Sandars 1993 - Later Aegean Bronze Swords

Later Aegean Bronze Swords Author(s): N. K. Sandars Reviewed work(s): Source: American Journal of Archaeology, Vol. 67, No. 2 (Apr., 1963), pp. 117-153 Published by: Archaeological Institute of America Stable URL: http://www.jstor.org/stable/502611 . Accessed: 08/03/2012 05:39 Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Archaeological Institute of America is collaborating with JSTOR to digitize, preserve and extend access to American Journal of Archaeology.

http://www.jstor.org

Later

Aegean

Bronze

Swords*

N. K. SANDARS PLATES

21-28

This was a fragile attachment of ivory, faience or riveted to the blade and lower In an earlier article' I attempted to sketch the semi-precious stone, the splay of the shoulder "horns" being degrip; origins, and the range in time and space, of the to deflect the enemy's point from the hand. first Aegean swords. These origins led back to Asia signed The "pinched-up"flanged shoulder of the B sword in the later third and early second millennium, on served a similar purpose. the one hand, and to a native Aegean invention Both these swords were still in use in the Aegean of the Middle Minoan period, on the other. The in the fifteenth century, though finds of weapons result of this fusion was the A type sword of Crete earlier part of the century are few. There and the Greek mainland; a weapon of unsurpassed from the is no need to enter into the problem of the bottom size and considerable magnificence; also a less date proposed for Schliemann's grave-circle at splendid but potentially more serviceable weapon, for without recourse to it we have the Mycenae, the B type sword concentrated in the Argolid. from Kakovatos (LH II A), and fine A sword These were the two swords in the hands of the from tombs in Messenia, of like date; swords tholos sixteenth century Aegean warrior. for while evidence survival of these swords into The A swords, sometimes over a metre long, had second half of the fifteenth century is provided the a fatal weakness in the hafting, a slender tang that is often found snapped, and would have left the by the rather untypical A sword in grave 44 of swordsman unarmed, with a useless hilt in his the Zapher Papoura cemetery Knossos, and a quite hand. The shorter, stouter B swords, with their typical B in the "King's Tholos" at Dendra (LH flanged tangs, should have been less vulnerable, IIIA i final date). There is also less direct evidence but there too the great size of the rivet-holes left in chamber tombs at Mycenae and in the Dodecathe metal weak, and almost as liable to snap. An nese, which will be discussed below. In the earlier article I concluded that the eviattempt was made to protect the swordsman's hand from an antagonist's blade, in close fighting, by dence pointed to the Argolid as the home of the the adoption of a horned type of grip on A swords. B type swords, while the A type were produced PART 1

* I owe a great debt of gratitude to the Principal and Fellows of St. Hugh's College, Oxford, for the award of the Elizabeth Wordsworth Studentship 1958-61, which enabled me to travel in Greece and Turkey, and to complete the research necessary for writing this essay. I have received very valuable advice and a number of suggestions from Miss Dorothea Gray, and from Dr. Hector Catling, to both of whom I owe especial thanks. I have had helpful discussions with others, including Mr. M.S.F. Hood, Director of the British School at Athens, Mr. John Boardman, Mr. A. Snodgrass, Mr. George Huxley, Mr. V. d'A. Desborough and Mrs. Maxwell-Hyslop. In Greece I received valuable help from Mr. and Mrs. Karouzou at the National Museum, Athens; from Dr. N. Platon, the Ephor for Antiquities in Crete, and Mr. S. Alexiou, present Ephor; also from Dr. S. Dakaris at Ioannina, Professor Morricone in Cos, Dr. P. Astrom of the Swedish Institute in Athens, Mr. S. Iakovides of the Greek Archaeological Service, and from the Director and Secretary of the American School in Athens. For assistance in Turkey I am grateful to Mr. Seton Lloyd, then Director of the British Institute in Ankara, to Bay RacI Temiazer, Director of the Archaeological Museum, Ankara, and to Dr. Nezih Firath of the Classical Museum in Istanbul, to Bay Muzaffer Ramazanoglu and Bayan Seyyide Celikkol at Adana, and to the Secretary of the Archaeological Museum in Jerusalem and the Director of the

Damascus Archaeological Museum. I have to thank for much help and facilities Dr. Marie-Louise Buhl, Keeper of the Antiksamlingen of the National Museum, Copenhagen,and Mrs. H. Salskov Roberts of that museum, also Dr. O. Vessberg of the Museum of Mediterranean Antiquities, Stockholm, to Dr. Z. Vinski of the Archaeological Museum, Zagreb, and to Dr. Gabrovec at Ljubljana; Professor Mikov gave me much help at Sofia, also Dr. Detev at Plovdiv as did Drs. Horedt and Rusu of the Archaeological Institute, Cluj; also Professor B. Brea and Mrs. M. Guido at Syracuse. Finally I am grateful for help and facilities at the British Museum to Mr. D. E. L. Haynes, Keeper of the Department of Greek and Roman Antiquities, and Mr. R. Higgins of that department, and to Mr. R. V. Nicholls of the Fitzwilliam Museum, Cambridge. I am particularly grateful to the Directors and Keepers of all museums referred to for permission to draw and publish the swords illustrated on plates 21-28. I also have to thank ProfessorE. Vogt of the Swiss National Museum, Ziirich, Mr. J. D. Cowen, and Mr. L. Grinsell of the City Museum, Bristol, for advice, and I am grateful to Mrs. E. M. Cox for the finished drawings. The drawing of the horse-bit (pl. 26:48) is by Audrey Corbett, to whom I owe thanks. 1 N. Sandars, "The First Aegean Swords and Their Ancestry" AJA 65 (I96I) 17-29.

OKG

O

Ci

* Cii Add Greek 2 Mainland" L Di Add Crete 5, Rhodes I, France ?1 A El Add Add Greece Crete 21 A Eli 0 F

Add EnglandI

XH K KG KG Bulgarian BulgarianSpears

Note: untypicalD i etc. swords not shown; in addition to swords Bulgarian spears of Aegean type and "Sian

1963]

LATER AEGEAN BRONZE SWORDS

119

and perfected in Crete, though later adopted on the Mainland. Although at that time Aegean foreign relations were widening, especially those of Crete, touching Lipari, Miletus, the Troad, the Levant and Egypt (MM III-LM IA), yet the A swords have a very limited range, and are not shown among Aegean (probably Cretan) gifts in Egyptian wall-paintings of the beginning of the fifteenth century (tombs of Senmut and Amenuser).2 Probably their manufacture, requiring as it did immense skill, was a jealously guarded secret; the armourers of Egypt and the Levant could produce nothing to compare with them. In the following pages an attempt will be made to trace the history of the bronze sword in the Aegean from the fifteenth to the twelfth century, within the framework of a general classification of types; but without detailed descriptions of individual weapons. These will be found listed, together with their references, at the end of this article. As before the system of classificationis based on the shape of the grip, since this shows both greater stability and more readily definable characteristics than the blade.'

aggerated A swords. Strength and a deadly slenderness were achieved through the new cast grip, and the high, finely ridged, midrib. The metal was concentrated where it was most wanted for the thrust; and these swords always feel well-balanced in the hand. The area of their dispersal is proof of their success, for beyond the Aegean they are found from Palestine to the neighbourhood of Plovdiv in Bulgaria, and of Skopje in Yugoslavia (map). In so scattered a group it is not surprising to find variations due either to differences of date, or to the styles of different workshops. In fact two principal sub-types may be distinguished, a C i and C ii.

2 N. de G. Davies, BMMA 2 "The Egyptian Expedi(1926) tion 1924-25," pp. 41-51 Tomb of Senmut T 71 and Amenuser

3 Since it seemed preferable to include all Aegean bronze swords (except the Type II swords already fully described and

T 131; Davies, "The Tomb of Rekh-mi-fte T Ioo at Thebes,"

discussed by Dr. Catling in ProcPS 22 [1956] 102, and Antiquity 35 [1961] 115) in one classified system, it has not been

Ci

In their main characteristicsboth sub-types are alike, but C i has invariably two rivet-holes in the blade at shoulder level, or rather lower, and usually from one to three in the grip. There is also usually a narrow unflanged extension at the pommel-end, with a small rivet-hole for securing the pommel; always a separate unit. The relative position of the blade-rivets and the end of the midrib vary, being either on a level with each other, or Type C Swords one higher than the other (pl. 21:I-2). The fifteenth century saw the invention of two The C i class includes a group of outstandingly new swords in which the swordsmiths attempted large and handsome weapons, three are over o.gom. to combine the best points of both the earlier (Zapher Papoura 36 and 44 and no. 12 in the weapons. The result was the "horned" sword, "King's Tholos," Dendra), while no complete which I shall classify as type C, and the "cruci- sword is less than o.6o m. (apart from an uncharform," type D. The essential characteristicsof the acteristic one from Cos). Splendidly ornamented horned sword, in which it differs little from A hilts are found on most of these swords, with a swords, are a slender blade with a high midrib, free use of gold-plating, elaborate repouss6 spirals and a horned protection for the hand, no longer and feather patterns, and often richly ornamented as a separate attachment, but carried out in the midribs. Rivet-holes tend still to be large, and the same casting as blade and tang. The tang itself is rivets to be capped with gold. These princely provided with substantial flanges, such as were swords come from Knossos, the Zapher Papoura found on some of the B swords, from which they cemetery; (pl. 21I:i) Phaestus, and the "King's were probably adopted. These weapons were long, Tholos," Dendra; this tomb, in other respects also, but they never reached the length of the most ex- stands close to the material from Crete. Rather

2 MMA "The Egyptian Expedition 1935" (1943) passim; see also J. Vercoutter, "L'Egypte et le Monde tgean (1956) 207-II.

I have used again the Egyptian dates as given by E. Driotan and J. Vandier "Les Peuples de l'Orient Mediterranean II, I'Egypte," Clio (1946) with CambridgeAncient History (1962) Revised Edit. Vol. I, Ch. II "Chronology" p. 17ff, W. Hayes; also M. Rowton p. 67 and F. Stubbings pp. 69-77 for the Aegean; in particular for lowering the date of LH IIIC from Furumark's 1230 to ca. 1200.

possible to follow that adopted by A. Furumark, Chronology of MycenaeanPottery (1941) 93-96; and used by S. Dakaris, ArchEph (1958)

114-53.

Where the systems 'overlap, Furumark's

a 2 is very broadly my D ii; his bi is Fii; his b3 is Eii and his c 2 is G. I have on the whole followed the dates of Chronology p. 10obut allowing more liberal margins and some lowering in the 13th and later centuries, see also n. 2.

N. K. SANDARS

120

plainer, but essentially similar C i swords come from a recently excavated tomb at Knossos, found with a silver cup, and from tombs at Mycenae (pl. 21:2) and Prosymna. That from Chamber Tomb 81 at Mycenae is a slighter, slenderer weapon, with small rivet-holes, therefore typologically later than some, since the reduction in the size of rivet-holes is only gradually introduced. The illustration given by Tsountas shows well the oval-shaped opening in the hilt-plate through a discolouration of the metal (barely visible today); this is a further development of the kidneyshaped opening found in the hilt-plates of some A swords. An incomplete sword with ivory and gold grip from Argolis is closer again to the Dendra and Cretan swords, its date will be discussed below. Where discoverable the dates are all middle or latter half of the fifteenth century, before the destruction of the LM II Palace at Knossos. It is possible that most of the mainland swords are slightly later than the Cretan, falling at the beginning of the fourteenth century. Chamber Tomb 81 at Mycenae cannot be dated, but the sword is more developed than that from Prosymna T 37, dated early LH IIIA. Blegen has suspected the presence of other fine swords in the Prosymna cemetery because of the number of large detached rivets, some gold-capped, which may be debris of swords left by plunderers. The Dendra tholos is the richest unplundered tomb of this date either in Crete or on the Mainland (apart possibly from one of the tholoi in Messenia, Myrsenoch6rion II, from which however few of the weapons have been published, only some of the daggers).4 The description of the great armoury of weapons which accompanied the Dendra tholos burials reads more like the account of the fourth and fifth ShaftGraves at Mycenae than any of the later chamber tombs, there or at Prosymna, or even the "warrior graves" of the Knossos cemeteries. The objects collected in the large burial pit with two skeletons, called by Persson "The Burial Gifts of the King" and of "The Queen," need not all be contemporary, but are in any case remarkable enough, including the famous octopus gold cup, gold and silver cups 4S. Marinatos, "Excavations Near Pylos," Antiquity 122 (1957)

97-1oo;

also ILN (April 6th, 1957)

PP. 540-43.

5 Prof. Persson uses a different notation to that of Karo, which I followed in my earlier article, Persson's A type is our B, his B is our D, the C's are the same. 6 Kakovatos, "Tholos Tombs," AM 34 (1909)

298-99;

My-

cenae, IV Shaft-grave, Karo, Schachtgrdbervon Mykenai (1930-

[AJA 67

with hunting scenes, a fine collection of gems and the arms. Of these there were five swords, four spears and two small knives.5 One of the swords beside the skeleton is a perfectly typical B with handsome gold covered grip (Persson's No. io), providing a link with the Shaft-Graves, while the technique of decorating the grip with minute gold bars, attached to the ivory parts, strengthens the link both with the IV Shaft-grave, with Kakovatos Tholos B, and with Knossos, where at least one sword was decorated in this technique. A detailed description of the technique employed has been published recently with reference to a sword from Argolis.' The whole complex of finds suggests the later fifteenth, rather than Persson's fourteenth century, with the IIIA i pots perhaps dating the interments but not the manufacture of the fine metal work which would then be a little earlier. The swords from the Argolid provide a typological and spatial link with a small group in the northwest: three swords were found in Epirus, two in a grave (pl. 21:3-4), and one near Skopje, Yugoslavia. These swords share the peculiarity of a second pair of rivet-holes in the blade, either at the base of the horns (Perimatos near Ioannina, and Tetovo near Skopje) or in the horns themselves (Dodona). The break in the horns of the Dodona, and one of the Perimatos swords, shows that this was a fault in design: with one exception it is not found elsewhere. The larger of the two swords from Perimatos (pl. 21:4) must have been an exceptionally long and heavy weapon, for without the grip it still measures o.85m. The existence of such a tomb not far from Ioannina is sufficiently surprising, and has led Dr. Dakaris to an extended and interesting discussion of the problems of northern relations. Still further north is the sword from near Tetovo, Skopje, in the Vardar valley, not far from the watershed of the Danubian river system.7 It looks to be a genuine Aegean product, and typologically closest to the Perimatos swords at Ioannina. It is possibly as long as the heaviest from Perimatos, or longer. The condition of these weapons shows that they, at any rate, had seen 33) no. 396, p. 97, no. 435, P. 103; A. Evans, Palace of Minos IV, ii, 854, for Knossos; see also ArchEph (1897) (1921-35) 123; and see Prof. Elisabeth Treskow in Ars Antiqua AG Auktion III (1961)

p. 30.

7This sword is shortly to be published by Dr. Vinski of Zagreb to whom I am indebted for additional information.

1963]


121

heavy use, unlike the luxury weapons further south. Before turning to the next small group, mention should perhaps be made here of blade fragments (there are no hilts) from Levkas, in the so-called "Royal-Graves."These I illustrated as class A in my earlier article; they could equally well be C, to judge by the finely grooved midribs; but without the grips it is impossible to be sure either way. There is a still stranger extension of Aegean traditions of bronze-working, beyond their usual range, in a small group of swords and spears in Bulgaria. Though diverging much from the other C i swords in the grip, they share the use of rivets in blade and grip; while in one of the three swords a duplication of the pair of blade rivets links it more specifically to the Epirus group and Skopje. The two better preserved, from Dolno Levski near Pazardjik and Perushtitsa near Plovdiv (pl. 22:5, 7, pl. 28:64, 61), are said to have been found in graves, and were associated with spears of nonEuropean, probably Aegean types (pl. 22:6, 8, pl. 28:65, 62). The main difference between these and all other horned swords is the great length of the horns, giving a width of up to iocm., or exactly twice that of the Mycenae chamber tomb 81 sword. They extend at a much more open angle, in one case not far from 90o degrees, while the construction of the hilt is simplified. The bronze horns, instead of being flanged or "folded" to secure plates of perishable material, are cast solid, and the perishable plates could have been simply two parallelsided strips. A tendency in this direction existed in some of the C ii swords and it reappears in an altogether later group (classes G and H, see infra). Somewhat eccentric though these Bulgarian swords appear when viewed from the Aegean, they are totally unlike the native bronze-work, both as to the quality of metal and the form and character of the founder's work; in particular the grooved midribs, the great length, strength and narrowness of the blade. In some respects they give the impression of being even more efficientweapons than their relatives further south. The horns are stronger, so is the grip, due to the reduction in the amount of perishable and fragile inlay material. On the other hand the angle of the horns lacks the graceful curve which seems to have been de-

signed to deflect, rather than to hold, the opponent's blade. Perhaps these more horizontal horns needed to be tougher to receive the jar of a forward thrust, and the design was adapted for the use of swordsmen less skilled in the finer points of combat, as practiced in the Aegean? The spears found with the Dolno Levski and Perushtitsa swords, and two other single finds from Kritchim near Plovdiv and Krasno Gradiste, Turnovo (pl. 22:10-Il and map; pl. 28:67, 63) are as important as the swords for European archaeology. Here it must suffice to say that those from Perushtitsa and Krasno Gradiste are a very common Aegean and Near Eastern type, and absolutely foreign to Europe, especially on account of the slit socket. They can be matched in the Ayios Ioannis chamber tomb (A.J. 4) and the Acropolis tomb at Knossos (with C ii sword pl. 23:14), and in the 1929 hoard from Ras Shamra (fourteenth century). The spear found with the Dolno Levski sword has a faceted blade that is not easy to match in the Aegean, though similar faceting occurs on a spear of a rather different shape in the same Ras Shamra hoard and on a spear from Cos.8 The great single spear found at Kritchim, near Plovdiv, is perhaps the most interesting of all, for it is the same exceptionally long and narrow type which has been found frequently at Knossos (Ayios Ioannis, Chamber tomb, No. ii, Hospital Site II 4, III, 14, V, 7 etc.) and the Dendra tholos, all of which are between 0o.47m.and 0o.57m.;Kritchim at 0.502 is surpassed only by one from Crete (New Hospital II, 4, 0.46m.) and one from Dendra (No. VI, from Bulgaria are 0o.57m.).' These three spears identical with their Aegean counterparts, but further discussion of them is better postponed to another occasion.

8 C. Schaeffer, Ugaritica III (I956) fig. 224, 5 and fig. 226; for Ayios Ioannis see lists infra in Catalogue. 9 A. Persson, The Royal Tombs at Dendra (1931) King's

Tholos, nos. vI-vII, pl. 20, p. 37. I am deeply indebted to Prof. Mikov of Sofia for photographs of the Bulgarian swords and spears and much help and information in connection with them.

C ii Horned swords of the second sub-classare unlike the first in having no rivets in the grip, and with one exception, no pommel-tang extension. The horns are formed by a characteristic "folding" of the metal, which appears to be beaten together, in place of the cast flanges more usual on C i swords; the "Silver Cup" tomb sword is an exception. They are usually of medium length, between o.6om. and o.7o0m.,and the midrib is seldom so pronounced as

122

N. K. SANDARS

[AJA 67

in the C i class. The "folded"horn brings them The other Aegean pottery in tombs at Gezer is: a closer to the class B swords of Shaft-GraveVI at very little LM IB, and rathermore LH IIIA, so Mycenae,and some from the Levant and Dodeca- that if the bronze sword is to be connected with nese."0As a class they are simplerand more uni- the events which brought this pottery, it is more form than C i. Only two come from Crete;one of likely to be linked with LH IIIA. The rather shaky these,from the "AcropolisTomb,"Knossos,is one associations point possibly to the beginning of the of the few that can be given a date (pl. 23:13-15). fourteenth century, the same time as the Thermi This appearsto havebeen a LM II "warriorgrave" sword. Schaeffer prefers a date a little earlier ("belike ZapherPapoura36 and 44, and graves in the fore 1425 or 1400"). New Hospital and Ayios Ioannis cemeteries.I Two swords from the Dodecanese fit neither inshall returnto these graves,after giving some ac- to the C nor the B class but lie somewhere between, count of the cruciformswords which they held with some features of both. One is from the (see map). Asclepeion site, on Cos, but this material is unMost of the other C ii swords are uncertainof published. The horned sword is without pommeldate and of provenance,and this means unfortu- tang extension and has the same rivet arrangement nately most of the mainland ones. Many have as the B sword in the VI Shaft-Grave at Mycenae labels of varying degrees of probability. The (No. 905), to which it stands quite close in other "MountOlympos"of a fine sword in the British respects, as also to its Levantine forerunners, with Museumis suspect,though a northernprovenance shoulders midway between the horned and the is made rathermore plausiblein view of the sword merely "pinched-up." The second is a very large from Grevena,whichis verylike it (pl. 23:17).The weapon in New Tomb IV, Ialysos, Rhodes. If sword from Thermi is in poor condition and is the Cos sword had more B features, this is closer not typical,as it evidentlyhad at least one rivet- to a true C, only differing in not appearing to have hole in the grip; its value is its date, for it was flanges on the riveted tang, and in its great size, found in a room with LH IIIA pots.On the other which at over a metre would make it the longest hand a sword from Gezer in Palestine(pl. 23:16) of all horned swords. This tomb is quite exceptional is quite typical,apartfrom the small pommel-tang at Ialysos on account of the number of weapons extensionwhich is shown in Macalister'sdrawing, it held, including a cruciform and a round-shoulthough it no longer exists. The range of possible dered sword of a later type, along with LH IIIA dates is unfortunatelymuch wider. It was found 2 pots: I shall return to it below. in a pit in the floor of the large and often used "Tomb 30"in the Gezer cemetery.The pit held a A sword in the "Tomb of the Tripod Hearth" ratherheterogeneouscollectionof bronzesandsome Before horned swords mention must be pots, mostly Cypriot.The other bronzescannotbe made of leaving the unique short sword or dirk in the dated closely, they include a type of flesh-hook which is MiddleBronzeAge in Cyprus,and a large "Tomb of the Tripod Hearth," Zapher Papoura and fine scimitaror "harpe"which is very close 14 (pl. 26:45). A curiously eclectic weapon, its to one from Ras Shamradatedby Schaefferto the length, 0.42, is too short for a sword, though the fourteenthcentury;but these objectswere known rivets in the blade suggest the conventional horned from at least as early as the "Royal Tombs" of C i; the midrib is high, but in place of the usual Byblos (twelfth-thirteenthdynasty), and one is section, with a smaller ridge or rib down the midcarriedby a Syriantributebearer,with othermore dle, it has two ridges, giving an almost square Aegean-lookingobjects,paintedon the walls of the section. The horns first rise, and then droop in a tomb of Menkheperrasonb at Thebes,a little after downward direction which suggests the otherwise quite different (and much later) swords of Classes 1450." The Cypriot pots are, with one exception, fourteenth century or even beginning of thirteenth; G and H (see infra). The horns of the shorter of so the range of possible dates covers the whole the two Dendra "King's Tholos" swords, No. 15, fourteenth century and extends a little beyond it.12 may be rather like it, but I have not handled it, 10 Sandars, op.cit. (supra n. I) pl. I8 etc. 11 N. de G. Davies, "The Tombs of Menkheperrasonb,"T 86 Egypt Exploration Society: The Theban Tombs Series (1933) pl. I.

