Word Search

9 pages
6 views

Archaeomagnetic dating of a vitrified wall at the Late Bronze Age settlement of Misericordia (Serpa, Portugal)

of 9
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Share
Description
The age of cooling of a vitrified wall at the Late Bronze/Second Iron Age settlement of Misericordia (Serpa, Portugal) has been determined using archaeomagnetic dating. Vitrification occurred in the Late Bronze Age (842–652 BC), in agreement with
Transcript
  Archaeomagnetic dating of a vitrified wall at the Late BronzeAge settlement of Misericordia (Serpa, Portugal) Gianluca Catanzariti  a, *, Gregg McIntosh  a , Anto´nio M. Monge Soares  b ,Enrique Dı´az-Martı´nez  c , Peter Kresten  d , M.L. Osete  a a Grupo de Paleomagnetismo, UCM, Universidad Complutense de Madrid, 28040 Madrid, Spain b  ITN, Instituto Tecnolo´gico e Nuclear, Sacave´m (Portugal) c Geological Survey of Spain (IGME), Calera 1, 28760 Tres Cantos, Madrid, Spain d  Kresten GeoData, To¨va¨dersgatan 18, S-754 31 Uppsala, Sweden Received 5 February 2007; received in revised form 21 September 2007; accepted 8 October 2007 Abstract The age of cooling of a vitrified wall at the Late Bronze/Second Iron Age settlement of Misericordia (Serpa, Portugal) has been determinedusing archaeomagnetic dating. Vitrification occurred in the Late Bronze Age (842 e 652 BC), in agreement with archaeological constraints basedon the style of the potteries recovered at the site. This demonstrates the suitability of vitrified structures for archaeomagnetic dating and thepotential for developing absolute chronologies for similar structures in Iberia and across Europe as a whole. Magnetite, low Ti-content titano-magnetite and, to a lesser extent, metallic Fe carry the archaeomagnetic signal, thus representing phases formed during the heating event. NativeFe was preserved due to it being isolated within the glassy matrix. The vitrified structure underwent at least one strong heating event which led topartial melting of the rocks used in its construction. Both microprobe and archaeomagnetic data support a single heating event, although multipleheating to temperatures  > 600 e 800   C cannot be excluded on archaeomagnetic grounds.   2007 Elsevier Ltd. All rights reserved.  Keywords:  Archaeomagnetic dating; Bronze Age; Vitrified structure; Portugal; Metallic iron 1. Introduction Throughout Europe, more than 200 hill-forts and settle-ments show evidence of having been subjected to intensiveheating, leading to vitrified or calcined structures (Kresten,2004). In silicate rocks (e.g., granite, gneiss, basalt), vitrifica-tion occurs through partial or complete melting of the primaryminerals and the formation of a glassy phase. Typical meltingtemperatures are between 1050 and 1235   C, with some crys-tallisation occurring down to temperatures of about 650   C. Incarbonate rocks (e.g., limestone and dolomite), heating resultsin calcination d the formation of burnt lime d which requireslower temperatures (  800   C).Vitrification of timber-laced or timber-clad ramparts maybe constructive, producing a more solid rampart, or destructiveif it was set on fire and burnt down. Alternatively, vitrificationmay have been incidental, due to intense, localised heating.For example, metallurgical processes reach temperatures suffi-ciently high to cause vitrification of the furnace walls, as wellas producing large amounts of vitrified material (scoria andslag) with high percentage of metal components.Late Bronze Age (12th to 8th century BC) and Iron Age(8th to 2nd century BC) settlements can be found throughoutthe Iberian Peninsula, some of which present vitrified struc-tures (Dı´az-Martı´nez and Soares, 2004). Vitrification of forti-fied settlements in Portugal was first identified at MonteNovo, a Late Bronze and Iron Age settlement near E´vora * Corresponding author.  E-mail addresses:  gcatanza@fis.ucm.es (G. Catanzariti), gregc@fis.ucm.es(G. McIntosh), amsoares@itn.pt (A.M. Monge Soares), e.diaz@igme.es (E. Dı´az-Martı´nez), geodata@hotmail.com (P. Kresten).0305-4403/$ - see front matter    2007 Elsevier Ltd. All rights reserved.doi:10.1016/j.jas.2007.10.004Journal of Archaeological Science 35 (2008) 1399 e 1407http://www.elsevier.com/locate/jas  (Burgess et al., 1999), and there are also descriptions of vitri-fication at a Neolithic dolmen near Viseu (Abrunhosa et al.,1995). Three other sites with vitrified walls are known inPortugal; the Late Bronze Age settlements of Passo Alto(Dı´az-Martı´nez et al., 2005) and Cerro das Alminhas-Sarnadi-nha (Dı´az-Martı´nez and Soares, 2004), and