vesuvius area, because the 1631 eruption has been taken as representative of the maximum expected hazards posed by the volcano in the mid-to-long term (Barberi et al., 1995). In order to assess the potential hazard of the Vesuvius, it is also necessary to ascertain whether eruptions were uniformly distributed in time or clustered in climaxes with very high eruption rates, alternating with periods of much less intense activity (Heiken, 1999). Such information exists for the relatively short period from 1631 to 1944 (e.g. Carta et al., 1981 and Arrighi et al., 2001), but is lacking for the much longer time interval from A.D. 79 to 1631 . The aim of this paper is to fill the gap mainly through critical reading of historical records and archaeomagnetic dating of the "medieval lavas".

Historical and geochronological data
Historical sources It is commonly believed that the information provided by the historical sources on the Vesuvius' activity from A.D. 79 to 1631 is scarce and fragmentary. Such a belief arose in comparison with the large amount of documentation published after the 1631 eruption (Cerbai & Principe,

1996). However, the written sources available on the Vesuvius' activity between A.D. 79 and 1631 provide useful records of the main events and at least some details on the types of volcanic activity. All the post-1631 accounts, dealing with Vesuvius' history or its "latest" eruption, contain more or less complete lists of the volcano's past eruptions, as well as a number of references to printed sources that deal with the Vesuvius as far back as Greek and Roman times. Perhaps the most complete and reliable of these lists is presented in the milestone work "Bibliography of Vesuvius" by Furchheim (1897). An accurate critical collection of the written sources on Vesuvius' activity from A.D. 79 to 1631 was given by Alfano (1924). Other authors have subsequently dealt with the chronology of the Vesuvius' eruptions between A.D. 79 and 1631 (Preusse, 1931; Figliuolo & Marturano, 1998; Nazzaro, 1998, Gasparini & Musella, 1991) without substantial changes to Alfano's conclusions. Recently, Stothers & Rampino (1983) considered not only written and associated archaeological sources on Vesuvius', but also on all the volcanic eruptions that occurred throughout the Mediterranean area before A.D. 630. The investigated printed sources are rather different from those usually quoted by people working on Vesuvius and include not only Latin and Greek texts, but Byzantine, Syrian, and Armenian sources, as well. It is important to note that all these studies have been performed based on the more common printed sources. Detailed work on manuscript and ecclesiastic documents found in archives and libraries is still required in order to


705 Table 1 The activity of Vesuvius between A.D. 79 and 1631 as inferred by historical sources. Citations refer to the printed sources. "Activity" refers reported weak, more or less continuous stromEruptions AD 203 222­235 379­395 AD 472 512 AD 536 AD 685 AD 787 AD 968 AD 991 AD 999 1006 or 1007 1037 1139 1150 1306? 1500?
1 2

bolian activity, with possible emission of intracaldera lava flows. See Principe (2003) for a detailed discussion and the original texts

Activity 172

Described Phenomena Violent strombolian activity1 Explosive eruption, with a plinian fallout phase2 More or less continuous strombolian activity3 At least fumarolic activity4. A phase of lava effusion possibly happened5 5­6 November: large explosive eruption, with a plinian fallout phase, dispersed to NE (ashes as far as Costantinopoli), darkening and heavy rains . Huge nuüe ardentes and severe destruction are described all around Vesuvius'6. From this eruption starts the cult of San Gennaro as protector of Naples against Vesuvius' More or less continuous strombolian activity7 Explosive eruption, darkening in Mesopotamia lasted all the winter of this year.8 February-March: earthquakes, plinian column and related fallout9. A not completely reliable source gives an account of lavas entering the sea10 Autumn-Winter: lava fountains and lava (or pyroclastic?) flows "six miles long" 11 Eruption with lavas into the sea12 Earthquake and explosive eruption ("...erupting flames and vomiting ashes ...") 13 Sudden lava fountains and lava emission ("...piceas atque sulphurea repente flammas erumpere..."), fine ashes 14 At least explosive eruption, with rumble, vigorous throwing of blocks (3 miles from the crater) and unbreathable gas emissions 15 27 January or February: Effusive eruption with lavas into the sea ("...eructavit incendium, ita ut usque ad mare discurreret ...") 16 From 29 May to 5 June: explosive eruption, with black lapilli fallout and reddish ashes ("... pulverem ferruginei coloris et subrubei ..."), falling during 30 days as far as Napoli, Capua, Salerno and Benevento17 Violent strombolian activity18 Unlikely parasitic eruption with lavas into the sea19 Fumarolic activity and unlikely phreatic explosion20

Claudio Galeno Pergameno, De Medendi Methodo Libri XIII, liber V, cap.12 Dione Cassio Niceo, Historiae Romanae, liber LXXVI. 3 Dione Cassio Niceo, Historiae Romanae, liber LXVI. 4 San Paciano, Paraenesis ad poenitentiam. 5 Ammiano Marcellino, Indictione decima. 6 Marcellino (Conte), Chronicon ; Procopio from Cesarea, De Bello Gotico, liber II, cap IV ; Carlo Sigonio, Historiarum de Occidentali Imperio Libri XX, liber XIV & XVI; Flavio Magno Aurelio Cassiodoro, Opera Omnia quae extant, Liber IV Variarum, Epistola 50. 7 Marcellino (Conte), Chronicon. 8 Procopio from Cesarea, De Bello Gotico, liber II (cap IV), liber IV (cap XXXV); Giovanni from Efeso, Ecclesiastical History (other references are listed in Stothers and Rampino, 1983). 9 Paolo Diacono, Historia Longobardorum, liber VI, cap. IX; Giorgio Agricola, De re metallica. De natura eorum quae effluunt ex terra, liber IV. 10 Appendix to the holy legend of San Gennaro (in Alfano, 1924). 11 Gregorio Monaco, manuscript in greek, University of Messina library. 12 Damianus Petrus, Opera Omnia, Opusculum XIX. 13 Pompeo Sarnelli, Cronologia dei Vescovi et Arcivescovi Sipontini, 1730, pg.112. 14 Damianus Petrus, Opera Omnia, Opusculum XIX ; Ubaldo Monaco, Chronicum Ducum Neapolis, in Raccolta di varie croniche, diarii ecc, appartenenti al Regno di Napoli, Napoli, Perger, 1781. 15 Ridolfo Glaber, Historia Francorum. 16 Anonymus Cassinensis, Chronicon; Anonymus Cavensis, Chronicon; Romoaldo Salernitano, Chronicon. 17 Anonymus Cassinensis, Chronicon; Anonymus Cavensis, Chronicon; Falcone Beneventano, Chronicon; Romualdo Salernitano Chronicon. 18 El Eldrisi, L'italia descritta nel libro del Re Ruggero, quoted in Baratta, Il Vesuvio e le sue eruzioni, Roma, Dante Alighieri, 1897. 19 Pighius Stephanus Vinandus, Hercules Prodicius seu Principiis juventutis vita et peregrinatio, Antuerpiae, Cristoforo Plautinus, 1587, pg 466; Andrea Schottus, Itinerarium Italiae, Amstelodami, Jansson, 1655, pg 526 (into the Italian edition, dated 1669, the name of the Author became Francesco Scoto). 20 Leone Ambrosio, De Nola, Venetiis, 1514, liber I, cap.I: De Agro Nolano, deque montibus Vesuvio et Abella.

produce a fully reliable picture of the activity of the volcano during this long historical period. Table 1 (Principe, 2003) provides a summary of Vesuvius' activity between A.D. 79 and 1631 deduced from critical reading of the original sources proposed by previous authors. The number of confirmed eruptions and periods of activity is much smaller than that reported in the post-1631 literature on the Vesuvius, as is the number of the truly primary sources, meaning contemporary (or nearly so) to the described events. Such a misleadingly large number of sources and eruptions is due, as already

noted by Stothers & Rampino (1983), to the proliferation of vicarious "second-hand" citations. This common practice gave rise, for instance, to various reports of the "great eruption of A.D. 512", which was hypothesized by Alfano (1924) on the basis of a letter from Magno Aurelio Cassiodoro to King Teodorico of Italy and uncritically accepted in all the subsequent literature. However, a careful reading of Cassiodoro's letter reveals that it was simply a request for fiscal exemptions for a population still suffering the effects of a previous devastating eruption in A.D. 472.


706

No lavas outside the Mount Somma caldera are recorded by historical sources during the first centuries after the A.D. 79 eruption. On the contrary, Vesuvius' activity only involved a series of explosive events of variable violence until at least A.D. 685. The A.D. 472 eruption was by far the most important of these events. Plinian fallout and huge nuües ardentes can be clearly identified in the historical accounts of this eruption (Rosi & Santacroce, 1983). An explosive eruption must have occurred around A.D. 536. This has been indirectly documented in the writings of Procopio from Cesarea, Giovanni from Efeso, and other authors in the western Mediterranean, who reported the darkening of the sky over Constantinople for many months, accompanied by an unusually harsh winter in Mesopotamia (Stothers & Rampino, 1983). The eruption of A.D. 685 surely had a strong explosive component as well. Analysis of the historical sources does not clarify whether this eruption also involved lavas as popular tradition implies the lava underlying the old Palazzo Baronale of Torre del Greco originates from this event, and does not crop out anymore at present (Principe, 2003). A number of the subsequent eruptions (e.g. A.D. 787, 968, 999, and 1037) had large effusive components. The A.D. 991 event was apparently exclusively explosive, as well as those in 1006­7 and 1139 for which no effusive episodes are described. However, the last two eruptions were documented by authors living outside the Vesuvius' area. For this reason, in the absence of great damages to inhabited areas, the emission of lavas during these eruptions remains possible (Principe, 2003). Although the summit of the volcano is described by Dione Cassio as a rounded depression "like an amphitheatre", without mention of any internal cone (Alfano, 1924) during the third century A.D., historical data suggest the gradual building of the Gran cono inside the Mount Somma caldera, starting from ~A.D. 172, due to intracaldera lava emissions and strombolian activity (Table 1). After the 1139 eruption, the Gran cono was certainly present inside the Mount Somma caldera, with an altitude lower than that of Mount Somma, and a large crater "resulting from the last eruption" (Boccaccio, 1353, in Alfano, 1924). This is the same general morphology of the volcano after the main explosive eruptions of the 1631­1944 period (Nazzaro, 1998; Arrighi et al., 2001). After the 1139 eruption, the next description of the Vesuvius in an active state, characterised by "emission of incandescent materials from the main crater" (strombolian activity), seems to be presented in the text of the Arab writer El Eldrisi, dated 1150 and quoted by Alfano (1924), who refers to Baratta (1897). The numerous sources subsequent to 1150 listed by Alfano (1924) only report that a mountain near Naples had occasionally erupted in the past, without providing further observations on eruptive activities coeval to the time of writing. There are, instead, some references to flames in the crater, visible especially at night, and probably attributable to the ignition of fumarole gases, rather than strombolian activity from the main crater.

