Excavation at Assiros Toumba 1986: A preliminary report, Annual of the British School at Athens, 82, 1987, 313-329

Excavations at Assiros Toumba 1986. A Preliminary ReportAuthor(s): K. A. WardleSource: The Annual of the British School at Athens, Vol. 82 (1987), pp. 313-329Published by: British School at AthensStable URL: http://www.jstor.org/stable/30103097 .

Accessed: 29/10/2014 14:35

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 ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

.

British School at Athens is collaborating with JSTOR to digitize, preserve and extend access to The Annual ofthe British School at Athens.

http://www.jstor.org

This content downloaded from 147.188.128.75 on Wed, 29 Oct 2014 14:35:55 PMAll use subject to JSTOR Terms and Conditions

http://www.jstor.org/action/showPublisher?publisherCode=bsa

http://www.jstor.org/stable/30103097?origin=JSTOR-pdf

http://www.jstor.org/page/info/about/policies/terms.jsp


EXCAVATIONS AT ASSIROS TOUMBA 1986. A PRELIMINARY REPORT

(PLATES 50-2)

EX CAVAT IO N was resumed at Assiros Toumba in Central Macedonia after an interval of five years.' In the meantime, study of the finds from the first series of excavations (1975-81) has been completed and these will be published in a Supplementary Volume in the British School Series.

This season was the first of three whose principal aim is the exploration of the uppermost three metres of deposit on the summit of the mound. In this way it should be possible to uncover the major part of the settlement plan for each of the building phases between c. I300 BC and c.750 BC (Phases 9 to I). Examination of the distribution of all kinds of material, including plant and animal remains, in each phase, should help to clarify the function of different areas of the settlement and the changes which take place between each phase. Thanks to the generous support of sponsors,2 to whom I am very grateful, excavation was possible on a larger scale than before, and good progress has been made towards these objectives.

The principal discoveries this year were architectural and are illustrated here by plans of each phase (FIGS. I-8) and photographs of some of the main features (PLATES 50-2). With few exceptions, the range of finds was very similar to that published in the earlier report,3 and will not be described in detail here. I should like to express my gratitude to all who assisted with this season's work on and off site, supervisors, and assistants.

WORK PRIOR TO 1986 By 198I three main areas had been opened on the mound. On the eastern slope a long trench 2 m wide4 had been cut to investigate the structures

which supported the edge of the settlement. The base of the mound had been reached at the edge and the settlement proved to have been defended from its earliest phase with a bank of clay.5 The earliest material recovered dates to the Middle Bronze Age c.2000-I800 BC,6 while higher up the slope massive terrace banks had been repeatedly reconstructed.

In the north-eastern quarter of the mound an area of approximately I8o m2,7 was opened to a depth of 3 to 4 m revealing the substantial mud-brick structures of the later Late Bronze Age, Phases 9-6.8 In the deepest level at the north-east end of the area, parts of three crop storerooms destroyed by fire were uncovered (Phase 9).9 Elsewhere excavation had stopped on the Phase 7 floor levels, c.I I50 BC, to show the small rectangular rooms and parallel streets of this end of the settlement. In 1984, at the request of the Ephor of Antiquities, the northernmost part

1 K. A. Wardle, 'Excavations at Assiros Toumba 1975- 1980', BSA 75 (I98O) 229-60. 'Assiros: a Macedonian Settlement of the Late Bronze and Early Iron Age', Archaia Macedonia 3, 291-305. Archaeological Reports (1975-6) i9; (1977-8) 44-7; (1980-I) 30-2; (1981-2) 36-7; (1982-3) 40-I.

2 Grants for the 1986 season were generously made by the British School at Athens, the British Academy, the Institute for Aegean Prehistory, the National Geographic Society of America, the Society of Antiquaries, the Twenty-Seven

Foundation, the University of Birmingham, and the Univer- sity of Cambridge.

3 BSA 75 (1980) 244-53, 256-61. 4 Trenches JA, JC, and KA. 5 Ibid. figs. 4 and 5. 6 Ibid. 250. 7 Trenches JB-JL. 8 Ibid. plate i9c and d. 9 G. Jones, K. Wardle, P. Halstead, D. Wardle, 'Crop

Storage at Assiros', Scientific American (March 1986) 96-I03.