121 have received valuable help from Dr. Catling with the difficulties of this tomb, particularly the dating of the Cypriot pots, for which he has emphasized the very wide bracket possible.


1963]

and the photograph'3 is too small to be certain. The blade of the Zapher Papoura dirk has a sharp taper to the point, and a flanged and T-shaped pommel-extension which, in the Aegean, is not known to occur before the fourteenth century, and which will be discussed later. The ivory hilt-plates of the grip were in place, and show a skeuomorphic thongmoulding, immediately below the pommel. The absence of vases has made close-dating of this tomb impossible, though it is the most important of all the Zapher Papoura cemetery on account of the great collection of bronze table and kitchenware which it held. The dirk was in close association with a one-edged knife, a spear and an ivory casket, none of which is any more helpful, but the date is not likely to be very long after 1400. Considerable doubt must attach to a sword, now lost, apparently of C ii type, said to come from Adliswil, Switzerland.

Class D Just as at the beginning of the fifteenth century there were alternative swords, A and B, so too in the second half of the century there were alternatives: the usually longer C horned, and the cruciform D swords. Like Class C these too can be subdivided into D i and D ii; but in this case the division is chronological as well as typological. Only D i is contemporary with C.

Di D i swords are fairly uniform. The length is usually between 0.60 and o.7om., corresponding to the length of the shorter and more utilitarian C ii swords. The only exceptionally long D swords are in the same "King's Tholos" at Dendra, that held the very long C sword. Five out of ten in the Knossos cemeteries are between and o.63m. o.6im There is usually a well-marked midrib and, with one exception, there are two rivet-holes, rather low in the blade, as in class A and many class C. The grip and shoulders are always flanged like C i; but instead of the "horns"the shoulder extension takes the form of an angular or lobed broadening, from which comes the name "cruciform." In fact the outlines show two opposite tendencies. One, best represented in the Zapher Papoura cemetery, is for very rounded projections or lobes (pl. 24:19); Persson, op.cit. (supra n. 9) xx, no. v. 14 F. Chapouthier, Mallia, Deux Epdes d'Apparat (ttudes Cretoises 5, 1938) pls. 8-20. 15 Persson describes the opening as "kidney-shaped," but if 13

123

the other, found in the graves of the New Hospital site and Ayios loannis, has a more angular outline, and may be slightly earlier.It certainlyhad a shorter life-span. The width of the blade below the shoulder is much the same as in the C type, usually about 4 cm., but owing to the blade being shorter the D swords look stouter, and more tapered. The small pommel-tang extension is almost invariable. The swords come from Crete, the Argolid, and the Dodecanese (see map). This is the sword par excellence of the Knossos cemeteries, with the socalled Warrior Graves accounting for ten of the thirteen Cretan weapons, the other three are without provenance; against this the mainland has five (from the Argolid three others are uncharacteristic, one also untypical from Attica), and Rhodes and Cos have three between them. This class holds some of the finest of surviving bronze swords, and one at least (Zapher Papoura 36) is an outright masterpiece with naturalistic ornamentation of the gold hilt covers, showing wild goats and lions, in a style recalling the galloping griffins and horses of the Mycenae Shaft-Graves, and nearer home the "acrobat" looped round the gold pommelcasing of an A sword from Mallia." The more routine spirals of the midrib and flange are almost identical with those on the C sword in the same tomb, only a little smaller. Larger spirals are on the midrib of the richest of the New Hospital site swords (the Shaft-Grave, Tomb, II, 3), and there are linked spirals on the gold sheet that covers the wood of the grip. These again are found on the best-preserved of the two D swords in the "King's Tholos," Dendra (No. ii),"5 and apparently on the great horned sword (No. 12), where only a morsel of ivory remains. The outline of the shoulder projection is the same on the Zapher Papoura and Dendra swords, whereas the New Hospital and Ayios Ioannis shoulders are more angular. Where the hilt-plates of these rich swords survive they all have the same oval opening, as do a number of the C swords. The technique of gold-plating over wood is so little practical, where a sword is to be jarred with hard use, that it strengthens the general conclusion that these weapons were designed for show rather than for blows. As well as the two handsome Dendra "King's Tholos" cruciform swords, that from Mycenae, we follow the distinction drawn by Evans between earlier kidney-shaped and later oval openings, this opening is certainly oval.

124

N. K. SANDARS

Chamber Tomb 78, is well made, and possibly was once an equally grand weapon, since the only surviving rivet was gold-capped (pl. 24:21). It has telltale breaks at two of the large rivet-holes in the grip. The much damaged and incomplete sword from Grevena may have been like these, but most of the mainland swords are uncharacteristic and rather poor; some lack midrib: Prosymna Tomb XXV and the two much smaller weapons in Mycenae, Chamber Tomb 91, which also lack grip-rivets, and have sunk lines in place of the midrib, like D ii dirks and daggers (see infra). An incomplete fragment from Corinth without midrib is also probably D ii (pl. 24:28), but if its grip was originally like these in Tomb 91 it would have to be reckoned among D i's. A short sword or dagger from Eleusis is only very loosely "cruciform," having slightly lobed shoulder projections; but the rivet pattern is quite different and it seems to belong to another tradition of flat, broad-bladed knives and daggers, perhaps related to our class E and to some untypical blades in Chamber Tomb 82 at Mycenae. The three swords known from the Dodecanese are all very much alike, short, with an angular or intermediate shoulder, and not very high midrib (pl. 24:20). They are particularly close to the New Hospital Tomb V sword. The link with Knossos is strengthened by a spearheadfrom the Asclepeion site on Cos, which is remarkably like the spear A.J. 3, found with another angular D i in the Ayios Ioannis Shaft-Grave. Both spears have the same decorative moulding at the base of the blade, and the same section; the only difference is that Ayios Ioannis has sharper wing-ends to the blade. Before discussing dates and workshops there are two features of D i swords which must be referred to. First there are the hilt-plates of semiprecious material, unused and sometimes even unfinished, which are found (rather surprisingly) in tombs as well as in occupation sites; and second the "skeuomorphic" thong-mouldings which decorate these hilt-plates and the bronze grips as well. Chamber Tomb 81 at Mycenae, which had a C i sword, also had two cruciform hilt fittings; one of veined agate is unfinished, hollows are already sunk for the rivet-heads, but the holes have not been bored for the pins. Gold rings with granulated edges are in place around only two of them. 16 It is not certain that all the small objects came from this tomb and not from one nearby, see JHS 24 (I904)

322.

[AJA 67

Tsountasfound anotherhilt-plateof faience,rather more lobed in shape, unassociatedon the Acropolis, and yet another very like it apparentlyin Chamber Tomb io2.." A crystal fragment was found at Knossos,while separategold sheet casings for B type swords were found in the same Chamber Tomb 81 at Mycenae where the unfinished agate hilt-plate was found, and another in the dromosto Tomb 88. This last is very like the casingof the B swordin the "King'sTholos," Dendra,exceptfor having only one rivet.As well as showing the overlapbetween lingering B type swordsand both the new C and D i classes,these detached, sometimes unfinished fragments pose the questionas to whethersuch objectsof skilled and luxuriousworkmanshipwere obtainedfrom a few sources and then put by till needed, or whether they were plunder from some despoiled workshop;or whetheragain a travelingcraftsman would leave the spare fittings behind him when he moved on to his next job. We know so very little about the mechanicsof bronze-workingand other allied craftsin the secondmillennium. The detached hilt-plates and grip-casings, whetherof agate,faience,crystal,or gold on wood, as well as those found in position on swords in the Knossoscemeteriesand at Dendra,all have in commona simulatedthong, usuallydouble,at the base of the grip. The fine hilt of the derivativeA sword in ChamberTomb 78 at Mycenaeshows what is probablyan earlierstage with the lowest grip-rivetplacedwithin the "loop"of the "thong," as though the latter had been hitched around a once protrudingrivet-head.The "lily-dagger"in the V Shaft-Gravetakes us a stage furtherback, for it has a separategold ribbonin place of the naturalcord or leatherthong."7The lowest rivet is still more or less within the loop on the D sword from New HospitalTomb II, as also on the shell inlay from the Knossosthrone-room;but in other cruciformswordsthe rivet has moved up the grip and the "loop"has becomepurely ornament(Zapher Papoura36, i; and Dendra "King'sTholos" No. I1). Even where the hilt-plateshave not survived, the existence of the moulding can be inferred from "ripples" in the bronze of the flange (Zapher Papoura 55, pl. 24:19). This decoration was used on swords of much later types also, and sometimes on different parts of the grip. 17 Karo, op.cit. (supra n. 6) no. 764, V Shaft-Grave.

1963]


125

It seems very likely that the origins of the cruciDiscussion form sword are to be found (as Evans thought) Taken as a group C i swords are more variable in those flanged-hilted, round-shouldered long dag- than C ii. No two of the rich and princely swords gers and short swords of MM III and LM I, from Zapher Papoura 36, Phaestos Tomba dei found at Knossos itself (in the North House in- Nobili 8, and the "King's Tholos," Dendra, are complete) and at Gournia, the latter with a broad, quite the same. In spite of over-all similarities they rather flat midrib. These in turn grew naturally out vary in the details of arrangement and placing of of the flanged daggers without midrib, like that from rivets, in length and section of midrib, and in the the Hagia Triada tholos and various stray finds; the motifs of ornament. Each one is an individual finely decorated dagger from "Lasithi" with fight- piece of skilled workmanship produced to please ing bulls and a boar-killing is another. Evans called the taste and requirements of some individual pathem the typical dagger of LM Ia'8 beginning per- tron or potentate, and seeming to bear the hallmark haps as early as MM II. I have suggested elsewhere of a particular "shop." The sword in the "Silver that these daggers may also have had a part in the Cup" Tomb, Knossos, the shorter of the two from formation of the class B sword, though less influDendra, and the two from Prosymna and Mycenae ential than Syrian and Palestinian flanged daggers are more utilitarian relatives, which may or may and short swords. However the absence of true not be from this shop. Nor is a date in the second B swords from Crete makes it more likely that half of the fifteenth century in doubt for the they were the Cretan alternative to that sword, whole Cretan group, which belongs to the last for use when a weapon shorter than the great epoch of the LM II palace. Cretan A sword was required. The protruding The Epirus-Yugoslavia group and the Bulgarian shoulder of Gournia brings them much closer to are utilitarian weapons from two, perhaps several, the D type than to B, while the midrib is more hands or shops. They pose a curious problem, for "advanced" than the Eleusis weapon already re- it is unlikely that workshops operating in one of ferred to. On the other hand this did have a pom- the establishedcentres of Aegean civilization should mel-tang extension, while another, otherwise ex- produce a particular style of weapon for the northactly like it, had a T-flange, which reduces the ern barbarians. On the other hand, supposing the likelihood that they could be ancestral to crucigrave at Perimatos near Ioannina and those in form swords. Bulgaria to have held the bones of returned adIf the Gournia sword is probably one parent, venturers, or temporary exiles, who had served a the other parent of the cruciform sword is quite term at one of the Aegean courts, then it is difficult clearly the A type, from which it inherits the long to see why the weapons they brought back with fine midrib, the low-set blade-rivets, and the hilt- them should have been so unlike any in use in plate opening. These are features also common to metropolitan Greece or Crete. A possible alternaC swords, and in fact the two types, the C sword tive is that both groups were the work of craftsand D i, are exactly contemporary, found together men trained in one or other of the great centres, in graves (Zapher Papoura 36 and the "King's Cretan or mainland, selling his skill in the northern Tholos," Dendra, and the new Argolis find, also mountains, where in the course of time he may Chamber Tomb 81 Mycenae, horned sword and have fallen into personal idiosyncrasy, or simply cruciform hilt-plates). Both make their first ap- have tried to meet the special demands of the barpearance about the middle of the fifteenth century barian; this is only surmise. The Cretan palaces and last into the fourteenth century. One of the give evidence of bronze-working on the spot; so latest D type is probably that in Mavrospelio do houses at Malthi and at Trianda, Rhodes. In XVIII, (pl. 24:22), a grave without pottery but barbarian lands there is no such evidence, only the with a T-flanged knife. The sword, not much more scatteredfinds of merchants' hoards and, at a much than a dirk, is as uncharacteristicas the C dirk from later date, founders' hoards. The existence of some Zapher Papoura 14, it has no midrib and is in very link, however tenuous, between this part of southpoor condition. A typical D hilt said to come from eastern Europe, geographically not far from the the Rh6ne requires authenticating. watershed between the Aegean and the Danubian 18Evans, op.cit. (supra n. 6) I, fig. 541, p. 719; IV ii, p. 851; see also Sandars, op.cit. n. 22. (supra

I)

126

N. K. SANDARS

[AJA 67

river systems, and the centres of Aegean civilization, in the middle of the second millennium, is fact. Compared with the idiosyncratic C i swords, the unriveted C ii look like the product of one workshop (except for an untypical sword labelled "Dendra," if that is its correct provenance, and the two from Cos and Rhodes). On the evidence available it is not possible even to guess where this centre was, unless the influence of the predominantly Mainland B sword argues for the Argolid; but too many of the swords are without find spots, and the absence, so far, of any from Mycenae itself, makes this explanation less satisfactory.The Gezer sword is more likely to be connected with Mycenaean ventures of the early fourteenth century (LH IIIA) than with Cretan enterprise; the same date holds for Thermi, where no Mycenaean pottery is later than LH IIIA and local wares imitate LH II. Some mainland centre might be thought most likely to have produced the Grevena sword, and perhaps the British Museum "Mount Olympos," but the postulated connections of the fifteenth century rulers of Thebes with Crete could provide a possible point of entry for earlier Cretan bronzes to Boeotia."9A terminus ante quem for the C ii swords, in the middle or second half of the fifteenth century, is given by the pots in the "Acropolis Tomb," Knossos. So the earliest dated C ii correspond well enough with the fine swords of C i; the peripheral C i and C ii groups may be a little later, probably the first half of the fourteenth century. In each case, where approximate dates can be advanced, somewhere near the beginning of the century is preferable for the casting of the C ii swords, because of their likeness to the one from Knossos. In summary, the position points to at least one workshop capable of producing luxury weapons, working between 145o and 1400; and another, contemporary with it, producing the unriveted C ii and perhaps the plainer riveted C i as well, all thoroughly workmanlike weapons which found a more extended market or a more enterprising clientele. The smiths who produced the fine Epirot, Macedonian and Bulgarian swords must have learned their trade in one of these metropolitan workshops during the period when Class C swords were in active production. In course of

time they became independent, but their work is not likely to have found its way into barbarian graves much after 1350, unless kept for an inordinate time as heirlooms. However, the evidence of hard wear might be taken to favour a fairly long useful life. If sufficient evidence for tracing the style of individual workshops be conceded for many of the horned swords, the same would hold good for the cruciform,where in fact the evidence is stronger. This is particularly the case with the swords from cemeteries at Knossos: Zapher Papoura, the New Hospital site and Ayios Ioannis. I have suggested that the more angular shoulders, as at Ayios Ioannis and New Hospital, were slightly earlier than those at Zapher Papoura. Most are well-dated by pottery, and there is nothing to suggest use of these two cemeteries after the catastrophe to the LM II palace. Zapher Papoura, on the other hand, does seem to have been in use after that event and the tombs are harder to date with precision, but there is no reason to think that the two important "warrior graves," 36 and 44, were later; in fact they are generally placed within the last years in the life of that palace. It is rather that D ii with pommel flanges, typologically the subsequent development, is always given a lobed, not an angular shoulder; also the probably later D i dirks without midrib are lobed (except for Mavrospelio XVIII which also appears to be late). The workshop that produced the finest C swords was certainly equipped technically to produce the best D i. Although nothing like the naturalistic decoration of the Zapher Papoura 36 D i hilt has been found on a C sword, hilt plates have so rarely survived that an argument ex silentio is not permissible. The Mallia workshop that produced the "Acrobat" was certainly equal to the draughtsmanship required for tracing the goat and lion; in fact the acrobat is superior work, and probably more than a century earlier. Chronological continuity of surviving work can be traced from Mallia through Mycenae, Myrsinoch6rionand Vapheio, back to Zapher Papoura and New Hospital, Knossos, on the one hand, and Dendra on the other. There are particularly close links between the Dendra "King's Tholos" and Crete, as Persson, Furumark and others have agreed. Persson describes the gold octopus cup as "a perfect master-

19 LH IIIA i floor deposit in the "House of Cadmos," probably destroyed at the same time as the LM II Palace at Knossos,

A. Furumark, "The Settlement at Ialysos and Aegean History," OpusArch 6 (1950)

264.