A presumed eruption in 1306, with lava reaching the sea, is referred to by Leandro Alberti (1568), and Pighius (1587) as well as by Schottus (1655), who copied from Pighius. This eruption was also confirmed by Masculo (1633) on the basis of these sources. However, Alfano (1924) denied the existence of such an eruption, demonstrating on the basis of historical arguments that the 1306 date is a typographical error in the text of Alberti (1568) who intended to discuss the 1036 event (Principe, 2003). The occurrence of at least one flank eruption at the Vesuvius during the period between A.D. 79 and 1631 is possibly recorded by a fresco, which once adorned the vestibule of the Basilica of San Gennaro Extra Moenia in Naples. The work has been attributed to the 15th century school of Andrea Da Salerno. Reprinted in Alfano (1924), it depicts the Vesuvius as observed from Naples, with two volcanic plumes, one issuing from the main crater (the elevation of which is lower than the peak of Mount Somma) while the other emerges from one or more vents situated on the southern flank of the Gran cono. The date of the eruption to which the fresco refers to is unknown. In other words, this fresco might simply be either a generic view of the erupting Vesuvius, or a copy of a previous work. Historical accounts describe a number of parasitic cones, which by age and position may represent the vents of some lavas emitted between A.D. 79 and 1631. These include the Viulo and Fossamonaca cones, the morphology of which is still well visible today, the two cones at Tironi and the three cones at I Monticelli ("little hills"). Sorrentino (1734) described two scoria cones and associated lavas from the site known as "i Tironi", north of Torre del Greco, as well as three small scoria cones, called "I Monticelli", aligned in a roughly east-to-west direction behind the Camaldoli hill. All these cones are attributed to medieval eruptions by Sorrentino (1734). The cones of I Monticelli have been buried by the products of subsequent eruptions and only a few morphological traces remain at present, whereas the Tironi hills were already almost completely quarried by Sorrentino's time. In 1514 the Venetian Leone Ambrosio described an event that took place in the year 1500. This could be a modest phreatic eruption. Apart from this source, however, no trace remains of the 1500 event in the various manuscripts, annals, and chronicles of the time. For this reason, its occurrence had already been disputed in the 17th century (Braccini, 1632). Summing up, on the basis of available historical data, it seems probable that from the early 12th century to 1631, the volcano was in a state of repose, apart perhaps from more or less continuous fumarolic activity. Since the Vesuvius was practically inactive for such a long time, it caught most of the inhabitants unaware at the moment of its awakening on December 16th, 1631 (Rosi et al., 1993). Tephra chronology The main stratigraphical markers of the entire post-A.D. 79 succession are the A.D. 472 deposits (Rosi & San-


707 Table 2 Available C data for charcoal and paleosols (*) of the post A.D. 79 ­1631 period. The most probable (bold type) eruption dates were defined in this work on the basis of calibrated age
1 Sample R 715 / L 371E PFSV 86A R939* Rome 643 PFSV 83 PFSV 86A PFSV 306* WW - 595 Rome 644 WW-6327 PFSV 101 Rome 653* Rome 641 Rome 652* WW-596 PFSV 86B 2 Locality 3 Age b.p. 181540 175060 163050 160535 160060 1590190 158060 155050 144060 135555 129040 128050 126555 123055 116655 114060 105085 95090 95050 47055 33055 4 Calibrated age A.D. 135­320
14

intervals, stratigraphic position, and physical characteristics, as given in the literature, and the historical data listed in Table 1. (**) Calibrated age given in Andronico et al. (1995)
5 Literature attribution Santa Maria AD 472 AD AD AD AD AD 472 472 472 472 472 6 Reference Andronico et al. (1995) ArnÐ et al. (1987) ArnÐ et al. Andronico ArnÐ et al. Rolandi et ArnÐ et al. ArnÐ et al. Rolandi et Andronico Rolandi et ArnÐ et al. Andronico Andronico Andronico Rolandi et ArnÐ et al. Rolandi Rolandi ArnÐ et ArnÐ et et et al. al. (1987) et al. (1995) (1987) al. (1998) (1987) (1987) al. (1998) et al. (1995) al. (1998) (1987) et al. (1995) et al. (1995) et al. (1995) al. (1998) (1987) al. (1998) al. (1998) (1987) (1987) 7 Most probable eruptions A.D. 203

LAGNO AMENDOLARE ARCIPRETE LAGNO di POLLENA TERZIGNO LAGNO AMENDOLARE CASE TRAPOLINO TERZIGNO OTTAVIANO LAGNO di POLLENA

A.D. 415­535 A.D. 59­861

A.D. 472 A.D. 472

A.D. 537­692 A.D. 645­760 A.D. 661­863 A.D. A.D. A.D. A.D. 685­885 690­935 780­980 775­1017

WW - 597 PFSV 308 * PFSV 315

TERZIGNO LAGNO AMENDOLARE BOSCO REALE TERZIGNO W of OTTAVIANO between POLLENA and CERCOLA

AD 512 II eruption PM5 II eruption AD 787 PM5 PM6 PM6 III eruption AD 787 IV eruption IV eruption 1631 1631

A.D. A.D. A.D. A.D.

536 685 685 685

A.D. 685 A.D. 787 A.D. 787 787/991/999/1006­7

A.D. 895­1279 1003­1279 1485­1650**

991/999/1006­7/1037/1139 1006­7/1037/1139 1631

tacroce, 1983, Mastrolorenzo et al., 2002) and the 1631 deposits (Rosi et al., 1993), both resulting from smallscale plinian events. On this basis, the tephra succession from A.D. 79 to 1631 can be divided into two parts. Andronico and coworkers (Andronico et al. 1995, 1996a,1996b) called the pyroclastic sequence prior to the A.D. 472 eruption the " Santa Maria cycle", which includes the tephra of the A.D. 203 eruption, as already recognised by Johnston-Lavis (1884) and ArnÐ et al. (1987). Andronico et al. (1995, 1996a, 1996b) identified six main explosive-component eruptions subsequent to A.D. 472; their tephra deposits are interlayered with seven minor tephra layers. The first main tephra layer was attributed by Andronico et al. (1995, 1996a, 1996b) to the A.D. 512 eruption, however, this eruption has been shown to be an unlikely source (see discussion above). Therefore, we propose to attribute this first main tephra layer, starting the post-A.D. 472 succession, to the explosive eruption of A.D. 536. In contrast to Andronico et al. (1995, 1996a, 1996b), Rolandi et al. (1998) identified only four main explosive eruptions in the post-A.D.-472 stratigraphy, and also attributed different ages to the layers. The last of these eruptions is represented by a lapilli fallout deposit which was attributed to the 1139 eruption by Rolandi et al. (1998). The description and stratigraphic position of this deposit correspond to the "ochraceous" lapilli fallout described by Principe et al. (1987), which is encountered

in pockets above the medieval-age lavas in an area southeast of the Vesuvius, and beneath the 1631 pyroclastic flow deposits. With the aim of identifying possible correlations between historically-defined eruption ages and the different tephra age attributions proposed by various authors, Table 2 reports all the 14C ages available in the literature for the period from A.D. 79 to 1631. The last column in Table 2 reports the historically substantiated eruptions or eruptive periods in the age interval obtained by calibrating the 14C ages. Of the eruptions deemed "possible" through study of the historical sources, the most probable ones (also based on stratigraphic position and physical characteristics) are presented. A newly determined radiocarbon age (see Appendix 1), based on charcoal fragments found under a lapilli fallout deposit of the A.D. 79­1631 period, confirms the occurrence of an explosive episode around A.D. 685. Comparing Tables 1 and 2, it is evident that, despite the rather low precision of the 14C age determinations, a good agreement exists between the 14C ages and the historically documented eruption ages for the events with an explosive component, a fact that underscores the importance of historical data.


708 Fig. 2 a-b Sector I. (a) The Granatello headland on the Topographic map of Mount Vesuvius compiled and drawn in six sheets at the scale of 1:10.000 by students of the Military Topography Institute of Florence (1875­86). The small promontory of Villa d'Elboeuf (1) is formed by the A.D. 79 pyroclastic flow and debris flow deposits. A branch of the nuüe ardentes of the 1631 eruption (Rosi et al., 1993) extended south of this, leaving Palazzo Mascabruno (2) untouched, but destroyng the ancient Granatello gardens (4) and partially covering the pre1631 lava that gave rise to the main coastal morphology. Construction of the Royal Palace of Portici (3) and surrounding gardens began in 1738. (5) The Herculaneum excavations. (b) Aerial view of the current Granatello and Royal Palace areas; P25 refers to the sampling site of the Granatello lava

Chronostratigraphy of lavas The problem of the absolute age and stratigraphic position of lavas from A.D. 79 to 1631 was raised for the first time by Johnston-Lavis (1884,1891b) who believed that the lava effusions reported in historical sources for the period from A.D. 79 to 1631 were much less frequent than the explosive events. Moreover, he maintained that the lavas were confined inside the Mount Somma caldera surrounding the Gran cono (Fig. 1), which had not yet been filled. This convinction was due to the fact that, following

Fig. 3 Sector I, immediately south of the Granatello area. (1) The La Favorita country house and (2) its gardens that reach the sea at Scogli della Favorita. (3) A few hundred meters south, a huge lava crops out extensively on the headland of Scogli della Scala (now Punta Quattro Venti), sampled at P18. Topographic base as in Figure 2


709 Fig. 4 a-b Sector II. (a) Aerial photograph of Torre del Greco as it appears today. (b) Map of Torre del Greco drawn by G. Morghen after the 1794 lava destroyed four-fifths of the then existing town. (1) The lava headland of Calastro (sampling site P10). (2) The Monastery of the Minor Franciscan Order and the annexed church of Madonna delle Grazie. The Monastery was spared from the 1631 pyroclastic flow (Rosi et al., 1993). (3) Palazzo Brancaccio and its gardens. (4) The Torre del Greco Town Hall housed in the one-time Palazzo Baronale. All these buildings lie on a morphological high spared by 1631 nuüe ardentes and then by 1794 lava, that corresponds to an old lava, unfortunately not cropping out anymore, attributed by historical records to the A.D. 685 eruption. (5) The area affected by the pyroclastic flow deposits of the 1631 eruption ("mare seccato", which is Italian for "dried sea") and by the lava of the 1794 eruption afterwards