314 K. A. WARDLE

of the site (chiefly Room I4) was roofed to protect the mud-brick walls from the weather whilst leaving them visible.

To the west, another area of 8o m2,'0 had been dug, first disclosing a range of rooms of the Early Iron Age (Phase 2), secondly to reach the Late Bronze Age destruction level (Phase 6) at a depth of c. 2.5 m below the surface.

In addition, a small trench of some 40 m2,11 had been opened in 1975 to a depth of little more than 50 cms uncovering an Iron Age room (Phase 2, Room I) in yet another burnt destruction level. The only work undertaken in the south end of the mound was a small trial on the very edge and the cleaning of an exposed section lower down.

Since the first phase of work at Assiros was completed, excavation has been started by the University of Thessaloniki at the contemporary site of Toumba, within the limits of Thessaloniki itself.12 The publication of the plant and animal remains and the hand-made pottery from the site of Kastanas on the Axios, have provided a great deal of comparative material.13

THE 1986 EXCAVATION14

Four new trenches 3 m wide were opened in line southwards from the area of the previous excavations to reach the break in the slope at the south of the mound. These were intended to define the extent of preservation of the Iron Age remains and locate the underlying Late Bronze Age walls where possible.15 The Iron Age destruction level was rapidly found at the northern end where parts of buildings could be seen and the two northern trenches were at once widened to 5.5 m. In the south, however, it soon became apparent that both trenches lay within an enormous pit and excavation was carried to a surprising depth of 3 m without reaching its base. Two narrow trenches were cut to east and west of this pit to try and define its limits.'6

Four further trenches totalling 95 m2,v7 were opened to the west of the area which contained the Iron Age destruction. In the northern part of this area stone foundations and a cobbled alley were found immediately below the surface. These represent the only comprehensible surviving structure of Phase I known on the site. A small test was made in the south-west corner of the area at the end of the season to determine the depth of the underlying Iron Age deposits (Phase 3) in which stone walls like those discovered in 1975 and 1980 were found.s8

The small trench on the western edge of the mound was cleaned of ten years growth, which had only moderately affected the mud-brick walls. It was enlarged to 9 m north/south and

Io m east/west to include the whole of the area of Room I and extended westwards to the very edge of the mound to investigate the boundary of the settlement. The Bronze Age walls were reached by the end of the excavation. Work continued in the area in the west central part of the mound in order to explore the mud-brick structures of Phase 7 and to reach their floor levels.

In the north-eastern area'9 excavation was carried out below the Phase 7 floors in Rooms

10 Trench NA/HC. " Trench HB. 12 Directed since 1984 by G. Hourmouziades. Cf. Arch.

Repts. (1984-5) 4I; (1985-6) 58. 13 B. Hansel, 'Ergebnisse der Grabungen bei Kastanas in

Zentral Makedonien 1975-78', JRGZM 26 (1979) 167-202. Prdhistorische Archdologie in Siidosteuropa: 2 Helmut J. Kroll, Kastanas, die Pflanzenfunde; 3 Alix Hochstetter, Kastanas die handgemachte Keramik; 5 Cornelia Becker, Kastanas, die Tier-

knochenfunde. 14 David Smyth, surveyor and Graham Norrie, photo-

grapher, and Diana Wardle have contributed illustrations to this report.

15 From north to south, trenches NB, NC, ND, and SB. 16 Trenches OB, NI. 17 Trenches NE, NF, NG, NH. 18s Cf. BSA 75 (1980) pl. 2oa. 19 Renamed trench JO.



EXCAVATIONS AT ASSIROS TOUMBA 1986 315

6-8 and in Room I I. After about half a metre the expected destruction level was reached. In the largest room, to the south, the destruction level was thick with burnt seeds and fragments of storage containers of different kinds which were cleared gradually and innumerable samples of the plant remains were collected. The whole floor area has been covered with fine nylon mesh (sita), a layer of fine sand, impermeable sheeting, and 30 cm of back-filled earth. Since it was too dangerous to continue with the narrow trench on the eastern slope damaged by earthquake in 1978, spoil from the adjacent area has been used to fill it gradually.