1963]


127

piece of the finest Cretan workmanship in the marine style." It is one among several objects probably of Cretan provenance. He also considered that the long horned sword probably came from the same workshop as that from Zapher Papoura 36, though it is not clear why he thought this meant both were manufactured on the mainland.20 The unfinished hilt-plates of semi-precious materials in graves at Mycenae point to Mycenae not being the site of their manufacture, for who would treasure incomplete and unfinished objects if he had only to go down the road to procure complete ones? Perhaps the hilts did not come from the same workshop as the bronze swords.21 The finds of crystal hilt-plate fragments, and others with gold nail decoration within the Palace at Knossos suggest that this work was done on the spot, as at Mallia in earlier centuries, and at Trianda in Rhodes in the fifteenth century; so the great workshop at Knossos may well have been destroyed with the LM II palace and never reopened. After the last of the splendid horned and cruciform swords there were no more luxurious weapons. A little gold sheet casing with the simplest of ornaments, or a gold rivet-cap, seems to have been the most that later generations were prepared to spend on their weapons.

military stress which led to the invention of the C and D i swords? The date for the invention, around the middle of the fifteenth century, is well grounded, linked with LM II pottery which in turn can be linked with Egyptian chronology.2 Mr. Hood, the excavator of some of the most important graves of the period immediately before the destruction of the LM II palace at Knossos, has called them "warriorgraves," the burials of a military aristocracyconcentratedon the court at Knossos and serving, as Evans believed, a new militarist dynasty. Whether this new dynasty was of Greekspeaking conquerors from the mainland, or thrown up by an internal Cretan political revolution with mainland support, does not concern us here, only the existence of the warriors and their weapons. Whatever its origin the aristocracy seems to have been thoroughly "Minoanized." It is possible also that the change appears more dramatic than it was, for lack of LM I graves leaves a gap in our knowledge of personal equipment which cannot be filled from other sources. To the question, what caused the rulers of Knossos to surround themselves with a military caste, the usual and most plausible answer is that which connects it with the destructions at rival Cretan sites, Gournia, Mochlos, Palaikastro, Phaestus and Hagia Triada. The Phaestus cemetery is poor in bronzes and has The Military "Aristocracy" only one really rich military tomb, that with the The swords themselves are among the most im- C i swords; but in the Knossos cemeteries there portant evidence of the high standard of the fif- are at least fifteen graves with swords, so far known, teenth century craftsmanship and metallurgy, and of which only three probably date from after the at the same time tell us a little about the society fall of the LM II palace-a fact which suggests that used them. It is a fair supposition that the that the organization of the military survived the invention of a new weapon, or the radical altera- fall of the dynasty which it had served. This in tion of an old one, only happens in answer to turn may be a hint at a situation well known in military need. One may well ask what was the the history of empires, from the emperor-making 20 S. Hood, "Late Minoan Warrior-Graves,"BSA 47 (1952)

alike the variations could be accounted for as the hand of different scribes, one of whom tended to stylize more than the others. The variants with a straight top to the blade might be intended for the more angular D i type of Ayios Ioannis and (M. Ventris and J. Chadwick, Documents in Mycenaean Greekl New Hospital, as against the lobed shapes from ZapherPapoura, or they may represent extremes of stylization or mere symbols p. 360, nos. 261 and 262) "so many swords: 5o"; and "Kukalos the cutler, 3 swords fitted with (bindings?)" is established, this that could as well be applied to any of the other classes (see might refer to the work of assembly necessary to complete infra). The cruciform D i sword shown diagrammatically in sword-grips of this type and make them usable. It is however Documents 347, fig. 24, to illustrate Pylos Ta 716, had gone well to remember Miss Gray's warning "archaeological com- out of use long before the Pylos tablets were deposited, and mentary . .. is bound to begin from ideograms which belong cannot be used to support the interpretation "studs" for the to the same physical world as the monuments" (Institute of qi-si-pe-e of this tablet. D ii which survived (as scrap) into Classical Studies, Bulletin 6, 1959, p. 55). Some of the sword the I3th century is not known to have had gold rivets. Weapideograms from Knossos, which according to Evans showed at ons current in the 13th century were classes F-H with pommel least two distinct types (Evans op.cit. [supra n. 6] IV, 835), flanges and much simplified methods of fixing the haft. are clearly meant to represent the protruding shoulders of cruci22 Recently through the LM II tomb at Katsamba, S. Alexiou form swords of type D, but others are less distinctive. Miss Gray KPHTIKA XPONIKA 6 (1952) 9-41. has suggested to me that since no two ideograms are quite 267, n. 121; Persson, op.cit. (supra n. 9) 45-47; see also A. Furumark, op.cit. (supra n. I9) 264. 21 If the suggested interpretationof the tablets from Knossos

128

N. K. SANDARS

[AlA 67

of Roman legionariesto the plots and revolutions Melos was conqueredand Cretanswere puttingin at Miletus and Samos. Trianda still flourishedin of the Janissaries. A studyof these graves,as well as of the swords the earlyfifteenthcentury,but thereis no evidence themselves,suggests that whereasthe sword was that the great A class sword found its way either certainlya pretty thing, and probablya prestige to Rhodesor Cos. Forms relatedto the B and C symbolof a militarycaste,the spearwas the more class have been found; one quite characteristic, effective and lethal weapon. From fifteen tombs and all showing links with the Asiatic flanged at Knossos we have 15-16swords and 18 spears, weapons.28On the other hand the D swordsfrom and the bronze heads alone of some spearsalmost Rhodes and Cos are perfectly characteristicand equal the length of the lighter swords.They must especiallycloseto the KnossostypefromNew Hoshave been tremendousweaponsbothin war and in pital and Ayios Ioannis,alsothe spears(see supra), the gems and the pottery.24Furumarkhas said25 the chase. These too were the years when the chariotwas that LM II vases in IalysosTombs stem not from introduced to Crete, when the fresco of "The the local LM IA style,but from the CretanPalace Captain of the Blacks"was painted (these inci- style; if this be concededthen they are evidence dentally were spearmen) and only a little later of the continuationof directlinksbetweenKnossos than the military"miniature"of the Hagia Triada and Rhodes in the second half of the fifteenth steatitevase. Some of the soldierycame from dis- century.This was the periodof coexistence,when tant parts,perhapssome of the Captainsdid also, Trianda still supportedits Minoan colonists and though they are hardly likely to have come from the first Mycenaeanswere arriving,the period of as far as Epirusand Macedoniain the north. We the earliest IalysosTombs and of the introduction have alreadyseen that Minoans and Mycenaeans of the Class D sword, the Knossossword par exseldom made presentsof their swords, and those cellence.Whether it came as a simple import,or found in the north could not be explainedin this in the baggage of membersof the "militaryarisway. tocracy,"or even of an expatriateswordsmith,it The usuallyacceptedpictureof a troubledCrete is not possibleto guess.Triandacertainlyhad supsuits the evidenceof the tombs, and explainsthe ported bronze-workers,and the B and C swords regularpanoplyof spearsand swords which they look like localproducts,as do the latertypes.Indeed contain; the "Chieftains"vase of Hagia Triada the importanceof the Dodecaneseas a centre of and the "Captainof the Blacks"fresco tell the the industry in bronze-workwill emerge more same story.The qualityof the swordsgives some clearly with these later types, and in particular idea of the high social standingof their owners, the immediatelyfollowing D ii. Then at Miletus while the basic alterationsin design-cast bronze there is a destructionof the first settlementwith horns, improvedhilts-betray an underlyingneed its apparentlypeaceful Minoan-Mycenaeanconfor military efficiencywhich is also evidencedby tacts.26The influenceof the Cretanstyle on mainthe more utilitarianswordswhich have survived. land vases diminished (LH IIB) and there are Outside Crete Minoans were still apparently clearsigns that the Minoanworld is shrinking.At prosperingin the laterfifteenthcenturyat Trianda the same time there was no lack of enterpriseon on Rhodes,on Cos and at Phylakopion Melos.The the partof the mainlanders.Perhapsexcludedfrom Dodecaneseare especiallyimportantin the history Melos till after the fall of Knossos,they were exof swords,principallybecauseof their r8le as half- ploringfor alternativesourcesof obsidian,exploraway house and entrepotfor Aegean and Levantine tions which took them westwardsto Lipari,where goods. The Minoan colony at Trianda,and prob- LH I and II/III pots accountfor 6o%of imported ably that on Cos, were foundedin the greatperiod pottery found.27 It was possiblyin the course of of Minoan enterprise in the sixteenth century when 23 Asclepeion

sword, Cos and Cameiros Rhodes, see Sandars, op.cit. (supra n. I) p. 28 and pl. 19, 6. 24 Gem of well-known type with goddess flanked by griffins, compare particularly New Hospital Tomb III and Ialysos N.T. XX, Hood op.cit. (supra n. 20) pp. 272-73 with further references, and A. Miauri, "Jalisos," Annuario 6-7 (1926) fig. 50, pp. 128-33; fig. 62, p. 139. 25 Furumark, op.cit. (supra n. I9) pp. 200-01; F. Stubbings, The Mycenaean Pottery of the Levant (1951) 20.

these westward explorations that Epirus and the 26 C. Weickert, Neue Deutsche Ausgrabungen im Mittelmeergebiet . .. . (959) 181-96, and IstMitt 7 (i957) Io2; (I959/ 6o) 67. 27 Cretans had found Lipari too, but their explorations may belong to the earlier phase of friendly Aegean coexistence for which there is so much evidence in the Dodecanese. B. Brea, M. Cavalier, BPI N.S. Io (1956) 47. Late Helladic, but see W. Taylour, Mycenaean Pottery in Italy (1958) pp. 16, 48.

1963]


129

"The Craftsman God of Ugarit" Before turning to the next developments a not entirely irrelevant picture may be glimpsed of the functioning, the international reputation, and the influence of certain great craft-centres, as it is reflected in the Ugaritic texts found in the ruins of Ras Shamra, and particularly in the collection called the Baal Epic."1 The Baal tablets were not written until the reign of Niqmadu of Ugarit, in the middle of the fourteenth century, but the epic material itself is probably considerably older. One of the principal protagonists is the craftsman god and architect,Kathirwa-Khasis, also called Hayin: sometimes a single individual, sometimes dual, whose names mean skilful and clever, combining manual skill or dexterity with intelligence, also deft or handy. Not only his name links him with Daedalus, but also his function as worker in metals (including the production of animal figures), furniture-maker, and architect.Like Hephaestus he is visited by am-

bassadors who, coming from Syria, must cross the sea to find him. They are told: then of a "Start away O two fishermen ... truth do you set your face toward h(q)kpt:" towards Kathir-and-Khasisgod of it all; Crete (kptr) the throne on which he sits (hkpt), the land of his heritage. From a thousand furlongs ten thousand leagues do homage and fall down ... tell to Kathirand-Khasis, repeat to Hayin the handycraftman the message of the victor (Baal)." Syrian Baal and Cretan (?) Hayin were allies throughout. Later we hear "Hayin has gone up to the bellows (or forge) the tongs (or handles) are in the hand of Khasis, he may smelt silver, purify (?) gold, smelt silver up to thousands ... smelt ore . .. and beaten work (or vessels) for a god, a seat for a god, a rest for his back, a footstool for a god which is covered (or spread) with leather (?) a litter (?) for a god with a canopy (?) thereon, whose carrying-poles are gold, a table for a god which is full with varieties of game from (the foundations of the earth), bowls for a god whereon the like of Syrian (?) small cattle, the like of wild beasts of (Yemen?), wild oxen up to ten thousand are the decoration.""3Reduced from the divine to the human plane this reads like a business order for silver and silver-gilt vessels, some shaped as or decorated with animals, chiefly oxen, also for chairs, litters, footstools and other furniture. In another place Kathir-and-Khasis "fetches down" a mace and names it for Baal; that is to say he supplies it already provided with magic potency."' In another epic he made and delivered (in person) a bow.3" There is no mention of swords, but this is not surprising if the production of the Aegean swords was kept secret, as appeared probable on other grounds such as its absence from the Egyptian tomb-paintings. In other respects these descriptions could serve as text for several of the foreign "embassy"or "tribute"scenes in the tombs at Thebes. We need not concern ourselves very far with the extremely illuminating account in the Baal epic of the activities of an international architect, the same Kathir-wa-Khasis. It begins with the unease of a Syrian ruler (in this case Baal, god of rain

28 H. Catling, V. Karageorghis, "Minoika in Cyprus," BSA 55 (1960) 109-27. 29 Furumark, op.cit. (supra n. I9) 270. 30 F. Stubbings, op.cit. (supra n. 25) passim. 31 G. Driver, Canaanite Myth and Legend, O.T. Studies III (1956) p. 91 passim.

32 Not Egypt, as earlier thought, but "perhaps some place in Crete," see Driver, op.cit. (supra n. 31) corrigenda p. 169, see Gordon JNES 7 (1948) 263. -3 Driver, op.cit. (supra n. 31) p. 91. 34Driver, op.cit. (supra n. 31) p. 81. 85Driver, op.cit. (supra n. 31) p. 53.

Ionian Islands were visited. The mainland bronzesmith may have had greater freedom of movement than his more orientalized Cretan counterpart,and so been persuaded into working for the barbarians. In these years too the Mycenaeans were discovering Cyprus which, however, continued true to its Cretan ties for some years to come,28 and Aegean weapons were not introduced there (only the Middle Minoan daggers are known). From the Levant there is ample evidence for the presence of Mycenaeans before the Hittite advance about 1375, since LH IIIA pots were found in a building at Katna destroyed by Suppiluliumu about that date.29 The rare finds of LH II pots (mostly IIB) from Syrian sites (Atchana, Ras Shamra, Byblos, Lachish; and Tell Ajjul in Palestine), show the beginnings of a really massive penetration, that was to take LH IIIA pottery to at least nineteen different sites in the fourteenth century.30When the "peace of the Egyptians," which had been so favourable to Cretan enterprise, was shattered by the Hittites and by local Syrian uprisings, it was the Mycenaeans of the mainland and the islands who benefited from the change.

130

N. K. SANDARS

[AJA 67

and thunder, but equally applicableto a mortal are singularly mundane. Cilicia has produced nothprince) who, after consultation,and as a resultof ing comparable, and the rich coastal cities of Syria reporteddisparagementsby his "brothers"(other hardly fill the requirements of the Ugaritic and rulers of like standing), sends for the craftsman- Mari texts. By the thirteenth century things are architect.When he arriveshe is entertainedwith altered and kptr may have become no more than a feast and told what the rulersupposeshis palace a commercial term. should be like. He then agrees to build it with certain stipulationswhich are overruled,but in PART 2 the end the architectis proved right and has his way. All these are, of course,incidentsof myth, D ii the actionis superhumanand the characters divine; Unlike the expiring horned sword, a further debut reduced to human scale the building of a house for Baal,like supplyingtablewareand furni- velopment took place in the cruciform. It lost its ture, gives a fair enough picture of the methods midrib, and it gained, in place of the unflanged and functionsof the palacearchitectand designer pommel-tang, a T-shaped flanged extension, into of internationalreputationand sphereof influence which the plates of the hilt and pommel were in the middle second millennium,and one that riveted. These plates of ivory, bone, or wood were might providethe wanted link betweenthe great now presumably made in one piece for each side, Syrian palaces,Mari and Tell Atchana,and those hilt and pommel together. The cruciform shoulof Crete.One ratherobscurepassagemay hint at ders are lobed like the Zapher Papoura D i swords. an expeditionmade from the workshopin Crete It is a small group, essentially a continuation of D i; and there are only two weapons large enough (kptr) in searchof mineralsand other raw materials. "Kathir-and-Khasis answered. . . I do (quit to rank as swords, both from Ialysos, Rhodes, and Crete) for the most distant of gods hkp (for the measuring o.6o and o.52m. (pl. 25:24). The rest are most distant of ghosts) two layers beneath (the in the 3ocm. range, or even less, with blades and springs of the earth, three spans) the rocks."Or shoulders correspondingly narrower, that is to say again this may be an entirelysupernaturalvisit to dirks and daggers. The intermediate 30-40cm. group comes exthe "CosmicMountain"(mountain of trial after death) to gain advicefrom El, the old high-god."8 clusively from Crete. The dirk from Palaikastro I do not find Furumark'sargumentfor rejecting is typical at o.33m. with an abrupt taper towards Crete as the home of the Kephtiu of Egyptian the point, clearly a stabbing weapon. fifteenth centurytexts very convincing.Consider- The dates range from the early or mid-fourable geographicalhaziness on the side of writers teenth century (LM or LH IIIA 2), after the fall from "outremer"may leave the exact locationof of the LM II Palace at Knossos, down into thirkptrat any one time uncertain,as betweendifferent teenth century (LM IIIB); possibly even later, Aegean centres,Cretan,mainlandor even Dodeca- for they survived long enough to appear in nese;but in the fifteenthcenturythereis the strong- founders' hoards which are dated LH IIIB "at est likelihood that Crete, and probablyKnossos, earliest."The earliest of this class are the dirk from was intended. At least one Cretan embassy is Palaikastro and the sword from Ialysos N.T. IV, allowed by Furumark,to accountfor certainob- which has a swelling in the blade section but no jects on the Senmut and Amenuser walls, vases, midrib. This is the same Ialysos Tomb that had, rhytaetc.; these are known to have been made in with its two burials, a B or C i sword and an E type Creteand are describedin Ugaritictexts and those knife, belonging probably to two periods. The diof Mari as coming from kptr,which the Baal epic mension and outline of the dirk or knife from a LM further describes as across the sea and at a considerable distance from Ugarit. It would be difficult to find anywhere on the coast of Asia Minor sufficiently prosperous in the fifteenth century to be exporting luxury works. The archaeological remains from sites like Miletus, though plentiful, 86

Driver, op.cit. (supra n. 31) p. 75.

IIIB Tomb on Carpathos are so like the damaged knife-dagger in the small Mycenae founders' hoard that they almost certainly came from the same workshop, and a fragment from Corinth is like them (pl. 24:26-28). The condition of both the mainland ones is too poor to tell whether they

1963]


131

also had the lines down the centre of the blade which can be seen on the example from Carpathos. The T-shaped, flanged pommel and the flattish blade with grooves down the centre, or very fine ribs outlined by fine grooves, are features radically new to the Aegean. The handle has been taken as the determinative feature of this sub-class,and was to become invariable on all sorts of weapons and tools; but the flat blade with its decoration also had a long life, so it is worthwhile trying to trace both features back to their source. The T-handle is the easier to follow, but both take us in the same direction. The flanged pommels are as old as the Hyksos dynasties, and were used in Egypt and Palestine from about the seventeenth century (Apophis dagger). In Syria and Palestine they remained popular and are found frequently at Ras Shamra, from the sixteenth century.37These hilts generally belong to knives and daggers and only occasionally come on sword-length weapons. The shape is less sharply expanded than in the Aegean, a semicircle rather than a T, and they are often without any grip rivets. The Aegean bronzeworkers adapted an oriental idea, they did not slavishly copy a foreign weapon. As soon as it reached the Aegean the T-flanged pommel was transferred to a number of different tools including one and two-edged knives,38 since it was an obvious improvement. An early adaptation is probably the curious dirk in Tomb 14 Zapher Papoura (Tripod Hearth) already described. The one-edged knife from a LH II or early III deposit in a house at Mycenae is earlier, but although the flange is carried round the top of the grip, this is hardly yet a T. It is very like a great number of Levantine knives, and may even be an imported piece.39 The ribbed or grooved blade is probably also Levantine. Fine ribbing applied to dagger-blades had been within the competence of the Middle Minoan bronzesmith; then there is a break of one or two centuries. In the Levant there was no such break and flat or flattish blades, often with this decoration, continue from the Middle to the Late Bronze

Age, with several dated in the fifteenth and fourteenth centuries. One is shown on an ivory from Ras Shamra, very possibly, according to Schaeffer, in the hand of King Niqmadu, a contemporary of Amenophis III and of Suppiluliumu.40 The "Syrian" weapons carried by tribute-bearers on the walls of the mid-fifteenth century Rekhmire Tomb are also T-hilted and appear to be flat.4 The importance of the Dodecanese as a commercial entrep6t for the Aegean and the Levant has been mentioned. During the period spanned by the burials of Ialysos N.T. IV, with their assortment of bronzes, Rhodes was flooding Egypt (Tell el Amarna), and to a lesser extent the Levant, with pottery containers of characteristic and rather monotonous types, chiefly the LH III A2 piriform stirrup jar. This jar is not common on the mainland but came to Rhodes directly from Crete in LM III Ai. It has been thought to have been mainly used in the trade in oil, which was exported from Crete and later probably from Rhodes. Perhaps the time lag noted by Stubbings, before the mainland adopted this form of pot, and by him put down to the difficulty experienced by the mainlanders in throwing and finishing a completely closed pot, may rather be accounted for by its specialized use in the overseas oil trade in which the mainland was not yet engaged-whereas Rhodes in the fourteenth century could have taken over the shipping of oil to the Levant? Recently light has been thrown on a similarly specialized trade between Crete and Cyprus.42At the same time Rhodian commerce was active as far west as Italy, where there seems to have been a settlement from Rhodes at Taranto. Rhodian pots of twenty different types have been identified there and again the monotony of some types is suggestive of the commercial carrying trade. Lord William Taylour thinks the murex trade was an important commodity there.43 Oriental objects are not infrequent in Ialysos Tombs, including the well-known Syrian duck-headed ivory ointment boxes, also seal-stones, beads and faience.44In view of all this activity, east and west, north and south, it is far from improbable that

3 F. Schaeffer, op.cit. (supra n. 8) I, 69, fig. 63, see also pp. 65, 67 etc. "from Hyksos graves"; also F. Schaeffer, Stratigraphie Comparee (1948) fig. 44 Ug. R6c. 2, fig. 45, u. also from Atchana Level III, at Antakya no. 729, At/8/54. 38 Sandars, "The Antiquity of the One-edged Knife," ProcPS 21 (1955) 179, Class 3. 39 Sandars, op.cit. (supra n. 38) fig. 3, 2. 40 F. Schaeffer, op.cit. (supra n. 8). 41 N. de G. Davies, op.cit. (supra n. 2) pls. 21-22.

42 Stubbings, op.cit. (supra n. 30) I6; Catling, op.cit. (supra n. 28) I2I; Cyprus was of course in close touch with Egypt and the Levant, and is probably responsible for some of the "Mycenaean" pots at Tell el Amarna. 4STaylour, op.cit. (supra n. 27) 135. 44 Furtwiingler-L6schke, Mykenische Vasen (1886) Text fig. 3, pls. B.26, E.I-3, pp. 14-15 from O.T. 31 etc. also N.T. with cylinder seal Miauri, op.cit. (supra n. 24) 127, fig. xvwI 47 no. 71 etc.