Le Hon (1865,1866), Johnston-Lavis attributed all the lavas cropping out along the coast to the 1631 eruption. The belief that all the lavas along the coast belong to the 1631 eruption has been recently supported by the work of Vittozzi & Gasparini (1965), who performed absolute age determinations via the radioactive 226 Ra-238U disequilibrium method on the lava of Granatello and on the three lavas of the Villa Inglese (see Figures 2, 3, 4,5, 6,7 for the locations). On the basis of these data, Burri & Di Girolamo (1975) ascribed the Granatello lava to 1631, and the Villa Inglese lavas to 1631, 1480 and 1139. Other lavas cropping out along the coast between Portici and Torre Annunziata (Scogli della Scala, Calastro, Ponte di Riveccio, Torre Bassano) have

been attributed by Di Girolamo (1970) to the 1631 eruption, always on the basis of the data of Vittozzi & Gasparini (1965), without any stratigraphic support. At least for Torre Bassano, historical data do not support such attributions, since a watchtower was built between 1559 and 1567 over this lava (Principe, 2003, Fig. 5). Recently, Rolandi & Russo (1989) ascribed both the lava under Masseria Donna Chiara and the lava cropping out along the coast near Torre Annunziata between Cantiere la Perla and Palazzo Monteleone to the 1631 eruption. Finally, although the emission of lavas during the 1631 eruption can be ruled out, based on volcanological evidence (Principe et al., 2003) and historical documents (Rosi et al., 1993), most of the lavas cropping out along


710

Fig. 5 Sector III. The Torre Bassano headland exhibits the morphology and stratigraphy typical of all the sector of the coastline from Portici to Torre Annunziata. In this site, the tower (1) was built between 1559 and 1567 on the promontory generated by a medieval lava that flowed into the sea. This lava crops out under the tower (2), along the road inland from the tower (P15), and along

the railway (2), and covers a stratigraphic succession (see inset) including the A.D. 79 eruption-related debris flow and previous pyroclastic deposits. The pyroclastic flow deposits of the 1631 eruption (3) extend on both sides of the mophological high induced by the lava and are now covered by numerous greenhouses and vineyards

the coast of the Vesuvius are still attributed to the 1631 eruption by many authors. We intend to definitively clarify this subject with our archaeomagnetic data. Synthesis of available chronological data On the basis of the available chronological information, at least 11 main eruptions took place at Vesuvius' between A.D. 79 and 1631. The most important of these events is the small-scale plinian eruption of A.D. 472. An effusive component is described by historical documents for at least four of these eleven eruptions (Table 1), with lava fronts spreading outside the Mount Somma caldera rim, and reaching the sea in A.D. 787, 968, 999, and 1037. According to historical, radiocarbon, and stratigraphic data, significant explosive episodes took place at least in A.D. 203, 472, 536, 685, 787, 991, 999, 1006/7, and 1139, resulting in a sequence of lapilli fallout deposits. Within this sequence, a number of minor tephra deposits give evidence of additional minor explosive episodes not reported by printed chronicles and still not dated by radiocarbon or other methods.

Archaeomagnetic sampling and measurements
This archaeomagnetic study of the Vesuvius is based on the precise method developed at Saint Maur Laboratories from the Institut de Physique du Globe de Paris (IPGP), France (Thellier, 1981; Tanguy et al., 1985, 1999, 2003). Sampling and measurement procedures used in this work

refer to the so-called "big sample plaster method" of Thellier (1981) for archaeological purposes and routinely applied at the Saint Maur Laboratories. Application of this methodology for dating of young volcanic products, its advantages, and limitations are given by Tanguy et al. (2003). The sampling method involves the collection of several big samples, each weighing 0.5­1 kg, distributed over some tens of meters. Each block is first detached with a hammer and then replaced in its original position. Plaster of appropriate fluidity is poured on the block, to support a plate carrying a spherical spirit level, thus making a horizontal surface 6­7 cm in diameter on which the sun shadow is marked for calculation of the geographical north. In the laboratory, the samples are replastered in square moulds 12 cm in size. Measurements are performed by means of a large rotating induction magnetometer and an AF demagnetization device for big samples (Le Goff, 1975). The advantages of this method with respect to traditional core-drilling are the extremely precise orientation of the sample in the field and the possibility of obtaining high-precision data on lavas, welded scoriae, and hightemperature-emplaced, non-welded, pyroclastic deposits (Hoblitt et al., 1985; Genevey et al., 2002). Another fundamental advantage is that the plaster method avoids the effect of any parasitic magnetization acquired during the drilling process (Lauer, 1978; Audunsson & Levi, 1989). On the basis of their stratigraphic positions and additional field data (e.g., age of the oldest buildings sitting on the lavas, changes of locality names after the main


711 Fig. 6 a-b Sector IV. (a) Topography of the Villa IngleseLe Mortelle area (topographic base as in Fig. 2) showing the smooth morphology induced by the medieval lava flows covered by subsequent tephra deposits, contrasting with the rough morphology of the of the 1760 lava surface cropping out in the eastern sector of this figure. (1) The old lava quarry of Villa Inglese. (2) Masseria Donna Chiara. (3) Villa Prota. (4) Le Mortelle. (5) Torre Scassata. The quarry has been progressively enlarged and today has reached the lava emitted in 1760. (b) Present aerial view of the Villa Inglese quarry. Numbers refer to the sampling sites of the pre-1631 lavas flows extensively cropping out both inside the quarried area and on the coast beneath and inland of the Torre Scassata tower

eruptions, etc. ...) (see Appendix 2), a total of 26 sites have been selected and sampled for archaeomagnetic measurements, giving a total of over 400 analysed samples, covering nearly all the lava outcrops attributable to the A.D. 79­1631 period. Fig. 1 shows the location of the sites which are grouped in geographical sectors, whereas Table 3 gives the geographical coordinates and a description of each site. Some sites are located in correspondence to newly built road segments or work yards for the construction of new buildings, which have brought to light portions of the lavas covered by the 1631 nuües ardentes. Some of the sampled lavas cropping out along

the coast do not appear on the 1:25,000-scale CNR (1986) geological map. All the lavas of this period are petrographically very similar (Joron et al., 1987; Belkin et al., 1998; Raia et al., 2000; Paiotti, 2000) (Table 4). The main phenocrysts are two clinopyroxenes (diopside and augite), which are both present in all lavas, and leucite. Olivine and a dark mica (phlogopite) are the other most important phases. The groundmass is composed of plagioclase (An80), leucite, clinopyroxenes, and oxides. Modal analyses (Table 5) show the fact that correlations between different sites cannot be performed on the basis of petrographical


712 Fig. 7 a-c Sector V. (a) The shore of Torre Annunziata (topographic base as in Fig. 3). (1) Scogli di Prota. (2) The old "Forte dell'Oncino" fortress. (3) Villa Filangeri. (4) Terme Nunziante-Lido Azzurro. (5) Palazzo Monteleone. Originating from a vent near Bosco del Monaco (label `vent' in this sketch map), a huge lava reached the sea forming the La Storta promontory, where the pier of Torre Annunziata is nested. Note the ancient lava quarries (q) marking the lava position between the vent and the sea. (b) Present aerial view of the Torre Annunziata waterfront. Some deposits attributed to the A.D. 79 and 472 eruptions are exposed under the aforementioned lava in the scarp along the Litoranea Marconi road, from Palazzo Monteleone (5) to Lido Azzurro (Cantiere La Perla) (4). (c) Stratigraphic section of the excavations for the new Torre Annunziata Hospital at Rione Boscomonica (P23 in map a). Ozf 467 refers to a sample of charcoal, dated by means of radiocarbon method (see Appendix 1). Its calibrated age is 639­687 A.D

composition. In fact, samples coming from the same lava, as in the case of P01 and P20 sampling sites, may differ in mineral abundance (Lc and Cpx) and vesicularity, as already found for the similar lavas of the 1631­1944 Vesuvius activity (Principe, 1979). For each site, an average of 12 archaeomagnetic analyses were performed on samples originating from

different portions of the selected lava. The obtained mean value of the semi-angles of the 95% confidence cone of the average paleodirections (a95) resulted in 1.2, with a range from 0.6 to 1.8 (Table 6).


713 Table 3 Details of the archaeomagnetic sampling sites. Geographical sectors are defined in Fig. 1 Site P01 P02 P03 P04 P05 P06 P09 P10 P11 P12 P14 P15 P16 P17 P18 P19 P20 P21 P22 P23 P25 P26 P27 P28 P29 P30 79 472 Coordinates 4045' 20" N 1424' 34" E 4045' 21" N 1424' 34" E 4045' 05" N 1424' 51" E 4046' 23" N 1425' 57" E 4046' 20" N 1426' 18" E 4048' 24" N 1427' 37" E 4046' 44" N 1425' 48" E 4047' 21" N 1421' 27" E 4049' 05" N 1421' 25" E 4045' 39" N 1426' 12" E 4045' 07" N 1425' 38" E 4046' 13" N 1422' 59" E 4045' 30" N 1424' 08" E 4046' 42" N 1429'57" E 4047' 34" N 1421' 14" E 4045' 44" N 1423'37" E 4045' 20" N 1424' 34" E 4045' 08" N 1425' 10" E 4048' 06" N 1427' 47" E 4046' 00" N 1427' 07" E 4048' 29" N 1420' 12" E 4048' 00" N 1422' 55" E 4045' 30" N 1424' 17" E 4045' 24" N 1426' 39" E 4048' 27" N 1421' 42" E 4049' 13" N 1421' 07" E 4048' 21" N 1420' 59" E 4051' 24"N 1423' 18"E Sector IV IV IV V V VI IV II I V V III IV VI I IV IV IV VI V I II IV V I I Samples 10 (21­30) 10 (11­20) 15 (31­45) 16 (46­54, 106­107, 301­305) 7 (55­60, 62­63) 9 (64 ­72) 10 (108, 110­111, 113­114, 195­199) 17 (115­131) 18 (132­142, 247­253) 9 (143­151) 11 (154­164) 15 (165­179) 15 (180­194) 15 (200­209, 334­338) 17 (210­226) 14 (227­239) 7 (240­246) 18 (254­271) 12 (306­317) 14 (320­333) 8 (339­346) 10 (347­356) 12 (357­368) 13 (369­381) 12 (382­393) 17 (395­406) 17(JCT, 1­17) 13(P08, 88­105) Sample Type Lava flow Lava flow Lava flow Lava flow Lava flow Lava flow Scoria cone; juvenile scoriae Lava flow Lava flow Lava flow Lava flow Lava flow Lava flow Lava flow Lava flow Lava flow Lava flow Lava flow Lava flow Lava flow Lava flow Lava flow Lava flow Lava flow Locality Villa Inglese, near Torre Annunziata, upper lava on the W corner of the quarry Villa Inglese, near Torre Annunziata, lower lava on the W corner of the quarry Torre Scassata, along the coast, near Torre Annunziata New road towards W from the cemetery of Boscotrecase (900m from the crossing) New road towards W from the cemetery of Boscotrecase (200m from the crossing) Terrioni quarry, W of Terzigno Fossa Monaca, NW of Torre Annunziata Calastro, Torre del Greco NA-SA highway, Ercolano gate (Napoli direction) Torre Annunziata soccer field Villa Balke--Club La Vela (Scogli di Prota), along the coast, near Torre Annunziata Torre Bassano, along the coast road, back from the ancient tower S.Vincenzo Postiglione, along the coast, near Torre Annunziata Passanti, near Terzigno Scogli della Scala, 4 Venti hotel beach Lido Incantesimo, near Torre Annunziata Villa Inglese, near Torre Annunziata, uppermost lava on the E corner of the quarry Masseria Galassi, along the coast, near Torre Annunziata Pozzelle, uppermost quarry, near Terzigno Masseria Bosco del Monaco, N of Torre Annunziata Granatello, Portici Vesuvius' Cafü, San Sebastiano street, N of Torre del Greco Masseria Donna Chiara, near Santa Maria La Bruna Torre Annunziata harbour, under Palazzo Monteleone Cupa dei Monti, N of Ercolano NA-SA highway, Portici -Bellavista gate Ercolano ancient city, inside the Archaeological excavation

