bronze age metalwork

8/7/2019 bronze age metalwork

1/67

A M P UR I A S (Barcelona)t. 43 , 1981, pagines 113- 170

A Study of the Bronze Age Metalworkfrom the Iberian Peninsula in the Brit ish MuseumBy RICHARDJ. HARRISONand

PAULT. CRADDOCKwith an appendix by MICHAELJ. HUGHES"

The collections of prehistoric bronzemetalwork discussed here have accumula-ted in a fortuitous manner for over125 years, with a succession of gifts largeaild small, purchases, and transfers fromother collections. The opportunity arose tocatalogue and examine the entire collec-tion, and to make it available for inclu-sion into the synthetic works that are nowactively in progress.

The year 1849 marks the acquisition ofthe first prehistoric bronze implementfrom Spain, donated by Mr. S. P. Pratt,F. R. S., and published the same year byMr. James Yates. For a long time thisdouble-looped palstave was one of theFew published prehistoric bronzes knownErom the Peninsula, and it has been illus-trated many times since then. It has theadded interest of being the only palstaveto be recorded as hafted, since it was dis-covered in sailcient coa1 workings in An-dalusia~,attached to a straight wooderihandle.But the largest number of pieces camefrom the famous collection amased by

the Very Reverend Canon Greenwell ofDurham Cathedral. He was actively co-llecting bronzes in the latter part othe nineteenth century, and through hisfriendly contacts with Horace Sandars, hewas able to acquire type-specimens fromthe Iberian Peninsula. The part of hiscollection containing these items wasbought by the American financier JamesPierpont Morgan, and presented outrightto the Museum as a gift in 1909.

Other important pieces were acquiredin 1964 when the Wellcome Collection waspresented to the Museum. The includeda dozen Early Bronze Age pieces fromGeorge Bonsor's old excavations aroundCarmona (Prov. Sevilla) in the period1890-1910, and which seem to have comefrom the site of El Acebuchal. Al1 of themcarry bold but illegible catalogue num-bers written in Bonsor's own hand.

The interest of the collection lies inits range of types and the secure prove-nances for many of the pieces. Not in-cluded in this study are the Siret Collec-tion of antiquities, purchased in 1889, and

* Richard J. Hamison: Dept. of Classics and Arihaeolagy University of Bristol; Paul T. Craddock: Re-aearch Laboratory, Bri tich M u s e u m , London; Michael J . Hughes: Research Laboratory, British Museurn, London.This manirscript was submitt ed in . Spring, t977, and there has been no opportunity to revise it since then .8


2/67

114 R. J. HARRISON, P. T. CRADDOCK AND M. J. HUGHES

Fig. 1 . - Location of al1 provenanced bronzes. Tlie numbers are th e same as those i n the Catalogue.

fully published by the Sirets' themselves A ) Catalogue and Figures.in 1887, nor the Iron Age finds. The co- B ) Tables of Analyses.llection of Bronze Age Gold from the Pe- C) Discussion,ninsula has been recently published afresh(Almagro Gorbea, 1974; Harrison 1977; The present location of al1 the piecesHawkes 1971), thus making al1 the Bronze is in the Department of Prehistoric andAge metalwork widely accesible. Romano-British Antiquities, of the British

The paper is divided into three parts: Museum, London WClB 3DG.


3/67

The catalogue is arranged in a stan-dard format, a nd one set of numbers areused throughout. The num bers of the pie-ces are retained both on the Figures andin the Tables of Analyses.

Abbreviations used in the Catalogue:Al1 measurements are in centimetres.L.: Length.Th.: Thickness.W.: Width.Registration Numbers:WG. : William Greenwell Collection.

(Al1 other Registration Numbers followthe standard British Museum system).

No. 1. (Fig. 2, no. 1)Provenance : Prov. Huelva.Registration: WG 909.Donor: Greenwell Collection, presentedby J. P. Morgan.Description: Flat ase, cast in an openmould with small casting irregularities onal1 surfaces. The cutting edge has beenground and hammered, and the butt, sidesand blade have been blunted at a later date.Hump-backed section. with squared sides.Irregular green patina. Good condition.Copper.Size: L. 12.5; W. 4.9; Th. 1.3.Unpublished.Comments: Pre-Millaran Type, c. 3000 -2500 B.C. One of a small number of thirdmillennium copper ases of this type from

S.W. Spain.No. 2. (Fig. 2, no. 2)

Provenance: Moguer, Prov. Huelva.Registration: WG 908.

Donor: Greenwell Collection, presentedby J. P. Morgan.Description: Tbin flat axe with a rectan-gular cross section, cast in an open rnould.One face is covered with casting irregula-rities, the other has been roughly smoothed.Tbe butt and cutting edge are both finelyground on one face, and covcred by a darkgreen-brown patina. Arsenical copper.

Size: L. 11.1; W. 4.1 ; Th. 0.6:Unpublished.Comments: Millaran Type, c. 2500 -2000 B.C.

No. 3. (Fig. 2, no. 3)Provenance: Spain.Registration: WG 911.Donor: Greenwell Collection, presentedby J. P. Morgan.Description: Flat axe with squared sides,lightly hammered a t butt, and witli a bluntedcutting edge. Blade edge chipped. One facewith finely pitted surface (from casting inan open mould), tbe other hammered andground to a smooth finish. Even, dark brownpatina, in excellent condition. Copper.Size: L. 11.95; W. 5.55; Th. 1.0.Unpublished.Comments: Millaran, or Earliest BronzeAge, c. 2200 - 1800 B.C.

No. 4. (Fig. 2, no. 4 )Provenance: Prov. Jan.Registration: WG 922.Donor: Greenwell Collection, presentedby J. P. Morgan.Description: Thick, rectangular-sectionedaxe with highly concave surfaces. Thins at

butt, roughly worked cutting edge, bluntedinto an assymetrical sbape (? through use).Pitted surface, bright green corrosion onone face.Size: L. 10.8; W. 3.1; Th. 1.8.Unpublished.


4/67

116 R. J. HARRISON, P. T. CRADDOCK AND M. J . HUGHES

Fig. 2. - Flat Axes (Pre-Beaker).1 , Prov. Huelua; 2, Xo~kc r ,Huc1i.a: 3, eSpain; 4.6, Prov. Jan


5/67

A STUDY OF THE BRO NZE AGE METAL

Comments: Probably Pre-Millaran or Mi-Ilaran, c. 3000 - 2000 B.C. (Compare with num-ber 5). Profile is similar to axes from theAlentejo in S. Portugal (Blance 1971, Taf. 24,nos. 1-2). but those have noticeably thinnersections than nos. 4 and 5 here. Copper.No. 5. (Fig. 2, no. 5)

Provenance: Prov. Jan.Registration: WG 921.Donor: Greenwell Collection, presentedby J. P. Morgan.Description : Thick, rectangular-sectionedcopper axe with slightly concave sides. Thin-ner towards butt, with a roughly hammeredcutting edge, blunted into an assymetriealshapc. Butt roughly finished, but main faces

are ground smooth. Signs of working on oneface, with vertical s tress lines. Copper.Size: L. 10.4; W. 2.8; Th. 1.8.Unpublished.Comments: Probably Pre-Millaran or Mi-Ilaran, c. 3000-2000 B.C. (Compare withno. 4.)

No. 6. (Fig. 2, no. 6)Provenance : Prov. Jan.Registration: WG 907.Donor: Greenwell Collection, presented

by J. P. Morgan.Description: Heavy, square-sectioned flataxe. with a very thick, rough, unfinishedbutt. Both sides havc heen heavily hamme-red, raising two low flanges. The cuttingedge is damaged. The original patina was adark, even green, which only survives inpatches, and which was largely flaked offwhen the piece was subject to heavy ham-mering of recent date. The flanges and ham-mer-marks are not original features. Copper.Size: L. 12.8; W. 3.9; Th. 1.85.Unpublished.Comments: Pre-Millaran or Millaran,c. 3000- 2000 B.C.

No. 7. (Fig. 3, no. 7)Provenance : Prov. JanRegistration: WG 915.

Donor: Greenwell Collection, presentedby J. P. Morgan.Description: Thick, rectangular-sectionedflat axe, with a shallow, irregular groove

down each face. One face and side have anetwork of fine surface irregularities, cau-sed by casting in an open mould. The otherface and edge are ground smooth. The buttand cutting edges have both been recentlyhamered. Even, deep brown patina, excellentcondition. Arsenical copper.Size: L. 15.8; W. 5.3; Th. 1.45.Unpublished.Comments: Probably related to the EarlyArgaric triangular-shaped axes, c. 1800 - 1500B.C.

No. 8. (Fig. 3, no. 8)Provenance: Villacarillo, Prov. Jan.Registration: 1909. 7-14. 4.Donor: Purchased.Description: Flat rectangular-sectionedaxe used as a wedge in antiquity. Very hea-vily hammered at butt, deeply scarred onboth faces, and with a distorted hlade. Deepbrown patina, specks of lime concretion, andsome modern corroison. Copper with lowarsenic.Size: L. 11.8; W. 6.4; Th. 1.6.Unpublished.Comments: Argaric Bronze Age, c. 1500 -1100 B.C. related to the Argaric (B) axeson basis of its triangular shape.

No. 9. (Fig. 3, no. 9)Provenance: Prov. Granada.Registration: 1909. 7-14. 10.Donor: Purchased.Description: Very thin flat axe with pa-ralle1 sides, and a rectangular section. Thin-ned at butt and cutting edge, which is blun-ted. Traces of a deep green patina remain,but the surface is now pitted and corroded.Arsenjcal copper.Size: L. 14.4; W. 3.5; Th. 0.5.Unpublished.


6/67

118 R . J. HARRIS ON, P. T. CRADDOCK AND M. J . H U G H E SComments: Probahly Early Bronze Age, mered cutting edge, and ancient damage atc. 1800-1500 B.C. %milar axes in Blance butt end below the patina. The surface is(1971, Taf. 13, nos. 12, 19, 21). well finished, although the cutting edge hasbeen damaged in modern times. Good con-

No. 10; (Fig. 3, no. 10)Provenance: Spain.Registration: WG 910.Donor: Greenwell Collection, presentedby J. P. Morgan.Description: Small, thick flat axe with aroughly rectangular cross section and wideblade. Thick butted, with modern damage,and a chipped cutting edge. One face isslightly concave, suggesting it was cast inan open mould. Pitted and corroded surfacewith remains of a very thick green patina.

Copper with small amount of tin.Size: L. 9.0; W. 5.9; Th. 1.08.Unpublished.Comments: Early Bronze Age, c. 1700 -1400 B.C. It is comparable to one from Bu-jos, Vila Real (Harhison 1967, fig. 6, no. 5)and another from Santa Cruz, S. Tiago doCacem (Scbubart 1975, Taf. 48, no. 419).

No. 11. (Fig. 3, no. 11)

dition wi th a light green patina. Illegibleink inscription on one face. Tin bronze.Size: L. 6.85; W. 4.95; Th. 0.65.Unpublished.Comments: Early Bronze Age, c. 1800B.C., prohably associated with the Beakercomplex.

No. 13. (Fig. 4, no. 13)Provenance: Probahly Prov. Granada.Registration: 1909. 7-14. 12.Donor: Purchased.Description: Small flat axe with a rec-tangular section. The cutting edge has twoground, hevelled edgcs. The butt has beenhammered, then partially cut through witha chisel and snapped off. Polished al1 over,with a smooth surface, light scratching onone side. Deep brown patina, excellent con-dition. Arsenical copper.Size: L. 6.3; W. 3.8; Th. 0.5.Unpublished.