132

N. K. SANDARS

[AJA 67

the T-flangedpommel and flat Syrianblade were richly ornamented relatives were too long for everyadoptedfirst on Rhodes, where the only sword- day, as well as being too grand, and must have been length D ii weaponshave been found (both from for ceremonial occasions only. Contact with the Ialysos Tombs). From Rhodes they may have Levant probably played no small part in the abanreachedCrete and the mainland,though the dat- donment of the Aegean long sword, for there long ing evidencedoesnot excludea contrarymovement weapons were never really popular. The few very from Syria via Crete to Rhodes. The cruciform long swords known, like the so-called Shardana shape was never really popularon the mainland, sword in the British Museum from near Gaza, which is unlikely to have played any important had a short life. part in its evolution and developmentsince D ii Along with the disappearance of great luxury as in there much later metal appears only scrap weapons, and of dependence on the Levant, came hoards. a popularization of the short sword, which was The Class D ii weapons take us into a new now apparently made in many more centres. This world of drab armament;even the ivory plates change will be more apparent when we deal with of the Tripod Hearth dirk are unique. Weapons the round and square-shouldered dirks and dagare serviceableand short with strong hafts. The gers, classes E and F. Is it pushing conjecture too reasonis not far to seek. A long sword cannotbe far to see in these circumstancesevidence of changed worn convenientlywhen carryingon the ordinary social conditions which now required a majority businessof living. In the Menkheperrasonb tomb- of the male population to go about their business painting at Thebes, dated a little after 1450,we armed? This would be a prelude to the age of see a short sword in a scabbardslung from the piracy and trouble that was not far off. shoulder,and a long sword resting on the other Class E i shoulderof one individual,who wears the Syrian dressand carriesalso a Syrianscimitaror "harpe." Although none are swords, certain round-shoulIn this, and the slightly earlier Rekhmirepaint- dered flanged knives and knife-daggers must be ings, short swords with T-shaped pommels are given a place in this classification on account of shown slung over the arm for presentationin their their r81e in the evolution of other types. All the scabbards,while long swords are carriedresting implements of this small class have a flanged grip, on the shoulder;and this is exactlythe positionof usually but not always with an unflanged pommelthe young soldier'sswordon the steatitevase from tang extension (like many C and D i swords). The Hagia Triada, where he stands in front of his flange is carried round the shoulder and down princeor captain,as thoughon parade,with sword towards the blade, which is short, broad and alon shoulder.The type of sword is not clear,but, most flat. The longest, which is also the earliest, is being the same length as its owner's leg, it was only 0o.355m.,and comes from the Ayios loannis probablyaboutone metre:definitelylong. It would Chamber Tomb, with a D i sword and LM II have been as awkwardto walk aroundwith as the pots; most are around 3ocm., with an average jingling sabres of nineteenth century cuirassiers. breadth of 6cm. The mainland group is very uniThe "chieftain"or captain, on the other hand, form around 27cm. Daggers, dirks or knives, they wears a handy short daggerstuck in his belt, and were never an alternative to the C and D swords, holds a spear.45Vases painted in the Levanto- but were auxiliary to these, as in the Ayios loannis Helladic pictorial style from Cyprus show a Chamber Tomb. They range in date from midmedium to long weapon apparentlyslung from fifteenth century (LM II at Ayios loannis) to early the shoulderand (ratherawkwardly) tucked un- fourteenth century (Prosymna Tombs III and der the arm.46 XLIII), being therefore exactly contemporary with The long sword and the spear were military arms to be taken up only when needed for use in fighting, the chase, or when on parade. Their

C and D i. In origin they stem from the same round-shouldered dagger of MM II-LM I as do the cruciform

45 Davies, op.cit. (supra n. 11) if drawn to scale the swords carried on the shoulder in Menkheperrasonbpl. I, and pl. 5, would not fit into the scabbards;see also supra n. 41 pls. 18-I9, also J. Pendlebury, The Archaeology of Crete (1939) pl. 37, 2.

Sj6qvist, Problems of the Late Cypriot Bronze Age fig. 20; Furumark, The Mycenaean Pottery (1941) fig. 25, LH III A2 and IIIB. I am indebted to Dr. Catling for suggestions and the Cyprus reference. 46E.

(1940)

1963]


swords; but where the latter grew longer with higher midribs, these were if anything shortened and certainly flattened in the (now knife-like) blade. In the better-preserved examples the blade is decidedly U-shaped (pl. 25:29), but in others a ferocious whetting of the edges has sharpened the profile, and incidentally increases the suitability of the label, "carvers," given to them by Miss Benton. In all probability they were general purpose knives, one of whose functions was carving47 (see map). Class E ii Just as cruciform D i weapons adopted T-flanged pommels, these knife-daggers appropriated the same hafting improvement, retaining flatness and broadness of blade, the round shoulder and the blunt end. There is however a slight increase in angularity in the blade profile (pl. 25:31). This is still a very small group which is fairly uniform with lengths between 0.30 and o.4om. and a breadth of 6cm. With one exception all dated examples come from the first half of the fourteenth century (LM or LH IIIA). One of the earlier may be that in Mavrospelio XVIII at Knossos, found with an uncharacteristic,because nearly flat, D i sword. But the tomb cannot be exactly dated for, like Zapher Papoura 14, it had no pottery (pl. 25:30). The D i sword must be one of the latest of its class, having already lost the midrib, while the other bronzes, which include a two-edged razor, would be quite well suited by a date soon after the fall of the LM II Palace. The same uncertainty, though for different reasons, applies to Ialysos N.T. IV, which, with its two burials, its C (or B) and its D ii sword has already engaged us more than once. Unfortunately the bronzes in the great "Cenotaph," Chamber Tomb 2 at Dendra, cannot be dated by the LH IIIB pottery in that tomb, so that E ii survival into the thirteenth century is uncertain. E ii is in part the contemporary of D ii (Palaikastro and Ialysos) but it is not likely to have survived as long as the latest of these, which were found not only in LM IIIB tombs (Carpathos and Gournes T 2, Crete), but in the even later "founders" hoard at Mycenae. It cannot be said with certainty where the E i knives picked up their flanged pommels, but the 4'

S. Benton, "The Pelynt Sword-Hilt," ProcPS 17, 2 (1952)

237-38.

133

late arrival of E i on the mainland of Greece makes it unlikely that it happened there. It could equally well have been in Crete or in the Dodecanese, due to the same oriental impulse that produced Tflanged pommels on the other knives and daggers. Class F In the later fourteenth century the categories of dirks, daggers and knives were still fairly stable (D ii and E ii). The early thirteenth century sees both main divisions, the more pointed dirks and the broader knives continuing; only now we have a square-shouldereddevelopment of E ii: Class F, which tends more and more to meet both needs, while the heirs of the D ii dirks (and remotely of horned swords as well) show various bizarre developments. Finally in the later thirteenth and in the twelfth century there is a complete breakdown of categories and every weapon appears to be sui generis (see map). Some of the round-shouldered E ii knives were already tending to squarer shoulders, and this squareness makes a chronological as well as a morphological division between the classes. Blades are still flat, flanges are if anything deeper, and the T of the pommel narrower and straighter. The numbers are too few for subdivision but there are two main groups; one with a broader blade with two horizontal rivets in it (Zapher Papoura AE 472, no tomb no.), and a narrower blade with one central rivet hole in it (Mycenae Acropolis hoard, 2548, and Zapher Papoura 95 e, pl. 25:32, 36). The two from Zapher Papoura are probably among the earliest. Miss Benton suggests fourteenth century.48The Chamber Tomb 95 weapon is certainly a dirk, it is o.37m. long but the shoulders are only moderately square and it is on the E ii-F borderline. On the slightly shorter, broader, knife-dagger from the same cemetery (Ashmolean AE 472 35.om.)

the skeuomorphic thong-moulding

is conspicuous, though not at the base of the grip, it has moved up to below the pommel, which is its position also on the Zapher Papoura 14 (Tripod Hearth) dirk (cf. pl. 26:45). If the relationship of F to E ii is explicit, there are also borrowings from D ii shown by the groove lines down the blade of another dagger from Episkopi (Pediada, Crete). Probably typologically early is a longish weapon 48

Benton, op.cit. (supra n. 47).

134

N. K. SANDARS

from a grave on Cos at Langada (principally a LH IIIB cemetery which however lasted on till a later date); the pommel is broad and the length, o.41m., makes it a real weapon, on the other hand the moulding below the pommel has degenerated into a mere contraction. On the whole it is the later weapons that are the more interesting and that means those from Mouliana, Crete, from Perati, Attica, the hoards, and the group from the far northwest. Some of these are swords, and with them it may be said that the tendency to reduce length is arrested; some of the latest are also the longest, like a sword from Kephallenia which is nearly half a metre, and dated to twelfth century (see infra).

[AlA 67

Tholos A at Mouliana had two F swords, one very damaged and incomplete, the other complete and measuring 0o.58m.(pl. 25:33-34). The dating of the objects in this tomb has for long been a troublesome problem. It has been discussed by Catling, Dakaris, Desborough and others and some advance has been made, but it is unlikely that at this remove of time any greater certainty will be achieved concerning the different groups of gravegoods, and their relationship to the cremations and inhumations. The conditions are such that each object has to be considered on its own merits, independently of its possible associations. Furumark has grouped the pots under two heads, an earlier at the end of LM IIIB 2a, about 1200 according to his table of absolute dates,49but a decade or two later if his final date of 1230 for LH IIIB is thought too high; also a later LM IIIB 2c group at 1125-1075,or even later; while Tholos B he dates LM IIIB 2b or 1200-1125, again perhaps beginning a little later. Desborough inclines to a considerably later, possibly Geometric date for the cremation and some of the bronzes of Tholos A. The ninth or even tenth century is out of the question for the F class swords, and in fact all writers have agreed in placing them with the earliest group of pots, therefore somewhere not far from 1200, and after rather than before. If width of pommel were a valid criterion of date, the more complete Tholos A sword is typologically late, being narrow, though not quite as narrow as one in a tomb at Perati, Attica, a pre-

dominantly twelfth century cemetery. The length of this sword, 0.585,is the same as one of the "Type II" swords in Tholos B (Catling's No. 14, formerly No. ii) and a little longer than the other (No. 15 formerly 12). Unfortunately none of the other Mouliana swords from either tomb is in a state from which one can estimate its length. The more complete F sword had some pretensions to grandeur. The rivets were capped with gold and the hilt-plates were of ivory, there is fine ribbing on the flanges and a moulding below the pommel and the blade has "blood-channels":a fine line running the length of the blade at about 8mm. distance from the edge. This is unique for F, but is not uncommon on Type II swords, and is exactly like Catling's No. 14 in Tholos B, while the slight uniform swelling in the section of the blade, though rather broader here, is again matched on sword 14. The very damaged F sword in Tholos A has the same blood-channels and the same section. There is an incomplete Type II sword in Tholos A and a fragment of another (Catling's Nos. 16 and 17, formerly 13 and 14), but the sections are quite different from our swords, having a rounded midrib 9mm. broad. Catling has now regrouped the Type II swords to accord with the scheme worked out for Europe by Mr. Cowen.50This classification,like ours, depends on the hilt-form and not the blade section, so that Type II sword No. 16 in Tholos A, and No. 14 in Tholos B, are both Catling's Group III, while No. 15 in Tholos B is now his Group I. Of the four Type II swords in both tombs, it is only with No. 14 from Tholos B (Group III) that the F swords show any affinity. Even then the hilt is totally unlike, having a century's long history in the Aegean and the Orient, while the other is seemingly a comparatively recent European invention. Similarity depends on the blades, the section, the blood-channels, and the (for this class and period) unusual length. This section of blade does come on some other F swords, but the length and blood-channelsare unique, and the coincidence with sword 14 can hardly be fortuitous. It is therefore my belief that both types of sword, the "foreign" Type II and the "native" class F, should be considered together as complimentary or alternative armament. The lengthening of the blade of the ordinary class F dagger or dirk (as found at Zapher Papoura) is an immediate response to the

49 A. Furumark, "The Mycenaean III C Pottery and its Relations to Cypriot Fabrics," OpusArch 3 (I944) 262.

50 H. Catling, op.cit. (supra n. 3), J. Cowen, "Eine Einfiihrung ... ," Bericht R-GK 36 (1955) 52-155.

Mouliand' Tholos A

1963]


135

appearance of the foreign sword, and specifically to Catling's Group III, for which he gives a date range of 200oo-I25,with which there is no need to quarrel. If Furumark's date of LM IIIB 2c (11251075or even later) be acceptedfor the second group of pots in Tholos A, rather than the more usual "tenth century or later,"it would be possible to link the F swords with the beginning of this period, that is to say the end of the twelfth century, but a date in the first half of the century is the more likely. Along with the lengthening of swords went a shortening of spearheads.There are no more great bronze spearheads in the Aegean of 30-50cms., as in the fifteenth and earlier fourteenth centuries. The latest (not from the Aegean) are probably those in the LC III hoard from Enkomi.5 Mouliand Tholos A has a very damaged small spearhead with a faceted socket, which is descended from an Aegean type, and in Tholos B there are two very short (9.5cm.) spearheadsof foreign types with entire sockets and bronze rivets in the shaft. These changes in armament mean another change in the tactics of combat. One of the vases in the late group in Tholos A, the krater decorated "in a late Minoan tradition" that persisted almost into the Geometric period, shows on one side a horse-riding warrior armed with a shield and spear. In spite of the uncertainty of its date, the scene may provide a clue to the change of tactics. A mounted spearman would not have been able to manipulate the great spears of earlier ages. The shield bosses in Tholos B seem to belong to a different pattern from that portrayed, but may have been used in the same way.52

small iron knife with bronze rivets, not unlike the iron knife, also with bronze rivets, in Tomb VII of the cemetery at Gypsades, Knossos, which is probably twelfth century.53 S. Iakovides states that in the Perati cemetery there are no more than three generations of burials in any tomb; so as the pottery is almost entirely "Granary"or "CloseStyle" with a very little that is possibly earlier, the finds should be broadly contemporary, predominantly of the twelfth century. They include Egyptian scarabs, faience beads and carnelian "pomegranate"beads like those that became common in the Levant, the Dodecanese, and elsewhere in the Mycenaean world in its last phase. There is even a little gold, so that Iakovides is probably right in claiming this as the last flowering of Mycenaean civilization, at a moment of comparative prosperity just before the final collapse, when the sea lanes were still open to the southeast for the importation of small luxuries. Like Ialysos and some other LH III cemeteries, there are a few cremation graves among a preponderance of inhumations in Chamber Tombs; while the few small fibulae found may point to northern links (but not necessarily new links, for they are of later types than the great violin-bow fibulae in Chamber Tombs at Mycenae, from which they may be descended)." Speaking very broadly, the Mouliana Tholos A and Perati Tomb 38 swords appear to be contemporary with each other.

51 Dr. Catling, who drew my attention to these spears, considers them Aegean work. 52 V. d'A. Desborough, ProtogeometricPottery (I952) 270. L. Lorimer, Homer and the Monuments (1950) I55. ~ LM III B2-Sub-Minoan ca. 200-10o75;M.S. Hood-Huxley-Sandars, "A Minoan Cemetery on Upper Gypsades," BSA

53-54 (I958-59) 194-262. 54 A double spiral ring in one tomb does look genuinely northern. 55Tsountas,ArchEph(189I) 25; Catling,op.cit. (supran. 3) 1956, p. iio, for the Athenshoardsee also Furumark,op.cit. (supran. 3) 95.

Mycenae Hoard Before leaving Attica and the Peloponnese for the north and west, mention must be made of founders' hoards (apparently a new phenomenon Perati Tomb 38 in the Aegean), since at least four F weapons have Another original weapon is a short sword with F- come from them. There is one in the Athens type hilt in Tomb 38 of the important and recently Acropolis hoard. Montelius' argument for dating excavated cemetery at Perati Portoraphti, Attica. this hoard by certain sherds of a LH IIIA 2 pot Like Mouliana Tholos A the Perati sword has very found near it is hardly tenable, and the contents deep flanges and a narrow pommel, but the grip is of the hoard can be matched in the others: Myshorter. There is no thong moulding, no blood cenae, Anthedon, Ithaca; so there seems little grooves, and the blade has four narrow ribs running reason to date it any earlier. The largest of the down the middle in the place of a midrib, probably hoards is the first of two, found by Tsountas in derived from the much finer ribbing or grooving 1890, on the Mycenae Acropolis.55 It had two F on the blades of D ii dirks. Tomb 38 also had a dirks, both under o.4om. so not long enough for

136

N. K. SANDARS

[AJA 67

swords (pl. 25:36).56 There were also Type II swords, and a sword and a dirk with hooked shoulders,Class G, to which I shall return. This hoard deserves detailed study in itself, for some of the bronzes in it are of great interest, and its size is remarkable.5" Most of the other bronzes: sickles, one-edged knives, flat and double axes, and chisels, are met in the small "porous wall" hoard found by Professor Wace in a house at Mycenae,58in the hoards from Anthedon, from Polis, Ithaca, and most recently in the wreck of a trading vessel discovered off Cape Gelidonya, in southwest Turkey. Tools are common to all, but weapons are rare and this big Mycenae hoard is alone in having so many. In addition to the half-dozen (at least) swords and daggers, there was a "northern" type of spear (pl. 25:37) (entire socket and broad, leaf-shaped blade), and also arrows of more than one type. The date of this hoard has been discussed by Catling, who finds no evidence for anything earlier than I300,59 while at the other end only the foursided arrows, Tsountas' rerpdirXEvpot,and the horse-bits have caused serious questioning. None of the other objects, including razors, tweezers and a great variety of one-edged knives, large and small, are later than LH III. The arrows which have caused most trouble are probably those with the number 2560 in the National Museum, and possibly the much longer and heavier ones (if indeed they are arrows) No. 2554 (pl. 25:39-40). The arrows No. 2560 are of the same shape as certain bone arrows known from Mesolithic times and usually called fowling or club-ended arrows; they have a small weighted head. In these bronze ones both head and tang are four-sided. Arrowheads very like them were found in the Menidi

Tholos and at Zygouries.60Club-ended bronze arrows, not four-sided but functionally similar, were found in the big tomb, No. XXX at Gezer, that had the horned sword. Of the two horse-bits I have only seen one, the other was evidently in very poor condition and possibly smaller (pl. 26:48). The one bit in good condition is of the same general type as a pair of bronze bits found in a "late Mycenaean tomb" at Miletus. This was presumably in the cemetery on the east slope of Deirmentepe, with rectangular chamber tombs with typical Mycenaean dromos. I have failed to discover the circumstances of the find; but at the date of their publication the earliest settlement at Miletus was thought to be LH III, and the pottery exposed in Berlin before the war was exclusively IIIB and IIIC.61 The extraordinary size of the mouthpiece, and in fact all the dimensions of the Mycenae bit, are the same as a pair of bronze bits found at Assur and called "late Assyrian," but in fact of unknown date pending publication in Assur. These bits belong to Potratz' "Class I," for which the range of dates is from mid-second millennium to a little after 8oo. They appear on reliefs of Assurnasirpal II, 883-859,but no longer on those of Assurbanipal, 668-626. The Class I bits have in common the twisted bronze "rope" of the mouthpiece, in two connected sections, with its extremities drawn through the cheekpiece and "knotted." The bronze at Mycenae is a single rod, twisted on itself; at Assur it is triple. The most likely bit to be datable is one from Gezer, which belongs to the period between the end of the eighteenth dynasty and the Hebrew monarchy."2 So although nothing in the hoards requires a date much before i20o all the objects could have been

56 Dakaris, op.cit. (supra n. 3) 139, n. 4. Following Furumark, Dakaris puts them in different classes, b i and b 3, because of the different outline of the blades, but the grips are practically identical, while the difference in length (o.373m. and o.33m.) or in breadth (the longer is also slightly narrower), does not seem to me important enough to separate them. In the essentials of square shoulder, pommel-flange with contraction below, and rivet pattern they are identical. 67Some idea of the quantity may be gained from the numbering in the National Museum where the catalogue numbers of objects run from 2530-2560, and these 30 individual numbers cover groups of, for example, 12 flat axes (no. 2540) or 9 double axes (no. 2541). It is possible that some objects from the smaller hoard, found by Tsountas in the same year, have been grouped with the larger hoard. 58A. Wace, BSA 48 (1953) pl. 2, pp. 6-7, and Stubbings, BSA 49 (1954) 292, Mycenae; AJA 6 (189o) 99, Anthedon; S. Benton, BSA 35 (1934-35) 71, Polis (?); J. du Plat Taylor, "Underwaterexpedition off Cape Gelidonya," AnatSt II (1961)

26; G. Bass, "The Cape Gelidonya Wreck," AJA 65 (I96I) 267. 59 Dr. Catling tells me he would now lower this estimate slightly. 60 0. Montelius, La Grace PrIclassique (1924)

I64 nos. 562,

564; also from a LH level at Zygouries, C. Blegen, Zygouries 203, fig. 191. (1928) 61 Fimmen, Die Kretisch-Mykenische

Kultur

(1924)

I6;

Stubbings, op.cit. (supra n. 30) 23. The bits are illustrated by H. Potratz, AOF 14 (1941-44)

1-39, fig. 11 Mycenae, and fig.

io as "Asia Minor west coast"; Potratz accepts the "Miletus" bits as from a late Mycenaean context "like the Mycenae bit itself," AFO Io (1936) 317-40, fig. 13, M 151 (Berlin). 62 Potratz, op.cit. (supra n. 61) figs. 2 and 7, see also AFO Io (1936)

334, fig.

II,

14277=VA

7284 (Berlin)

and Prze-

worski, Die MetallindustrieAnatoliens (I939) pl. xmII,3. I am grateful to Mrs. Maxwell-Hyslop for advice on the dates of these Asiatic bits. R. Macalister, The Excavations of Gezer (1912)

vol. II, fig. 214, pp. 13-14-

1963]


hidden during the following century; individual

pieces,however,weredoubtlessmadeearlier.