Lava flow Lava flow Unwelded pyroclastic flow; single juvenile clast Unwelded Pollena quarry pyroclastic flow; single juvenile clasts

Results and errors Fig. 8(a, b) displays a plot of the archaeomagnetic data obtained in this work in relation to the curve of secular geomagnetic variation proposed by Tanguy et al. (1999) for Mount Etna (Sicily), and Bucur (1994) for France, both reduced to the latitude and longitude of the Vesuvius (4048'N, 1425'E). With the purpose of checking the applicability of these curves to Vesuvius, we have also sampled the juvenile fraction of the pyroclastic flow deposits of the A.D. 79 and 472 eruptions (Table 3). The position of these two pyroclastic samples on the secular curve at the expected age (Fig. 8a) suggests that this curve

can be used confidently for dating Vesuvian lava samples (see also Tanguy et al., 2003). Fig. 8(a, b) shows that the reduced curves for France and Etna are nearly straight lines that run closely side by side from about A.D. 800 to 1200, where all but two of the analysed units are distributed. These units (P05 and P22) fall on a different curve referring to instrumental paleomagnetic record (Cafarella et al., 1992), indicating that they belong to the 1767 and 1701 lavas, respectively. The low a95 values result in small confidence circles and in relatively small uncertainties in age determination. For some sites, the positions of the confidence circles on the


714 Table 4 Synopsis of eruptions between A.D. 79 and 1631, based on historical, radiocarbon and archaeomagnetic data. (A): plinian and small-scale plinian eruptions. (B): violent strombolian to subplinian eruptions. (C): more or less continuous activity with strombolian to violent strombolian eruptions. Grey rectangle: mixed, effusive and explosive eruptions. (*): eruptions with lavas dated in the present work by means of the archaeomagnetic method

Table 5 Modal analyses of selected lava samples. Values are expressed in vol. % on 1000-point-analysis for each thin section Sites P01 P02 P03 P04 P06 P11 P12 P14 P16 P17 P19 P20 P21 P27 Locality Villa Inglese, upper lava Villa Inglese, lower lava Torre Scassata W cemetery of Boscotrecase (900m) Terrioni Quarry NA-SA Ercolano Torre Annunziata, soccer field Scogli di Prota-Villa Balke San Vincenzo Postiglione Passanti Lido Incantesimo Villa Inglese, upper lava Masseria Galasspi Donna Chiara Cpx (%) 6.8 16.4 7.9 1.7 8.0 21.1 1.5 5.7 6.7 7.3 5.0 10.7 7.2 4.3 Lc (%) 1.4 0.4 5.0 1.0 4.9 3.6 1.2 2.6 4.5 3.8 0.3 4.1 0.7 0.1 Cpx xenoliths (%) 0.9 0.9 0.9 0.9 0.2 0.9 1.0 1.2 0.6 2.4 1.9 4.7 1.9 0.5 0.1 Ol (%) Plag (%) Vesicles (%) 1.2 4.8 1.9 15.1 5.4 7.2 15.8 14.7 0.6 3.9 21.9 0.9 1.9 4.7 Groundmass (%) 90.5 78.81 84.3 82.0 79.4 63.3 81.1 76.6 85.9 84.3 71.5 82.9 89.2 88.5


715 Table 6 Archaeomagnetic data. N= number of samples analysed for each site; a95 = semi-angle of the 95% confidence cone of the average paleo-directions; K = precision parameter from Ficher (1953) statistics; I = mean inclination; D = mean declination Sites P01+P20 P02 P03 P04 P05 P06 P08 P09 P10 P11 P12 P14 P15 P16 P17 P18 P19 P21 P22 P23 P25 P26 P28 P29 P30 79 Locality Villa Inglese, upper lava (PA01), (PA20) Villa Inglese, lower lava Torre Scassata W cemetery of Boscotrecase (900m) W cemetery of Boscotrecase (200m) Terrioni Quarry Pollena quarry (472 AD pyroclasts) Fossa Monaca scoria cone Calastro Na-Sa Ercolano Torre Annunziata, soccer field Scogli di Prota-Villa Balke Torre Bassano San Vincenzo Postiglione Passanti Scogli della Scala Lido Incantesimo Masseria Galassi Pozzelle quarry Masseria Bosco del Monaco Granatello Vesuvius' cafü, N of Torre del Greco Torre Annunziata, palazzo Monteleone Cupa dei Monti, N of Ercolano NA-SA Portici-Bellavista Ercolano ancient city, inside the Archeological escavations N 15 9 14 13 8 5 13 7 15 9 8 10 14 10 7 14 11 15 11 9 7 10 11 10 9 11 a
95

K 943 1040 1199 853 1412 2563 2596 1650 1094 654 1784 634 1385 1454 1284 520 1183 1374 4870 1523 1395 711 1484 1093 2085 1573

I() 59,4 60,4 60,6 56,7 59,7 66,7 56,7 60,2 62,7 58,6 63,4 60,4 63,1 66,9 56,8 64 59,1 67,2 66,4 62,4 64 60,8 63,8 58,1 63,4 58,7

D() 20,1 16,8 15,8 17 -16,7 7,0 4,8 18,8 13 16,9 12,3 19 17 12 19.4 13,6 16,7 3,3 -12,1 13,7 11,7 17,7 12,3 19,3 11,4 -4,8

1,2 1,4 1,1 1,3 1,3 1,2 0,8 1,3 1,1 1,8 1,2 1,8 1 1,2 1,4 1,6 1,2 1 0,6 1,2 1,4 1,7 1,1 1,3 1 1,1

Table 7 Archaeomagnetic ages of the pre-1631 lavas analysed in this paper. Bold character: most probable eruption dates assigned in this work on the basis of both historical and stratigraphic considerations

Site P21 P06 P16 P10 P30 P25 P18 P12 P28 P23 P15 P09 P03 P02 P14 P26 P01+P20 P19 P11 P29 P17 P04

Locality Masseria Galassi Terrioni Quarry San Vincenzo Postiglione Calastro NA-SA Portici-Bellavista Granatello Scogli della Scala Torre Annunziata, soccer field Torre Annunziata, palazzo Monteleone Masseria Bosco del Monaco Torre Bassano Fossa Monaca Torre Scassata Villa Inglese, lower lava Scogli di Prota-Villa Balke Vesuvius' cafü, N of Torre del Greco Villa Inglese, upper lava Lido Incantesimo NA-SA Ercolano Cupa dei Monti, N of Ercolano Passanti S flank W of Boscotrecase

Archaeomagnetic age 80020 85040 87030 90030 90020 90030 90040 90020 90030 92030 94030 100040 100030 100050 100050 100050 105030 105030 108060 110040 110040 114040

Attributed eruption AD 787 9th century 9th century 10th century 10th century 10th century 10th century 10th century 10th century 10th century 10th century, 968 968, 999, 1006­07 968, 999, 1006­07 968, 999, 1006­07, 1037 968, 999, 1006­07, 1037 968, 999, 1006­07, 1037 999, 1006­07,1037 1006­07, 1037 1037, 1139 1139 1139 1139

curve are not significantly different, suggesting that these lavas belong to the same eruption. Table 7 shows the archaeomagnetic ages and their range of error. Following the approach of Tanguy et al. (2003), it has been possible to obtain age determinations within 40 years, with a higher precision of 20 years in the best cases. These results are substantially better than those of the radiocarbon method (Table 2), at least for the considered time span, a fact that underscores the importance of archaeomagnetic dating of young volcanic suc-

cessions. Only one site yielded unacceptable analytical results, the Donna Chiara lava (P27), owing to the poor outcropping conditions (only small lava outcops are accessible and most were probably displaced after cooling). Short-period palaeomagnetic techniques (secular variation curve) have previously been applied by Carracedo et al. (1993) to investigate (1) some of the Vesuvian lavas, interpreted in the literature as products of the 1631 event, (2) the pyroclastic flow deposits of this same eruption and (3) other lavas unquestionably emitted after this eruptive


716 Fig. 8 a-d (a) Geomagnetic secular variation curves for Etna (Tanguy et al.,1999) and France (Bucur, 1994) (heavy and light dashed lines, respectively) reduced to the latitude and longitude of Naples, and geomagnetic curve based on the direct magnetic field measurements performed since 1640 (solid line; Cafarella et al., 1992 and Thellier unpublished data, see in Bucur, 1994). Confidence circles associated with the analyses performed in this work (P) are also plotted. (b) Enlarged section of the previous plot for the medieval period. (c) Data set for lavas from A.D. 79 to 1631 from Carracedo et al. (1993) (V) and (d) from Gialanella et al. (1993 and 1998) (G) and Hoye (1981)

event. The data obtained for the lavas attributed to the 1631 eruption suggested that these lavas were the result, not of a single eruptive event, but of several different eruptions, with a palaeomagnetic correlation with the medieval curve (Carracedo et al., 1993). Most of these lavas have been newly analysed in the framework of the present work (Table 8). There is evidence that substantial differences exist between this classical paleomagnetic method and the "big sample" dating procedure (Tanguy et al., 2003), mainly due to rotation of the steel, diamond-tipped corer used for traditional palaeomagnetic sampling which induces secondary magnetisation in lavas (Lauer, 1978; Audunsson & Levi, 1989). In spite of this fact, a number of authors have extracted samples of Vesuvius' lavas by coring and performed paleomagnetic analyses by spinner (Hoye, 1981; Carracedo et al., 1993; Gialanella et al., 1993 and 1998). Figures 8c and 8d report the results obtained by these authors, as well as the curves of secular geomagnetic variation drawn in Figures 8a and 8b. Table 8 shows the correspondence between the various analysed sites. Another source of error lies in the fact that Hoye (1981) based his attribution of the lava he analysed to the