Provenance: Probably Prov. Jan. Comments: Argaric Bronze Age, c. 1500-1100 B.C. (cf . Blance, 1971, Taf. 23, no. 21Registration: 1909. 7-14. 1. from Mas de Menente).Donor: Purchased.Description: Slender flat axe with a rec-tangular section, with one face deeply pitted,the other fairly smooth. The cutting edge isslightly hammered and the butt thinned. Dullgreen-brown patina, with modern ahrasions.Arsenical copper.Size: L. 13.6; W. 4.75; Th. 1.0.Unvublished.Comments: Probably Argaric Bronze Age,c. 1500- 1100 B.C.

No. 12. (Fig. 3, no. 12)Provenance: Probahly from El Acehu-chal, Carmona, Prov. Sevilla.Registration: P1964. 12-1. 355.Donor: Wellcome Collection.Description: Small flat axe with an ham-

No. 14. (Fig. 4. no. 14)Provenance: Prov. Jan.Registration: WG 919.Donor: Greenwell Collection, presentedby J. P. Morgan.Description: Small, thin, flat axe with arectangular section. The cutting edge washammered and the butt thinned, althoughhoth are now damaged. Scratched into theeven brown patina, the word ~ B a u t e n ~or~Bailenn.Excellent condition. Arsenical cop-per.Size: L. 7.3; W. 2.2; Th. 0.4.Unpublished.Comments: Probably Argaric Bronze Age,c. 1500- 1100 B.C. (cf. Blance, 1971, Taf. 23,no. 19. from El Argar).


7/67

Fig. 3. - Fiat Axes. 7, Prov. JaBn; 8, Viilacariilo, Jadn;9>Prov. Granada; 10, Spain, 11, probably Prov. Jadn, 12, probably Ei Acebuchal, SeviBa.


8/67

120 R . J . H A R R I S O N , P. T . CRADDOCK AND M . J . H U G H E SNo. 15. (Fig. 4, no. 15)

Provenance: Prov. Jan:Registration: WG 920.Donor: Greenwell Collection, presentedby J. P. Morgan. -Description: Small, flat axe, with a thinbutt and rectangular cross section. Cuttingedge hammered, and worn to an assymetri-cal shape. The axe has been chemically clea-ned, and the vertical stress lines are clearlyvisible on each face. Arsenical copper.Size: L. 7.3; W. 3.6; Th. 0.6.Unpublished.Comments: Argaric Bronze Age, c. 1500 -1100 B.C.

No. 16. (Fig. 4, no. 16)Provenance: Probably Prov. Granada.Registration: 1909, 7-14. 11.Donor: Purchased.Description: Small flat axe of rectangu-lar section, with a slightly thinned butt andhammered cutting edge. Original surface ofgreen-grey corrosion intact on both faces.Good condition. Arsenical copper.Size: L. 7.5; W. 3.4; Th. 0.6.Unpublished.Comments: Argaric Bronze Age, c. 1500-

1100 B.C.No. 17. (Fig. 4, no. 17)

Provenance : Prov. Jan.Registration: WG 917.Donor: Greenwell Collection, presentedby J. P. Morgan.Description: Thick, rectangular-sectionedflat axe with a thick butt (lightly hammered)and a slightly expanded cutting edge (re-cently blunted). Surface lightly pitted from

being cast in an open mould. Deep, evenbrown patina. Good condition. Arsenicalcopper.

Size: L. 9.9; W. 3.95; Th. 1.0.Unpublished.Comments: Argaric Bronze Age, c. 1500 -1100 B.C.

No. 18. (Fig. 4, no. 18)Provenance: Prov. Sevilla.Registration: 1916. 6-5. 175.Donor: Presented by the Rt. Hon. Lord

Avebury, from the collection of the lateLord Avebury (Sir John Lubbock).Description: Slender flat axe with a thick,rectangular cross-section. Butt slightly thin-ned and hammered, cutting edge hammeredand hlunted. Uneven green and brown pa-tina. Good condition. Arsenical copper.Size: L. 13.8; W. 5.35; Th. 1.0.Unpublished.Commerit: Argaric Bronze Age, c. 1500-1100 B.C. Typical Argaric type, rather fartherto the west tban would be expected.

No. 19. (Fig. 4, no. 19)Provenance: Prov. Granada.Registration: 1909. 7-14. 3.Donor: Purchased.Description: Slender, thin-sectioned flataxe with neatly hammered cutting edge, un-thinned butt. Modern scratchings on eachface. Ground and polished on each face.Green and brown patina. Tin bronze.Size: L. 11.9; W. 5.2; Th. 0.9.Unpublished.Comments: Argaric Bronze Age, c. 1500 -1100 B.C.

No. 20. (Fig. 4, no. 20)Provenance: Prov. Jan.Registration: WG 916.Donor: Greenwell Collection, presented

by J. P. Morgan.Description: Slender, rectangular-sectio-ned flat axe, with a thin and well finishedbutt, and a polished cutting edge. Surfaceground and polished smooth, with traces ocoarser grinding on one face. Deep brown

patina with some green patches. Excellentcondition. Arsenical copper.Size: L. 10.6; W. 3.9; Th. 0.65.Unpublished.Comments: ~ r g a r i cBronze Age, c. 1500 -1100 B.C.


9/67

F i g . 4 . - Argaric Flat Axos . 13 , 1 6 , probably Prov. Granada; 14, 15 , 17, 2 0 , Prov. J a n ;18 , Prov. Seviila; 19, Prov. Granada; 21, possibly Prov. Granada: 22, Prov. Ccliz.


10/67

122 R . J . H A R R I S O N , P. T . CRANo. 21. (Fig. 4, no. 21)

Provenance: Possibly Prov. Granada.Registration: 1909. 7-14. 13.Donor: Purchased.Description: Small, slender flat axe, witha rectangular section, a slightly expandedcutting edge and cutting edge. At some time,the axe has been cleaned, then coloured darkhrown to imitate a bronze patina. Arsenical

copper.Size: L. 7.9; W. 2.6; Th. 0.65.Unpublished.Comments: Argaric Bronze Age, c. 1500 -1100 B.C.

No. 22. (Fig. 4, no. 22)Provenance: Prov. Cdiz.Registration: WG 918.Donor : Greenwell Collection, presentedby J. P. Morgan.Description: Small, slender flat axe witba rectangular cross-section, thinned, well fi-nished hutt, and a hammered cutting edge.Ground and hammered smooth, with a deep,even brown patina. Some pitting on oneface. Good condition. Arsenical copper.Size: L. 8.8; W. 3.45; Th. 0.8.Unpublished.Comments: Argaric Bronze Age, c. 1500-1100 B.C.

No. 23. (Fig. 5, no. 23)Provenance: Velefique, Sierra de los Fi-labres, Prov. Almeria.Registration: WG 913.Donor: Greenwell Collection presented byJ. P. Morgan.Description: Flat axe with a rectangularsection, and a slightly thinned, hammered

butt. Cutting edge hammered, then ground.Well finished and ground on one face, lightlypitted and hammered on the other face. Deepgreen patina. Excellent condition. Arsenicalcopper.Size: L. 15.6; W. 6.3; Th. 1.15.Unpublished.

DDOCK AND M . J . H U G H E SComments: Argaric Bronze Age, c. 1500.1100 B.C.

NO. 24. (Fig. 5, no. 24)Provenance: Prov. Jan.Registration: WG 914.Donor: Greenwell Collection, presented byJ. P. Morgan.Description: Well proportioned, thick-sectioned flat axe. Butt is tapered and squa-red off, and the cutting edge hammered andground smooth. Hammering (below patina)on one face. Deep brown patina. Excellentcondition. Arsenical copper.Size: L. 13.9; W. 6.8; Th. 1.45.Unpublished.Comments: Argaric Bronze Age, c. 1700-1200 B.C.

No. 25. (Fig. 5, no. 25)Provenance: Prov. Toledo.Registration: 1909. 7-14. 20.Donor: Purchased.Description: Large flat axe with a rectan-gular section, thinned and lightly hammeredbutt. and a ground and hammered cuttingedge. Well finished, both faces pitted withsome corrosion. Deep green patina on one

face. Arsenical copper.Size: L. 19.6; W. 8.3; Th. 1.4.Unpublished.Comments :Bronze Age, c. 1500 - 1100 B.C.

No.26. (Fig. 5, no. 26)Provenance: Sierra de los Baos, Almo-naster, Prov. Huelva.Registration: WG 912.Donor: Greenwell Collection, presentedby J. P. Morgan.Description: Flat axe with a rectangularcross-section, parallel sides and splayed cut-ting edge. Butt is squared and well finishedand the cutting edge hammered flat, thenground in recent times. Find spot scratchedon one face. Tin bronze.Size: L. 14.3; W. 7.4; Th. 1.0.


11/67

tiig. 5 . - Flat Axes. 23 , Velc f iquc , ri1meri.z; 24 , Prov. Jan;25, Prov. Taledo; 26, Sierra de Los Baos, Huelva.


12/67

124 R. J . H ARRI SO N, P. T . CRADDOCK AND M . J . H U G H E SUnpublished.Cominents: Early or Middle Bronze Age,c. 1600-1200 B.C. Others from Almonasterare illustrated by Schubart (1975, Taf. 44).

The piece was either found in 1871 or soldto Greenwell in that year, since the dateis written in ink on one face.No. 27. (Fig. 6. no. 27)

Provenance: Villacarillo, Prov. Jan.Registration: 1909. 7-14. 2.Donor: Purchased.Description: Flat axe with a rectangularsection, a squared, well-finished butt, ham-mered cutting edge, and two incipient ridgeson each side. Cutting edge slightly damaged,

but in good condition with a deep, evengreen patina. Tin bronze.Size: L. 13.3; W. 8.4; Th. 1.15.Unpublished.Comments: Earlier part of Bronze Age,c. 1600- 1200 B.C. The splayed cutting edgeand thick butt are more characteristic ofaxes in the S.W. of the Peninsula (Schubart1975 ; Tafeln. 47-49).

No. 28. (Fig. 6, no. 28)Provenance: Prov. Len.Registration: 1909. 7-14. 21.Donor: Purchased.Description: Flanged axe with a splayedcutting edge. The butt is thinned, and boththe flanges and blade seem to have beenhammered in modern times. Surface islightly pitted, and now patinated a deepyellow-brown. Blunted cutting edge. Copper.Size: L. 10.8; W. 6.7; Th. 2.15.Unpublished.Comments: Early Bronze Age, c. 1600-1400 B.C. This is essentially a French orSouthern British type, common in the Arre-ton Tradition in Britain, where tin bronze,not copper, is the normal metal (ApSimon,1954; Britton, 1963, Tables 6 . 7 ; Butler, 1963).

A few similar pieces are known from Lan-guedoc (Guilaine 1972, fig. 28, no. 3). Al-though at least 14 flat axe findspots areknown from the neighbouring provinces of

Zamora (including an apparent hoard of4 flat axes from Linares de Riofrio) and Bur-g o ~ ,no similar axes are mentioned (MartinValls and Delibes de Castro. 1975; Delibesde Castro, 1973). The flat axes from Reino-so and Retuerta (Prov. Burgos) were shownby analysis to contain 5 % and 15 % tin res-pectively, making the composition of theLen piece even more peculiar (ibid.).

No. 29. (Fig. 6, no. 29)Provenance: Probably from El Acebu-chal, Carmona, Prov. Sevilla.Registration: P 1964, 12-1. 356.Donor: Wellcome Collection.Description: Tanged dagger, with clear

traces of hammering and cold working atthe edges. The hilt is heavily hammeredi'upinto a low, irregular tlange on each side,with the individual blows clearly marked.The butt is thinned to a chisel end. Goodcondition, with traces of green patina. Arse-nical. copper.Size: L. 13.5;W. 4.55; Th. 0.4.Unpublished.Conzments: Early Bronze Age, c. 1800B.C., of Beaker type.