As in previous years, systematic recovery techniques were used for collection of plant remains from all likely deposits. In order to process the larger quantities of earth, a new water sieve using water pumped under pressure was constructed to replace the flotation cell used in the earlier excavations.20

Pottery was washed, sorted, and recorded concurrently with the excavation. Paste-filled incised pottery was removed before washing wherever possible and stored separately as was the Mycenaean. Small objects have been registered and photographed for the record, while most of the vases recovered have already been mended and registered.21

The excavation team, which included a number of students from the University of Thessaloniki, was housed and fed in makeshift rented accommodation in the village of Assiros itself, 3 km from the Toumba. Once more especial thanks are due to the villagers of Assiros who have supported the work in many different ways as well as tolerated with excellent good humour the curious strangers in their midst.22

PHASE I: EARLY IRON AGE C.800-750 BC23 (FIG. I, PLATE 50a, b) Remains of occupation debris and buildings of this phase were found in a strip 5 m

north-south and 15 m east-west, immediately below the modern ground surface. To the south, erosion had removed further evidence of buildings apart from a number of groups of stones in the top of the Phase 2 destruction debris which suggested post settings. Two rough stone foundation walls running east-west were found 6o-Ioo cm apart. Between them was a small cobbled alley which sloped down steeply from east to west. The foundations were built more deeply at the west end to avoid being undermined by rainwater running down the alley. At the eastern end it opened into a larger cobbled area whose eastern limit is still to be discovered. At the western end the alley runs out from between the walls, indicating that a path may continue from that point. A number of stones suggest that structures may lie in this direction but this is beyond the present limit of excavation.

The northern wall proved to join with the curved stone alignments noted in 197724 at its eastern end and at its western it made a right-angled return northwards. To the north of the wall was a more or less level floor surface with a number of pots smashed on it, and an area of hard, fire-reddened clay which may have been a cooking area. The whole length of wall indicates an apsidal building approximately 14 m in length, and, to judge by the curvature of the apse, some 8.5 m in width.

No walls were identified to the north in 1977, but the whole area was strewn with pithos sherds and there were a number of small vessels which could be restored. In one place a hard

20 Water sieving was carried out by a team under the control of Dr Glynis Jones of Sheffield University.

21 The potshed was directed by Diana Wardle, while Sara Parfitt stood no nonsense from either the sherds or the Director.

22 Particular thanks are due to lannis Papadopoulos of Nea Nikomedeia who once again gave up his leave to be our foreman.

23 The dates suggested for each phase are notional, based where possible on Mycenaean pottery etc. as discussed in the first report. The additional evidence from the new excavation, particularly from Phases 8 and 9, of which only small areas had previously been dug, cannot yet be assessed in detail.

24 BSA 75 (1980) pl. 2od.



316 K. A. WARDLE

b

/

/

/

/ /

/

/

A

0 5 0om

ASSIROS TOUMBA 1986 PHASE 1 c.750 B.C.

Stone walls

FIG. I. Plan of Phase I.




burnt clay surface formed the floor and in another there were sherds set in clay. A number of isolated stones were recorded at that time which can now been seen to fit into the reconstructed plan of this building. To the north west, similar areas of broken pithos sherd and fragments of wall indicate the presence of buildings but not much sense can be made of their plan. It is hoped that excavation in 1987 may reveal the northern part of the apse since the depth of deposit over the Phase i floor at this point is about 30 cm.

Four concentric rows of large stones had been discovered in 1977 but only the innermost two align with the foundations to the west. The outer rows are covered in part with cobbling. Thus the building may have been remodelled at this end for some unknown reason.

The southern wall also proved to have a curve at its eastern end, indicated by a few stones, but it had clearly been eroded away beyond this point. At the western end, a short return to the south was observed, which aligned with two groups of stones further south which are presumed to be post settings. The length of this building is a little greater than that to the north but the surviving length of the apse is too short to establish its width accurately.