Epirus, the Ionian Islands and Sicily The E class of weapon never found its way to the northwest. The F class did, like C before it, and probably outstripped C, as I shall hope to show. There is one from Dodona, and another in a cist-grave, Tomb I, at Kalbaki, near Ioannina, not far from the site where the two C swords were found. This grave had, as well as a rather peculiar F dagger (pl. 25:35), a spear with entire socket and a tiny hump-backed knife, both probably northern types, and some local pots. The group is discussed in some detail by Dakaris, along with other graves at Kalbaki, which had bronze wire bracelets with spiral twisted ends of unequivocally northern pattern, also an amber bead. Another find from the district, without further details, is an ogival spear of a type particularly frequent in Yugoslavia, but known also in the Ionian Islands, Italy and Central Europe. Dakaris has investigated the background to these "northern" objects, amongst which the F knife seems curiously out of place. The blade is not unlike the Carpathos D ii dagger with its narrow lines down the centre (pl. 24:26), but the grip is unlike any other, for it has the flanges on one face only. An old mend is witness to the hard service it had seen before it was buried. The Ionian Islands are very rich in amber beads, often crudely shaped. On Kephallenia there are graves with fibulae and Late to Sub-Mycenaean pottery. It is in this setting that the two F swords from an inhumation Tomb 2, Diakata, make their appearance. Miss Benton dates them in the twelfth century or later, and if the pottery is as late as she believes (Sub-Mycenaean-Proto-Geometric) the eleventh century would seem equally appropriate (Furumark IIIC 1-2). The pommel of the shorter sword is among the narrowest known and typologically very late; both are over o.4om. long and are without exotic features. Another sword was found at Lakkithra in Tomb A, Grave 6. This was the only swordsman's grave in a multiple tomb of the peculiar Kephallenia 63 See Benton, ProcPS 18 (1952) 237; the Lakkithra sword I owe to Dr. Catling. A tholos tomb said to contain LH IIIB sherds recently dug by Dakaris near Parga, Epirus, is interesting in this connection, see Archaeological Reports I960-61, Hellenic Society, p. 15. 64 This does not appear in photographs but I have examined

137

type. Marinatos compared the sword to those from Diakata. Damage to the hilt and at the shoulders leaves some uncertainty about its classification. The pottery is not identifiable as between individual graves but as a whole it shares the peculiarities of much Kephallenia pottery of LH III and later, the difficulty of dating being much increased by the absence of a true Proto-Geometric from the island.63 Sicily-Cornwall In the second millennium T-flanged pommels were (with two exceptions, see Appendix) unknown in Europe outside the Aegean area. It is usually thought that the Italian Iron Age series must derive from the Aegean. A difficultyhas been the gap to be bridged between the last of the one and first of the other. In view of this difficulty two finds in Sicilian cemeteries are of particular interest. The one is an incomplete dagger or dirk in Tomb 44 of Monte Dessueri, the other a miniature version of a sword in Pantalica North, Tomb 48. I believe that both are related most nearly to the F class of the Aegean. The Monte Dessueri weapon is incomplete and much damaged (pl. 25:41, pl. 28:68). The grip is broken off,64 and one side is badly worn. It is not possible to say how the grip ended, but that both shoulders were originally square and slightly flanged can be seen from the one still intact; there is a rivet-hole in the grip, the blade is flat and was probably once longer. If the damaged shoulder is restored, the width is the same as that of the Kalbaki dagger. If this fragment had been found in the Aegean I would not have hesitated to complete it with a T-flanged pommel. The Kalbaki dagger had been broken, and anciently repaired, at the same point in the grip; and a dagger almost certainly of F class from Oros, Aegina, is again broken near the beginning of the hilt, as is the Pelynt, Cornwall, fragment (pl. 25:42, 44), SO that the broken grips may explain those apparent gaps in the spread of T-flanged pommels. The square shoulder rules out "Peschiera"daggers, on which the shoulder is always sloping. In fact a miniature replica of what this once probablylooked the weapon in the museum at Syracuse where the kindness of Prof. Bernab6 Brea made it available, I am also most grateful to Mrs. L. Guido for much help and for photographsand drawings. For dating see Taylour, op.cit. (supra n. 27) 74; and Maxwell-Hyslop ProcPS 22 (1956) 126, pl. io.

138

N. K. SANDARS

[AJA 67

like is to be found not far off in the Pantalica cemetery. It is in grave (or Chamber Tomb) 48 of the earlier north cemetery, and though only 13.2cm. long it reproduces all the features of the F sword or dagger: the square flanged shoulder, the T-shaped pommel, and a rivet-hole where one is so often placed, in the centre of the upper part of the blade (pl. 25:43, pl. 28:69). Miniatures of this sort, surrogates for real weapons, are known in other graves at Pantalica, at Dessueri and in Sicilian hoards. It has sometimes been said that miniatures are found in the (smaller) Tsountas' Mycenae and the Athens Acropolis hoards, but this is very unlikely. The Athens "miniature" shows the "blade" thicker than the "grip" in section, and Tsountas is no doubt right in calling the Mycenae one a tool.65 A far more convincing object is the miniature in a tomb at Gezer, Palestine, dated by Macalister to his "Fourth Semitic Period IIoo-9oo"; being round-shouldered it is more like some of the Sicilian "miniatures" than anything in the Aegean.66Dessueri 44 and Pantalica North 48 are not far separated in time, for the Dessueri Tomb is one of the earliest of that cemetery, judging by its "Pantalica type" pots and the other bronzes, though none are closely datable.67 No more Aegean pottery was imported in the Pantalica cemetery, but dates, based chiefly on the fibulae found in some graves, suggest just before or just after iioo for the end of the Pantalica I phase, to which grave 48 belongs, and the beginning of Dessueri, with grave 44 a very little later. They may well be contemporary with the three swords from Kephallenia.68 Mrs. Maxwell-Hyslop has warned us against a too facile acceptanceof an Aegean origin for exotic Sicilian and Italian metalwork of the end of the second and beginning of the first millennium, and amongst others the T-flanged pommel of the Italian Iron Age swords."6The temptation to over-

simplify the problem is great, and probably many diverse influences were at work in the disturbed years following the great sea raids on Egypt. There was opportunity for unexpected encounters, and bizarre cross-fertilizations abounded. Some features of the later Italian swords have no parallel in the Aegean; I suggest that with Aegean outposts in the Ionian Islands in the twelfth and eleventh centuries, and Aegean swords or daggers in Sicily probably as late as the eleventh, there is not so great a chronological gap as has been sometimes supposed." Class F daggers or swords in Sicily provide a useful pointer to the date of, and the route taken by, that otherwise curiously isolated dagger said to have been found over a hundred years ago in a barrow at Pelynt, Cornwall71 (pl. 25:44). It is quite typical with its square shoulders and wellmarked flanges. The end of the grip is missing, so that it is not possible to apply width of pommel as a criterion of date; nor is the position of the single surviving rivet at the base of the grip any help; but this provides a clue to the shop or hand that manufactured it. Others with the rivet in this position are the two from Diakata, Kephallenia, one from Langada, Cos, Tomb 46 (the widths are exactly the same at the shoulder) and Dessueri Tomb 44. The route westward taken by the daggers may be the same as that, in reverse, of the later dispersal of amber beads, often very crude, that have been found on the Mycenae Acropolis at various times and places, in Crete and the Dodecanese (including tombs on Cos), in Epirus and in the Ionian Islands (Ithaca and Kephallenia very plentiful), and of course in Sicily. The amber is not necessarily Baltic, and generalizations as to its ultimate source cannot be made without analysis of each individual example." The impression gained from this class as a whole

6 sMontelius, op.cit. (supra n. 60) I54, fig. 498 Athens; Tsountas, ArchEph (1891) 26, pl. 3, 6, Mycenae; I owe to Dr. Catling an interesting suggestion that they may be unfinished rough-outs for small tools. 66 Macalister, op.cit. (supra n. 62) fig. 531 b. Gezer; H. Miiller-Karpe,Beitriige zur Chronologie der Urnenfelderzeit . . . (1959) pl. 2, H-I Pantalica. 67 The longest of the weapons, a parallel-sided blade with no grip extension but three large rivets, may belong to the same family as unclassified blades in Chamber Tomb 82 at Mycenae, and one like them from Steno, Aulis, at Chalkis; see Tsountas' excavations Mycenae, and Theocharis, Aulis, not published. 68 Brea, Sicily Before the Greeks ('957) 151, Taylour, op.cit. (supra n. 27) p. 69 n. I. 69 Maxwell-Hyslop, op.cit. (supra n. 64) 135, fig. I, 3, Modica

hoard, miniature sword with flanged pommel, probably the earliest from Italy; see also Miiller-Karpe,op.cit. (supra n. 66) text fig. 32, Pantalica "phase 2," iith century; also figs. 46-47 from Tarquinia and Terni 2; 9th century. See Wainwright, AnatSt 9 (1959) 197 on other Asiatic traits. 7o Miiller-Karpe,op.cit. (supra n. 66) absolute chronology pp. 226-27, the evidence for dating Bronze "C" seems based on Aegean objects by no means restrictedto the I4th century, i.e. the flanged pommel of Hammer for which see Appendix; flanged pommel swords must bridge Pantalica II-9th century in Italy. 71I am indebted to Mr. L. Grinsell for advice on the reliability of this old find. 72 Hood-Huxley-Sandars, op.cit. (supra n. 53) 237, 261 (A. Werner).


1963]

139

The new tendency seen in the F class dagger, dirk, and sword of the thirteenth and twelfth centuries, to break away from standardized forms in favour of the uniquely peculiar specimen, is even more marked in the G class. The chief distinguishing mark is the shoulder with downward hooked horns (or "quillons," to give them a more technical name). This is Furumark's class C2. Unlike the F class, some of the shorter of which may have served as knives or double-purpose knife-daggers, the G class is exclusively one of weapons, as may be seen from the narrow, pointed, dirk-like blades. All have the T-flanged pommel and either a distinct midrib, or grooves, or raised lines down the centre of the blade (there is one possible exception). The curious form of horns or quillons leads one naturally to look back to the C swords, the latest of which was dated to the first half of the fourteenth century. The only possibly later horned sword from the Aegean was the Tripod Hearth (Zapher Papoura Tomb 14) (pl. 26:45) dirk, which is also typologically midway to class G, since it has a downward bend to the broad ends of the horns. Even so this does not carry the tradition of horns below the fourteenth century, unless we date the Tripod Hearth Tomb considerably later than is justified by its other contents. There are however G swords for which we have no dating evidence at all, and which might help to bridge the gap. One comes from Siteia, and has horns

very like those of the "folded" C ii swords, though of course now hooked downwards (pl. 26:50), also a moulding below the pommel and groove lines (either incised or produced by the mould) down the blade that hark back to Class D ii. At the other extreme there are examples from graves in Attica and Delphi that take us down to LH IIIC and so into the twelfth century (see map). The longest and most interesting of the G swords is one from the Perati cemetery, from which also came a massive class F blade (Tomb 38). The hooked sword was in Tomb ii, it measures o.6om., longer even than the Mouliand F sword, and the same as some Type II. The pommel is very narrow, and immediately below it there is still a ribbon of gold with tiny punched bosses. This, with the gold-capped rivet of one of the Mouliand F swords, is the only surviving gold from later weapons. The blade also is unique, being truly leaf-shaped with the greatest width in its lower third. Gold and carnelian beads in the same Perati Tomb confirm an impression of comparative riches; another interesting object is a bronze one-edged knife with the grip-end in the form of a bird's head, turned back to face the point. The date of most of the tombs at Perati is given by the Close and Granary style pottery. A knife extraordinarily like the bird-head knife in this tomb was found in a hoard of bronzes at Spalnaca in Transylvania,73 but the date of the hoard is Hallstatt B, therefore possibly as much as two hundred years later. In fact the Perati bird-head is more likely to have relatives among what appear to be bird-head scabbard-ends of Asiatic swords. These have not survived in metal but one seems to be worn by the BoghazkiSy warrior at the gate, by a "weather-god" at Sinjerli, and a "Baal" figure on a stele from Ras Shamra.74The style is the same as that of the Syrian ointment boxes, and there are many oriental luxury goods with this motif. The longest of all G weapons is a sword in the big Mycenae Acropolis hoard to which we have already given some space. It is 0.623 long and has a strong midrib; there is also a short sword very

73 Now in the Museum of the ArchaeologicalInstitute in Cluj, Rumania, I am indebted to Drs. Horedt and Rusu for information and facilities for study of this hoard. See Holste, Hortfunde Sildost Europas (1951) pl. 47, 25. I am also indebted to Mr. Cowen for advice. 7 The existence of a bird head is not in all cases equally

clear, owing to the weathering of the stone, but in all the scabbardend is bent back so that the tip faces the grip and the bird head is more or less strongly suggested: 0. Gurney, The Hittites (1952) pl. 4; Schaeffer, op.cit. (supra n. 8) Ugaritica II (1949) 23-24; Ausgrabungen in Sendschirli x (1895); Mitteilungen aus OrientalischenSammlungen XI pls. 40-41.

is no longer that of the products of big prosperous workshops, with a long tradition behind them, like those of Crete and mainland Greece and even the Dodecanese, in the fifteenth and fourteenth centuries. It suggests instead work turned out by foundry smiths of no great ambition or particular skill, who were capable of producing a rather dull utilitarian weapon on the one hand, but who under some exceptional stimulus could also produce the large, clumsy, ill-conceived weapons of Perati and Mouliani. After a brief account of the two remaining classes, G and H, I will return to this change and make tentative suggestions as to some of its causes.

Class G

140

[AJA 67

N. K. SANDARS

like it measuring0.471"(pl. 26:47,46). The shorter bronzes,from an importantfind now at Copenweapon looks quite well-balanced,a handy dirk; hagen. Details of the find are not ascertainable but the longer is most unwieldy and eccentric, but there seems no reason to doubt the associamoreso thanthe Peratisword,and may be grouped tion of the three bronzes,a short sword or dirk, with it and with the Mouliana F sword as ex- a knife and a spear, which were purchasedtoamples of inexpertexperimentation. gether in Rhodesand are said to have come from I that sword a Mycenaeantomb at Siana (pl. 27:53-55).The suggested Regardingthe Mouliani of class H sword or dirk is essentiallyidenticalwith the reasonfor its manufacturewas the appearance Type II swords, introduced from north of the one in the Ashmolean, Oxford, left by Arthur Pindus.The sameMycenaehoardhad one of these Evansand saidto comefromPergamon(pl. 27:52). new style weapons,now incompletebut which is The lengths are very nearly the same, also the estimatedto have measuredabouto.6om.,like the grip and the formof the horns.The only difference longer of the classG. Dr. Catling has arguedper- is that the lines down the bladecentreare, in one suasivelyfor the introductionof Type II swords case,sunk and in the other raised.Most of all they mercenaries,enlistedby hard-pressed resembleeach other in the rivetlessflanged grip by "barbarian Mycenaeanprincesat a time of greatdisturbance.""7which narrowsto a rodlikeextensionof rectanguIf this were the true explanationa sight of these lar section, the hallmarkof this workshop.It is eminently efficientweapons might have spurred found also on the knife in the Sianagroup,which, on local smithsto attemptto betterthem by longer though slightlybroaderand heavier,is in all other and bigger weapons in an old tradition,before respects like a knife from Ialysos found in a themselvesgoing over to the new fashion. gravewith a spear,a pin and potsdatedLH IIIB-C A G dirk, probablyfrom Ithaca(pl. 26:49), has (pl. 27:56). This in turn is so close to one from a the blade decoratedwith finely outlined ribbing "Mycenaean Tholos" at Colophon that they like the Siteiasword,but it is the grip that is more could have come from the same mould (pl. interesting,for in placeof the usual skeuomorphic 27:57). Both are now in the British Museum. moulding and its debasedproduct,a single con- Unfortunatelythe pots from the ColophonTomb striction,there are two constrictions,one imme- have not survived,but a silver pin or needle with diatelybelow the pommeland the other abovethe a loop-headis a fairly common type, usually in shoulderwith a parallel-sided, broader,middle sec- bronze (Tarsus LH II, Troy etc.), and a faience tion between. There is also deep channellingon spacer-bead,which has been illustratedwith the the outside edge of the flange rather like the pin, is of a kind that occursfrequentlyat Ialysos Dodona F dirk. These featuresgive it a most un- and elsewherein the late Mycenaeanworld.77Yet Aegean aspect, nor is it Italian either. Oriental anotherof these knives comes from Troy, from handleslike that on a two-edgedknife from Gezer the VII city. It appearsto be slightly heavierand broaderthan the Ialysosand Colophonknivesand provide a more likely source (pl. 26:51). so closer to Siana.78 Class H swords: Siana GroupBronzes The third of the Siana bronzes,the spear (pl. This history of bronze swords in the Aegean 27:55),is characterized by the facetingof the socket must end with a small group which lies on the and the markedsplay of the facets at the base of borderlinein time and in space of the Minoan- the blade.This is an Aegean feature;it is known Mycenaeanworld. These swordsor dirks are part also in Rhodes,at Ras Shamraand at Tarsus (pl. of what I proposeto call the "Siana Group"of 27:59).79 The last is perhapsthe closest and may 75 There has been some confusion in the past between these weapons as only the shorter is figured by Tsountas. This has been often copied with the position of the top rivet wrongly placed: it should be central. 76 Catling, op.cit. (supra n. 2) I96I, I20. 77 Furtwingler-L6schke, op.cit. (supra n. 24) pl. D, 9 and pl. C, Io and ii from tombs 13-38; lalysos O.T. XXVII had, as well as the knife (no. 72. 6-20. I) in the British Museum, the pots illustrated op.cit. Atlas pl. 9,51 and another "like" pl. 30, 270; the spear mentioned op.cit. pp. 14 and 75 was not identified by the excavators and cannot now be traced. I am very grateful to Mr. R. Higgins of the British Museum for this in-

formation. British Museum Bronze Age Guide (I92o) i66, fig. 177 and AJA (1923)

17.

78 Unfortunately not datable, it came from room VII e and could be either from the VII a or b city, W. D6rpfeld, Troja und Ilion I, p. 396 fig. 384. 79 Furtwdingler-LSschke, op.cit. (supra n. 24) pl.

D

14; Schaef-

fer, op.cit. (supra n. 8) III, fig. 224; Tarsus at Adana Archaeological Museum no. 38-1645 3681 and H. Goldman, Excavations at Gdzlii Kule Tarsus II (1956) pl. 427, 97, and p. 55 where the reference to "a beautifully preserved lancehead" (no. 96) evidently is intended for this.

1963]


141

be LB IIA (before the conflagration), but the house in which it was found had a squatters' occupation over it and may equally have been introduced from above and belong to the LB IIB period, the sea raids. This very small group of bronzes is so sharply distinguished from all others, the knives as well as the dirks, that it seems reasonable to treat it on its own. At the same time I believe the dirks may be legitimately related to certain Syrian-Levantine weapons which are, at first sight, rather different. They do not have horns but have a projection which must have served the same purpose. First there is a sword or dirk from Ras Shamra which is a little longer than Siana, but the blade is similar in shape and there is central ribbing like Pergamon, and the same rodlike extension of the flanged grip (pl. 27:58). This flange however is deeper and is scalloped in outline, and there is a separate metal collar (a feature the Pergamon and Siana dirks almost certainly also had if they were to keep pommel and grip together, but it was very likely not of metal). A possible local forerunner, also at Ras Shamra, appears on an ivory in the hand of a ruler who Schaeffer thinks may be Niqmadu II, the contemporary of Suppiluliamu and Amenophis III, and which has already been referred to in connection with the adoption of central ribbing on Aegean blades."8 The Ras Shamra dirk is undoubtedly very closely related to one from Atchana, belonging to levels I or II, which has a similar grip, but this time the wide lunate metal pommel has been preserved. This hilt has a long Asiatic ancestry going back to Ur and the Royal graves at Alaca H6y6k.81 The blade is generally similar though a little thicker in section. The dates too are broadly contemporary: before the destructions by the Sea Raiders early in the twelfth century. The only points of similarity to the much longer sword with the cartouche of Mineptah, from Ras Shamra, is the parallelism of part of the blade, and the section with central ribbing. This applies too to the problematic sword from Samos which, though said to have been found with a sixth century

bronze statue, must certainly belong to an earlier epoch.82A fragment that might possibly have belonged to a dirk of Atchana type is in the Ashmolean Museum (pl. 27:6o). Finally, the Medinet Habu reliefs seem to show our swords in the hands of the Sea Raiders.83A great variety are shown and among them are some with marked horns of the Pergamon dirk type, while others are more like the Atchana sword; the lines down the blade centre are clearly shown. We should not perhaps put too much weight on this evidence, or expect exact archaeological documentation from such a source, but on the whole it appears that the "horned" or semi-horned sword is oftener in the hands of figures with feathered headdresses (called Peleset, Thekel, Shekeles and Denyen), while the figures with horned helmets (called Shardana "of the Sea") have swords with sloping shoulders which are certainly not Aegean. Swords which might represent Type II are in the hands of Egyptian mercenaries. If the Siana and allied bronzes are admitted to be a distinctive group their distribution is interesting, comprising as it does the coastal fringe of Asia from the Troad to Syria, with the best authenticated examples coming from Rhodes (two gravegroups), particularly if the comparative isolation of Rhodes from the Greek mainland in the twelfth century is remembered (see map). Typologically the swords can be explained by a marriage of Aegean and Orient. The type of blade is not originally Aegean though Aegean weapons had been evolving in that direction, particularly in the Dodecanese, since the fourteenth century (D ii from Rhodes and Carpathos). The horns are undoubtedly Aegean and closest to the Siteia class G sword, but the pommel extension is not, and is confined to this small group. However rod-handles of a sort are Asiatic and Cypriot rather than Aegean. There is unfortunately little material evidence from inside Anatolia and particularly from the Hittite Empire, but the "Dagger-god" of Yazilikaya stands embodied in a dirk with a pointed blade and ribbed or furrowed centre lines.84 The Siana group are not the only bronzes that

so Schaeffer, op.cit. (supra n. 8) 276-77, fig. 239; Schaeffer dated the dirk 1250-1200. 81 Maxwell-Hyslop Type 9, "Daggers and Swords in Western Asia," Iraq 8 (1946) 1-65. 82 Schaeffer, "A Bronze sword from Ugarit," Antiquity 29 226-29; Samos sword L. o.5o5m. including I2cm. (1955) tang JHS (1909) i9h. I am indebted to Mr. A. Snodgrass for drawing my attention to this sword.