No. 30. (Fig. 6, no. 30)Provenance : Prov. Jan.Registration: 1909. 7-14. 8.Donor: Purchased.Description: Three riveted halberd withone rivet remaining, on a fow. oval sectio-ned midrib running from the butt to the tip.The hilt is damaged, and the tip is bent dou-ble. The rivets are al1 driven through fromone face. Areas of dark green and brownpatina. Arsenical copper, with rivet havingless arsenic than the blade.Size: L. 15.3; W. 6.3; Th. 0.5. Original,L. 22.0. Rivet: L. 2.5; Th. 0.6.Unpublished.Comments: Early Bronze Age, c. 1700 -1400 B.C. This halberd is related to Schu-bart's Carrapatas Type in northern Portugal(Schubart, 1973; Harrison, 1974a).


13/67

Fig. 6. - farly Bronze Age Metalwork. 27, Villacarillo, Jakn; 28, Prov. Len:29 , probably El Acebuchai. Sevil la; 30, Prov. Ja4n.


14/67

126 R. J. HARRISON, P. T. CRAI>DOCK AND M. J. H U G H E SNo. 31. (Fig. 7, no. 31)

Provenance: Possibly from the IberianPeninsula.Registration: 1925. 10-17. 2.Donor : Purchased.Description: Large tanged dagger with aftat cross section and extensive cold workingand hammering al1 along the cutting edges.The hilt is thinned, and hammered up intoa low flange on each side. Very faint tracesof an omega hilt remain in the surface cor-rosion. Tip broken off. Originally finelyground al1 over, but now speckled with cor-rosion and a thick green patina. Arsenicalcopper.Size: L. 25.6; W. 6.7; Th. 0.35. Tang, Th.0.5.Unpublished.Comments: Beaker type, c. 1750- 1400B.C. The closest paraltels in the Peninsulaare found in the Montelavar Group of EarlyBronze Age single burials. concentrated inthe western half of Iberia (Harrison, 1974;1974a).

No. 32. (Fig. 7, no. 32)Provenance: Probably from El Acebu-chal, Carmona, Prov. Sevilla.Registration: P 1964. 12-1. 104.Donor: Wellcome Collection.Description: Palmela point with hamme-red cutting edge. The tang is square-sectio-ned. Well finished and ground al1 over, andin excellent condition with a green patinaon one face. In ink on the other face aretwo illegible catalogue numhers. Arsenicalcopper.Size: L. 10.9; W. 2.95; Th. 0.25. Tang, Th.0.35.Unpublished.Comments: Early Bronze Age, c. 1800 -

1500 B.C., of Beaker type.No. 33. (Fig. 7, no. 33)

Provenance: Probably from El Acebu-chal, Carmona, Prov. Sevilla.Registration: P 1964. 12-1. 105.

Donor: Wellcome Collection.Description: Palmela point with neatlyhammered cutting edge, carried evenlyaround the whole blade. The tang is square-

sectioned. Well finished and ground al1 over,with most patina scraped off. Illegible inkinscriptions on each face. Arsenical copper.Size: L. 11.8; W. 3.4; Th. 0.18. Tang, Th.0.35.~ n ~ u b l i s h e d .Comments: Early Bronze Age, c. 1800 -1500 B.C., of Beaker type. The shoulders areunusually sharply-angled on this piece.

No. 34. (Fig. 7, no. 34)Provenance: Probably from El Acebu-

chal, Carmona, Prov. Sevilla.Registration: P 1964. 12-1. 106.Donor: Wellcome Collection.Description: Palmela point with hamme-red edge (only visible on one face). carriedal1 around the blade. The tang is square-sec-tioned. Well finished, with faint stress linesrunning vertically in the centre of one faceof the blade. Traces of patina on each face.Good condition, with an illegible ink inscription on one fa'ce. Arsenical copper.Size: L. 10.2; W. 2.95; Th. 0.35. Tang, Th.0.4.Unpublished.Comments: Early Bronze Age, c. 1800 -1500 B.C., of Beaker type.

No. 35. (Fig. 7, no. 35)Provenance: Probably from El Acebu-chal, Carmona. Prov. Sevilla.Registration: P 1964. 12-1. 107.Donor: Wellcome Collection.Description: Palmela point with faintsigns of hammered cutting edge. The bladehas been bent, and shows faint stress lines

and surface irregularities in the centre ofone face. The tang is square-sectioned. Tra-ces of patina on each face. Reasonable con-dition, with an illegible ink inscription onone face. Arsenical copper.Size: L. 12.8; W. 3.4; Th. 0.2. Tang, Th.0.3..


15/67

Fig. 7. - Bell-Beaker Metalwork. 31, possibly from th e Iberian Peninsula;32-38, probably from El Acebuchal, Sevilla.


16/67

128 R . J . H ARRI SO N, P. T. CRADDOCK AN D M . J . H U G H E SUnpublished. tion, with an itlegible inscription in ink onComments: Early Bronze Age, c. 1800 - blade. Arsenical CoPPer.1500 B.C., of Beaker type. Size: L. 10.0; W. 3.4; Th. 0.3. Tang, Th.0.35.

No. 36. (Fig. 7, no. 36)Provenance: Probably from El Acebu-chal, Prov. Sevilla.Registration: P 1964. 12-1. 108.Donor: Wellcome Collection.Description: Palmela point with veryfaint traces of an hammered cutting edge.Tbe square-sectioned tang has been beatenup, making irregular low flanges on eachside. Remains of patina on eacb face, withan itlegible inscription on one face. Arse-

nicat copper.Size: L. 10.85; W. 2.8; Th. 0.25. Tang, Th.0.35.Unpublished.Comments: Early Bronze Age, c. 1800 -1500 B.C., of Beaker type.

No. 37. (Fig. 7, no. 37)Provenance: Probably El Acebuchal, Car-mona, Prov. Sevilla.Registration: P 1964. 12-1. 109.Donor: Wellcome Collection.Description: Palmela point, with a deeplypitted surface. Square-sectioned tang. Illegi-ble ink inscription on one face. Arsenicalcopper.Size: L. 10.6; W. 2.4; Th. 0.25. Tang, Th.0.3.Unpublished.,Comnzents: Early Bronze Age, c. 1800

1500 B.C.. Beaker type.No. 38. (Fig. 7, no. 38)

Provenance: Probably El Acebuchal Car-mona, Prov. Sevilla.Registration: P 1964. 12-1. 110.Donor: Wellcome Collection.Descriptions: Palmela point with cuttingedge hammered on each face. Square-sectio-ned tang. One face patinated. Good condi-

Unpublished.Comments: Early Bronze Age, c. 1800 -1500 B.C., Beaker type.

No. 39. (Fig. 8, no. 39)Provenance: Presumed to be fromSpain.Registration: 1909. 7-14. 24.Donor: Purchased.Description: Palmela point, with a rec-tangular sectioned tang and only very faint

traces of cold working. Tip and point bothbent over, and surface damaged and corro-ded. Arsenical copper.Size: L. 10.6; W. 3.45; Th. 0.2. Tang, Th.0.3.Unpublished.Comments: Early Bronze Age, c. 1800 -1500 B.C. of Beaker type.

No. 40. (Fig. 8, no. 40)Provenance: Arjona, Prov. Jan.Registration: 1964. 12-6. 106.Donor: Trechmann Bequest.Description: Variety of Palmela pointwith an elongated, hammered blade andtang. Patches of a thick concretion and cor-rosion cover much of the surface. Arsenicalcopper.Size: L. 13.4; W. 2.6; Th. 0.2. Tang, Th.0.3.Unpublished,Comnzents: Early Bronze Age, c. 1700 -1400 B.C. Typologically more developed thanthe Palmela point (nos. 32-39).

No. 41. (Fig. 8, no. 41)Provenance: ~Presumedn lo be fromSpain.Registration: 1909. 7-14. 25.Donor: Purchased.


17/67

Fig. R . - Spearheads and Arrowheadc. 39 , 41-44. al1 ePresumed to be froin Spainr;40, :irjanal Jan; 45-46, Prov. Jan; 47 . Prov. Crdoba: 48, Prov. Granada.9


18/67

130 R. J . HARRISON, P. T. CRADDOCK AN O M . J. HUGHESDescription: Developed variety of a Pal-mela point with a slender blade with tracesof working and a square sectioned tang. NOedge working on blade, although there aresome fine grinding marks. Even brown pa-tina. Excellent condition. Arsenical copper.Size: L. 9.7; W. 2.1 ; Th. 0.25. Tang, Th.0.3.Unpublished.,Comments: Early Bronze Age, c. 1600 -1400 B.C. Typologically more developed thanthe Palmela point (nos. 32-39).

No. 42. (Fig. 8, no. 42)Provenance: *Presumed. to be fromSpain.Registration: 1909. 7-14. 26.Donor: Purchased.Description: Tanged spearhead, relatedto Praganca type. Rectangular sectionedtang, twisted at end. Surface cleaned in mo-dern times. Deep brown patina, and in ex-cellent condition. Arsenical copper.Size: L. 6.1;W. 2.05; Th. 0.25. Tang, Th.0.28.Unpublished.Comments: Early-Middle Bronze Age,c. 1400-1200 B.C.

No. 43. (Fig. 8, no. 43)Provenance: Presumed to be fromSpain.Registration: 1909. 7-14. 27.Donor : Purchased.Description: Smalt flat tanged arrow-head, with flat tang, lightly hammered oneach side. Thinned at edges, but no hamme-ring visible on surface. Green and brown pa-tina with fine surface pitting. Too small tosample for analysis.Size: L. 4.5; W. 1.65; Th. 0.15.Unpublished.Comments: Probably Middle Bronze Age,c. 1400-1200 B.C.

No. 44. (Fig. 8, no. 44)Provenance: Presumedu to be fromSpain.

Registration: 1909. 7-14. 28.Donor : Purchased.Description: Tanged spearhead with twoharhs. The tang is split at the end and

tigbtly curled up. Tbe barbs have been bentinwards. Dark brown patina. Arsenical cop-per.Size: L. 4.6; W. 1.35; Th. 0.15. Tang, Th.0.25.Unpublished.Comments: Early-Middle Bronze Age,c. 1400-1200 B.C. (cf. Blance, 1971, Taf. 2414from Praganca).

No. 45. (Fig. 8, no. 45)Provenance: Prov. Jan.Registration: 1964. 12-6. 108.Donor: Trechmann Bequest.Description: Barbed and tanged arrow-head, with a shouldered tang. One barb bro-ken off. Deep, even brown patina. Excellentcondition. Arsenical copper.Size: L. 6.45; W. 1.4; Th. 0.25. Tang, T&0.25.Unpublished.Comments: Middle Bronze Age, c. 1400 -1000 B.C. (cf . Blance, 1971, Taf. 2415, fromVila Nova de So Pedro).

No. 46. (Fig, 8, no. 46)Provenance: Prov. Jan.Registration: 1964. 12-6. 109.Donor: Trechmann Bequest.Description: Cast barbed and tanged ar-rowhead, hammered al1 over, and prohablycast in an open mould. Rih on one side only.Tip and one barb bent. Dark green patina.Excellent condition. Too small to sample foranalysis.Size: L. 5.05; W. 1.95; Th. 0.25. Tang, Th.0.1.Unpublished.Comments: Prohably Late Bronze Age,c. 900-700 B.C. It has a similar profile to cer-tain arrowheads from the Ra de Huelvahoard (Almagro, 1958, E. 1, No. 188).