Lines of stones running southward from the wall indicated internal partitions but these, too, were cut off by erosion. Two pithos bases were found set upright within the presumed area of the building and may belong to this phase rather than to the underlying Phase 2 destruction level. At the eastern end in the approximate centre of the apse was found a horseshoe-shaped structure of mud-brick and clay 30 cm across. This used burnt mud-bricks salvaged from the Phase 2 destruction and rested on the debris of that phase. Within the horseshoe was a similarly shaped 'platter' of clay mixed with chaff which had been fired moderately hard, perhaps by use of the whole as a hearth. When this was removed a small clay bowl was found immediately beneath it.

There was little evidence to show the nature of the upper structure of these two apsidal buildings. The foundation walls seemed too irregular to have carried further stones or courses of mud-brick. Some of the gaps could represent the position of timber uprights against the line of the wall.25 The debris over the floors included numbers of hard fired lumps of clay with impressions of reeds or branches. These may have come from the roof but could also indicate walls built with wattling daubed with clay.

These two large apsidal buildings are placed so close together that provision for drainage was needed down the cobbled alley between them. This suggests that the summit of the mound was crowded with buildings at this period.

Apsidal buildings were relatively common in Middle Bronze Age Greece, but seem to disappear during the Mycenaean period. They reappear in the Late Protogeometric period, as in the case of the 'Heroon' at Lefkandi.26 Part of a small building of this shape was found at Assiros in I980 in the earliest Iron Age level (Phase 3),27 but this was constructed of mud-brick and was only 4 m long. By the late Geometric period large buildings of this type are known at many sites. Some are houses at Lefkandi or Smyrna.28 Others are religious such as the early temples at Eretria, Perachora, and perhaps Antissa.29 The two buildings at Assiros are among the largest with the exception of the Eretria temple.

At Assiros, I feel that the evidence of the pithoi and small vessels suggests domestic rather than any other function, but clearly these would have been the most important buildings in

25 Cf. the 'bat hut' at Eretria, Antike Kunst I7, 70 fig. I, and the Heroon at Lefkandi, Arch. Rept. (1982-3) 12-15, and Antiquity (November I982) 170 fig. 2.

26 Loc. cit. and BSA 77 (1982) 247 n. 47. 27 Arch. Repts. (1980-I) 31. 28 Lefkandi, M. R. Popham and L. H. Sackett, 'Excavations

at Lefkandi Euboea, 1964-66', fig. 68; Smyrna, BSA 53-4 (I958-9)-

29 Eretria, Antike Kunst loc. cit.; Perachora, H. Payne, Perachora I, The Sanctuaries of Hera Akraia and Limenia 27 ft.; Antissa, BSA 32 (1931-2) 42 pl. i8.



3P8 K. A. WARDLE

the settlement at this period. Unfortunately the preservation of bone and plant remains is minimal so close to the surface and these do not assist with the definition of the function of these buildings.

PHASE 2: EARLY IRON AGE C.950-900 BC (FIG. 2, PLATES 50C, d, 5Ia) The well-defined destruction layer of Phase 2, with masses of burnt mud-brick and other

debris has been found to extend over Io m south of the area excavated earlier, but beyond this has clearly been removed by erosion. Only the area to the south of the Phase I wall foundations has so far been excavated.

Four rooms have been discovered in a row running east-west about 8 m south of the rooms uncovered in 1975 and 1977 (Rooms 2-5). Their dimensions, where these can be determined, are similar, c. 6 x 4 m. The walls are of mud-brick which has been disturbed by root action and burrowing insects and are often difficult to define precisely. Floor surfaces do not appear to have been specially prepared and can only be distinguished as the change at the base of the destruction debris. As in the rooms to the north, each room contained several pottery vessels of different kinds.

In the west, Room 12 was very poorly preserved, with only part of its east wall and fragments of the north wall. A number of pots were found against the north and east walls, but the pithos base further out in the room, may date to Phase I. To the east room ii was better protected from erosion, and could be clearly defined on the east and west. To the north the stone footings of Phase I almost exactly coincided with the wall alignment below, while that to the south was missing. Here, too, were a number of vessels and a pithos base.

The best preserved room, Io, had a small fragment of the south wall preserved to confirm its length, while the walls on the other sides were preserved to a height of nearly 20 cm. In addition to the whole pots found in different parts of the room, there were two rows of poorly fired conical loomweights lying together in the northern part of the room (PLATE 50c). Fifteen in all, these must have fallen from an upright loom in use at the time of the destruction. One timber support was noted against the north wall, but it may not be possible, given the conditions of preservation so near the surface, to locate any others.