the Iron Age settle-ment of Castelo Velho de Safara (Soares, 2001). In Spain, twoother Iron Age settlements with vitrified walls are alsoknown d Pico del Castillo near El Gasco (Dı´az-Martı´nez,2004a,b) and Pico del Pozo de los Moros near Villasrubias(Dı´az-Martı´nez and Soares, 2004).The dating of vitrified structures has been problematical.Whilst archaeological constraints (e.g. ceramic remains) canbe used to date the occupation of the site, the application of physical methods has been difficult. In particular, the use of thermoluminescence (TL) to date the vitrification event maybe adversely affected by the high temperatures reached duringvitrification (Kresten et al., 2003). The same high temperaturesimply the complete destruction of organic material duringcombustion, limiting the application of radiocarbon datingfor the same event.Gentles and Tarling (1987) describe a Scottish vitrified dunthat carried a stable remanent magnetisation, from which theyestablished the thermal history of the dun (contemporaneousheating/vitrification of the different parts of the dun). In addi-tion they were able to date the firing of the dun using a prelim-inary reference archaeomagnetic curve for pre-Roman toRoman period in the UK (placing the firing between the late1st century BC and the late 1st century AD). Since then therehas been little new work carried out on such structures, prob-ably due to the lack of well-defined reference archaeomagneticdating curves.Vitrified material is a good candidate for archaeomagneticdating, since it may acquire a thermoremanent magnetisation(TRM) on cooling that can be used to date the age of cooling.For material found in situ, the direction of the TRM may bedated by comparison with a regional reference curve of the Fig. 1. (a) Location of the Misericordia site, and (b) the studied vitrified structure.Fig. 2. Cross-section of a hand specimen from Misericordia, length about20 cm. To the right, red-burnt schist (here sample MIS02UN) that towardsthe left turns into completely vitrified material (sample MIS01A).Fig. 3. Back-scattered electron micrograph of sample MIS01UN showing me-tallic iron (rounded or oblong, white) and iron oxides (light grey dots) withina glassy matrix. Magnification according to scale bar at lower left.1400  G. Catanzariti et al. / Journal of Archaeological Science 35 (2008) 1399 e 1407   geomagnetic secular variation. Such a curve is now availablefor the Iberian Peninsula (Go´mez-Paccard et al., 2006), appli-cable over the time interval between 850 BC and 1960 AD.This study presents the results of an archaeomagnetic investi-gation of a vitrified wall from the Misericordia archaeologicalsite in Portugal. Detailed archaeomagnetic and rock magneticstudies have been carried out in order to determine the proper-ties of natural remanent magnetisation (NRM) and the natureand stability of the phases carrying the archaeomagnetic sig-nal. Supplementary microprobe analyses, together with rock magnetic studies, have been used to characterise the vitrifiedmaterial and establish its thermal history. 2. The proto-historic settlement of Misericordia(Serpa, Portugal) The Misericordia archaeological site (Lat 37.9  N, Long352.4  E) is located close the city of Serpa (southern Portugal,Fig. 1a), on the left bank of the River Guadiana. Good naturaldefences at the river side of the settlement are complementedby a rampart along the easiest southern approach of the settle-ment. Based on several artefacts, namely pottery, collectedduring several archaeological surveys, the site was occupiedduring two periods separated by a hiatus: the Late BronzeAge (12th to 8th century BC) and the Second Iron Age (5th/ 4th to 2nd century BC) (Soares, 1996). Characteristic col-lected artefacts of these periods are pattern-burnished potteryfor the second phase of the older period (10th to 8th centuryBC d Cardoso, 1995; Gibson et al., 1998) and fragments of ce-ramic vessels decorated with stamps or Iberian-style redpainted bands from the Iron Age period.Vitrified material (scoriaceous and pumiceous debris), met-allurgical debris and stone structures have been identified atthe settlement. The stone structures observed at the settlement,including the southern rampart, are made up of schist andmetagraywacke blocks from the local substrate rock. A wellpreserved vitrified wall, 5.5 m long and 2 m high, with a largernon-vitrified stone base, is located at the western side of thesettlement. The wall may have been part of a furnace or a de-fensive tower. Other vitrified structures can be observed in theexterior of the