19/67

No. 47. (Fig. 8, no. 47) No. 49. (Fig. 9, no. 49)Provenance: Prov. C6rdoba. Provenance: Guadix, Prov. Granada.Registration: 1964. 12-6. 110. Registration: 1909. 7-14. 16.

: Donor: Trechmann Beques t. Donor: Purchased.1 Description: Arrowhead with long tang Description: Two riveted dagger, with

and two small barbs. Roughly hammered one square sectioned rivet still in place.ipto shape, with two shallow grooves on Clear traces of a straight-ended hafting-plateare visible on each side. The tip is chipped,Of blade' Edges Of blade and tang and the whole surface of the hlade is badlybevelled and hammered into Dark corroded. Rivet and blade made of Assenicalgreen.brown patina. Toa small to sample for copper,analysis. Size: L. 18.0; W. 4.5; Th. 0.4. Rivet: L.Size: L. 5.0; W. 1.6; Th. 0.22. Tang, Th. Th.0.25. Unpublished.Unpublished. Comments: Argaric B. c. 1500/1400-1100Comments: Probahly Final Bronze Age, B,C,c. 900-700 B.C. Identical oieces are knownfrom Obaga de les ventoses, Montmell (10 %tin) and Serra de YEspasa, Capcanes (10 %tin), both in Prov. Tarragona (Vilaseca, 1973,pp. 172-3 and lam. 6911).

No. 48. (Fig. 8, no. 48)i Provenance: Prov. Granada.' Registration: 1964. 12-6. 107.l' Donor: Trechmann Bequest.

Description: Cast arrowhead from a two-piece mould, showing the unevenly matchedhalves of the mould. The rib ends at the baseof the blade in a small button, probablyto prevent the shaft from ,splitting. , Sur-fnces and edges well ground and finished.al1 covered by a deep green patina. Tinbronze.

Size: L. 7.6; W. 1.8; Th. 0.38. Tang, Th.0.4.Unpublished.

No. 50. (Fig. 9, no. 50)Provenance: Guadix, Prov. Granada.Registration: 1909. 7-14. 17.Donor: Purchased.Description: Three-riveted dagger, withthree square sectioned silver rivets in place.Clear traces of a straight-ended hafting plateare visible on each face. Tip missing, andwhole surface heavily corroded and pitted.Painted green ( !), and not sampled for ana-lysis owing to mineralization.Size: L. 15.8; W. 3.95; Th. 0.3. Rivet:L. 1.45; Th. 0.38-0.62.Unpublished.Comments: Argaric *B., c. 1500/1400 -1100 B.C.

No. 51. (Fig. 9, no. 51)Provenance: Guadix. Prov. GranadaComments: Probably belongs to the endof the Final Bronze Age, c. 900-700 B.C. Si-n3ilar bronze arrowheads with a bulb of me-tal at t he base of the blade ar e known fromvarious sites in Languedoc, datid by Guilai-ne (1972, fig. 125) to


20/67

132 R. J. HARRISON, P. T . CRADDOCK A ND M . J . H U G H E S

Fig . 9. - Riveted Knives or Daggers. 49-61, Guadix, Granada; 63 , Villacarrillo, Jan: 54, Prov. Jan.


21/67

' Comments: Argaric


22/67

134 R. J. HARRISON, P. T. CRADDOCK ANO M. J. HUGHESComments: Earlv Bronze Aee, c. 1600 - Size: L. 10.9; W. 0.45; Th. 0.2.1400 B.C. Very simiiar to a dagger fronl Sa- Unpublished.lir, Conc. Loul (Portugal) (Schubart, 1975, comments: Found together ~ i t hnos. 58,Taf. 11, no. 115). 60.

No. 57. (Fig. 10, no. 57) No. 60. (Fie. 10, no. 60)Provenance: Prov. Jan.Registration: 1909.-7-14. 9.Donor : Purchased.Description: Dagger with a prominentmidrib and four large circular rivet holespunched into the hilt from the same sideof the blade. The midrib continues into thehilt but tapers off before reaching the hutt.The end of the hilt has been carefully thin-ned and ground to shape, and the wholesurface of the blade polished to a high fi-nish. Deep redbrown patina; excellent condi-tion. Tin hronze, with some lead.Size: L. 18.3; W. 3.4; Th. 0.45.Unpublished.Comments: Probably to be dated c. 1200 -1000 B.C. The type is unusual for the Pe-ninsula, and can be more closely paralleledin France.

No. 58. (Fig. 10, no. 58)Provenance: Prov. Crdoba.Registration: 1964. 12-6. 99.Donor: Trechmann Bequest.Description: Thin bronze tube with a so-lid end, probably used as a thong terminal.Made from thick bronze wire, hammered flatat one end and rolled over into a cone.Brown patina with some encrustations. Toosmall to sample for analysis.Site: L. 7.4; W. 0.55; Th. 0.25.Unpublished.Comments: Undatable. Found togetherwith nos. 59, 60.

No. 59. (Fig. 10, no. 59)Provenance: Prov. Crdoba.Registration: 1964. 12-6. 100.Donor: Trechmann Bequest.Description: Thong terminal. Same asno. 58. Not sampled for analysis.

Provenance: Prov. Crdoba.Registration: 1964. 12-6. 101.Donor: Trechmann Bequest.Description: Round sectioned awl or longpin with a spot of modern solder at one end.Not sampled for analysis.Size: L. 9.0; Th. 0.25.Unpublished.Comments: Found with nos. 58. 59.

No. 61. (Fig. 10, no. 61)Provenance: Almucar, Prov. Granada.Registration: 1909. 7-14. 14.Donor: Purchased.Description: Square sectioned chisel, wellground and finished al1 over, and not dama-ged by use. Casting seam on one face. Deep,rven brown patina. Arsenical hronze.Size: L. 13.7; W. 1.28; Th. 1.0.Unpublished.Comments: Argaric type, not especiallydiagnostic of either phase A or B. c. 1700-

1000 B.C. Compare with similar examplesfound shors des tombeaux, at El Argar (Si-ret and Siret, 1887. P1. 26, nos. 68-74).No. 62. (Fig. 10, no. 62)

Provenance: Prohably Prov. Jan.Registration: 1909. 7-14. 7.Donor: Purchased.Description: Square sectioned chisel, so-mewhat damaged at cutting edge, althoughbutt has not been hammered. Originally wellground and finished, but the deep green pa-tina is no* damaged. Arsenical copper.Size: L. 15.9; W. 1.3; Th. 1.15.Unpublished.Comments: Argaric, c. 1700-1000 B.C. (seecomments for no. 61).


23/67

1 Fig. 10. - Argaric a n d Othcr Xetalwork. 55, 57, Prov. Jan;56 , Prov. Sevi l la; 58-60, Prov. Granada; 61,Almucar, Granada; 62 , probably Prov. Jan


24/67

136 R . J . H A R R IS O N , P. T . CRADDOCK AND M . J . HUGHESNo. 63. (Fig. 11, no. 63) off to show a very porous casting. Originallywell finished and smoothed, now evenly pati-Provenance: Andalucia, Spain ([[Found nated deep &reen,yin bronze with lead.in ancient coal-workings in Andalusia). Size: L. 17.7; W. 4.4;. Th. 3.05.Registration: 1849. 2-20. 1. Unpublished.Donor: S.P . Pratt, Esq. F.R.S. Comments: Atlantic Bronze Age, c. 1000 -Description: Slender double looped pal- 700stave with a cast trident motif on one faceand a low thin rib on the other. One sideloop broken, butt damaged. Casting seams 66. (Fi&. 11. no. 66)visible on each side. Badly corroded and Len,pitted, but originally well finished and smoo-thed. Leated tin bronze. Registration: 1909. 7-14, 22.

Size: L. 37.7; W. 4.15; Th. 2.35. Donor : Purchased.Published: Yates 1849: 369, 1881 : D e s c r i ~ f i o ~ :Slender single looped palsta-97 and fig. 89. ve with cast decoration inside one flange,and on each face below the stop ridge. StubsCotnments: Atlantic Bronze Age, c. 1000- of a second loop on one face,

700 B.C. According to Yates (1849, P. 369). . from !he patination over them and their neat~. . .whenfirst discovered (it was) firmly atta- finish, it likely that only one loop wasched by thongs to a straight handle of wood ever cast on to the palstave.(see no. 64, which it closely resembles). Despite some modern damaee and sur-No. 64. (Fig. 11, no. 64)

Provenance: Langreo, Prov. Oviedo.Registration: WG 932.Donor: Greenwell Colection, presentedby J. P. Morgan.Description: Slender 6ouble looped pal-stave with a cast trident on one face and a

low thin rib on the other. Btade and buttchipped, with a neatly finished casting seamon each side. Porous casting, hui well fi-nished and evenly patinated to a deep green.Leaded tin bronze.Size: L. 16.3; W. 4.15; Th. 2.15.Unpublished.Comments: Atlantic Bronze Age, c. 1000-700 B.C. (see no. 63 which it closely resem-bles).

No. 65. (Fig. 11, no. 65)Provenance: Lena, Prov. Oviedo.Registration: WG 931.Donov: Greenwell Collection, presentedby J. P. Morgan. :Description: Slender double looped pal-stave with a single low rib on each face.Edge ground recently. Butt roughly finished

face crrosion, the axe was originally wellfinished. Tin bronze with some lead.Size: L. 15.7; W. 4.55; Th. 2.5.Unpublished.Comments: Atlantic Bronze Age, c. 1000-700 B.C. The decoration is unusual.

No. 67. (Fig. 12, no. 67)Provenance : Spain.Registration: WG 930.Donor: Greenwell Collection, presentedby J. P. Morgan.Description: Double looped palstave withlow flanges. One loop is crushed, but thecasting has been properly trimmed and fi-nished. Heavily cleaned, with only traces ofpatina surviving. Tin bronze with lead.Size: L. 18.2; W. 5.4; Th. 3.35.Unpublished.Comments: Atlantic Bronze Age, c. 1000-700 B.C.

No. 68. (Fig. 12. no. 68)Provanance: Almagro, Prov. Ciudad Real.Regi.;tration : 1964. 12-6. 104.Donor: Trechmann Bequest.


25/67

A S TU DY O F T H E B R O N Z E AGE M E T AL W OR K F R O M T H E I B E R IA N P E NI N SU L A

C'ig. 11. - Double-Loopcd Palcta.ires.63 , eAndaluciar; 64, ~ a r i ~ r e o ,Oviedo; 65, Lena, Oviedo; GF, Prov. Len.


26/67

138 R . J. H A R R I S O N , P. T. CRADDOCK AND M . J. H U G H E SDescription: Double looped palstave witha shallow flange and thin blade. Butt ham-mered and edge blunted, but casting iswell finished and trimmed. Tin bronze with

Lead.Size: L. 15.9; W. 4.6; Th. 1.95.Unpublished.Comments: Atlantic Bronze Age, c. 1000 -700 B.C.

No. 69. (Fig. 12. no. 69)Provenance: Spain.Registration: WG 929.Donor: Greenwell Collection, presentedby J. P. Morgan.Description: Heavy double looped palsta-ve with very short hafting flanges and a pro-minent stop-ridge, below which are two sha-llow grooves. The square sectioned butt andthe exterior of the palstave have been very

well finished and polished al1 over. Deepbrown patina, excellent condition. Tin bron-ze with lead.Size: L. 18.7; W. 6.0; Th. 3.9.Published: Smith 1920, fig. 164.Comments: Atlantic Bronze Age, c. 1000-700 B.C. (see no. 89).