To the east, only part of the fourth room (9) has been exposed, and this has not yet been completely excavated. It had been assumed in 1977 that the area to the south of the original row of rooms, where the destruction debris seemed comparatively light, was an open yard area.30 This year, however, walls were noticed below the Phase I level, which continued the east and west walls of Room io, suggesting that part, at least, of the area was occupied by structures.

On the western edge of the Toumba, Room I was now completely cleared and proved to have a group of pots in its south-west corner as well as another pot in the north-west. The north wall of the room was partly built of stones, with a doorway 1.2 m wide in it (PLATE

5od). This wall was very close to the trench edge and there were masses of fallen stone to the north. It could possibly be the southern wall of a narrow alley. To the west, outside the room, was a roughly laid stone surface which petered out towards the edge of the Toumba.

Showing through this surface were alignments of stones which appeared to follow the line of the Toumba edge, curving to the east at the north end where several courses were preserved (PLATE 5Ia). Excavation of these below the Phase 2 surface showed that they were the last phase of a rough circuit wall, rebuilt several times in Phase 3. In places these alignments

30 BSA 75 (1980) 253 fig. 15, 'rooms' 6 and 7.




door

1

2 3

8

J5 4

7 6

Not yet excavated

12 11 10 9 /

;F 0

/ N

I

/ /

8

a0 5 1o0

ASSIROS TOUMBA 1986 \ PHASE 2 c.900 B.C. \

Mudbrick walls Stone circuit wall

FIG. 2. Plan of Phase 2.



320 K. A. WARDLE

FIG. 3. Sketch of spindlewhorl.

hardly merit the title of wall and can surely not have been defensive. They may simply have marked the perimeter and been sufficient to prevent children or animals straying over the steep edge of the mound. Similar groups of stones had been found outside the Late Bronze Age buildings on the east of the mound but the 2 m width of that trench prevented further definition.

EARLY IRON AGE, PHASE 2 OR 3 The enormous pit discovered at the southern end of the site (see FIGS. 4-5) cannot yet be

precisely dated, since its edges have not been defined in relation to other strata. The pottery is of the first Iron Age style and there are none of the distinctive features of Phase I. It was certainly cut after the end of the Bronze Age but the fill contains a high proportion of Bronze Age pottery washed in. Parts of the northern and western edges have been discovered, while masses of hard mud-brick towards the base in the south-east indicate its limit in that direction. It must have been approximately oval in shape and measured over 13 m north-south and 9 m east-west. The base has not yet been reached, 3 m from the ground surface, but the presence of many layers of water deposited silt suggest that it cannot be much deeper.

At first it was thought that this might be an open area used for cultivation rather than for buildings, or a hollow formed by the continual use of the same path to give access to the settlement from an early date. However, the silt shows that it must have been an enclosed pit which often contained standing water. It must also have been open for some considerable period before it was allowed or encouraged to fill, presumably before the start of Phase I.

One possible function is that of a 'borrow' pit where clay from the mud-brick debris of earlier phases was dug after spring or autumn rains had softened it. This might then be used for new bricks, or, more probably, the mortar used to bind them. Iron Age mud-brick walls are not well enough preserved to observe the use of 'dirty' mortar, but the practice was clearly common in the bronze age (see p. 323 below).

Some of the most intriguing finds from this year's excavation came from the fill of the pit, but it is not always clear whether they were thrown into it while it was open, or were washed out of the earlier strata at the sides. One biconical clay spindle whorl (FIG. 3, PLATE

5IC) is

incised and white-filled on both surfaces with random patterns as if the makers had seen some form of script but did not understand it. Schliemann found a number of similar spindle whorls at Troy, where they are of Bronze Age date.31 Another unusual find was a large iron double axe (PLATE 51b).

31 H. Schliemann, Troy and its Remains (1875) pl. xxii-lii esp. 357, 375, 426, 492. H. Schmidt, Heinrich Schliemann's Sammlung Trojanischer Altertiimer Taf. i-ix esp. 5220.