83 N. de G. Davies, Medinet Habu, University of Chicago Oriental Institute Publications, I pl. 39, detail showing Ramses III defeating the Sea Raiders,the warrior in the left-hand "swanboat," see also pl. 32. It should perhaps be noted that the bow was the weapon that saved the Egyptians-proving superior to the sword; see also G. Wainwright, JEA 47 (1961) 71ff. 84 Seton Lloyd, Early Anatolia (1956) pl. 13 b, frequently illustrated.

142

N. K. SANDARS

straddle the Aegean and Asiatic coasts. The horsebits in the Mycenae Acropolis hoard with their connections with Miletus and Assyria have been referred to. There is also a rather insignificant little one-edged knife with a twisted rod of bronze for a handle, which has escaped notice so far, but which also has an interesting distribution. There is one in the same big Mycenae hoard (pl. 25:38), one in a tomb at Perati, another from Karphi and yet another from Boghazkay.8" With these and with the Siana group of bronzes we have moved into a world of confused frontiers and probably still more confused politics. The great powers are no longer land-empires and principalities but small groups scattered among islands and along a coastal fringe, inhabiting the nests of pirates; they are almost certainly raiders, footloose and on the make.

[AlA 67

Historians of the end of the Mycenaean age have, with the help of Egyptian and Hittite documents and Greek traditions, pieced together a reasonably convincing chronicle of events. There is very little that a study of swords can do to deepen knowledge of those troubled years. On the other hand the historical findings do help considerably in the interpretation of the archaeological material. On the whole, I have preferred in these pages to restrict myself to presenting the material evidence (as far as that may be done with regard to a single arm), and have left interpretation aside. Yet here and there a few conclusions are likely to be forced on any objectively archaeological observer. Enough evidence exists for each of the latest weapons we have been discussing, the F, G and H swords, to date them as contemporary with one another. They belong to the thirteenth and twelfth centuries. We have seen that by 1300 the last of the products of the great fixed workshops, the C and D i swords, had disappeared. Their traditions were now carried on in a number of smaller workshops in places farther removed from each other, or by traveling bronze founders and smiths who produced the D ii and E ii weapons and tools; and who also owed a good deal to fresh contacts with the Levant, in which the Dodecanese had an

important r61le.By the later thirteenth century disintegration had gone much farther, and though some of the F class daggers and short swords retain a semblance of traditional uniformity, the majority of weapons are strange hybrids. There is a return to pointed dirks with flat blades and blade decoration borrowed from Asiatic sources; and occasionally to longer blades with midribs. There are significant new factors that begin to make themselves felt at the end of the thirteenth century (probably very little if at all before 1200) and which have nothing to do with the Aegean tradition. There is the appearance of Type II swords and of short spearheads with entire sockets. At the same time metal seems to be more plentiful, or at any rate more accessible (not necessarily the same thing), for which fact the large numbers of sickles, knives, and flat axes in hoards are evidence. The hoards themselves point to new methods of trade and transport, and the shipwreck off Cape Gelidonya with its vast weight of copper ingots, and its bronzes clearly linked to those of the hoards, show how metal was traveling in bulk between the Aegean and the Levant, and how easily simple types of tool and weapon were dispersed abroad. The Type II swords themselves appear to be an exception to the general disintegration of types. The eight weapons of Catling's Group I are widely scattered but reasonably alike; they come from Mycenae, Crete, Naxos, Cos and Cyprus, while his second group is confined to the mainland of Greece."8The northern origin of this sword and the relationship of Group I to Cowen's Nenzingen type will hardly now be disputed. The short spearhead with entire socket is no less northern in appearance; the only other possible source would be the Caucasus. I have no exhaustive list of the Aegean find-sites of these spears, most of which are of a simple leaf-shape, but some (with a distinctly northwesterly grouping) are ogival. Of those known to me there is a not insignificant coincidence with sites from which Type II swords have come. Mycenae, Moulian~i,Anthea, Langada Cos, and Enkomi have both. At Cos and at Anthea they were in the same grave. Other notable finds of spears are in the Ionian Islands, in Epirus, and Delos.87 There is at least a strong likelihood that

85The National Museum (no. 2744) group of knives includes this type; S. Iakovides excavations, Perati Attica, preliminary report To Ergon ... 1959 (1960) 9-12, fig. 9; BSA 38 (1937-38) pl. 28, 540, 687, 645, Karphi; K. Bittel, Boghaz-

kdy I, pl. Io, 9-10. 86 Catling, op.cit. (supra n. 3) 1961. 87 MetaxataKephallenia, Marinatos,ArchEph (1933) 92, with amber; Polis Ithaka, Benton, op.cit. (supra n. 58) 72; Dakaris,

Discussion

1963]


the mercenaries of Dr. Catling's article came armed with sword and spear; what else they brought with them is more problematical. Neither the numbers nor the distribution of Type II swords (Group I) nor of "northern"spears suggests more than the movement of highly mobile individuals. The next question is naturally: did they bring their own bronzesmiths with them? I have suggested that the lengthening of F and G swords into the o.6om. range, and the occasional adoption of so-called "blood-grooves" was a first reaction of Aegean smiths and their clients to the Type II swords. Dr. Catling has made a very interesting suggestion that the addition of the pommel-spur took place in the Aegean (Group II) and was then introduced into Central Europe, where it is typical of Hallstatt A swords (Erbenheim Group). If this is accepted it shows a second stage in which Aegean smiths have learnt to copy and modify an alien weapon, just as earlier they had adopted and adapted the T-flanged pommels of the Orient. Thereafter the southern swords diverge further from the European ones. With these mercenaries we are perhaps in the presence once more of a military caste, with its demand for a degree of uniformity of armament, and its patronage of smiths and workshops, such as existed when Class C and D swords were in use. The absence now of luxurious weapons show there was no rich court circle behind this demand. I do not believe there was a spectacular eclipse of Aegean metalwork in the "Dark Age." The production of all except the earliest Type II swords was probably in the hands of local smiths. By the time that iron came into general use the old outmoded types like F, G and H had been superseded in Aegean workshops. As far as concerns the swords after the twelfth century, the transition to new styles is partly blanketed by the simultaneous translation into the new, more perishable metal: iron. But it is an inescapable fact that the iron swords trace their ancestry back to Type II, with very little indeed of Aegean or Levantine tradition behind them. The little that can be deduced from the outgoing swords of the twelfth century concerns two important and apparently unconnected developop.cit. (supra n. 3); Delos Museum, B 4004, 10365; see also V. G. Childe, "The Final Bronze Age ...," ProcPS 14 (1948) 185; D. Gara'anin, Katalog Metala, National Museum Belgrade

(1954) pl. 7, 2; pl. 38 etc. For Greek sites with northern spears see also Catling, op.cit. (supra n. 3) p. 111 Anthea, 114; Cos

143

ments. First a very close link with Anatolia, particularly between the coasts and the islands: Siana type bronzes (dirk, spear and knife); bronze horsebits (Mycenae-Miletus-Assur), the small knives with twisted stem, the Anatolian-type boot-vase from Voula, Attica;88 and secondly the foreign, northern armament, sword and spear, found lightly scattered over the entire Aegean and Levant, including Egypt. A third factor, less spectacular but not without importance, is the influential r61leof the Dodecanese: Cos and Rhodes outstandingly."8 This had been building up for a long time, but, whereas in the fourteenth century it was as a relay-point and centre of exchange between Crete or mainland Greece and the Levant and Cyprus, now with the Siana Group we find those close ties with the Anatolian coast which were strangely lacking before. This then is the material: its interpretation in terms of Hittite correspondence, of sea raids and land raids, of Philistines and Achaeans is not the business of this essay. APPENDIX:

HAMMER DRNOVO BORODINO

At present the best discussion of the sword from Hammer, Bavaria: Bronze C tumulus burial with sword and pin, is probably E. Lomborg, ActaA Copenhagen 30 (1959) 136, where it is compared to one very similar from Dollerup, Aalborg, Jutland, found in a double grave with small bronzes characteristicof Northern Bronze Period II (IIbc), op.cit. 136-37. Both swords are extraordinary in Europe north of the Alps, having flanged grips without rivets, and T-shaped flanged pommel ends like Italian Iron Age and Aegean and Oriental Bronze Age swords. There are rivet-holes in the T, placed in a way never found in the Aegean. It will have been seen from these pages that the T-flanged pommel could have been taken from the Aegean any time after 1400 and down to the twelfth century, but not much later. At that time it was beginning to appear in Sicily and then in Italy. In the Orient its time range was altogether greater. These two swords are not quite as isolated as they appear if account is taken of one from Drnovo in Western Yugoslavia now at Ljubljana (no. P 3333). I am indebted to Dr. Gabrovec of that Langada T. 21 etc. 88 Praktika (I955) 24, fig. 20. 89 Dr. Catling has pointed out to me the curious situation in the I2th century when links between Cyprus and the Greek mainland are more obvious than between the latter and Rhodes.

144

N. K. SANDARS

museum for information and for supplying me with a drawing. It is a single find; the length is 0.47 and, though slighter, the blade and shoulders are very like Dollerup. The slightly flanged grip has two rivet-holes and the top is unfortunately missing. It does not belong to any of the betterknown European sword types. Drnovo may be compared with another rather similar sword found in the Danube at Nagytiteny, south of Budapest, which may never have had a T pommel projection, but the shoulders are similar. This I saw through the kindness of Dr. A. Mozsolics of the Budapest National Museum. None of these swords are Aegean work but all in greater or less degree show some southern inspiration, and may be in-

[AJA 67

mania 40 (1962) 255-87, appeared too late for comment here. CATALOGUE C Crete ZapherPapoura36, "Chieftain'sGrave,"Shaft-Grave. L. 0.945m. (without pommel), grip has pommeltang or spur, three small rivet-holesin grip, one in spur, rivets gold-cappedand pommel of ivory, two rivets in blade level with top of midrib, high rounded

midrib with decorationof fine spirals,also doublespiralson flangesof grip, and quadrupleon shoulder flange. Evans, "Prehistoric Tombs at Knossos" (hereafter PTK) Archaeologia 59 (1906)

51, figs.

38 and iio. In direct associationwith cruciform sword. Other bronzes, spears and tableware on Awith much hoard silver important very gold, slab above the single body. A very important hoard with much gold, silver and precious material, found at Borodino in South ZapherPapoura44, shaft-grave.L. o.955m., pommelRussia, held a silver dagger with a broad flattish tang with small rivet-hole, grip with three large rivet-holes, two large rivet-holes in blade, high gold-plated middle section, not so much a midrib strong midrib, with fine spirals, double-spiralson as a central panel. The hoard is discussed by Dr. At present pommel-tangand horns incomflanges. Gimbutas in ProcPS 22 (1956) 143-72,see especially plete. PTK, p. 62, fig. 66, AshmoleanMuseum"AE and terlinked.

p. 144 and pls. 12 and 13. The dagger is ca. o.275m., has square flanged shoulders and a flanged grip

462." Found with single burial, untypical A sword and stirrup-jar: date LM II/IIIA i (pl. 2i:I).

with three rivet-holes and no projection, there are Knossos, "Silver Cup Tomb," small ChamberTomb, sword found with first burial under skeleton. L. no blade rivets. Dr. Gimbutas, like earlier writers, o.613m., in three pieces, three rivets in grip, small connects the dagger with Mycenaean Shaft-Grave rivet-holes,pommel-tangextension with rivet-hole, be derived from Aegean weapons, but if it is to be derived from Aegean two small rivet-holesin blade, high angular midrib. Hutchinson, "A Late Minoan Tomb at Knosprototypes it is actually closer to later types. The B class swords and long daggers of the Shaftsos,"BSA 51 (1956) 68-73,pl. 8, e, f; fig. 2, 16, no. 2. Found with bronze, silver-plated,stemmed cup, Graves seldom have a flanged grip of this type, Cretan type silver pin, two alabastra,blossom vase: and where they do there are always rivet-holes date LM IIB (Hutchinson), LM IIIA I (Furuin the blade as well. The square shoulder, flanged mark). grip and rivetless blade are most like class F, but Phaestus, ChamberTomb 8 of "Tombe dei Nolili," these always have the T pommel, while E i, which two swords found but only one is illustrated,condition poor, in five pieces, lower part missing. Surdoes not, has a more rounded shoulder. In fact viving L. 0o.43m.,grip with pommel-tang,number is a Borodino mixture of characteristicsdrawn from of rivets uncertainbut one survivingwith gold cap, several different types of dagger. The "idea" of the high midrib ending at level of shoulders,gold sheet decorated centre panel of the blade might come on flanges with feathered pattern. MonAnt 14 from daggers like Vapheio

and Myrsinoch6rion,

(I904)

pp. 508-35, figs. 2o, 2oa, found with traces

of burning, and other bronzes including tableware but the decoration is only distantly Mycenaean. like Z.P. 36. The decorated pin has its best analogues in the Hungarian Bronze Age, where Mycenaean motifs ZapherPapoura 14, see under class H. are found much filtered through the local style. Mainland Greece The closest to this decoration which is actually Dendra, "King's Tholos," found in pit I of Chamber from the Aegean is found, perhaps suggestively, on near "King's"body. L. o.94m.,grip with three small the pommel of the sword of uncertain classificarivet-holes,gold-cappedrivets, two similar in blade, pommel-tangnow lost. Fragments remain of ivory tion from the island of Skopelos (possibly a fusion hilt-platesdecoratedwith spirals and probablythe of A and B class weapons), see Zenakis, KPHTIKA small gold bars of which some 5,000 were found, XPONIKA (hereafter KChr) 3 (1949) 534. see infra, Argolis sword for descriptionof this techThe important article by H. Miiller-Karpe, Gernique, high rounded midrib to level of end of

1963]

LATER AEGEAN

shoulder-flanges. Persson, Royal Tombs at Dendra (1931) p. 35, pls. xx no. III, xxII 2, xxIII 1-3 (sword no. 12); found with two cruciform swords and one class B, also spears and knives, gems, the octopus cup, and silver and gold cups. LH III A i (Furumark); the swords may be earlier. Dendra, "King's Tholos," from pit I found at the "King's" feet in a heap with other bronzes. L. o.682m. (incomplete), three gold-capped rivets in grip, two in blade, no pommel-tang surviving, horns unusually thick and short, strong rounded midrib to shoulder-level. Persson, Royal Tombs p. 36, pl. xx no. v (sword no. 15); found in a heap with spears, knives and two small lead horns (probably for bulls) date same as above. Mycenae, Chamber Tomb 81. L. o.62m., pommeltang, one small rivet-hole in grip, two in blade, the outline of the oval hilt-plate opening shows as a stain on the metal, well-marked rounded midrib. The sword is unusually light and slender, width at horn-tips only 6cm. Undated tomb, it also held an uncharacteristic class A sword and two unfinished hilt-plates of cruciform swords, one of banded agate, a bronze "drum" and small bone objects. ArchEph (1897) pl. 8 no. 2, pp. 107-o8. Nat. Mus. 3118, Tsountas dig of 1895 (pl. 21:2). Prosymna, Chamber Tomb XXXVII, two or more period tomb, the sword came from the earlier stratum. L. 0o.75m.,three gold-capped rivets in grip, one in pommel-tang, two in blade, fragments of wood and ivory from grip and gold wire-binding, Blegen, Prosymna (1937) 329 and 123-28, fig. 298; the lower stratum had at least ii burials much mixed, the sword was found in a heap with 6 skulls, bones and pots and a two-edged leaf-shaped razor. The pottery is LH IIIA. Argolis. Incomplete, lower part of blade missing, surviving L. o.40m., three rivet-holes in grip, goldcapped rivets with small pins, grip originally had ivory plates, ivory pommel preserved with gold covering, spiral decoration on gold which is formed of numerous tiny staples hammered flat and incised (see full and interesting account by Prof. Elisabeth Treskow of Cologne). There is a high strong midrib and two rivet-holes in blade. Sword typical of the fine quality C i swords of Crete and Dendra. There is a strong probability that this and the class D sword (see infra) came from the recently plundered tomb 12 at Dendra, which held a fine corselet and bronze vessels; gold-capped rivets, probably from a sword and vases of LH IIB-III AI. This sword is described and illustrated in Ars Antiqua AG Auktion III (1961) p. 30 No. 70 and coloured frontispiece. Dr. Astrom, Director of the Swedish Institute in Athens, who with Mr. Verdelis of the Greek Archaeological service has recently excavated further tombs at Dendra, has very kindly communicated the following information in connection with this, and the second sword of cruciform type. "Tomb 12 at Dendra was plundered early in January 196o, before the excavation took

BRONZE SWORDS

145

place in May of the same year. Some gold-capped rivets were found in the plundered part of the tomb (see Verdelis in ArchEph [1957] i6ff.). I believe that it can be proved in the forthcoming publication of the finds that the rivets belong to one of the two swords that were sold in Lucerne in April 1960. The tomb contained two Mycenaean IIB and two III AI pots. Only a single burial took place in the tomb and that in the Mycenaean III AI period, at which date the swords were placed in the tomb." See also JHS (1961) Archaeological Reports I9606I, pp. 9-Io for a general note on Tomb 12 at Dendra. Dodona. L. 0.58, "from the ruins," much damaged and incomplete, most of grip missing and tip, stout midrib, small rivet-holes low in blade, two more within and at base of horns, Carapanos, Dodone et ses Ruines (1877) pl. no. 140. LVII Perimatos, Ioannina. From a tomb, broken and incomplete, L. o.58m., no pommel-tang, two small rivet-holes in grip, two in blade, low down, and two at base of horns, finely ribbed high midrib, nicks for hilt-plates at end of grip and horns. Dakaris, ArchEph (I958) 131 figs. 6, 7. I received much assistance from Mr. Dakaris at Ioannina (pl. 21:3). Perimatos, Ioannina. From a tomb like above, very damaged, almost the whole of the grip and horns missing. L. o.857m., two rivet-holes low in blade alone survive, strong midrib. Sword probably like the other but longer. Dakaris, ArchEph (1958) 131, n. 7, fig. 6 (pl. 21:4). islands Ialysos, Rhodes. New Tomb IV, two burials in Chamber Tomb, L. I.ogm., quite untypical sword, halfway between C and B, grip appears to be unflanged, and without pommel-tang extension, 3 (perhaps 4) rivets in grip, two in blade, outline of oval hilt opening on blade, shoulders slightly flanged, and less horned than sharply pointed, strong high midrib, Maiuri, Annuario 6-7 (1926) No. 3622 pp. 93ioo, fig. 15, no. 18. The tomb had also a D ii sword, an E ii knife-dagger, and a one-edged razor, also pots LH III A2, but association not certain. Asclepeion, Cos. L. o.4Im., untypical, a cross between C i and the Syrian relatives of B. Three large rivetholes in grip and two in blade, stout faceted midrib, shoulder more pointed than horned, no details, unpublished, kindly shown me by Prof. Morricone. Nidhri, Levkas, "Brennplatz" R 7 and "over grave 24." Two very incomplete blade fragments, no trace of hilt, but with strong lobed midribs; they belong to the lower part of two different swords, but may be class A. D6rpfeld, Alt Ithaka (1927) pp. 229, 241, Beil 62, 3; AJA 65 (1961) 26 and pl. 17, 5, 6. Yugoslavia Tetovo, Skopje, Yugoslav-Macedonia. L. ca. Im. No pommel-tang, six rivet-holes in grip, two pairs of rivet-holes in blade, well-marked midrib. An excep-

146

N. K. SANDARS

[AJA 67

tionally fine sword perhaps closest to Perimatos. Dr. Vinski of Zagreb is publishing it shortly. I am indebted to him for information.