No. 70. (Fig. 12. no. 70)Provenance: ~Foundin Beariz, mines ofBalcobo, near Carballino., Prov. Orense.Registration: P. 1964. 12-1. 112.Donor: Wellcome Collection.Description: Double looped palstave withprominent rib on blade. Butt end well fi-nished, although the hafting flanges are notcarried through to the end, but blocked off.Well finished and ground down. Some cast-ing imperfections on one face. Heavily leaded

tin bronze.Size: L. 21.8; W. 6.2; Th. 2.5. (Original L.)24.4.Unpublished.Comnzents: Late Bronze Age, c. 1000- 700B.C. Blade broken off and restored withplaster.

No. 71. (Fig. 13, no. 71)Provenance: Northern Portugal, abelievedto have been excavated N. of Oporto.Registration: 1959. 7-11. 1.Donor: Purchased.Description: Double looped palstave withpart of a hollow casting jet a t but t. Edgechipped. and whole surface deeply pittedand corroded. Continuous, low hafting flan-&es, low rib on each face of blade. Leadedtin bronze.Size: L. 19.9; W. 5.25; Th. 3.1.Unpublished.Coinments: Almost certainly associatedwith No. 72, since they have identical patinas

and similar metal compositions. AtlanticBronze Age, c. 1000 - 700 B.C.

No. 72. (Fig. 13, no. 72)Provenance: Northern Portugal, abelievedto have been excavated N. of Oportou.Registration: 1959. 7-11. 2.Donor: Purchased.Description: Double looped palstave withpart of a hollow casting jet at butt. One loopbroken. Continuous low hafting flanges, witha short , low rib on each face of the blade.

Corroded and pitted. Leaded tin bronze.Size: L. 21.5; W. 4.65; Th. 2.95.Unpublished.Comments: Atlantic Bronze Age, c. 1000.700 B.C. (see notes with no. 72).

No. 73. (Fig. 13, no. 73)Provenance: La Baeza, Prov. Len.Registration: 1964. 12-6. 103.Donor: Trechmann Bequest.Description: Broad double looped palsta-

ve with a thinned butt, well finished surfaceand smoothed casting seams, and short, highflanges near the stop-ridge. On each face be-low the stop-ridge is a short, thick rib. Deepgreen patina, excellent condition. Tin bronze.Size: L. 21.0; W. 7.8; Th. 4.6.Unpublished.


27/67

Fig. 12. - Double-Looped Palstaves. 67, .69 , Spain; 68, .41rnagro, Ciudad Real; 70 , Bcnriz, Oronse.


28/67

Fig. 13. - Double-Looped Pnlstaves. 71-72, Northern Portugal, North of Oportoa;73, La Bacza, Lcn; 74, Spain.


29/67

A STUDY OF T H E BRONZE AGE METALConzments: Atlantic Bronze Age, c. 1000-700 B.C. The two halves of the mould werc

not well aligned, hence there are traces ofa thick casting seam where the halves ofthe mould overhung each other.

No. 74. (Fig. 13, no. 74)Provenance : Spain.Registration: 1964. 12-6. 105.Donor: Trechmann Bequest.Description: Fragrnent from hlade of apalstave, prohahly of douhle looped type.

Prominent central rib and wide blade. Welliinished, with an even green patina. Tinbronze.

Size: L. 10.3; W 5.45, Th. 1.65.Unpublished.Comments : Atlantic Bronze Age, c 1000-700 B.C. Fragment broken in antiquity. sincethere is a deep patina over the break.

No. 79. (Fig. 14, no. 75)Provenance: aMontfort>,,Prov. Lugo. (Pos-sibly Monforte de Lemos?)Registration: 1964. 12-6. 102.Donor: Trechmann Bequest.Description: Double looped palstave witha wide haft with low flanges. Butt hroken

off to show a very porous casting. Rib oneach face of hlade. Edge chipped. Castingwelt cleaned of irregularities, and coveredwith a thick green patina, partly scrapedaway. Tin hronze with lead.Size: L. 24.2, W. 6.3; Th. 3.2.Unarblished.Comnzents: Atlaniic Bronze Age, c. 1000-

700 B.C.No. 76. (Fig. 14, no. 76)

Provenance: ~Palencia..Registrat ion: 1891. 9-5. 2.Donor: Presented by A. W. Franks.Descr~ption:Douhle looped palstave, bro-ken off at butt, showing a very porous cas-ting. Casting flashes on loops, hut axe gene-rally well trimmed. Low rib on each face of

blade, edge of hlade ground. Atl the contoursof t6 is axe are unusually rounded comparedto the sharper edges on other Palstaves.Even, dark green patina. Exccllent condition.Tin hronze.

Size: L. 22.4; W. 6.1; Th. 3.7.Unpublished.Comnzents: Atlantic Bro&e Age, c. 1000 -700 B.C.

lNo. 77. (Fig. 14, no. 77)

Provenance: Northern Portugal,


30/67

142 R. J . HARRISON, P. T. CRADDOCK AND M. J. H U G H E S

Fig. 14. - Double-Looped Palntaves. 75, ~Montforta,Lugo;76 ,


31/67

@No. 79. (Fig. 15, no 79) No. 80. (Fig. 15, no. 80 and fig. 21)Provenance: Spain. Provenance: eFound at Oviedon, Prov.

, Registration: P1964. 12-1. 113. Oviedo.1 Donor: Wellcome Collection. Registration: 1876. 8-1. 1.' Description: Double looped palstave with Dono': Presented by A. W. Franks.low midrib on blade. The casting is unfi- Description: Double looped palstave bro-

Fig . 15 . - Double-Looped Palstavci, 79 . Spain, 80, .Found at Oviedoi,.:nished with one loop blocked, both castingjseams untrimmed, and the casting spme in:place. The haft ing flanges are incompletelycast and the surface is full of small air holes.:Leaded tin bronze.; Size:L.23.2; W.6 .0; Th. 4.9.

Unpublished.Cornments: Late Bronze Age, c. 1000 - 700:B.C.

ken into three pieces, with tbe casting jetadhering to it, filled with a ball of pur e lead.The surface is well finished and has a dullgreen patina. One loop missing, the otber isbroken. Leaded tin bronze, surrounding acore of pure lead and capped with a lead ballin the casting jet.

Size: L. 23.6; W. 4.45; Th. 2.85. Jet:Height : 4.8 ; diameter: 4.4-4.65.


32/67

3 44 R . J. H A R R I S O N , P. T. CRADDOCK AND M. J . H U G H E SUnpublished. Unpublished. bComments: Atlantic Bronze Age, c. 1000 - Comments: Atlantic Bronze Age, c. 900 -700 B.C. This is an extraordinary piece, ana- 650 B.C. The type is highly distinctive oflysed by Dr. M. J. Hughes at the end of tbe Central Portugal (fig. 33), and contemporaryCatalogue, where its remarkable technical with the double looped palstaves and Cas-aspects are discussed. It was cast in a two tropo1 type sickles.~ i e c emould with a resin core. then filled.~~~- ~ ~~with lead. The top of the casting jet is No. 83 . (Fig, 16, no, 83)sliebtlv analed, as occurs on No. 78. An axe- . -in the Museo Arqueolgico Provincial, Ovie- Provenance: Spain.

do, found at Laron, Cangas de Narcea (Ovie- Registration: WG 928.do) has al1 of the externa1 features of Axe G ~ ~ ~ ~collection, presentedno. 80, even down to the bronze cap sealing by M ~ ~ ~ ~ ~ ,in the ball of lead in the casting jet. Thisstrongly suggests that technical features so Descripti0.n: Single looped palstave, wellmarked in our axe are not unique or acci- finished, but a porous casting. Filed in mo-dental ones. dern times. Stubs for. a second loop whichwas ncver added are visible on one side.No. 81. (Fig. 16, no. 81)

Provenance: Spain.Regislration: WG 926.Donor: Greenwdl Collection, presentedby J. P. Morgan.Description: Single looped palstave withloop restored in modern times. Low haftingflanges and high stop ridge, with an irregu-lar midrib on each face of blade. Porouscasting, not well finished, but hammered atbutt and apparently used as a tool. Tinbronze.Size: L. 22.0; W. 5.82; Th. 3.8.Unpublished.Comments: Atlantic Bronze Age, c. 1000 -700 B.C.

No. 82. (Fie. 16. no. 82 )

Irregular, thick patina. Tin bronze.Size; L. 12.9; W. 3.5; Th. 2.05.Unpublished.Comments: Atlantic Bronze Age, c. 1000 -700 B.C.

No. 84. (Fig. 16, no. 84)Provenance: ~Found6 feet deep near iJa-velgas, Asturias, Spain, at the edge of an an-cient goldworkings (Prov. Oviedo).Registration: 1927. 12-6. 1.Donor: Presented by C. G. Gunthur.Description: Double looped socketed axewith an irregularly moulded rim, facettedhlade, and two deep grooves cut into eachface, not cast. The loops are worn, and theedge blunted. Well finished. with a finegreen patina over much of the surface. Tin. - . bronze with lead.Provenance : Spain. Size: L. 11.9; W. 4.4; Th. 3.1.Registration: WG 927. Unpublished.

Donor: Greenwell Collection, presented Comments: Atlantic Bronze Age. c. 1 ~ 0 0-by J. P. Morgan. 700 B.C. This axe Lies far to the north o theDescription: Flat backed palstave cast ina two piece mould with clear casting seamson each side. There is a low button cast onone face. The loop is roughly finished butthe but t is hammered and the rest of thepiece is well finished. There is a smallcasting - hoie near the butt. Fine conditionwith a deep green patina. Tin bronze..Size: L. 18.7; W. 3.95; Th. 1.8.

main distribution, which clusters in CentralPortugal (fig. 32). There is, however, thetruly remarkable three-piece bronze nr~uldfor a double looped socketed axe from LosOscos (Oviedo) nearby (Pericot, 1934, p. 223;Blas Cortina, 1975) which suggests thatothers may yet be found in the Asturiasand Galicia to modify the distribution pat-tern.


33/67

. A STUDY OF THE BRONZE AGE METALWORK FROM THE IBERIAN PENINSULA 145. _.

- : - ~ -

82

o o83

34

Fig. 16. - Late Bronze Age Atlantic Metalwork. 81-83, Spain; 84, Navelgas, Oviedo.10


34/67


35/67

A STUDY OF THE BRONZE AGE METALWORK FROM THE IBERIAN PENINSULA

00

--O, , , , ; 1

-'

,-:'/

~ ; ~ ./:;::/

~ ~ \ m "

t_ f # :.IIJI

1. - -

%:/;

8587

-S 86

/

l.. - . _ ~ __ .. ___Vd

1

:/7:/

88 89

Fig. 17. - Trunnion (or Lugged) AXes and Varieties. 85-88, Spain;86, Villacarrillo, Jan, 87, Prov. Jan; 89, probably Spain.

147


36/67


37/67

A STUDY OF TH E BRONZE AGE METALWORK FROM TH E IBERIAN PENINSULA 149

o

90

o

91

I5 -0-92 ~ 9 3 Fig. 18. - Late Bronze Age Atlantic Metalwork. 90, Castropol, Oviedo;91, Prov. Huelva; 92, Prov. Badajoz; 93, Prov. Granada.

Unpublished. No. 95. (Fig. 19, no. 95)PrQvenance: Prov. Jan.Registratian: WG 925.

Camments.: Bronze Age, c. 1500 - 1000 B.C.

I1I

o

The teeth are not set at an angle, and thesaw will not c1ear its own cut (Harrison,1974a). Danar: GreenweH CoHection, presentedby J. P. Margan.