1631 eruption on the thesis of Principe (1979), wherein lava sample PFSV 200 was erroneously described as a lava from the 1631 eruption on the basis of the Le Hon (1866) map of Vesuvius' lavas. This site corresponds (Table 8) to site V19 of Carracedo et al. (1993) and site P29 of this work, which obviously belongs to medieval times (1139 eruption, see below). Consequently, Hoye (1981) constructed an unreliable secular curve, of which the 1300­1600 part was confused with the medieval one (compare Hoye, 1981 and Tanguy et al., 1999 and 2003). Figures 8c and 8d show the fact that the results obtained with classical methods present a much larger spread on the secular variation curve with respect to our samples (Figures 8a and 8b). Moreover, the literature data have a95 values higher than our data and their analytical results are not always conform with the stratigraphic order of the samples. The best results with traditional methods were obtained by Carracedo et al. (1993) on the juvenile fraction of the 1631 pyroclastic flow, the circle of confidence of which is compatible with the first "direct" measurements of the Earth magnetic field (Cafarella et al., 1992). Such a good analytical result may be due to the fact that "tender" samples, such as scoriae, are drilled


717 Table 8 Correspondence between the ages given by Vittozzi & Gasparini (1965), Di Girolamo (1970), Burri & Di Girolamo (1975), Carracedo et al. (1993), Gialanella et al. (1993 and 1998),
Sampling locality This paper Di Girolamo (1970) Burri & Di Girolamo (1975)

and Hoye (1981) and those obtained in this work for the same lavas. Sample codes are in brackets. MA = Middle Ages
Rolandi & Russo (1989) Hoye (1981) Gialanella et al. (1998) Incoronato (1996) 1631 (GA8) 1631 (GA9) 1631 (GA10) Carracedo et al. (1993)

Lirer et al. (2001)

San Vincenzo Postiglione Masseria Donna Chiara Calastro Scogli della Scala Granatello Torre Annunziata soccer Palazzo Monteleone/La Perla NA-SA Portici Bellavista Scogli di Prota/Oncino Torre Villa Torre Villa Bassano Inglese, lower lava Scassata Inglese, upper lava

9th century (P16) ?9th century? (P27) 10th century (P10) 10th century (P18) 10th century (P25) 10th century (P12) 10th century (P28) 10th century (P30) 1006/7 (P14) 968 (P15) 999 (P02) 999 (P03) 1037 (P01, P20) 1139 (P11) 1139 (P29)

MA (V8) MA (V15) MA (V14) MA (V16) MA (V27)

1631 1631 1631 1631 1631 1631

1631(GA1) 1631(GA2) 1631 (GA12) pre-1301 (GA3) pre-1301 (GA13) pre-1301 (GA6) pre-1301 (GA4, 5)

1631 1480 1631

13th-14th century MA

MA MA MA MA MA

MA (V12) MA (V22) MA (V3, V9) MA (V11) MA (V2, V4, V10) MA (V20) MA (V19)

NA-SA Ercolano Cupa dei Monti

1631 (PFSV200)

1631 (GA11)

very quickly and consequently acquire less secondary magnetisation during mechanical core sampling than "hard" samples, such as lavas. Finally, we must point out that some authors (Gialanella et al., 1993, 1998; Incoronato, 1996, 1999; Incoronato et al., 2002) were led to incorrect conclusions on medieval lavas (Table 8) because they also used the wrong reference curve defined by Hoye (1981).

The interpretation of new archaeomagnetic data
The geographical distribution of the main lavas dated in this work and their stratigraphical relationships are sketched in Fig. 9, on the basis of both field evidence gathered during sampling and the cartography of lava outcrops given in the detailed geological map of Johnston-Lavis (1891a). In Appendix 2 the ages we obtained via the archaeomagnetic method are discussed in the framework of geological and archaeological constraints, for each of the geographical sectors defined in Figure 1. Unnoticed events As suggested by the comparison of historical (Table 1) and archaeomagnetic data (Table 7), several effusive eruptions occurred between the second half of the 9th century and the first half of the 10th century that were not documented in the surviving historical sources (Table 9). This fact may be explained by the limited circulation of documents in a historical period dominated first by the

repeated invasions of the Longobards and Franks, and then by the conflicts between the Byzantines and Saracens for the domination of the Dukedom of Naples (Girelli, 1994). It is possible that some of these effusive eruptions also involved an explosive phase, which might be responsible for the minor tephra layers of this time span. Apart from these eruptions, all the historically documented eruptions are corroborated by the archaeomagnetic evidence, confirming both the validity of the historical information, when available, and the power of archaeomagnetism as dating tool of young volcanic units. Mixed eruptions The current stratigraphic and geochronological findings confirm the attribution of the "ochraceous" lapilli fallout that lies between the 1037 lava and the deposits of the 1631 eruption in the Villa Inglese area (Fig. 6 and 7c) to the 1139 eruption, as already recognised by Principe et al. (1987) and Rolandi et al. (1998). A number of eruptions are of mixed (effusive and explosive) type, such as those of A.D. 787, 999, 1006­07, and 1139 (Table 9), based on existing radiocarbon ages, historical data, and our new archaeomagnetic data. Moreover, in the case of the eruptions of 1006/7 and 1139, archaeomagnetic age determination support a phase of lava discharge for which documentation was lacking in the historical chronicles (Principe, 2003).


718 Fig. 9 Sketch map, flow directions and ages of the medieval lavas dated in this work (topography as in Fig. 1).Bold dots refer to the sampling sites listed in Figure 1 and Table 3

Table 9 Mineralogy of selected lava samples. The presence of each mineral phase has been qualitatively estimated (Scarce = <5%, present =5­10%, quite abundant=10­15%; abundant >15%)

Effusions from central vent and intracaldera activity The lavas of the 1139 eruption are distributed at relatively high elevation on the western and southeastern slopes of the volcano, downstream from the Fosso Grande, Cognoli, and Cognoletto valleys (Fig. 9), which channelled the numerous post-1631 lavas (Arrighi et al., 2001). Contrary to the parasitic activity that generated lavas sometimes more than 10 m thick and several hundred meters across, the 1139 lavas have fronts which never exceed 2­3 m in thickness. Hence, the 1139 lavas were probably erupted from the main crater. In addition to the lavas of the 1139 event, few other lavas between A.D. 79 and 1631 (e.g., samples P06 and P29) were able to surpass the Mount Somma caldera rim. Overflow probably occurred when the Gran cono attained sufficient development inside the Mount Somma caldera, as testified by historical sources (see discussion above). To appreciate this fact it must be recalled that the Mount Somma caldera was an amphitheatre-shaped depression without any inner cone after the A.D. 79 eruption. Subsequent eruptive events determined the gradual growth of the Gran cono and, consequently, the progressive decrease in the caldera storage capacity for the lavas which erupted both from the top of the Gran cono and from fractures cutting its flanks. Finally, probably in the 9th century, lavas started to be dicharged from the caldera rim, which locally did not act as a morphological barrier anymore. A similar phenomenon occurred after the 1631 eruption that brought about the total destruction of the Gran cono. Again, lavas were able to surpass the caldera rim only after a century of activity when the Gran cono is


719

documented to have been fully reconstructed (Melilli, 1996; Arrighi et al., 2001). Parasitic vents The most evident of the Vesuvius' parasitic cones is the Camaldoli hill (Fig. 9), which is covered by the deposits from the the A.D. 79 eruption (CNR, 1986) and by those of the medieval activity, as described in detail by historical sources (Sorrentino, 1734) and confirmed by the stratigraphic surveys of Rolandi et al. (1998). Since the Camaldoli cone predates the A.D. 79 event, it is beyond the focus of the present work and, for this reason, archaeomagnetic sampling was not carried out. As already mentioned, the considerable thickness and morphological prominence of many "medieval" lavas cropping out along the coast depend on their emission from parasitic vents, thus suggesting activation or reactivation in this time interval of significant tectonic elements. Four possible sources for the eccentric lavas have been identified in this investigation on the basis of historical and field data (Fig. 9): Tironi, I Monticelli, ViuloFossamonaca, and Masseria Bosco del Monaco. Based on the archaeomagnetic analyses, it is reasonable to attribute the lower lava of Villa Inglese-Torre Scassata to the A.D. 999 eruption and set its discharge vent at Fossamonaca. The lava widely outcropping at Torre Annunziata along the coastline was discharged from the vent of Masseria Bosco del Monaco in the early 10th century. Another vent (possibly opened during the same eruption or another eruption occurring within40 years of the Masseria Bosco del Monaco event) was also responsible for the large lava of Granatello, the vents for which are very likely the two Tironi cones described by Sorrentino (1734). On the basis of both their position and the persistence of the related morphological anomaly, even beneath the lavas discharged on this slope after 1631, the three vents at I Monticelli may tentatively be attributed to the 1037 eruption, though this is by no means certain. The source vents of the Torre Bassano, Oncino and Le Mortelle lavas are still unknown. Timing of volcanic activity Based on our new archaeomagnetic data, the critical reading of historical sources, and taking into account the Vesuvius' literature for the post-A.D.-79 activity, the A.D. 79­1631 period of activity can be divided into two phases by the small-scale A.D. 472 plinian eruption. The first phase from A.D. 79 to 472 includes a main explosive event occurring in A.D. 203 and a number of minor eruptions and periods of activity, tentatively classified as strombolian to violent strombolian based on historical and available stratigraphical data. During this period, there is no evidence of lava effusions outside the Mount Somma caldera, although effusive intracaldera activity cannot be ruled out.

The second phase, from A.D. 472 to 1631, was initially characterised by a number of explosive eruptions as documented by available historical and stratigraphical data. However, a new kind of activity began in A.D. 787, with the occurrence of several mixed, explosive and effusive, eruptions. Lavas were erupted from the main cone and fractures situated outside the Mount Somma caldera, invading the southern flanks of Vesuvius' and shaping the coastal morphology. These products are relatively primitive, with a MgO content of generally between 3.5 and 6.5 wt% (Joron et al., 1987; Belkin et al., 1998; Raia et al., 2000; Paiotti, 2000). This activity lasted for about 360 years and ended with the 1139 eruption which preceded a 5-century-long period of quiescence. A number of features of the post-A.D. 472 phase, such as the initial occurrence of intracaldera events, the mixed style of eruptions suggesting an open-conduit system, and the MgO concentrations of the erupted products, are similar to the post-1631 behaviour of the volcano (Joron et al., 1987; Belkin et al., 1993). By assuming that the post-1631 period mimics the post-472 period, and that the last eruptive stage ended with the 1944 eruption (Principe, 1979; Carta et al., 1981), we are tempted to compare the long quiescence following the 1139 eruption and preceding the 1631 event with the present state of inactivity. However, we must point out that a repose period of less than one century preceded the A.D. 472 eruption, which is similar to the 1631 event, although the pre-A.D. 472 eruptive history is poorly known. It is clear that the present level of knowledge of the volcano behaviour during the A.D. 79­ 1631 period is not yet sufficient to draw sound extrapolations on its future state.