settlement, on its southern side, which may berelated to metallurgical activities. Many loose blocks of rock found close to these vitrified structures display evidence forpartial melting and clast welding.Scanning electron microscope (SEM) and microprobe anal-ysis of vitrified material show that they were subjected to hightemperatures and partial melting, while the analyses of a slagshows that it probably resulted from the smelting of iron ores(Soares, 1996). A stone mould discovered at the settlementsuggests that other metallurgical activities not related toiron, but probably with copper, were also carried out. 3. Sampling and methods The study was carried out on the vitrified wall (Fig. 1b) de-scribed by Soares (1996). The wall, trending approximatelyNE e SW, consists of partially melted and vitrified blocks rang-ing between 30 and 50 cm in size. Substrate lithologies found Table 1Representative electron microprobe analyses (weight-%) of silicate and oxidephasesArea no. SampleMIS01UN MIS02UN1a1 1a3 1a4 1a3SiO 2  7.90 43.94 70.52 61.47TiO 2  89.67 1.02 0.14 0.68Al 2 O 3  2.65 35.85 14.42 21.34FeO tot. 0.43 11.70 1.92 4.22MnO 0.00 0.08 0.03 0.02MgO 0.00 2.96 0.34 1.02CaO 0.12 0.18 0.28 0.13Na 2 O 0.25 0.65 1.03 0.59K  2 O 0.79 4.24 6.17 5.20P 2 O 5  0.08 0.00 0.04 0.05Cr 2 O 3  0.00 0.04 0.00 0.02CuO 0.00 0.00 0.00 0.00Sum 101.90 100.70 94.89 94.74Phase ox a-gl s-gl wrMIS01UN, vitrified material; MIS02UN, schist. Phases: a-gl, alumina-richglass; ox, oxide mineral (rutile?); s-gl, silica-rich glass; wr, tentative analysisof red-burnt schist.Table 2Representative electron microprobe analyses (weight-%) of iron droplets in the vitrified materialArea no. SampleMIS1-13 MIS1-08 MIS01UN1a1 2a1 1a1 1a2 1a5 1a6 2a2Si 0.01 0.02 0.01  e  0.735 0.758 4.922Fe 99.01 95.98 92.14 84.54 91.153 90.885 75.392Mn  e e e e  0.001  e  0.002P 0.24 0.33 2.07 9.73 0.167 0.14 7.319Cr  e e e e e e e Cu 0.01  e  0.17 0.07 0.929 1.011  e S  e e  0.07 0.10  e e  0.203Co 0.07 0.07 0.39 0.29 0.356 0.548 0.087Ni 0.04 0.04 1.75 2.58 2.585 2.086 2.216Sum 99.38 96.52 96.60 97.31 95.94 95.43 90.14Size 5  m m 10 m m 25  m m 25  m m 2  m m 2  m m 3  m m1401 G. Catanzariti et al. / Journal of Archaeological Science 35 (2008) 1399 e 1407   at the site and used as construction material include schist andmetagraywacke. The massive size of the structure, togetherwith its good state of preservation (i.e. no evidence for differ-ential movements, no cracked or disjointed blocks), suggeststhat it remains in situ. Twelve independently oriented archae-omagnetic samples were taken using a portable rock drill,from seven different blocks distributed across the whole of the exposed wall. They were oriented using both solar andmagnetic compasses. In addition, some (unoriented) handsamples were taken for complementary rock magnetic and mi-croprobe analyses.Microprobe studies were undertaken at the Institute of EarthSciences,UniversityofUppsala,usingaCamecaSX50electronmicroprobe. Wavelength-dispersive methods were used em-ploying a variety of natural and synthetic standards, usingacurrentof20 kVandcounting timesfrom5 e 30 s.Thearchae-omagnetic study was conducted in the Palaeomagnetism Labo-ratory of the Complutense University, Madrid. A total of 17specimens (diameter ¼ 2.5 cm, height ¼ 2.5 cm) were pre-pared from the 12 samples. The NRM and low field magneticsusceptibility (  K  ) of each specimen was measured using anAGICO JR5 spinner magnetometer and an AGICO KLY3susceptibility meter respectively. Stepwise thermal (TH) de-magnetisation of NRM was carried out using a Scho¨nsted In-struments TSD-1 thermal demagnetiser, in 30 e 60   C steps upto 580 e 800   C. Stepwise alternating field (AF) demagnetisa-tion was conducted using a Scho¨nsted Instruments GSD-5 tum-bling demagnetiser, in 2 e 20 mT steps up to the maximumavailable peak AF of 100 mT.Additional rock-magnetic measurements were made usinga Coercivity Meter (Jasonov et al., 1998), with a maximum ap- plied field of 500 mT. This instrument generates magnetic hys-teresis curves, along with stepwise acquisition and reverse-fieldacquisitionofisothermalremanence(IRM),fromwhichthefol-lowing parameters were derived; coercivity (  B c ), saturationmagnetisation (  M  s ), saturation remanence (  M  rs ) and coercivityof remanence (  B cr ). A variable field translation balance(MMVFTB) was used to measure thermomagnetic curves.Magnetisation was measured in an applied field of 107 mT,withheatingandcoolingcarriedoutinair.StepwiseTHdemag-netisation of orthogonal IRM was carried out following Lowrie(1990). Orthogonal fields of 2 T, 0.5 Tand 0.04 Twere appliedusing an ASC IM10-30 impulse magnetiser. TH demagnetisa-tion was carried out using the Schonsted Instruments TSD-1thermal demagnetiser, measuring the IRM remaining aftereach step with the AGICO JR5 spinner magnetometer. 