38/67

150 R . J. HARRISON, P . T . CRADDOCK AND M . J . H U G H E SDescription: Thin bar, hammered and cut Donor: Presented by A. W. Franks. ,at each end. Roughly rectangular section.Smooth surface, with a deep brown patina.Copper.Size: L. 21.8; W. 1.1; Th. 0.55.Unpublished.Comments: Probably Argaric, c. 1700- 1000B.C. Quite likely an ingot, or at least a pieceof worked copper used as an ingot, withpieces chopped off as needed.

No. 96. (Fig. 19, no. 9h)Provena,nce: Prov. Jan.Registration: 1909. 7-14. 5.Donor: Purchased.Description: Small flat axe casi in atwo piece mould with the casting seamsvisible on each side. Both faces are deco-rated in relief, each with a diferrent pattern.The butt shows the place where the castingflash has been snapped off. Poorly finished,with a deep green patina. Tin bronze.Size: L. 6.1; W. 3.5; Th. 0.3.Unpublished.Comments: Late Bronze Age, c. 1000-700B.C. These small axes were probably castin batches, and were not intended for prac-tical use.

No. 97. (Fig. 20, no. 97 )Provenance Tabernas, Prov. Almeria.Registration: 1876. 7-6. 1.

Description: Carp's tongue sword, lowerhalf broken away and now lost. Blade deco-rated with six rows of f ine lines on eachside of midrib, carrying into tbe hilt astriple rows of dots. Deep green patina, goodcondition.

Size: L. 31.4; W. 6.3; Th. 1.2.Published: Almagro, 1940.16m. VI, 9 ; Ha-rrison, 1976, fig. 4.Comments: Atlantic Bronze Age, c. 800-700 B.C. See Harrison 1976 for a full dis-cussion.

No. 98. (Fig. 20, no. 98)Provenance: Carcabuey, Prov. Crdoba.Registration: 1876. 7-6. 2.Donor: Presented by A. W. Franks.Description: Rod tanged sword with bro-ken tip. Roughly incised decoration on eachface of blade. Well finished and finely de-corated ricasso on hilt. The tang is a sirn-ple, irregularly sectioned rod hammeredand filed into shape. Blade widens slightlytowards the tip. Deep, even brown patina,excellent condition. Tin bronze.Size: L. 52.0; W. 3.4; Th. 0.55; Tang 1.00;

Tang Length 11.6.Published: Harrison 1976, fig. 1.Comments: Late Bronze Age, c. 1200-1100B.C. For a full discussion see: Harrison,1976.


39/67

A STUDY OF THE BRONZE AGE METALWORK FROM THE IBERIAN PENINSULA 151

95,\ . . ~ _ . I 94

96

o 5Hg. 19. - Metalwork: Various. 94, Niebla,f{uelv


40/67

R . J. H A R R I S O N , P. T. CRADDOCK AND M . J. H U G H E S

REPORT ON THE SCIENTIFIC EXAMINATION OF THE BRONZEPALSTAVE (No. 80) FROM OVIEDO, SPAlN (F IG S 15 + 21).

This palstave with its lead core is amost unusual object. Its main featuresare a thin outer layer of a copper alloycast to the shape of a palstave with a cast-ing jet still part of the palstave at thebutt end. The whole of the interior of thepalstave is filled with lead, and has anunusual feature in the cast ing jet in that

there appears to be a discontinuity bet-ween the lead filling the palstave and thatfilling the casting jet (in which the leadis almost spherical in shape). The castingjet has a flat copper alloy top with silaperture in the centre, and there appearto be joint marks in the copper as if itwas added subsequently (fig. 15).

A semi-quantitive estimate of the com- mic absorption spectrometry (AAS). Theposition of the different parts of the pals- location of: each of the 5 numhered sam-tave was obtained using X-ray fluorescen- ples is shown on fig. 21.ce, and quantitative analyses were then The results o the quantitative AAScarried out on drilled specimens of metal analyses were as follows, expressed intaken from the difference par ts using ato- weight percentage:

Sample N O . L e s d Tir, ron zincCooper-- -1. Casting jet, outer metal (top) . . . . . 66.0 22.3 6.18 0.04 0.0152. Casting jet, outer metal (sides) . . . . . 61.9 29.1 4.75 0.10 0.0113. Casting jet, lead core . . . . . . . . . 6.97 Rem* 0.76 - -4, Palstave, lead core . . . . . . . . . . 1.37 Rem* 1.47 - -5 . Pdstave body, outer metal . . . . . . . 66.8 22.9 4.75 0.18 0.011

(Rem* indicates that the balance of the analysis to 100% is lead)

a) Palstave sition. Similar alloys were used to makeRoman statuary bronze, and it is suitedThe outer metal of the palstave is a to large castings of that type rather than

heavily leaded bronze containing 22.9 % to smaller objects such as tools or wea-lead and 4.75 % tin. No other metals con- pons. Examination of the break which oc-tribute significantly to this alloy compo- curs at its mid-point suggests that the


41/67


42/67

154 R. J. HARRISON, P. T. CRADDOCK ANO M. J. HUGHESwould have contained a core. If the capwas cast on to the top of the castingjet with a further pouring of bronze, itwould have been necessary to leave a holeto remove the core, and this may be thesignificance of the hole in the cap.

C) Lead CoreAfter removal of the clay core from

the leaded bronze palstave molten leadwas poured in. The melting points of leadand leaded bronze are sufficiently diffe-rent that molten lead entering the palsta-ve would not cause the outer metal tomelt. The curious appearance of the leadin the casting jet, an almost sphericalglobule about 3 cm. across, and the appea-rance of a discontinuity at the bottom ofthe casting jet hetween the globule andthe lead core of the palstave suggests thatthe lead core was added in two stages.

Examination of some unusual materialin the top of the casting jet immediatelyabove the butt end of the palstave sug-gests a possihle reason for this, and thematerial itslef is of some importance. Thematerial in question is a brown-coloureddeposit which is readily scraped awaywith a scalpel to yield tiny fragments ofa brittle crystalline-like substance. Someinitial tests on this material made it ap-parent that it consists of a mixture of asandy deposit and an organic binder whichis readily soluble in acetone. Some of theorganic binder was obtained by leachingthe material with acetone and analysingthe dried extract by infra-red analysis. Thebinder is a resin, and although chemicaldegradation of the resin does not allowits precise identification, coming from aMediterranean country it is most probablyderived from pinus species of trees. Thecrystalline material left after acetone lea-

ching was examined in a mounted sectionunder the polarising microscope. It iscomposed predominantly of angular tosubangular quartz, roughly 100-200 mi-crons in diameter with some other minoramounts of minerals such as felspars andzircon. The fact that the quartz particlesare angular rather than.rounded suggeststhat they are not derived from a beachsand, but may possibly have been produ-ced by gentle crushing of slightly largergrain-size quartz sand. The angular shapeof the particles may assist the action ofthe hinder in holding the material to-gether.

From the occurence of the core mate-rial inside the base of the casting jet, thefollowing sequence of events in makingthe palstave may he suggested:

Stage1. The bronze casing is made in a

two-part mould with a core of a sand/re-sin mixture.

2. On cooling of the bronze, the coreis removed mechanically and also by war-ming the bronze, which caused the resinto melt and run out of the castifig jet;the process was not complete and somemelted core/resiil mixture remained in thecasting jet.

3. Molten lead was poured into thebronze casing. The heat caused the resi-dual core material to flow downwards un-ti1 it collected at the bottom of the cast-ing jet, sealing off the lead in the pals-tave. Further lead was then poured in, pro-hably later, to fill the casting jet.

The composition of the lead which fillsthe palstave (4) (6.97 % copper, 0.76 % tin)is different from that which fills the cas-ting jet (3) (1.37% copper, 1.47 O/o tin),


43/67

l A STUDY OF THE BRONZE AGE MErALWORK F R O M T H E IBERIAN PENINSULA 155and the presence of the melted core (Sta- The copper and tin contents o the leadge 2 above) suggests a reason for this core are so different for the two parts of7- the pouring had to be broken off be- the palstave that they probably representcause of the melting of the residual core. different batches of lead.

The palstave has a leaded bronze cas- mains, which proves to be of angularing and a lead core which was probably grains of quartz set in a binder of resin.added in two stages. In the casting jetsome of the original casting core still re- R. L. File No. 3697

1 E ) T A B L E ~OF ANALYSESI'RE-BEAKER AXE S

I3EAKER AXE

I3EAKER DAGGERS

IPALMELA POINTS


44/67

R . J . H A R R I S O N , P. T . CRADDOCK AN D M . J . H U G H E S

R J H

ARGARIC AXES

ARGARIC DAGGERS

ARGARIC SICKLE, CHIS ELS, AND SAW


45/67


46/67

158 R. J. HARRIS ON, P. T. CRADDOCK AN D M. J. HUGHESATLANTIC SWORDS

a) The Background impossible to comment upon their validi-ty or to use them.The most important analytical work since the 189Vs there have been enor-

on Iberian Bronze Age pieces is that Pu- advances in analytical chemistry,blished by Henri and Louis Siret in 1887, and it is now clear that the standard me-by Louis Siret in 1913, and by the Stutt - thods previously used for brbnze analysisgart team in 1960 and 1968 (Junghans, were in error (Caley, 19.67). The methodSchroder and Sangmeister, subsequently lhen used for determining the ti n contentreferred to as SAM). Apart from these re- in bronze was to treat *he sample withsults, there are little other reliable data nitr ic acid, which left the ti n as an in-her re bOth the analyses and also the de- soluble complex hydrated oxide; this wastails of sampiing, the condition of the me- ignited and weighed as the tin content.

the methOd analysis adO~ted+Unfortunately, some of the other metals,which elements were es ~ e c i a l l ~sought af- notably the copper, were also broughtter, the sensitivity of the method, and an down in small amounts, and this gave ti nestimate of the standard deviation of the results which were about 10 % high,results, are published. Without this infor- ~ h i ~may seem a small error, but themation the results by themselves are of implications can be seen if we take onelittle use. For example, recently the ana- the SiretsSanalyses for illustration (Si-lyses of five Late Bronze Age looped pals- ret and Siret, 1887; p. 232). The reportedtaves were published (Eiroa, 19731, which composition for a small ingot metalpurported to show that they contained from ~1 A~~~~ (no, 90) was:between 20.2 O jo and 48.2 % of tin, andfrom 0.6 % to 3.0 % of zinc. ~u rt he rm or e, copper 29.87 % (revised) 32.85 %the total metal contents lie between 197786 O/o and 95 %. When compa,red with tin 36.21% (revised) 32.59 %the values obtained here, and also by SAM 1977and by the Sirets, these tin and zinc va- lead 20.84 % ............ 20.84 %lues are very high indeed. UnfortunatelyEiroa provides no discussion of the re- If the results are 10 O /o too high forsults nor of the methods by which they tin and correspondingly too low for co pwere obtained, and in consequence it is per, the ingot contained about 33% of


47/67

Etoth copper and tin; a very different alloyfrom that reported. This analytical errormust be borne in mind when using theS,irets' otherwise excellent results.

h

b) Sampling,' The samples were taken with a 12 voltiportable jeweller's dril1 fitted with a size

60 hardened steel bit. These bits are appro-ximately 1 mm. in diameter and one bitcould be used to take about ten samplesb,efore becoming too blunt for further use.?ypically the bit penetrated about a centi-nietre into t+e metal and removed bet-ween 10 and 20 milligrams of sample.The surface metal and any corroded metalwas avoided, and the drillings were onlycollected when bright shining turnings ofnieta1 were produced. A few of the arti -fiicts included in this catalogue are tota-1ly corroded and thus no sample could beo:btained. It is very dangerous to analysectxroded metal and assume it has thes;ime composition as the uncorroded me-tal once had (Caley, 1964). The positionsfrom which the samples were taken arerecorded on the illustratioris of the bron-zes by an enlarged dot, following the ad-mirable practice of Siret who so markedhis sample positions in 1913. It will benoticed that the samples are taken fromcsntral regions of the casting where thenlost representative composition of thecpsting as a whole should be found.