Conclusions
The archaeomagnetic method has resolved the ages of Vesuvius' eruptions in the time period from A.D. 79 to 1631. The method affords high reliability and an analytical error which is generally less than 40 years. Archaeomagnetic dating highlights several lavas, erupted between the second half of the 9th and the first half of the 10th century, that have been unnoticed by the historians of that rather dull period. Conversely, historically documented ages of eruptions are consistent with archaeomagnetic and geological data. All the medieval lavas cropping out along the southern slopes of Vesuvius' at low altitudes were erupted in ~360 years, from A.D. 787 to 1139. The large lavas that shaped the morphology of the Vesuvius' coast were largely discharged from parasitic vents situated outside the Mount Somma caldera. A number of these eruptions were also characterised by an explosive component, recorded by radiocarbon-dated tephra deposits (A.D. 787, 999, 1006­ 07, and 1139). The dated lavas include all those attributed by various authors to the eruption of 1631, often on the basis of questionable data and misinterpretations. The chronology of the lavas and tephra younger than the A.D.


720

79 eruption confirms that the violent 1631 eruption was preceded by a long period of repose (lasting probably almost 500 years). In the early centuries of the first millennium A.D., Vesuvius' activity was dominated by explosive eruptions, the last of which was the cataclysmic event of A.D. 472. Then, as in the case of the post-1631 activity (Rosi et al., 1993; Arrighi et al., 2001), intracaldera lava effusion recommenced, after apparently no rest at all. Then, from 787 to 1139 (~360 years), at least nine eruptions of mixed kind, effusive and explosive, took place, and lava fronts moved outside the Mount Somma caldera rim. These lavas were mostly discharged from vents placed low on the volcano slopes, invading the coastal strip and shaping the present coastline. The last centuries before the violent explosive paroxysm of 1631 were quiet. On one hand, such a picture is somewhat reassuring, given the long period of quiescence that followed the 1139 eruption and similarities with the current situation following the 1944 eruption. On the other hand, a rest period of less than one century preceded the A.D. 472 eruption, which is very similar to the 1631 event. In addition, the presence of such recent eccentric lavas underlying all the main inhabited land along the Vesuvius' coast cannot but arouse a certain degree of apprehension.
Acknowledgements We are grateful to Mattia Barsanti and Ilaria Cerbai for their generous help in sample collection. We also wish to express our gratitude to the Library of the Department for National Technical Services of the Presidency of the Council of Ministers for kindly providing digital copies of H.J Johnston-Lavis' geological map with its topographical basis, which represent only part of its rich cartography resources. Clara Principe Ardu is acknowledged for her translations of the Latin and Greek texts we used. Mauro Rosi and Raffaello Cioni provided useful comments on an earlier version of this manuscript. We are also indebted to Vince Neall for his thoughtful review of an earlier draft of the manuscript and with Luigi Marini and Jocelyn McPhie for greatly improving the English. This work was supported by funding from the Institute of Geosciences and Earth Resources in Pisa, and the Institut de Physique du Globe de Paris (IPGP-CNRS UMR 7577, contribution IPGP n 1958). The archaeomagnetic analyses are part of the degree thesis of AP and of the PhD thesis of SA.

the AMS measurement. The 14C/13C isotopic ratio was measured relative to the internationally accepted HOxI standard material (Stuiver, 1983). Then corrections were applied for spectrometer background, the contamination during preparation of the graphite target and isotopic fractionation. Using the corrected radioisotope ratio, the conventional radiocarbon age of 137040 BP was calculated. A calibration performed using the CALIBETH software (Niklaus et al., 1992) and the tree ring data set from Stuiver et al. (1998) provided us with a (cumulative probability) age range of 639­687 A.D. This result, compared with historical data, suggests that the deposit belongs to the A.D. 685 eruption, confirming the presence of an explosive phase in such an eruption.

Appendix 2: Geological and archaeological framework
Sector I The geomorphology of this coastline is dominated (Fig. 2) by the small promontory of Villa D'Elboeuf, the seaward front of the Granatello lava (P25), and the two rockcliffs known as Scogli della Favorita and Scogli della Scala (Fig. 3, P18). Construction of the Royal Palace of Portici and its surrounding park, on the Granatello lava, started in 1738 in an area that was famous before 1631 for its marvellous pomegranate gardens (whence the name), which were completely destroyed by the 1631 eruption (Mormile, 1632). A lava tentatively correlated by various authors with the Granatello lava (Di Vito et al., 1998; Bellucci, 1998) lies above the pyroclastic flow and lahar deposits of the A.D. 79 eruption which are found into numerous wells drilled in the immediate surroundings of Portici and also in outcrops along the coast below and immediately west of Villa D'Elboeuf (Fig. 2, n.1). The coast to the south of Granatello maintains a low flat profile, partially filled by the 1631 nuües ardentes deposits up to Scogli della Favorita, and then rises to the promontory of the Scogli della Scala (now called Punta Quattro Venti), where a single large lava crops out (P18) in an over 150 m long sea cliff . Archaeomagnetic dating indicates that the Granatello and Scogli della Scala lavas are a single huge lava ("Granatello" lava), which was erupted between the 9th and 10th century. A lava attributable to the 1139 eruption (P11, P29) overlies the inland portion (P30) of the Granatello lava (Fig. 9). Both these lavas are intersected by the Naples-Salerno highway. Sector II The lava (P10) forms the from which the breakwater (Fig. 4) extends outwards. stretch of coastline (around promontory called Calastro, of the port of Torre del Greco This lava crops out on a long 250 m), with a mean thickness

Appendix 1: New radiocarbon dating
For internal control of the consistency of the archaeomagnetic data, the charcoal sample OZF467 found within the scoria fallout underlying P23 (Fig. 7c), was radiocarbon-dated at the Accelerator Mass Spectrometry (AMS) facility at ANSTO (Lawson et al., 2000). AMS makes use of a nuclear particle accelerator to detect single 14 C atoms in the overwhelming presence of stable 12C and 13C atoms. Compared to the traditional counting technique, an AMS measurement requires smaller mass samples (typically 1 mg of carbon is sufficient) and shorter measurement times. The standard AAA method (Mook & Streurman, 1983) was employed to remove contamination. The pre-treated sample was converted to CO2 by combustion and graphitised (Hua et al, 2001). The graphite was loaded into an aluminium sample holder for


721

of 7­8 m. The locality name Calastro is of Greek origin and the Calastro settlement precedes the founding of Torre del Greco which occurred in the 14th century on the site of one of the coastal towers built to defend the coast during the domination of the Suevians, in the 13th century (Orlandi, 1631). Since this tower was situated at eight Roman miles from Naples, the settlement was initially called Turris Octava. During their first extensive development, the two villages of Sola (south of the residential area, where the cemetery is now located) and Calastro were reported to be damaged by the eruption of 1037 (De Seta et al., 1980). In the 16th century, first a fortress, and then a mill were built on this lava because of the presence of a fresh water spring nearby. In spite of the very ancient origin of the name, the lava of Calastro is attributed to the 10th century based on our archaeomagnetic result. Sample P26 from this sector comes from a lava attributed on the CNR (1986) geological map to modern activity but, in fact, pertains to the 1006­07 eruption, based on our archaeomagnetic data. Sector III An outcrop (P15) of the same lava exposed on the seacliff under the tower (Fig. 5) is located in the stretch of the coastal road leading from Torre del Greco towards Torre Annunziata, just inland of Torre Bassano. Based on our archaeomagnetic measurements, the lava under Torre Bassano is attributed to the A.D. 968 eruption. Sector IV Near the coast, a few kilometers west of Torre Annunziata, two directly overlapping lavas (P01, P02) crop out in the large quarry of Villa Inglese (Fig. 6). On the northeastern side of the quarry they are overlaid by the pyroclastic flow and lahar deposits of the 1631 eruption. These, in turn, are covered by the 1760 lava. A third sample (P20) was collected from the uppermost of these two lavas, on the northwestern side of the quarry. Here, the upper lava is divided into two lobes, giving the false impression of the existence of a third lava, as erroneously stated by Gialanella et al. (1998). The "medieval", stratigraphically highest lava is of whitish colour due to weathering, very compact, and, because of its good physical and mechanical characteristics, has been quarried widely since the early 19th century. It is locally known as "il Villa Inglese" lava (Formicola et al., 1990; Penta & Del Vecchio, 1963). The "il Villa Inglese" lava is located directly below the second "medieval" lava, which was correlated by Rolandi & Russo (1989) with the seaward lava on which the Torre Scassata tower was constructed (Fig. 6, n.5). Archaeomagnetic data (P03) confirm such a correlation. There is no geologic evidence supporting Rolandi & Russo's (1989) contention that both medieval lavas cropping out in the quarry are present under the Torre Scassata, as only

one lava crops out widely below the tower and along the broad beach beyond. The two lavas of the Villa Inglese occupy an important position in the stratigraphy of the pre-1631 deposits, particularly in the so-called "medieval" lavas. The Villa Inglese lavas have been taken into account in practically all the studies on the Vesuvius' eruptive history between A.D. 79 and 1631 (Burri & Di Girolamo, 1975; Delibrias et al., 1979; Principe et al., 1987; Rolandi & Russo, 1989). We found that the two Villa Inglese lavas are archaeomagnetically attributable to the A.D. 999 and 1037 eruptions. This last eruption also produced the lava sampled at Lido Incantesimo (P19). In the quarry of Villa Inglese, below the two aforesaid lavas, another lava is present, though it is not visible at the present time (Principe et al., 1987; Rolandi & Russo, 1989; Formicola et al., 1990). On the basis of the previous discussion, the coastline morphology, the precise cartography by Johnston-Lavis (1891a) of the lava units cropping out at that time and their stratigraphic relationship, we can hypothesize that this third and oldest lava is the easternmost edge of the large lava that formed Le Mortelle promontory (P16, n.4 in Fig. 6). It is attributed to the 9th century based on the archaeomagnetic datum. The two Villa Inglese lavas extend along the coastline to the south, overlying the southern border of this large lava. Subsequently, the 1631 nuüe ardentes deposits filled the topography, leaving the upper surface of the Le Mortelle lava exposed in some morphological windows, on which the foundations of Masseria Donna Chiara and Villa Prota were laid (n.2 and n.3 in Fig. 7a, respectively). Rolandi & Russo (1989) attribute the lava of Masseria Donna Chiara to the eruption of 1631 and insert it between the 1631 pyroclastic deposits and the 1760 lava in the stratigraphy of the Villa Inglese quarry. The Masseria Donna Chiara lava is not present in the quarry, but occurs west of the quarry area, beneath the farm of the same name, the existence of which is documented since at least 1645 (Raimondo, 1985 and 1994). This lava (P27) is overlain by a characteristic bed of normally graded ochreous lapilli belonging to the "middle-age" tephra sequence (Principe et al., 1987). The territory surrounding the Masseria Donna Chiara was widely, though not entirely, covered by the 1631 nuües ardentes (Rosi et al., 1993), which spared the nearby Villa Prota of which the construction dates back to the 16th century (De Seta et al., 1980). Although we did not obtain a reliable archaeomagnetic age for the lava of Masseria Donna Chiara, it is older than 1631 and might probably belong to Le Mortelle lava (9th century, see above). The two scoria cones of Viulo and Fossamonaca are located in the northernmost portion of this sector.The Fossamonaca cone (P09) is a scoria cone open seaward, from where lava was seen to escape (Sorrentino, 1734). Viulo is another, smaller-sized scoria cone located just south of Fossamonaca and that was probably constructed soon thereafter. Due to the thick undergrowth covering the Viulo vent, only the Fossamonaca cone has been suitable for archaeomagnetic sampling. Archaeomagnetic