4. Results and discussion  4.1. Microprobe analysis The amount of vitrification d as described by the abundanceof glass d is variable, both between and within individualblocks. This probably reflects compositional differences andtheir relative proximity to the heat source. In hand specimen,the vitrified material resembles pumice with elongated pores(Fig. 2). Under the microscope, few of the srcinal minerals Fig. 4. Bi-plot of susceptibility versus NRM intensity, indicating isolines of equal Koenigsberger ratio ( Q n ).Fig. 5. Equal area projection of initial NRM directions.  D , declination;  I  ,inclination;  a 95 , semi-angle of confidence;  k  , precision parameter. Fisher(1953) mean direction shown in grey.Table 3Summary of archaeomagnetic data  N   ( n )  D   I   a 95  k r  sum NRM 12 (17) 20.7 54.5 3.4 160 11.931ChRM 12 (17) 21.4 56.5 2.5 305 11.964Relocated ChRM 12 (17) 22.4 59.7 2.5 305 11.964NRM, natural remanent magnetisation; ChRM, characteristic remanent mag-netisation;  N   ( n ), number of samples (specimens) used to calculate the meandirection;  D , declination;  I  , inclination;  a 95 , semi-angle of confidence;  k  , pre-cision parameter;  r  sum , resultant vector: all from Fisher (1953).1402  G. Catanzariti et al. / Journal of Archaeological Science 35 (2008) 1399 e 1407   have survived. Instead, the dominating phases are (1) glasswith variable compositions, (2) relic minerals, and (3) newlyformed minerals. Lighter glass with high silica and alkali con-tent is blended with heavier glass rich in iron and alumina(Fig. 3). Relic minerals such as rutile, zircon and ilmenite,have been tentatively identified. Occasional iron oxide grainsmay represent secondary minerals formed during heating,along with metallic Fe ( a -Fe) spherules of diameters 2 e 25  m m (Table 1). The Fe content of the spherules varies between 90% and 99%, the remainder being P, Ni, Cu and (in-ferred) C (Table 2). Phases permitting precise peak tempera-ture estimates are absent, although the composition of someof the Fe spherules indicates maximum temperatures exceed-ing 1200   C.Compositional variation, both in the Fe spherules and in theglass phases, indicates pronounced variations in temperatureand redox conditions, even on a centimetre scale. This pointstowards a fairly rapid (disequilibrium) heating process, fol-lowed by similarly rapid cooling, preventing any devitrifica-tion of the glass phase. The sample investigated displaysa texture that indicates expanding superheated vapour fromthe decay of micas. Preservation of this texture impliesa lack of prolonged or repeated heating. This, along with thelack of devitrification, supports a single, strong heating event.  4.2. Archaeomagnetic study 4.2.1. NRM  Initial K and NRM intensities vary between 0.001 and0.225 SI and 0.03 and 26.00 Am  1 respectively. Koenigs-berger ratios ( Q n ¼ NRM/   KH  , where  H  ¼ geomagnetic fieldintensity z 40 Am  1 ) of between 1 and 100 (Fig. 4) are typ-ical of an NRM that is a thermoremanent magnetisation(TRM), i.e., acquired on cooling in an ambient magnetic field.The initial NRM directions are well-grouped (Fig. 5, Table 3) and show no dependency on the intensity of NRM or on thesampled blocks. In particular, the dispersion of directions of samples from the same block is the same as that observed be-tween blocks (shown in Fig. 7 for ChRM directions, althoughthe same is true for initial NRM directions). All of this is con-sistent with an NRM acquired on cooling of the vitrified wall,with no subsequent differential block movements. Fig. 6. Orthogonal vector projections of NRM demagnetisation. (a) TH demagnetisation, (b) AF demagnetisation. Closed/open symbols refer to horizontal (N e W)/ vertical (N e up) plane.Fig. 7. Equal area projection of characteristic remanence directions (ChRM).The Fisher (1953) mean declination (  D ) and inclination (  I  ) is shown ingrey, and the statistical properties  a 95  and precision parameter ( k  ) are quoted.Individual sample directions are given in the accompanying table, which givesblock number, sample code, declination (Dec) and inclination (Inc).1403 G. Catanzariti et al. / Journal of Archaeological Science 35 (2008) 1399 e 1407 
Related Documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks
SAVE OUR EARTH

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!

x