The samples were analysed by AtomicAbsorption Spectrometry, using a PerkinE.lmer 306 double beam spectrometer. Theelements arsenic and bismuth were ana-lysed using the spectrometer with a fla-rieless atomiser with Deuterium arc back-gjround corrector. For a detailed descrip-tion of the use of Atomic Absorbtion Spec-trometry in the scientific examination of

antiquities and for full and exact detailsof the methodology followed in this worksee Hughes, Cowell & Craddock (1976).The results are expressed in percentages;most elements could be detected down to0.005 % in the metal, except for lead andtin which were only detected down to0.01 %. The standard deviation of the re-sults is & 1% for the major elements and-t 30 % for elemeilts present in amountsbelow 0.2 % in the metal. This may seeman unduly pessimistic estimate of the pre-cision of the results, especially for thetrace elements, but the results of recentinterlaboratory comparisons showed thatit is iil fact a realistic estimate oE theprecision (Chase, 1974).c) Vaviations in the composition of

bronzesThere are a number of inhomogenei-

ties in the composition of the bronzeswhich affect the reliability of the analy-ses, the main ones being caused by se-gregation and


48/67

160 R. J. .HARRISON, P. T. CRA.DDOCK ANO M. J. H U G H E Sfor gravity segregation. Three have simi-lar lead contents at each eild, suggestingfairty rapid cooling. Axes 71 and 79 bot'hdisplay gravity segregation with a higherlead content in the blade than at the butt.Axe no. 70, however, had a higher leadcontent a t the butt rather than at the bladeend of the casting. Unfortunately the bladeitself is missing, but it seems improbablythat the metal was poured from that end(fig. 12). Casting from the butt end is inany case shown by the sprue cups whichare still attached to several palstaves.

Normal segregation was investigatedin the two tin bronze swords (nos. 97-98)from samples taken along their lengths.Swords were cast from the hilt withthe blade pointing down. During the cast-ing of an alloy the solidification usuallyproceeds from the bottom and sides ofthe mould upwards, leaving the materialaround the sprue-cup to solidify last. NO-wadays, as in the Bronze age, moulds aredesigned to encourage this so that gasbubbles and solid impurities will be ca-rried up out of the casting and into themetal in the sprue cup which will be dis-carded. However, this does encourage thephenomenon of normal segregation. TheEirst metal to solidify tends to be rich inthe component of the alloy with thr !ti-ghest melting point, iil this case the cop-per (1,083"C.) and the last metal to soii-dify tends to be rich in the metal withthe lowest melting point, in this case tin(232 "C.). The decrease in the tin contentaway from the hilts of the two swordsshows normal segregati011 to have takenplace. The tin content for the axes alsoshows normal segregation, but to a lessc~degree, as they are considerably shorter.

It can be seen from these examples ofgravity and normal segregation that thecomposition of an alloy is often not uni-

form throughout the casting. Great cau-tion must be used in relating the compo-sition of a sample to the composition ofthe object as a whole.

The other common form of segrega-tion in bronze castings is inverse segre-gation where elements such as arsenic, tinor lead are locally forced to the surfaceduring the final cooling of the melt. Withbronze this phenomenon is known as tinsweatingn and no complete explanationwhy it occurs has been given. It seemsthat during the final cooling o the metal,the still molten, tin rich alloy left insidethe casting is subjected to enormous pres-sure from the contracting shell o the cast-ing, and the molten metal may be forcedthrough the forming metal dendrites tothe surface (Bailey, 1960). Errors in esti-matiilg the composition of an alloy fromthis cause can be easily avoided by notanalysing surface metal even if clean anduncorroded.

Discussiona) Pre-Beaker Industry (figs. 1, 22)

Of the six axes included in this indus-try on typological grounds, five are of re-latively pure copper and one is an arse-nical copper with 2-3% arsenic. These re-sults are similar to those obtained by theSAM team on objects from Los Millaresaild Alcala, and the use of unalloyed cop-per is of great interest. A similar patterilcan be discerned from other SAM analysesf ro~na comparable Chalcolithic industryin Italy, where the flat axes from Reme-dello contexts have strikingly low arsenicvalues. Arsenical copper is also known inRemedello contexts, but seems to havebeen retained only for used in knife anddagger blades (Barker 1971).


49/67

Blance (1961) and Almagro and Arri- likely to be related, since the Millaran andI~a s(1963) argued that the Millaran cultu- Remedello industries are much more pri-1:re with its early metallurgy and fortified mitive. Renfrew (1967a) has revived theisettlements was introduced from the eas- idea of an independent origin for Milla-i:ern Mediterranean, probably the Aegean. ran metallurgy, and the metal analysesIBut of the thirty Early Bronze Age axes lend strong support to the hypothesis.

Fig. 22. - Arsenic Contents of a) Millaran Flat Axes,b) Reincdello Flat Axes (Italy), c) Contemporary Flat Axes in th e Aegean.

from the Aegean published by Renfrew(1967), none are of pure copper, and thir-teen a re of arsenical copper with arseniccontents varying from 1.1% to 6.24 %(Hg. 22). This composition suggests thecleliberate addition of arsenic, and is clear-ly different from that of the Millaran andP:emedello axes, where the small amountsO arsenic almost certainly derive fortui-i,ously from the copper ore. Thus the Ae-gean axes are of alloyed metal, whereasthe Spanish and Italian (Remedello) axesare simply copper, which is a major tech-nological difference. This suggests thatthe two metal-working traditions are un-

11

b) Beaker and Early Bronze Age Indus-try (figs. 23-24)111 post-Millaran times, arsenic began

to be used as a regular alloy with copper,and arsenical coppers are especially cha-racteristic of Beaker and Argaric indus-trial traditions. AIthough there is no di-rect evidence to show how the arsenic wasintroduced into the copper, there is littledoubt that is was done deliberately (Char-les, 1967), and the final composition ofthe metal carefully controlied (Craddock,1976). To have simply relied on the for-tuitous arsenic content of the copper ores


50/67

162 R. J. HARRISON, P. T. CRADDOCK AND M. J. HUGHESsurviving the smelting processed (where per; if the arsenic compound was simplythe majority of ars2nic would inevitably added to the copper ore, and the two smel-have been lost during the oxidisation roas- ted together, there would he a possibili-ting of the copper sulphides prior to smel- ty of arsenic loss and no chance at al1 ofting) would not produce the regular series controlling the quality of the alloy. It isof arsenic contents seen in the metalwork. probable that tin bronzes were formed in

3 ig . 23 . - Arsenic Contents 01: a) Palmela Points; b) Argaric Flat Axen

The production of deliberate alloys forparticular types of artefacts such a s Pal-mela points (Harrison, 1974b) or daggers(fig. 23) suggests that either arsenic me-tal or a relatively pure arsenic compoundsuch . as Reaglar (AS&) or Orpiment(AS&) was used.

The use of arsenic metal is unlikelysinceno ancient metallic arsenic has beenfound;;and native arsenic, although foundin volcanic'deposits, is rare, and metallicarsenic is difficult to produce. Thus it isprobably that the arsenic copper alloyswere made by' adding some reasonablypure arsenic'compound to the rnolten cop-

the same way by adding Cassiterite (SnO,)to the molten copper, for although pre-historic artefacts of pure tin are knownfrom the Aegean, tin ingots or scrap tinfrom hoards or working areas are quiteunknown (Charles, 1975). 10 kg. of cassite-rite with copper ingots found with a cru-cible containing partly reduced cassiteriteand copper metal from Sardinia (Cambi1959) of Late Bronze Age date stronglysupports this hypothesis.

The flat axe from Carmona (no. 12),probably from the Beaker deposits onthe site of El Acebuchal, is a tin bron-ze with 5.9 % tin. Although rare, tin bron-

t'


51/67

A STUDY OF TH E BRONZE AGE METALWORK FROM TH E IBERIAN PENINSULA 163zes do occur in Early Bronze Age contexts in the Peninsula, and several Palme.la points of bronze and sorne awls wereanalysed by the SAM teams. Axe no. 12also contains 1.45 % arsenic, which is

another explanation has been offered byCharles (1967), who pointed out that arsenic is an excellent deoxidant of copper,as it readily combines with oxygen in themetal. The arsenious oxide so formed is

Fig. 24. - Distribution o Palmela Points in the Iberian Peninsula. Ringed Circles indicate several examples.

above the normal arsenic content found incopper to which it has no t been deliberately added. Alloys containing both tinand arsenic have been found elsewhere,particularly during the transitional periodfrom arsenic to tin alloying (Craddock,1976). The usual, and most likely, explanation for them is the fortuitous use ofarsenical copper scap as a source of copper in the making of tin bronze. However,

highly volatile, and consequently expelledfrom the metal. Tin, on the other hand,is not a good deoxidant, because the tinoxides formed are stable and stay in themetal. Modern bronzes therefore almostinvariably contain a small quantity of zincor phosphorus added to deoxidise the copper befo re the additional tin is added.Thus, while a tradition of working witharsenical coppers still persisted, sorne


52/67

164 R. J. HARRISON,P. T. CRADDOCK AND M. J. HUGHESsmiths may have continued to add arse-nic to deoxidise the metal.

The two Beaker tanged daggers(nos. 29, 31) and the ten Palmela points(nqmbers 32 and 41) are al1 of arsenicalcopper, with only traces of other metals.~ h karsenic content varies between 1.2 %and 4.6 %, averaging 2.3 % (fig. 23). Thisagrees closely with the SAM results wherea 2 % arsenical copper was the pre-ferred alloy for Palmela points, in spiteof their very wide distribution in the Pe-ninsula (fig. 24).

c) The Argaric ZndustryArsenical copper continues as the pre-

dominant alloy al1 through the Argaric pe-riod, and tin bronze is notably scarce. Itis one of the curious and interesting fea-tures of Argaric metallurgy that , in an areaof such abundant metal resources, arseni-cal copper alloys and single stone mouldsshould persist long after going out offashion elsewhere in Europe, and even inother parts of the Peninsula. Tin occursin small alluvial deposits around Cartage-na (Prov. Murcia), and around Cadiz, bothareas of occupation by early metal wor-kers (Allan, 1970). The main tin depositsare in the NW of the Peninsula, however,and they do not seem to have been wi-dely exploited until the later part of theEarly Bronze Age. It is possible thatbecause tin only occurs in Southern Spainas small, sporadic deposits, the supply wasliable to be erratic, and that most of thetime the smiths would have had to relyupon arsenic which, unlike tin, occurs withthe copper (Harrison, 1974b; p. 79).

The eight Argaric axes (n.OS 7-11; pp. 13and 26) are predominantly of arsenicalcopper, although there are examples of tinbronze with correspondingly lower arsenic

contents (nos. 19, 26). The arsenic con-tent o the sixteen flat axes varies from1.2% to 3.9 % averaging 2.0 % (fig. 23b),which is remarkably similar to the arse-nic content of the Palmela poiqts. Axesare characteristic of burials in phase B(1500/1400- 1100 B.C.).