722

dating allows us to attribute it to the eruption of A.D. 999, as is the lower Villa Inglese lava (Fig. 9). Sector V About 1.5 km north of Torre Annunziata, the top of a small hill (Fig. 7a) emerges from the 1631 pyroclastic deposits surrounding it. This is all that remains of the vent (Masseria Bosco del Monaco cone) for a large lava flow that expands towards the sea (Johnston-Lavis, 1891a), and the eastern edge of which crops out in the construction site for the new hospital of Torre Annunziata (P23). The 1631 eruption deposits are located above this lava (Fig. 7c), which is intercepted by a series of small quarries (Fig. 7a), up to the port area of Torre Annunziata, where it crops out extensively along the Litoranea Marconi road, between the thermal baths of Terme Nunziante and Palazzo Monteleone (Fig. 7a and 7b). The lava outcrops along the Naples-Salerno highway, near the Torre Annunziata-Nord exit, also belong to this lava flow, and were sampled for this work at the soccer stadium of Torre Annunziata (P12). The westernmost branch of this lava crops out opposite of the seaside garden of Villa Filangeri (n.3 in Fig. 7). The age of the large Masseria Bosco del Monaco lava is early 10th century, as indicated by our archaeomagnetic measurements. This is one of the eruptions which was unnoticed by historians. Another large lava dominates the morphology of the Vesuvius' coastline near Torre Annunziata. This is the rock-cliff lava of Scogli di Prota (P14, n.1 in Fig. 7a), known in the literature as the Oncino lava because of the 16th-century fortress called Forte dell'Oncino situated on the eastern side of the Scogli di Prota promontory (n.2 in Fig. 7a). In the stratigraphy proposed by Rolandi & Russo (1989), the Oncino lava is positioned below the lava of the Litoranea Marconi, which is attributed by these authors to 1631. However, no relation is recognizable in the field between these two lavas, just as there is no basis for attributing the Litoranea Marconi lavas to 1631. In the stratigraphic section of Cantiere La Perla (near Terme Nunziante), this lava merely occurs at the top of the succession. Based on our archaeomagnetic measurements the Oncino lava is attributable to the 1006­7 eruption (Table 7). Sector VI Two lavas, situated stratigraphically between the deposits of the A.D. 472 and the 1631 eruptions, were sampled in this sector at Terrioni quarry (P06) and at the village of Passanti (P17). They are attributed to the eruption of the 9th century (P06) and to the 1139 event (P17), based on our archaeomagnetic measurements. These lava flows have first been channelled into the narrow gorges of the easternmost, low-lying sector of Mount Somma, and have expanded downhill. As this area is outside the main

populated areas, no historical information is available for these lavas.

References
Alberti FL (1568) Descrizione di tutta l'Italia. Venezia, L. Degli Avanzi, p 504 Alfano GB (1924) Le eruzioni del Vesuvio tra il 79 ed il 1631 (studio bibliografico). Napoli, Scuola tipografica Pontificia per i figli dei carcerati, p58 Andronico D, Calderoni G, Cioni R, Sbrana A, Sulpizio R, Santacroce R (1995) Geologic map of Somma-Vesuvius Volcano. Per Mineral 64: 77­78 Andronico D, Cioni R, Marianelli P, Santacroce R, Sbrana A, Sulpizio R, (1996a) Introduction to Somma-Vesuvius. In "Vesuvius Decade Volcano" Workshop Handbook; Naples, September 17­22, p. 49 Andronico D, Cioni R, Sulpizio R, (1996b). General stratigraphy of the past 19,000 yrs at Somma-Vesuvius. In "Vesuvius Decade Volcano" Workshop Handbook; Naples, September 17­22, p.21. ArnÐ V, Principe C, Rosi M, Santacroce R, Sbrana A, Sheridan MF (1987) Eruptive history. In Somma-Vesuvius, ed. R. Santacroce, CNR Quad Ric Sci 114: 53­103 Arrighi S, Principe C, Rosi M (2001) Violent strombolian and subplinian eruptions at Vesuvius during the post 1631 activity. Bull Volcanol 63: 126­150 Audunsson H, Levi S (1989) Drilling-induced remanent magnetization in basalt drill cores. Geophys J Int 98:613­622 Baratta M (1897) Il Vesuvio e le sue eruzioni. Dall'anno 79 a.C. al 1869. "Dante Alighieri", Roma, p. 202. Barberi F, Principe C, Rosi M, Santacroce R (1995) Scenario eruttivo al Vesuvio nel caso di riattivazione a medio-breve termine. Aggiornamento al 20 gennaio 1995. Internal report National Group for Volcanology, Italian National Researches Council (GNV) ­ Civil Defence, p. 14 Belkin HE, Kilburn RJ, De Vivo B, (1993) Sampling and major element chemistry of the recent (A.D. 1631­1944) Vesuvius activity. J Volcanol Geotherm Res 58: 273­290 Belkin HE, De Vivo B, Torok K, Webster JD (1998) Pre-eruptive volatile content, melt-inclusion chemistry, and microthermometry of interplinian Vesuvius lavas (pre-A.D. 1631). J Volcanol Geotherm Res 82: 79­95 Bellucci F (1998) Nuove conoscenze stratigrafiche sui depositi effusivi ed esplosivi nel sottosuolo dell'area del SommaVesuvio. Boll Soc Geol It 117: 385­405 Braccini GC (1632) Dell'incendio fattosi sul Vesuvio a XVI di Dicembre MDCXXXI e delle sue cause ed effetti, con la narrazione di quanto seguito in esso per tutto marzo 1632. E con la storia di tutti gli altri incendi, nel medesimo monte avvenuti. Napoli, Secondino Roncagliolo, p. 104 Bucur I (1994) The direction of the terrestrial magnetic field in France during the last 21 centuries. Recent progress. Physics of the Earth and Planetary Interiors 87: 95­109 Burri C, Di Girolamo P (1975) Contributo alla conoscenza delle lave della grande eruzione del Vesuvio del 1631. Rend. SIMP, 30: 705­739 Cafarella L, De Santis A, Meloni A (1992) Secular variation in Italy from historical geomagnetic field measurements. Physics of the Earth and Planetary Interiors, 73: 206­221 Carracedo JC, Principe C, Rosi M, Soler V (1993) Time correlation by paleomagnetism of the 1631 eruption of Mount Vesuvius. Volcanological and volcanic hazard implications. J Volcanol Geotherm Res 58: 203­209 Carta S, Figari R, Sartoris G, Sassi E, Scandone R (1981) A statistical model for Vesuvius and its volcanological implications. Bull Volcanol 44(2): 130­151 Cerbai I, Principe C (1996) Bibliography of Historic Activity on Italian Volcanoes. Institut of Geochronology and Isotope Ge-