o.56m., Casson, Man (1923) 172, fig. 2 top; also Heurtley, Prehistoric Macedonia (1939) figs. Io4ff., where it is erroneously labeled "Karagliri, Bulgaria." Bulgaria "Bought in Athens," Stockholm, Museum of National Antiquities, No. 19293. L. o.675m., hardly a Dolnolevski, Prov. Pazardjik (Kalglari or Karaglhiri in older literature). L. o.77m. width of horns (point midrib but central thickening outlined by two channels 8mm. apart. to point) o.i2m., almost perfect, two rivet holes in grip and two pairs in blade, high grooved midrib, ?Dendra. Fragments of a sword in the National Muhorns nearly horizontal and cast solid. Found toseum, Athens, labeled "Dendra" but unpublished; if all fragments are from the same weapon approxgether with a spear but without other particulars. Guide to Archaeological Museum, Sofia (1952) pl. imately o.8om., horns flanged not "folded," high midrib with decorative ribbing, a curious sword in 9.1 (my pl. 22:5-6, pl. 28:64). some details more like C i. Perushtitsa, Plovdiv: L. o.76im., flanged grip with pommel-tang and slot opening, three rivet-holes in Islands grip and two in blade, nearly horizontal solid horns, round midrib, blade in two pieces and worn at Thermi, Lesbos. From room 4 in T; hilt and fragment of blade only. L. o.34m., uncharacteristic havedges; found according to information supplied at Plovdiv with a spear and a skeleton in a tomb (pl. ing one rivet-hole near end of grip, but no other rivet-holes and in all other respects a C ii type, rather 22:7-8, pl. 28:67). broad midrib, analysis of the metal gave a high Doktor-Iosifovo, Michaelograd. L. o.567m., worn condition, particularly top of grip and blade, which is tin content of i6.o tin to 83.0 copper, found with reduced to the midrib, one surviving rivet-hole in LH IIIA imported vases and dated by the excavator grip, two in blade, horns solid, midrib round, grip "just after 1400." See W. Lamb, Thermi (I936) pl. very like Perushtitsa (pl. 22:9, pl. 28:66). I am 32.63 and pl. XLVII, also BSA 31 (1933) 161-62. most indebted to Prof. Mikov at Sofia and Dr. P. Detev at Plovdiv for assistance with these Bulgarian Palestine swords. Gezer, Tomb XXX, broken and in poor condition. Fragment surviving today (in the Classical MuC ii seum Istanbul) is L. o.235m., the top of the grip is lost, there are no rivet-holes but Macalister shows Crete one in the part of the grip now lost, the shoulders are flanged not "folded," and the narrow midrib "Acropolis Tomb" S.W. of Aqueduct, Knossos. Ashmolean No. AE 489. Incomplete, L. o.694m., evihas decorative ribbing; found in 1909 with a number of bronzes including a scimitar or "harpe" dently from a grave, found as a result of kilnand a Cypriot type flesh-hook, also Cypriot pots digging. Entirely typical with a not high rounded midrib; poor condition, the tip missing. Found covering a wide range of dates but concentrated on with three spears, one exceptionally fine and very the i4th century; the tomb was used over a long long, also a fragment from the horn of what apperiod. Macalister, The Excavation of Gezer (1912) pears to be a second horned sword, and a squat iii, pls. lxxiv-v, and see Schaeffer, Stratigraphie alabastron dated by Evans LM I but probably LM Comparde (1948) 197-98 and fig. I58 (pl. 23:16). II; see Hood, BSA 51 (i956) p. 83, n. 5, and BSA 47 (I952) P. 245, n. 5, P- 54; A. Evans, Palace of Switzerland? Minos (1921-25) IV, ii, fig. 832, p. 85I. (pl. 23:I2Adliswil, Ziirich. A very doubtful sword published in a bad drawing by J. Naue, Vorrdmische Schwerter 15). Chersonesus, Giamalakis Collection No. 500. Hilt end (1903) pl. v, 2, and p. io; Heierli, Urgeschichte der and lower part of blade missing, existing L. 0.332m. Schweiz, 269, fig. 280. Prof. Vogt of the Swiss NaNo further information. Zenakis, KChr (1950) tional Museum, Ziirich, has kindly informed me III, pl. 3. that the Museum has no records of this sword its archives so that it is best ignored. amongst Mainland Greece Galaxidi, Nr. Itea on the Gulf of Corinth. Copenhagen Museum No. 3249; L. o.682m., good condition, very fine rilling on midrib. Bought in Athens, see J. Undset, ZfE (1890o) 15, fig. 23, P. J. Riis, Fortidens Kultur I, p. 27 (pl. 23:18). "Mount Olympos," British Museum I930, 12/15/o10. L. o.6i8m. Finely grooved midrib, some staining on grip from grip-plates. Bought (pl. 23:17). Grevena (in Naturhistorische Museum Vienna). L.

Di Crete Zapher Papoura 36, "Chieftain's Grave" (see Class C i). L. o.6Im., rounded shoulders, agate pommel, grip with gold plates and gold-plated rivets, three in grip, two in blade; pommel-tang extension with small rivet-hole. Originally there was a wood back-

1963]

LATER AEGEAN

BRONZE SWORDS

ing between the bronze grip and the gold of the hilt-plates;repouss6and incised decorationof lions and wild goats in a rocky landscapeon the gold hilt-plates,a double-thongmoulding round the base of the grip shown in the gold and the bronze, a well-markedmidrib decoratedwith fine spiralsand double-spiralson the flanges, fragments of linen adhere to the blade. Found in the lower part of the tomb directlyassociatedwith the C i sword and seals. PTK p. 51, figs. 53, 59, 109, 112, see also Evans P. of M. IV ii p. 853, and BSA 47 (1952) 265-67. Zapher Papoura 42, Shaft-Grave.L. o.62m. (Evans gives o.585m.which must be wrong) roundedshoulders, pommel-tangextension with small rivet-hole, one small rivet in grip, none in blade, rivets goldplated, no surviving hilt-plates, but gold collar, small spirals on flanges and midrib. Single burial with two one-edgedrazors and a hone. PTK p. 60. Zapher Papoura43, pit-cave.L. o.5om., no pommeltang surviving, rounded shoulders, no rivets?, no decoration surviving but fine fluting on midrib; burial disturbed by collapsed roof, sword found near body and one-edged razor, one-edged knife found higher up. PTK p. 62. Zapher Papoura 55, Pit-cave. Present length without tip o.655m., AE 481 (the measurementsgiven in PTK do not seem to tally with the present length or those shown in fig. io9), rounded shoulder, fragments of ivory hilt-platessurvive, pommel-tang extension with rivet-hole, small rivet in grip, two small holes in blade, fine ribbing of midrib, double thong-mouldingon flange.Found with single body, a spear,one-edgedknife, boar'stusks cut for mounting as a helmet, and a stirrupjar, LM IIIA I, PTK p. 66. See BSA 53-54 (1958-59) 245 for a similar

147

a painted coffin; large number of bronzes: spears, including one very long one, knives, daggers, twoedged razor, arrows, fragments of wire possibly from a helmet, a gold cup, gems (seal-stones) and a lamp, probably LM IB-LM II? Hood BSA 51 (I956) p. 83 n. 2, p. 95, 97, pl. 14,e, fig. 3, 5. New Hospital Site, Knossos, Tomb II, Shaft-Grave. L. o.6im. (with pommel restored) angular shoulder, ivory pommel (pommel-tang lost?) grip with three large rivet-holes for wooden pegs with goldcapped decorative studs, two small gold-capped bronze rivets in blade, wooden hilt-plates with gold sheet covering decorated with repouss6 spirals, also on the gold casing of top of midrib; gold thongmoulding at base of grip and gold ring and "cup" for pommel, sword much decayed, perhaps from contact with body. Found with exceptionally long decorated spear, a jug and alabastron, date LM III A i. BSA 47 (I952) pp. 265-67, fig- I5, a; pls. 50, b, 53, a. New Hospital, Knossos, V, Chamber-tomb. L. 0o.47m., angular shoulder, pommel-tang with small rivethole, one small rivet in grip and two in blade, broad flat midrib with grooves; found with very long spearhead, bronze helmet, lead disc, stone bowl and three alabastra, date probably LM II. BSA 47 (1952) p. 275, pl. 54,e, fig. 12, v, 6. Mavrospelio XVIII, Knossos, Chamber-tomb. L. 0.475m., very poor condition and now incomplete, angular shoulder, pommel-tang broken, no rivets in grip, two in blade, probably had broad flattish midrib but now very corroded, found with an E ii knife, a two-edged razor, a spear and small "cutter." BSA 28 (1926-27) 282 (pl. 24:22). Giammalakis Collection 356, no provenance, tip missing. L. o.365m., rounded shoulders, pommel-tang. Zenakis KChr (1950) 109, pl. 3Giammalakis Collection 437, incomplete. L. 0.415m., no pommel-tang surviving, moderately rounded shoulder, KChr (1950) 110, pl. 3Herakleion, no provenance, No. 409. L. 0.212m., tip missing but taper suggests never much more than a dagger, no pommel-tang, three large rivet-holes in grip, two small rivets in place in blade, broad flattish midrib.

stirrup-vasefrom Gypsades (pl. 24:19). Zapher Papoura 98, Chamber Tomb. L. o.6im. (Evans) slightly rounded shoulder, no pommeltang, some remainsof wood grip; ratherbroad,not so high midrib; two burials, the later in a larnax disturbed, the earlier had the sword, a "charcoal holder"or lamp, a stone vase and a one-edgedrazor. PTK p. 98 and pl. io9 (where it is numbered 97 a). Ayios loannis, Knossos,Shaft-GraveA.J. i. L. o.4om., Mainland angularshoulder,pommel-tangextensionwith small Greek rivet-hole,three large rivet-holesin grip, two small Dendra "King's Tholos" (see under C i). L. 0.765m., in blade; rather broad, ribbed midrib. Found with rounded shoulder, agate pommel, pommel-tang spear with short tang and slots "boar spear,"and probably present, grip with three rivet-holes and another with thong mouldings, and small spear gold-capped rivets, perhaps wooden pegs and bronze and two-edged razor. Hood, BSA 47 (1952) 261, studs with gold capping like New Hospital II? Wooden hilt-plates covered with gold sheet with fig. 8. spiral patterns incised on it, thong-moulding at base Ayios Ioannis, Knossos Chamber-tomb(A.J. 2). L. of grip; oval hilt opening and two gold-capped o.62m., angular shoulder, pommel-tangwith small rivets in blade, strong rounded midrib. Sword no. rivet-hole,three large rivets in grip, two small in blade; staining shows oval hilt opening, fragments ii found by "King's" body with other swords, no. of ivory pommel and of wooden hilt-plates also 12 etc. Persson Royal Tombs p. 35, pls. xx, I, no. n; XXII, I. found; blade in three pieces and tip missing; thick ratherbroad midrib with grooves;found in part of Dendra "King's Tholos." L. o.715m., tip missing, a destroyed tomb probably with single burial in moderately rounded shoulder, pommel-tang with

148

N. K. SANDARS

rivet-hole, three grip and two gold-capped blade rivets, ring of granulation rounded edges of gold sheet covering for grip-plates, oval hilt opening, strong rounded midrib, found beside the "King's" body, sword no. 9, Persson, Royal Tombs p. 34, pl. xx no. iv. Date LH III A i. Prosymna, Chamber Tomb XXV. L. o.445m., uncharacteristic sword, rounded shoulder, pommel-tang, no rivets in grip, two in blade, no midrib. From the cist in the north chamber of this uniquely complicated tomb; found with an E ii knife. Blegen, Prosymna 86-92, fig. 198, upper. Mycenae, Chamber Tomb 78. L. o.66m. (Tsountas) angular shoulder, grip broken and in poor preservation, pommel-tang, three large rivet-holes in grip and one large gold-capped bronze rivet, two small rivet-holes in blade; high faceted midrib. Found with quantities of gold nails, perhaps from the grip (see Dendra C i sword etc.), also an uncharacteristic class A sword, with handsome hilt, and two one-edged knives (class ia and class 2) bronze cup with wishbone handle etc. Nat. Mus. no. 3084, ArchEph (1897) pl. 8, i (pl. 24:21). Mycenae, Chamber Tomb 91. L. 0.502m., two similar swords, uncharacteristic, rounded shoulder, pommel-tang, no grip rivets, two in blade, no midrib, but grooved lines down centre of blade, unpublished? Most like Prosymna XXV. Argolis (perhaps from a Chamber Tomb at Dendra). Incomplete, pommel end of hilt and lower part of blade missing, surviving length o.387m., one rivet in surviving part of grip, flanges gold-covered, shoulder angular, two small gold-capped rivets in blade. Very probably from the recently plundered tomb 12 at Dendra, along with the C i sword (see above) dated LH IIB-IIIA I. Ars Antiqua AG Auktion III, p. 30 No. 70, and coloured frontispiece, see also JHS (I96I) Archaeological Reports I96O-6I, pp. 9-Io. Grevena, incomplete. L. o.418m., shoulders angular (if drawings accurate), hilt broken at first rivethole, two blade rivets, and strong midrib, NHM Vienna, Man (1923) 172, fig. 2 bottom, Heurtley, Pre. Mac. fig. 104, ee. Islands Ialysos, Rhodes, N.T. XLV. L. o.37m., angular shoulder, no pommel-tang, no grip rivets, pronounced broad midrib; found with two-edged razor and spear, Annuario 6-7 (1926) p. 199, no. 3601, fig. 124. Ialysos Old Tomb, no tomb number. L. ca. o.42m., moderately rounded shoulder, pommel spur with rivet at break, one rivet in grip, two in blade, broad not high midrib, top missing; Furtwlingler-L6schke, Mykenische Vasen (F. and L.) pl. D, 13. Rhodes or other Greek Islands. B.M. Walter's No. 3, L. o.52m., moderately rounded shoulder, pommeltang broken, no grip rivets, two blade rivets, narrow rounded midrib, tip missing. Conceivably the same as F. and L., D, 13, but not very likely (pl. 24:20).

[AJA 67

Eleona, Cos, probably from a tomb. L. o.418m., angular shoulder, pommel spur with rivet-hole, no grip rivets, two in blade, broadish low midrib. Unpub lished. France? Rh6ne at Lyons, fragment of grip and blade. L. 0.25m., angular shoulder, top of grip missing, two bladerivets, broad midrib, Chantre, "Jtudes Paleoethnologiques . .. Age du Bronze" Album (1875) pl. xv bis, 3; appears to be a characteristic D sword but provenance suspect? (pl. 24:23). Swords related to Class D i Crete Knossos, North House, in a hoard of bronzes. L. o.37m., top of hilt broken, incipient flange to grip, bulging rounded shoulder, possibly antecedent to class D i, P. of M., II, ii, pp. 627-28, fig. 392, no. i7; IV, ii, p. 851, fig. 835. MM III or LM IA. Gournia, F. 14. L. o.353m., rounded flanged shoulder, grip broken, has one rivet-hole, three blade rivets, broad low midrib. LM I? B. Hawes, Gournia (90oI04) pl. Iv, 50. Greek Mainland Eleusis cist-grave type "gamma," H. 10. L. ca. 0.34m. Round shoulder with slight bulge, pommel-tang with rivet hole, no grip rivets, blade has large rivets in a row, no midrib, ILN (Nov. 13, 1954) 840-43, fig- 7; Praktika (I954) 55-57, fig5. Class B and D i Type Hilts, Unattached Crete, Knossos Throne Room. Shell and crystal, P. of M., IV, ii, 931-33, fig. 904; PTK p. 10, n. D. Greek Mainland, Mycenae, Chamber-Tomb 81. Agate, unfinished, hollows sunk for rivets but not bored, gold granulation round edges, thong-moulding; with Class C i sword etc. ArchEph (1897) pl. 8, no. 5. Mycenae, Chamber-Tomb o02. Faience class D, and gold class B, perhaps associated with gold beads etc. and "palace style" pots. ArchEph (1897) pl. 8, 6; JHS 24 (1904) 322, pl. XIi. D ii Crete Palaikastro, The Beehive-Tomb. L. o.33m., very poor condition; broke up during cleaning, one rivet in pommel, one in grip and in blade, parallel-sided blade with sharp taper toward the point. Found probably with a knife and razor and the pots. Date LM III A 2, or possibly earlier. Furumark, Chronology 95, BSA Suppl. (I923) pl. xxv, I, p. II7; BSA (I901-02)

vIII, p. 303.

Gournes Pedhiada Tomb 2. L. o.37m., fragments of

1963]

LATER AEGEAN

ivory of hilt-plates still survived, blade with sharp taper to point; in tomb but bronzes found outside larnax, pots LM IIIB(?) Archaiologikon Deltion 4 (1918) p. 73 fig. I8 No.2 (photograph not at all clear). A sword or dagger which I have not seen reproduced is referred to in Deltion as being "like" the Palaikastro dirk, from "a late chamber tomb" near Canea. See reference under Gournes above.

Mainland Mycenae, from the Acropolis, from above the III Shaft-Grave, found with many "Hera idols." L. ca. o.33m., blade badly corroded splitting laterally, three rivets, H. Schliemann, Mykenae (1878) pp. 90o-91 N. 238. Mycenae, in large Acropolis hoard of 189o (see under class F infra). Two swords are referred to by Tsountas, ArchEph (1891) 25, as being "like" Schliemann's N. 238 (see above) and their measurements given as o.4om. and o.345m., a third mentioned as measuring o.37m. is evidently the class F sword illustrated ArchEph (1897) pl. 8, 4, for which see infra. Mycenae, "porous wall" small founder's hoard, No. 52. 409. L. o.29m., two breaks, two rivets in grip. LH IIIB or later. Wace, BSA 48 (1953) pl. 2, d; F. Stubbings, BSA 49 (I954) 292 (pl. 24:27). Corinth, MF 1271, a fragment. L. o.Iim., as hilt-end missing it could be D i but other proportions, width etc., identical with Mycenae No. 52.409 above. Corinth XII, pl. 91, American School of Classical Studies at Athens, 1523 (pl. 24:28).

Islands Carpathos, BM No. 46, from a tomb. L o.3Im., some ivory or bone of grip survives, two rivet-holes in grip, blade with two breaks, sharp taper to point, very fine decorative grooving on flanges and four parallel lines down centre of blade, better state of preservation, but slender dimensions like Mycenae founder's hoard and Corinth fragment. Found with pots of LM IIIB; JHS 8 (1887) 449, pl. 83, 3 (pl. 24:26). Rhodes, Ialysos, O.T. IV (B. M. Walters 2). L. o.52m., two rivets in grip, two in blade on same level with each other, staining visible from oval-shaped hiltopening, one moulding in bronze flange below pommel and two at base of grip, very broad lobed shoulder, one fine groove down flange and four fine ribs down centre of blade. A very rich tomb, the other bronzes including a long very narrow spear like some in the Knossos cemeteries (L. o.315) also bronze bowl fragments, gold sheet with sphinx motif, glass beads "oenochae" type, perhaps LH IIIB, Furtuw'ngler and Laschke, pl. D, 11, pp. 8 and 75. See also pl. A, I8, and pl. D, 6,8,I2,I6,I7 (pl. 24:

BRONZE SWORDS

149

or B sword, class E ii dagger and one-edged razor. Pots dated LH III A 2 include one (no. 2927, p. 97) of specifically Cretan type, Maiuri, Annuario 6-7, p. 93, fig. I5, No. I8, see also p. 97, pot no. 2927.

E i (Daggers) Crete Knossos, Ayios Ioannis, Chamber Tomb (gold-cup) see D i above. A.J. 6, L. o.355m., pommel spur with small rivet-hole, one large rivet-hole in grip and two in blade, found with other bronzes, gold cup etc. and D i sword. LM II; Hood, BSA 51 (1956) fig. 3, 6; pl. 15, a. Ayios Ioannis, Chamber Tomb, ditto. A.J. 7, L. o.2o5m., one small rivet-hole near base of grip. LM IB or II? BSA 51 (1956) fig. 3, 7; pl. I5, a. Giammalakis Collection No. 355. L. o.324m., two large rivet-holes in grip, two in blade on a level, pommel spur with small rivet-hole, blade narrower than usual (due to whetting down?), profile of shoulder squarer than usual in this class. Zenakis, KChr (1950) pl. 3, P. 109. Giammalakis Collection No. 357. L. 0.3I5m., pommel spur appears broken, one rivet hole in blade (looks very like A.J. 7 above), KChr (1950) pl. 3, P. iio.

Greek Mainland Prosymna III, Chamber Tomb, damaged end of grip and blade. L. o.266m., one rivet in grip, no pommel spur (very like A.J. 7 above), found in a great pile, presumably floor sweepings, including LH II-III pots, Blegen gives the date LH III but see Hood, BSA 51 (1956) p. 96, n. i, on possibility of a slightly earlier dating. Blegen, Prosymna, fig. 462, pp. 18384Prosymna, XLIII, Chamber Tomb. L. o.272m., one rivet in grip, one in blade, no pommel spur, the bronzes which were heaped together include twoedged razors, and a bronze dish. Pottery all LH III. Prosymna, fig. 485, p. i85. Galaxidi (Ashmolean AE 65 1895). L. 0.265m., one side badly damaged and mostly missing, one rivet hole in grip, one in blade (unpublished). "From near Patras" 1891 (Ashmolean Museum 1927. 1375). L. o.243m., two rivets, most of grip lost, blade much narrowed by whetting. "From near Olympia" I898 (Ashmolean Museum 1927. 1376). L. o.282m., pommel end of hilt slightly damaged, otherwise fair condition, one rivet hole in blade (pl. 25:29). Note: Where the end of the grip is missing, as in the dagger from "near Patras," the missing portion could have had a pommel flange, in which case it would belong to Class E ii.

24-25).

Ialysos, N.T. IV. L. o.6om., one rivet-hole in grip, one in pommel and two on a level in blade, three finely grooved lines down centre of blade (see type C i), rich tomb with two burials, a very long C i

E ii Crete Mavrospelio XVIII, Chamber Tomb (see also under

150

N. K. SANDARS

D i). L. o.275m., one rivet in pommel, one in blade, sides of blade much whetted down. BSA 28 (1926-27) 282; Herakleion Museum No. 2142 (pl. 25:30).