The Argaric daggers (Schubart, 1975a)and the sickle (no. 53) are also of arseni-cal copper, with the exception of no. 57which is a tin bronze. It is important tonote that tools which need to be hard toretain a sharp cutting edge tend to havemore arsenic than the axes or points,which need to be tough to withstandshock during use. Thus, the daggers havea high arsenic content from 2.2 % to4.7%, the average being 3.7 %. These ana-Iyses demonstrate not only that arsenicalcoppers were being deliberately produced,but that the smiths could control the finalalloy composition with considerable accu-racy. Tylecote (1962) showed that arsenichas little effect on the copper in the caststate, and only with hammering and wor-king is the strength of the alloy increased.It is unlikely to be a coincidence thatmost daggers and nearly al1 Palmelapoints show extensive signs of cold wor-king.

Dagger no. 49 has rivets of arsenicalcopper, but the rivets of dagger no. 55 andknife no. 53 are of silver.

Silver became popular in the later Ar-garic (B) period, and silver rivets andrings are quite common. Native silver me-tal only occurs naturally in insignificantquantities, although rather more does oc-cur as the natural gold-silver alloy, Elec-trum. The Sirets found deposits of nativesilver a t Herrerias (Prov. Almera) c...for-m de ramifications brillantes d'argentcristais.,, (Siret and Siret, 1887, p. 231)which, when analysed, yielded 89.62 % sil-


53/67

A STUDY OF THE BRONZE AGE METALWORK FROM THE IBERIAN PENINSULA 165ver and 10.20 % impurits (Sable, calcaire, etc.)>>. The most common ore of silveris the sulphide, silver glance. In Spainanother important ore is Cerargyite, orRom Silver (Silver chloride), which oc-

goes off as lead oxide. The silver is leftbehind .as a button of metal in the cru-cible.

I t is interesting to see that the rivetsof dagger no. 55 contain about 0.5 % lead

30 0

Fig. 25. - Distribution o Early Bronze Age


54/67

166 R. J. HARR ISON , P. T. CRADDOCK AND M. J. HUGHESwas a large increase in the amount of me-tallic lead in use. It is known from Arga-ric sites, and at El Oficio several frag-ments and a small, flat lead cingotn weigh-ing 52 gr. were found (in Houses E andP respectively (ibid. 192-4). Until theArgaric B period little silver or leadwas available, but after 800 B.C., leadedbronzes became common in AtlanticEurope.

The remaining Argaric pieces are thepair of chisels (nos. 61, 62) made of arse-nical copper in order to provide a hard,sharp edge, and a curious saw (no. 94),which is also of arsenical copper.

Otker Early and Middle Bronze AgeMetalworkThis is a mixed but interesting group

of pieces, including a riveted halberd ofarsenical copper where the blade has fourtimes more arsenic than the rivets(no. 30). It is noteworthy that the hal-berds in both Atlantic and Argaric Indus-tries are of high arsenical copper (Harri-son 1974a), and that the same alloy waspreferentially selected for halberds in theBritish Isles and Ireland (Coghlan andCase, 1957, p. 96). The distribution of hal-herds shows a scatter over much of thePeninsula (fig. 25).

The flanged axe (no. 28) from Len isunusual since axes of this form are essen-tially British o r French types, and usuallycontain 10% or more of tin. The purecopper composition is puzzling, since tinor arsenic would be expected t o be alloyedwith it, and although the axe has a wor-ked blade suggesting it was intended fornormal use, it would have been a poortool.

The small 4umber of arrowheads(nos. 42-45) are al1 of arsenical copper,

and emphasize the scaricity of tin bronzegin the Early Bronze Age of Iberia.

Tke Atlantic Bronze IndustryNearly al1 the pieces are of Late Bron-

ze Age date (1000-600 B.C.) and their pe-culiar interest lies firstly in their compo-sitions, and secondly in their provenances.

With one exception (no. 78) they areal1 of good tin bronze, with very variableamounts of lead. The looped palstaves(nos. 63-81) contain most lead, the othertypes (82-95) have rather less, while thearrowheads, spearheads and swords (nos.48, 91-93, 97-98) have almost none.

a) Palstaves (nos. 63-81)The tin content varies from 4.2 %-

13.7 %, with an average of 8.6 % (fig. 26),while the lead content for the same axesvaries from O %-33 % (fig. 27). It is ins-tructive to contrast the normal distribu-tion of the tin with the more scattereddistribution of the lead. Siret publishedsome very similar results as long ago as1913, where double looped palstaveseither had a low lead content, or one inexcess of 20 %.

The unprovenanced palstave (no. 78)does not contain any tin, but is a highlyleaded arsenical copper. The arsenic maybe derived from the copper, or from theuse of ancient scrap metal. In either case,the alloy is a most peculiar one and theunfinished nature of the blade suggeststhat the palstave was never intended forserious use.

Some of the axes analysed by Siret,and palstave no. 80 from Oviedo in ourseries, still have their lead-filled spruecups attachtid. This lead filling penetratesinto the body of the palstave itself, and


55/67

A STUDY OF THE BRONZE AGE METALWORK FROM THE IBERIAN PENINSULA 167Siret did sorne experiments to find outthe possible reasons for this. (Siret, 1913,p. 353). He made a duplicate palstavemould, poured in bronze, and whilst thelatter was still molten, poured lead intothe topo As a result, the lead not only fi-lled the sprue cup, but being heavier than

...............

...............

...............

...............

for use as functional axes. The reason forthe high leadcontent is puzzling.

Lead could also have been added tocopper in order to de-silver the latter. Thecopper and lead are melted together, andthe mixture cooled to about 500 oC. At thistemperature, the copper begins to solidi-

J. 'fn - , r - r - r - - - t ' - "...+' - : - - F : I , : i : : = I : : " " " : { : = : : : ~ : : F : : : : : : : " T ; : : : : 7 . : ' : : : : : T ~ : : : : : = : : : : : I - + : . . . . : . r ~ %'0 u Sn. o'0

iNo.' ~ ~ ~ ~ ~ : ; ' : ' : ' : : ' ; : : " : ' : ' : t - - . - . - . - r - - . . . . . - . - - - r - - ~ " ; ; ; ; ; " I - , . - , . - , : . . . . . , ~ ~ p . ; ; . . - l ~ ~ ~ - l : : : : : : : : : : : : : : : : : ' ~ 4 - + " ' 4 - - - r - : : - : - o o 10 1S 20 35%Pb.-

Fig. 26. Ti n Content Late Bronze Age Atlantic Looped Palstaves (upper).Fig. 27. - Lead Content o Late Bronze Age Atlantic Looped Palstaves (lower).

the bronze, i t sank into the still moltenbronze, reproducing the effect he had noted on the ancient palstaves. However, theOviedo palstave studied by Hughes(no. 80) was even more extraordinary. I thad a core of sand bound with resin,which was removed before the lead wasadded. 1t was only after the bronze hadse t that lead was poured into the cavityleft by the coreo Many of the heavily leaded bronzes are very rough castings, andit is obvious from their unfiilished statethat they could not have been intended

fy, bu t the lead remains molten, and thesilver alloyed with the copper passes intothe lead. The last stage is to cupellate thelead to recover the silver from it. Thismethod of silver recovery was practisedfrom Roman times onwards in Spain, andmay have been introduced by the Proenicians. But the highly leaded bronzes analysed he re do not seem to have been connected with de-silvering processes. Firstly,the lead does not have a high silvir content, which would be expected if the method had been used, and secondly, the


56/67

168 R. J. HARRISON,P. T. CRADDOCK AND M. J. HUGHEScopper left after the de-silvering should his alloy would be to make the moltennot have such a high lead content. metal flow more easily into the mould.Small amounts of lead have no appre- Highly leaded bronze alloys are rarelyciable effect on the workability of the me- met with in Europe until the Late Bronzytal, but the quantities such as those re- Age, and then notably in North West Ibe-

Fig. 28. - Distrib ution of Doublo-Looped Palstaves in the Iberian Peninsula . Ringed Circlcs indicate moulds:Salid Triangles indicate hoards; Open Symbols indicate uncertain provcnances(after information suppliod by R. Hardaker).

ported here (ie; over 20 %) make the me-tal almost impossible to use because thelarge areas of lead interspersed amongthe bronze form regions of weakness, andany attempt to work, or even use,,the me-tal results in tearing along the boundariesbetween the lead and the bronze. Sincelead does not dissolve in copper, the mainreason for a prehistoric smith to add it to

ria, Brittany and the British Isles (Bur-gess, 1968). In Galicia and Brittany leadedbroilze palstaves occur in large hoards,and the reason why such appareiltly non-utilitarian axes were made is difficult toappreciate now. The colour and appearan-ce of the castings would make it impossi-ble to deceive a potential customer intoaccepting them as ordinary tin bronzes,


57/67

A STUDY OF TH E BRONZE AGE METALWORK FROM THE IBERIAN PENINSULA 169

el'

oFig. 29. - Distribution of Double-Locped Palstaves in Western Europe.

(after Savory, 1949; 1958; Cordier & Gruet, 1975, fig. 58).


58/67

170 R. J. HARRISON,P. T. CRADDOCK AND M. J. HUGHESand as the lead-copper mixture was use- region, and some of them, notably theless to the Bronze Age smith, it is difficuli socketed spearheads and swords, are ac-to interpret them as ingots. Greek and complished pieces of work. Of al1 theEtruscan s tatuettes of the VII-VIth. cen- kinds of tools and weapons made by theiuries B.C. occasionally contained large Atlantic industry, only the looped palsta-

F ig . 30 . - Dis tr ihut ion ai Si~rgle-LoopedPalctaves i n thc iberian Peninsilla. Synibols as fo r f i g . 28.

amounts of lead, and later Hellenistic andRoman statuary bronze regularly had ahigh proporti011 of lead in order to helpwith the difficult casting (Craddock,1976). But double looped palstaves do notseem to fa11 into the carne category asfine statuary. and they ar e notably simplepieces to manufacture. Furthermore, hightin bronzes with no lead a t al1 were beingmade at the same time and in the same

ves seem to have been mass produced,and the incomplete state of so many ofthem suggests that the work was donehurriedly and without any interest in pro-ducing a well-finished article. In contrast,the careful burial of so many of thehoards suggests that they may have beenmade for votive deposit connected withsome ritual now lost to us.

The double-looped palstaves are wide-


59/67

A STUDY OF TH E BRONZE AGE METALWORK FROM THE IBERIAN PENINSULA 171ly scattered over Atlantic Europe (fig. 29),and have long been cited as evidence formaritime contacts between Iberia and theBritish Isles (Childe, 1939; MacWhite,1951). The few British pieces have been

cely more numerous but do seem to becloser to the Iberian types than are theBritish ones, and are very probably theresult of dlrect contacts with the IberianPeninsula.

Fig. 31. Distribution o Class II Castropol Sickles. Ringed Circle indicates hoard(after information supplecl by R. Hardakcr).

re-studied by Savory (1958, 1968), and dated italics than the Iberian ohes, since theBritish axes are known to occur in localMiddle Bronze Age horizons between1200-900 B.C. The hoard from West Buck- oland, Somerset, contained a pair of double-tooped palstaves, a ribbed braceletand a spiral-twisted bronze torc (Rowlands, 1976). The French pieces are scar-

The single and double-Iooped palstaves are largely contemporary, althoughthe doubh!-looped type appers to be themore numerous and widespread (figs. 28,30). In the small hoard from Covaleda(Prov. Soria), both types are associatedwith a late form of a lugged axe and aspear ferrule (Fernndez Miranda andBalbn Behrmann, 1971, fig. 2).


60/67

172 R. J. HARRISON, P. T. CRADDOCK AND

Documents

bronze age metalwork