723 ology, Italian National Researche Council (IGGI), Internal report n. 6/96, p. 686 CNR ­ Progetto Finalizzato Geodinamica, 1986. Geological map of Somma-Vesuvius volcanic complex. Roma, L. Salomone Delibrias G, Di Paola GM, Rosi M, Santacroce R (1979) La storia eruttiva del complesso vulcanico Somma Vesuvio ricostruita dalle successioni piroclastiche del Monte Somma. Rend. SIMP, 35(1): 411­438 De Seta C, Di Mauro L, Perone M (1980) Ville vesuviane. Milano, Rusconi Immagini, p. 370 Di Girolamo P (1970) Rilevamento petrografico-stratigrafico lungo il margine S-W del Vesuvio. Rend. SIMP, 26: 77­108 Di Vito MA, Sulpizio R, Zanchetta G, Calderoni G (1998) The geology of the South Western Slopes of Somma-Vesuvius, Italy, as inferred by borehole stratigraphies and cores. Acta Vulcanologica, 10 (2): 383­393 Figliuolo B, Marturano A (1998) The eruptions of Vesuvius from the 7th to the 12th centuries. In Volcanoes and History, ed. Nicoletta Morello, Genova, Brigati, 133­156 Fisher RA (1953) Dispersion on a sphere. Proc R Soc London 217 (A): 295­305 Formicola F, Pappalardo U, Rolandi G, Russo F (1990) Archeologia, geologia e vulcanologia nel territorio di Torre del Greco: tre discipline a confronto. PACT, 25, 125­182 Furchheim F (1897) Bibliografia del Vesuvio compilata e corredata di note critiche estratte dai pi autorevoli scrittori vesuviani da Federigo Furchheim. Napoli, p.297 Gasparini P, Musella S (1991) Un viaggio al Vesuvio. Napoli, Liguori Editore, p. 307 Genevey A, Gallet Y, Boudon G (2002) Secular variation study from non-welded pyroclastic deposits from Montagne Pelüe volcano, Martinique (West Indies). EPSL, 201: 369­382 Gialanella P, Incoronato A, Russo F, Nigro G (1993) Magnetic stratigraphy of Vesuvius products. I - 1631 lavas. J. Volcanol. Geotherm. Res., 58, 211­215 Gialanella P, Incoronato A, Russo F, Sarno P, Di Martino A (1998) Magnetic stratigraphy of Vesuvius products. II--Medieval lavas. Quaternary International, 47­48, 135­138 Girelli A (1994) Napoli dalle origini a Carlo D'AngiÐ. Milano, Fenice 2000, p. 95 Heiken G (1999) Will Vesuvius erupt? Three million people need to know. Science, 286: 1685­87 Hoblitt RP, Reynolds RL, Larson EE (1985) Suitability of nonwelded pyroclastic-flow deposits for studies of secular variation: A test based on deposits emplaced at Mount St. Helens, Washington, 1n 1980. Geology, 13: 242­245 Hoye GS (1981) Archeomagnetic secular variation record of Mount Vesuvius. Nature, 291: 216­217 Hua Q, Jacobsen GE, Zoppi U, Lawson EM, Williams AA, Smith AM and McGann MJ (2001) Progress in radiocarbon target preparation at the ANTARES AMS centre, Radiocarbon 43(2A): 275­282 Incoronato A _(1996) Magnetic stratigraphy in volcanic area. The experience at Vesuvius. In Palaeomagnetism and tectonics of the Mediterranean Region, edited by A.Morris and D.H. Tarling, Geol Soc Spec Publ 105: 367­371 Incoronato A (1999) Comment on the paper "Some considerations on the state of Vesuvius in the Middle Ages and the precursors of the 1631 eruption" by A. Nazzaro. Annali di Geofis., 42: 137­140 Incoronato A, Angelino A, Romano R, Ferrante A, Sauna R, Vanacore G, Secchione C (2002) Retrieving geomagnetic secular variations from lava flows: evidence from Mount Arso, Etna and Vesuvius (southern Italy). Geophys J Int 149: 724­ 730 Johnston-Lavis HJ(1884) The Geology of Monte Somma and Vesuvius, being a study in Vulcanology. Q J Geol Soc London 40: 35­149 Johnston-Lavis HJ (1891a) Geological map of Monte Somma and Vesuvius, 1:10.000, constructed during the years 1880­1888. London, George Philip & Son Johnston-Lavis HJ (1891b) A short and concise account of the eruptive phenomena & geology of Monte Somma and Vesuvius in explanation of the great geological map of that volcano. London George Philip & Son, 32, p. 8 Joron JL, Metrich N, Rosi M, Santacroce R, Sbrana A (1987) Chemistry and Petrography. In Somma-Vesuvius, ed. R. Santacroce, CNR Quad. Ric. Sci., 114, 105­171 Lauer JP (1978) Creation d'aimantations remanentes de sciage et forage au cours de la prparation d'echantillons de roche destins Þ une tude palomagntique. C R Acad Sci Paris, 287(D): 1­4 Lawson EM, Elliott G, Fallon J, Fink D, Hotchkis MAC, Hua Q, Jacobsen GE, Lee P, Smith AM, Tuniz C, and Zoppi U (2000) AMS at ANTARES ­ The first 10 years, Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms., 172: 95­99 Le Goff M(1975) Inductomtr Þ rotation continue pour la mesure des faibles aimantations rmanentes et induites en magntisme des roches. Master thesis (in french), CNAM, Paris Le Hon S (1865) Histoire complüte de la grande üruption du Vesuve de 1631, avec la carte au 1/25.000 de toutes les laves de ce volcan depuis le seizieme siecle jusqu'a aujourd'ui. Bruxelles, M. Hayez, pp.64 et Bull Acad Sci Lett Beaux Arts Belg 20: 483­538 Le Hon HS (1866) Carte Topographique des laves du Vesuve, a l'echelle de 1/25.000, 1631­1861, avec la coupe geologique du Rivage Napolitan. Bruxelles et Leipzig, C. Muquardt; Naples, A. Detken Leone A (1514) De Nola opusculum. Venetiis, Rubri Joannis, p. 210 Lirer, L., Vinci, A., Alberico, I., Gifuni, T., Bellucci, F., Petrosino, P., Tinteri, R., 2001. Occurrence of inter-eruption debris flow and hyperconcentrated flood-flos deposits on Vesuvio volcano, Italy. Sedimentary Geol 139: 151­167 Masculo GB (1633) De incendio Vesuvii excitato XVLJ Kal. Ianuar anno trigesimo seculi decimo septimi, libri X cum cronologia superiorum incendiorum ephemeride ultimi. Napoli, Secondino Roncagliolo, p. 350 Mastrolorenzo G, Palladino DM, Vecchio G, Taddeucci J (2002) The 472 A.D. Pollena eruption of Somma-Vesuvius (Italy) and its environmental impact at the end of the Roman Empire. J Volcanol Geotherm Res 113:19­36 Melilli A (1996) Tefrostratigrafia dell'attivitÞ eruttiva del Vesuvio compresa fra il 1631 ed il 1944: caratteri fenomenologici. Unpublished thesis, p.330 Meo F, Russo S (1997) Torre Annunziata, uomini e fatti. Torre Annunziata, D'Amelio, p.223 Mook WG, Streurman HJ (1983) Physical and chemical aspects of radiocarbon dating. The Proceedings of the First International Symposium on 14C and Archaeology. In: Mook, W.G., Waterbolk, H.T. (Eds.), PACT, 8: 31­55 Mormile G (1632) L'incendio del Vesuvio, e delle straggi, e rovine, che ha fatto n tempi antichi e moderni, infine a 3 di marzo 1631, con nota di tutte le relazioni stampate fino ad oggi del Vesuvio, raccolte da Vincenzo Bove. Napoli, Egidio Longo, p. 48 Nazzaro A (1998) Some considerations on the state of Vesuvius in the Middle Ages and the precursors of the 1631 eruption. Annali di Geofisica 41(4): 555­565 Niklaus ThR, Bonani G, Simonius M, Suter M, WÆlfli W (1992) CalibETH: an interactive computer program for the calibration of radiocarbon dates. Radiocarbon 34: 482­492 Orlandi G _(1631) Dell'incendio del Monte di Somma. Compita relazione e di quanto succeduto insino ad hoggi. Napoli, Lazzaro Scoriggio, p. 15 Paiotti A (2000) Archeomagnetismo di lave vesuviane: l'attivitÞ tra il 79 ed il 1631. Unpublished thesis, p. 126 Penta F, Del Vecchio S (1963) Lave vesuviane dei principali centri estrattivi. Napoli, Fondazione politecnica del mezzogiorno d'Italia, p. 108 Pighius SV (1587) Hercules Prodicius seu Principiis juventutis vita et peregrinatio. Antuerpiae, Cristoforo Plautinus, p. 653


724 Preusse P (1931) Considerations sur la forme du Vesuve et son activitü dans l'antiquitü. Bull Volcanol 27­30, 11­23 Principe C (1979) Le eruzioni storiche del Vesuvio: riesame critico, studio petrologico dei prodotti e implicazioni vulcanologiche. Unpublished Thesis, University os Pisa, p. 123 Principe C (2003) Cronologia dell'attivitÞ del Vesuvio fra il 79 A.D. ed il 1631. Submitted to Archivio storico delle province napoletane Principe C, Rosi M, Santacroce R, Sbrana A (1987) Explanatory notes to the geological map. In Somma-Vesuvius, ed R. Santacroce, CNR Quad. Ric. Sci., 114, 11­52 Principe C, Giosa P, Cerbai I, Crocetti S, Franceschini F, Marini L, Gambardella B, Buettner A, Rosi M (2003) A dynamic model for the 1631 Vesuvius eruption. Geophysical Research Abstracts, Vol.5, 05432 Raia F, Webster JD, De Vivo B (2000) Pre-eruptive volatiles contents of Vesuvius magmas : constraints on eruptive history and behavior. I ­ The medieval and modern interplinian sctivities. Eur J Mineral 12: 179­193 Raimondo R (1994) Itinerari torresi e cronistoria del Vesuvio. III ed., Ercolano, La Buona Stampa, p. 505 Raimondo R (1985) Uomini e fatti dell'antica Torre del Greco, opera postuma. Ercolano, La Buona Stampa, p. 587 Rolandi G, Russo F (1989) Contributo alla conoscenza dell'attivitÞ storica del Vesuvio: dati stratigrafici e vulcanologici nel settore meridionale tra Torre del Greco, localitÞ Villa Inglese e Torre Annunziata (Campania). Boll Soc Geol It 108: 521­536 Rolandi G, Barrella AM, Borrelli A (1993) The 1631 eruption of Vesuvius. J Volcanol Geotherm Res 58: 183­201 Rolandi G, Petrosino P, Geehin J (1998). The interplinian activity at Somma-Vesuvius in the last 3500 years. J Volcanol Geotherm Res 82: 19­52 Rosi M, Santacroce R (1983) The AD 472 "Pollena" eruption: volcanologic and petrological data for this poorly-known, Plinian-type event at Vesuvius. J Volcanol Geotherm Res 17: 249­271 Rosi M, Principe C, Vecci R (1993) The 1631 eruption of Vesuvius reconstructed from the review of chronicles and study of deposits. J Vulcanol Geotherm Res 58: 151­182 Schottus A (1655) Itinerarium italiae. Amstedolami, Jansson (cited by Alfano,1924, pg. 36) Sorrentino I (1734) Istoria del Monte Vesuvio. Divisata in due libri. Napoli, Giuseppe Severini, p. 228 Stothers BR, Rampino RM (1983) Volcanic Eruptions in the Mediterranean before A.D. 630 from written and archaeological sources. J Geophys Res 88: 6357­6371 Stuiver M (1983) Business meeting: international agreements and the use of the new oxalic acid standard. Radiocarbon, 25(2): 793­795 Stuiver M, Reimer PJ, Bard E, Beck JW, Burr GS, Hughen KA, Kromer B, McCormac G, Van der Plicht J, Spurk M (1998) INTCL98 radiocarbon age calibration, 24,000­0 cal BP. Radiocarbon, 40(3): 1041­1083 Tanguy JC, Bucur I, Thompson JFC (1985) Geomagnetic secular variation in Sicily and revised ages of historic lavas from Mount Etna. Nature, 318: 453­455 Tanguy JC, Le Goff M, Chillemi V, Paliotti A, Principe C, La Delfa S, Patan G (1999) Secular variations of the geomagnetic field direction recorded in lavas from Etna and Vesuvius during the last two millennia. C R Acad Sci Paris Earth & Planetary Sciences 329: 557­564 Tanguy JC, Le Goff M, Principe C, Arrighi S, Chillemi V, Paiotti A, La Delfa S, Patan G (2003) Archaeomagnetic dating of Mediterranean volcanics of the last 2100 years. Validity and limits. EPSL 6640(2003): 1­14 Thellier E (1981) Sur la direction du champ magnetique terrestre, en France, durant les deux derniers millenaires. Phys Earth Planet Inter 24: 89­132 Vittozzi P, Gasparini P (1965) Datazioni di lave vesuviane. Geof. Meteor., XIV, 3­4