Greek Mainland Dendra, Chamber Tomb 2, "The Cenotaph." L. 0.40m., parts of wooden hilt-plates survived, three rivets one each in pommel, grip, and blade. Found in a pit under the entrance to the tomb with a large quantity of bronzes including table and kitchen ware, spear, knives and razors; this collection may be the sweepings from several burials, so dating and associations are uncertain. Some pottery from the chamber has been identified as LH IIIB. Persson, Royal Tombs at Dendra, p. 97, no. 23 pl. xxxIII, 4; and pl. xxx which shows the 35 objects from the pit, see also Furumark Chron. 64. Dendra Chamber Tomb 7. L. o.38m., fragments of the wooden handle survive also ancient repairs, one rivet in pommel, two each in grip and in blade, shape of pommel rather flat. From shaft V; this tomb had five burials, the shaft was full of bronzes crushed together as though swept from several burials, it held bronze mirror, bowls with wishbone handles, one-edged razors, the knife with flanged pommel is very like that from Gypsades T. I, which is probably LM IIIA 1/2 (BSA 53-54 P. 232, fig. 32, 5). Nothing in Dendra Ch. T. 7 is earlier than LH III. Persson, New Tombs at Dendra (1942) 34-35, fig. 25, I. Prosymna XXV, Chamber Tomb. L. 0.31IIm., one rivet in pommel, one in grip, four in a line in blade, shoulder only moderately rounded, there is a moulding in the grip. Found in a cist in the north chamber with a D i sword and two beads, date LH IIIA 2? Blegen, Prosymna fig. 198. Prosymna XLII, Chamber Tomb. L. o.299m., one rivet each in pommel, grip, and blade, rather narrow pommel, shoulders squarer than that in T. XXV, decorative grooves on flange, date LH IIIA? Blegen, Prosymna fig. 377. Prosymna XLIII, very damaged, blade shown in fig. 487, i. In Prosymna described by Blegen as probably same type as the two above, the blade looks very thin and perhaps more like the two-edged razors with tang? Eleusis, cist-graves. L. o.3Im., one rivet in pommel, two in grip, five in a line in blade, four rivets survive and three have large heads, all have narrow shanks, from the same cemetery as the D i knife, but not illustrated, Praktika (i954) and ILN (i955) 67. Thebes, Ismenion Tomb 5. L. o.305m., one rivet in pommel, two in grip and in blade, LH III A2, Delt. 3 (1917) 80-98, fig. 69. Greece (provenance not known, John Evans collection, Ashmolean Museum 1927. 1445). L. o.35m., damaged tip missing, one rivet in pommel and in grip, two in blade, decorative grooves on flange, perhaps the sword referred to by Naue as from Corinth and in

[AJA 67

a privatecollection,J. Naue, VorromischeSchwerter, p. 10, pl. 5, 3 (pl. 25:31).

Islands Cos, Asclepeion (1939). L. o.383m., one rivet in pommel, two in blade; unpublished. Rhodes, Ialysos N.T. IV, see also under class C i and D ii. L. o.38m., one rivet in grip and in pommel, two in blade, very pointed top to pommel, date probably LH IIIA 2, Maiuri, Ann. 6-7, pp. 98-ioo, fig. 15, 20. F Crete Zapher Papoura, Knossos, Chamber Tomb 95. L. 0.37m., two rivets in grip, one in blade, very narrow pommel, blade triangular and pointed, undisturbed tomb with two burials, bronze mirror and stirrup-vase, LM III A2? Evans, PTK p. 83, figs. 94, I14. Zapher Papoura, no context (Ashmolean Museum AE 472). L. o.35m., two rivets in grip, two in blade, there is a moulding in the flange below the pommel. ProcPS (1952) pt. 2 pl. 27, 3 (pl. 25:32). Episkopi Pediadha, Heracleion Museum no. 4326. L. o.345m., one rivet in grip, two in blade, some ivory of haft adhering, pommel semicircular profile, four fine ribs or grooves down centre of blade, Praktika 1952 (1955) 619. Moulianai Tomb A, Heracleion No. 997. L. o.585m., one rivet in pommel, three in grip, fragments of ivory in grip, gold-capped, square-shafted rivets, pommel very narrow, moulding in grip just below pommel, upper part of blade much whetted down, fine decorative grooves or "blood-channels" following profile of blade 5 cm. inside the edge. Very complicated tomb with two period burials, inhumations and cremations. ArchEph (1904) 46; V. d'A. Desborough, Protogeometric Pottery (1952) 269; Catling, PPS 22 (1956) pt. II, p. 113 and Antiquity 35 (196i) 115; Lorimer, Homer and the Monuments (1950) pl. 19, 3. A. Furumark, Chronology 0O6-07 (pl. 25:33). Moulianai, Tomb A. Heracleion Museum No. 998, second sword of type F from the same tomb, much more damaged, top third of grip and pommel missing. Surviving L. 0o.435m., broken at a rivet hole, decorative ribbing down outside of flange, and grooves following profile of blade 8 m. from edge, blade a little thicker in section than no. 997: 5mm. at centre. Other bronzes include a Type II sword, Catling's Group III, and fragment of another, fibulae, pins, spearheads, bronze bowl handles with bull-heads, traces of iron and pots of rather uncertain date (LM IIIB-Geometric?) (pl. 25:34). Dikte, Heracleion no. 326, grip and top of blade only. Surviving length 9.5cm. of which 8cm. is grip, three rivets in grip, one in pommel which is very narrow, moulding below pommel, blade made

1963]

LATER AEGEAN

separately from grip. J. Boardman, The Cretan Collection at Oxford (I96I) p. 17, fig. 3, J. Dikte, Heracleion no. 327, grip only and top of blade surviving. L. Io.2cm. of which 8.2cm. is grip, three rivets in grip, two in blade, pommel very narrow, moulding in grip below pommel, blade made separately, Boardman, op.cit. p. 17, fig. 3, K.

Greek Mainland Mycenae Acropolis, large hoard found in I89o, National Museum no. 2547. L. o.35m., three rivet holes in grip, one in blade, contraction in grip below pommel, according to Tsountas found in a ruined house on the Acropolis along with other swords and daggers (Type II and class G swords, and possibly class D ii for which see above), also double axes, sickles, wedge-shaped tools, flat-axes, spears, arrowheads, horse-bits, tweezers, metal strips and embossed discs and gold wire. Tsountas, ArchEph (1897) Iio, pl. 8,4; Dakaris, ArchEph (1958) 138-39, fig. 8, a (where it figures as type /8 3) Catling ProcPS 22 (1956) pt. II, p. o09 (pl. 25:36-40 and pl. 26:47-48). Mycenae Acropolis, large hoard of 1890, National Museum no. 2548. L. 0.373m. three rivets in grip, one in blade, contraction below pommel, refs. as no. 2547 above, illustrated by Dakaris fig. 8 and called type P i (following Furumark). Mycenae Acropolis, large hoard of 1890, handle only survives. L. 0o.o094m.,for refs. see above. Mycenae Acropolis, staircase hoard. Two swords or daggers, each having three rivets in grip, were found in a hoard with knives, sickles and double axes in the LH III staircase, probably hidden at the final destruction. Mylonas, AJA 66 (1962) 406-08 and pl. 121, fig. 4. Athens Acropolis, hoard. L. 0o.374m., one rivet in pommel, two in grip, one in blade. The hoard has sickles, wedge-shaped tools, etc. like the Mycenae hoard. Montelius, La Grace Preclassique (1924) 155, fig. 498, the proposed association with sherds of LH IIIA 2 is unlikely. Perati, Attica, Chamber Tomb 38. L. 0.403m., two(?) rivets in grip, very narrow pommel, four narrow ribs down centre of blade, found (probably associated) with a small iron knife with bronze rivets and a bronze ring, LH III C lakovides, Praktika (1955) P. 100, pl. 30, P i and pl. 31 3 (knife) see also BCH (1960) 2, p. 661. Dodona. Length not known, one rivet in pommel, two in grip, one in blade, pommel with triangular profile, moulding under pommel, decorative channelling of flange. Montelius, La Grkce Preclassique pl. 13, 2; Dakaris, ArchEph (1958) 141, fig. 9, 15. Kalbaki near Ioannina, cist-grave A. L. o.328m. (restored as 0.352) (see Dakaris ArchEph [1958] 123), rivet holes damaged and reworked in antiquity, one in pommel, one in grip, two in blade; pommel very narrow, flanges of grip and pommel on one side only, three pairs of finely incised lines down centre of blade, signs of old mends, grip much damaged

BRONZE SWORDS

151

and broken. Found with an inhumation, a spear of European type (without slit in socket), small oneedged knife. Dakaris, ArchEph (1958) 123-84, where northern links are discussed (pl. 25:35).

Islands Oros, Aegina, Aegina Museum No. 969/o.15. Fragment of a blade only, with beginning of grip. L. o.1Im., one rivet in blade, one in grip at break, very corroded. G. Walter, Aigina (1938) 25; the dagger is not published (pl. 25:42). Rhodes. Incomplete, L. ca. o.40m., two rivets in grip, one in pommel, contraction in grip below pommel, section of blade shows slight thickening. Montelius, La Grace Preclassique pl. 13, I. Langada, Cos, Tomb 46, tip missing. L. o.41m., two rivets in grip, one in pommel, slight contraction below pommel, two fine decorative ribs along flange, unpublished but see JHS 65 (0945) 102, which refers to Chamber Tombs with dromos of usual type, containing swords and pottery of the "last Mycenaean period" at Langada. Langada Tomb 53, a sword possibly similar to Langada 46. Diakata, Kephallenia, Tomb 2. L. o.40m., tip missing, one rivet in pommel, three in grip, grip contracted below pommel which is moderately narrow, condition poor with two breaks; found with a pot described as "more Protogeometric than Mycenaean"; date 12th century or later, Benton ProcPS 18 (1952) pt. 2, pl. xxvII, I, p. 237; ArchDelt (i919) 118. Diakata Tomb 2, second sword better preserved, tip missing. L. 0.405m., one rivet in pommel, three in grip, pommel very narrow and flat; date and refs. the same as above. Lakkithra (Goekoop excavations) Tomb A 6. L. of grip missing and probably tip of o.4Im., top blade, two rivets survive in grip, two in blade, shoulders too damaged to see their shape clearly, traces of wood remain in the haft, blade flat. A 6 also had a spear which appears to have been deliberately bent, the type cannot be seen from the photograph. ArchEph (I932) Marinatos pp. 1-47. I owe this reference to Dr. Catling.

Sicily Dessueri Tomb 44. Syracuse museum no. 22064, very damaged and incomplete. Present L. o.223m., grip broken but with one large rivet surviving in hole, only one shoulder and one side of grip intact, and slightly flanged; found with arc fibulae, a sword of local type, a two-edged razor, one-edged knife and red "Pantalica" pottery. Date around or just after IIoo, see Maxwell-Hyslop, ProcPS (1956) 126, pl. x. I am indebted to Mrs. Margaret Guido for drawings and photographs of this and the following dagger and to Dr. Brea (pl. 25:41, pl. 28:68). Pantalica North, Grave 48. The only undoubted miniature known to me of this type; tip missing, L. o.o032m., one rivet at top of blade, slight flange on

152

N. K. SANDARS

shoulder, on grip and round the T-shaped pommel

(pl. 25:43, pl. 28:69). England Pelynt, Cornwall, from a barrow in the "Five Barrows" group, a fragment. L. o.1Im., strongly flanged grip and shoulder, rivet-hole in grip; an old find but probably reliable. ProcPS 17 (195i) pt. i, p. 95, pl. ii, also ProcPS 18 (1952) pt. 2, p. 237. I am indebted to Mr. L. V. Grinsell for further information about the circumstances of finding, according to which it reached the Truro museum between 1840 and 1850 along with other objects from the barrows (pl. 25:44). G Crete Zapher Papoura, Knossos, Tomb 14, "The Tomb of the Tripod Hearth." L. 0.42m., domed pommel with one rivet-hole, three in grip, four in blade, two larger above the two smaller, ivory of pommel and grip survives intact, two mouldings below pommel in the ivory and the bronze of the grip, horns rise, then thicken and drop to the point, there is a high double-lobed midrib down the blade's length, the blade tapers sharply to the point. This Chamber Tomb is described as the largest and most important in the cemetery, it held fourteen bronze vessels (cauldron, jugs, etc.), a plaster tripod-stand, fragments of an ivory casket, and bone-inlayed wooden box; and close to the dirk, a spear, razors (one-edged), mirrors, and a one-edged knife. Evans thought that the main burial had been removed soon after the funeral because of a fall of rock; redrawn from PTK, pp. 34-45 figs. 38a and b. Not dated but probably beginning of 14th century (LM IIIA)

(pl. 26:45).

Siteia, Fitzwilliam Museum, Cambridge, no. G.R. Dawkins collection. L. o.5om., tip miss94b-I9o6, ing, pommel rather flat, four rivet holes in grip, flattish blade with three fine ribs, spaced about 5mm. down its length, perhaps produced by chasing after the casting but this is unusual, the horns or quillons begin as horizontal projections and then turn down, the metal is folded forming a "seam." Benton, BSA 29 (1927-28) 114, fig. I, 3; Dakaris, ArchEph (1958) 141 (pl. 26:50). Purchased in 19o6 together with a Type II sword from the same provenancepublished by permission of the Syndics of the Fitzwilliam Museum, Cambridge, from a tracing by Mr. R. V. Nicholls. Mainland Greece Mycenae, Acropolis, large hoard of 1890, National Museum no. 2536. L. o.4Im., three rivets in grip, one (central) moulding below pommel. The drawing in Furumark, Chronology 94, fig. 4, c2, is evidently intended for this weapon, it shows the rivet-hole in-

[AJA 67

correctly placed to one side, redrawn from the correct drawing; see Tsountas, ArchEph (1891) pl. 2, 5, horns very much as in Siteia sword with "seam" visible, strong rounded midrib. For this hoard see under class F above; see also Benton, ProcPS 18 (1952) 2, p. 237; and Montelius, La Gr'ce Pr'classique, pl. 13, Io (pl. 26:46). Mycenae, Acropolis, large hoard of i89o, National Museum no. 2537. L. o.623m. (Tsountas gives 0.65; the sword has been sometimes confused with no. 2536); one rivet in pommel, two in grip, two in blade, grip very narrow, horns rather broad and with "seam," pronounced midrib with very narrow central flattening, tip now missing; for refs. see above (pl. 26:47). Perati, Attica, tomb 12. L. ca. o.6om., three rivets in rather broad grip, two in blade, pommel extremely narrow, remains of bone or ivory in grip and gold ribbon below pommel with small repousse dotted decoration, horns narrow and sharply bent down, they have a "seam" in the metal, one is broken near the tip, the blade is leaf-shaped with the greatest width in the lower half, three fine ribs run down each side of a slight central swelling shown in section. Also in the tomb was a one-edged knife with bird-head on handle, date probably LH IIIC. Iakovides, Praktika (1954) 88-10o3,figs. 5 and 6. Delphi, Temenos T. L. o.38m., part of wooden grip in place, one rivet in pommel, others if present not visible, horns like Siteia sword above, Fouilles de Delphes (90o8) V, pp. 6-io, fig. 19, this refers to nervure mediane but see Benton, ProcPS (1952) 2, p. 237, who has "checked that the blade is flat." The tomb also held a one-edged razor and leaf fibula, Furumark, Chronology 95, gives date as LH IIIC 1-2. Islands "Probably Ithaca," British Museum no. 38. I-10 342, 2753. L. o.43m., dome-shaped pommel, three rivets in grip, grip contracts below pommel then broadens remaining parallel till above the quillons where it contracts again, the horns or quillons have a downward curve and show a "seam," there are two decorative grooves in the pommel-flange and one in the grip-flange, three very fine ribs run the length of the blade at 5mm. apart. Benton, BSA (1927-28) 29, 114, fig. I, no. 2 (pl. 26:49). Gezer, Jerusalem Archaeological Museum, 971 (M76). L. o.i75m. Flanged grip and part of blade of knife or dagger, Middle-Late Bronze Age (pl. 26:51). H Siana, Rhodes, Copenhagen Museum no. 5668. According to the catalogue "found in a Mycenaean tomb near Siana together with a spear and a knife, and bought in Rhodes in 1904 from a dealer." L. o.345m., no rivet holes, pommel projection L. 3cm., flanged grip, horns horizontal with slight downward

1963]

LATER AEGEAN

BRONZE SWORDS

153

and projecting over the top of the blade, there is a inclination, pointed dirk-like blade with three distinct thickening of the blade in section, and four grooved lines down centre, ca. 7mm. apart at top and converging to the point. Found with a knife grooves run down the middle 4-5cm. apart, dated with similarly flanged grip and projection,also a Schaeffer,UgariticaIII (1956) pl. x, pp. 1250-1200, spear, with faceted thickening for socket within 277-78 (pl. 27:58). the blade. I am deeply indebtedto Dr. Marie-Louise Atchana, Hatay, S. Turkey. At 36/4, from above a level II floor, so could be Level I or II. L. ca. 0.50m. Buhl, keeper of the Antiksamlingen,Copenhagen, and to Mrs. Halle Salskov Robertsfor facilities to (given as "half a metre"). There appear to be no rivets, the grip is flanged, the broader curved flanges study and permissionto publish the "Siana"finds at top and bottom are like the Ras Shamra sword, (pl. 27:53-55). Evans Ashmolean Sir there is a wide crescentic pommel which appears to Museum, J. Pergamon, Mysia, be solid metal and complete, the grip flanges end at collection 1927-1385 (registration date 1904). L. the base in an extension, similar to, though less proo.358m., no rivets, pommel projection of 3cm., nounced than, the Ras Shamra one; the blade thickflanged grip, slightly raised horns with seam in ens in section at the centre and has either two metal, pointed blade with three very fine ribs down its length 5mm. apart at top and converging, ridges or four grooves outlining the ribs for the see A. Evans, ScriptaMinoa I (19o09)63, and Przelength of the blade which becomes very pointed and dirk-like at the bottom. L. Woolley, Alalakh worski, Die MetallindustrieAnatoliens (i939) pl. and p. 276, where it is dated beI8, 5. XVIII, v. An element of doubt must exist as (i955) pl. LXX, to provenance.The agreement in the date when tween 1275 and 1187. (Unlike the Ras Shamra the two weapons came on the market is a link sword I have not had the opportunity to handle this with Siana, added to their striking similarity (pl. weapon and the illustration and description leave some uncertainties.) 27:52). Ras Shamra Syria, R.S. 1954, Inv. 18.14, a single Egypt? Fragment of hilt? L. o.o6m. See "Egyptian Bronzes in the Collection of John Evans," Archaefind from the Palace,Court V. L. 0o.46m.,no rivets, flanged grip, the flanges curved and wider at both ologia 53 (1892) 83-94, pl. I, fig. i, incorrectly ends, pommel projectionof 4cm., there is a separate joined to blade of alien type from which it has now metal collar resting on the flanges at the base of been separated, no details (Ashmolean Museum) this projection,the collarhas doubleridge or mould(pl. 27:60). ing, at the base of the grip is a slightly horned extensioncast in one piece with the rest of the grip ST. HUGH'S COLLEGE, OXFORD

SANDARS

PLATE

0

2

OO

NO

3H

0

o

a

11

CMS

CMS2

I: Zapher Papoura 44, Ashmolean Museum and A. Evans. 2: Mycenae Ch. T. 81, National Museum, Athens. 3-4: Perimatos, Ioannina Museum

21

PLATE

22

SANDARS

0 0

o

0

0 WO

0

0

?

0

0:

10

11

'i /6

'1

I

i'9 II1

I -CMS~ CMSIIji

iif \4fl\

5-6: Dolnolevski, Sofia Museum. 7-8: Perushtitsa, Plovdiv. 9: Doktor-Iosifovo, Sofia. io: Kritchim, I: Krasno-Gradiste, Sofia Plovdiv.

SANDARS

12

PLATE

23

1718

"•~12A II

IFI

I,

;1 ") it t

trjg

\111~

~

\

16 it15

CMS

14

'15

Il

12-15: Knossos "Acropolis Tomb," Ashmolean Museum.

17: "Olympos." British Museum.

i6: Gezer T. 30, Istanbul Classical Museum. 18: Galaxidi, Copenhagen National Museum

o

0

O\on o N

.An

> r..,

"I

Q

Q

rD,CA

)

C)

CAC~ Q

21

24

2

25

"

0

c~

O

~1-

C)~

cP-t

O 's~(n

-:

o

00

I ro

CMS i

WW

CIA t31% t3i

•o

-aqSO

VO

34

CL,

n

0E? S-t "• ( E'AO

•

cJMs.

00

o

/

?

rD~ (-D C

D-

-t-'t3?

•:

•'0. i__?• --"

"

C•"

C Co ?

•>.*

.

41

43 ""

•

4 4

r51

\38

26

PLATE

SANDARS

4546o

50

0

47

0

-

0

00 0

i

51

48

45: Zapher Papoura 14, after Evans. 46-48: Mycenae Acropolis hoard, National Museum, Athens. 49: Ithaca? British Museum. 50: Siteia, Fitzwilliam Museum, Cambridge. 51: Gezer, Archaeological Museum, Jerusalem

SANDARS

52

54

53

PLATE

27

57

56

58 il

5

59

60

c i

,CMS

-

L•,.'III 52: Pergamon,AshmoleanMuseum. 53-55: Siana, CopenhagenNational Museum. 56: lalysos O.T. 27, British Museum. 57: Colophon Tomb A, British Museum. 58: Ras Shamra, Damascus. 59: Tarsus, Adana. 6o: Egypt? Ashmolean Museum

PLATE

28

SANDARS

62 68

6

61

63

64

65

67

69

61-62: Perushtitsa. 63: Krasno-Gradiste.64-65: Dolnolevski. 66: Doktor-Iosifovo. 67: Kritchim. 68: DessueriT. 44. 69: PantalicaN. 48. 68-69 not to scale

Sandars 1993 - Later Aegean Bronze Swords

Recommend Documents