PHASE 3. EARLY IRON AGE, C. I050-950 BC (not illustrated) This phase represents a number of building levels in different parts of the site. (It had at

first been thought that it could be separated into two clear phases but this attempt was abandoned and Phase 4 abolished.) The buildings may be of stone, mud-brick, or use post frames but their overall plan is not yet obvious. One common feature of the levels, wherever they occur, is the soft, loose deposit, usually grey in colour, which attracts burrowing foxes and badgers as well as smaller animals. Thus the stratigraphy is very disturbed.

No structures were discovered in the western trench apart from the circuit walls mentioned above, but the fallen mud-brick debris of the Bronze Age buildings had clearly been formed into a well-trodden surface. In the south west, where a test was carried out below the level of the Phase 2 rooms, a length of stone wall was found with roughly the same north-south alignment as elsewhere on the site. To the south of the Phase 2 Room 12, a similar test located dilapidated stone walls and a pithos smashed in situ which belonged to this phase. The same test discovered the edge of the pit, mentioned above, but unfortunately the stratigraphy is not yet clear enough to relate it to these features.

PHASE 5. LATE BRONZE AGE. C.I 100-1050 BC (not illustrated) The latest bronze age structures discovered so far were destroyed by fire. Walls of this phase

have been discovered on every part of this site but no coherent settlement plan can yet be suggested. It is clear, however, that the streets and the alignment of the buildings which had existed for a long period until the Phase 6 destruction, were completely ignored in the construction of the Phase 5 settlement. This may reflect a break in occupation, or a destruction over so large an area that the opportunity was taken to replan the settlement.

On the west of the mound part of one room was excavated this year which reused a Phase 6 wall, had a southern wall of mud-brick built over the Phase 6 street, and a northern wall with a good stone footing on which the mud plastering was still preserved in part. Some of the walls found in the southern part of the mound probably date to this phase but this cannot yet be proved.

PHASE 6. LATE BRONZE AGE. C.I 150-1 I oo BC (FIG. 4, PLATE 52a) So far it is this phase which has provided the most coherent settlement plan. The mud-brick

buildings of the preceding Phase (7) were rebuilt with only a small number of minor changes. The street system remained unaltered. It is not clear how long the phase lasted before the settlement was destroyed by fire. Numbers of pottery vessels were found shattered and burnt on the floors, while the fire itself helped to preserve several large storage containers, of wicker, or of clay mixed with chaff. This season three more wicker baskets were found, each in a different room (Rooms 23 (PLATE 52a), 27, and 32), while another smaller container was found in Room I I, when the corner of the baulk was removed there. So far the pattern of agricultural storage in this phase is of scattered pithoi and clay containers, with no more, in general, than one or two to each room. Some are clearly in areas regarded as interiors (Room I) while the larger group in Room II appears to be in a open area, perhaps partly sheltered against the north wall.

At the southern end of the site several walls were located which seem to have the same

alignment as those of Phases 6 and 7, and one clearly had burning on its face. This runs west-east just to the north of the great pit. Other walls appear to adjoin it on its south side, suggesting that if there was a street here it ran to the north of the wall. In this case there



322 K. A. WARDLE

171 16

14

12 13!

11 21.

20 10

y; 9 1 8 7'

6 15 5

3 4 "IC lil 236 24-

..

- 72y

c

28.

25 22

22,"

7,

/

/ I I Early Iron Age Pit

I I

/

N .

I

0 5 fOm L I 1


Posts * Yards Pithoi O Clay bins 0 Mudbrick walls





may well be larger rooms in the area to the north than any discovered before in this phase since the gap between this (hypothetical) street and the one to the north (22, 2 and 4) is 13 m, while the next block as far as (29 and Io) is only 8 m. Only one wall has so far been located on the west side of the area and this could perhaps mark the western limit of the buildings.

PHASE 7. LATE BRONZE AGE. C.I200-I 150 BC (FIG. 5, PLATE 52b) Most of the excavation in the west central part of the mound this season was concerned

with the mud-brick fill of the Phase 7 rooms in this area. Progress was slower than hoped because of the Phase 6 baskets mentioned above which had to be excavated with great care (one has in fact been left for next season), and the numbers of pits (? of Phase 5 date) which had disturbed the area.

It is clear that this phase has quite a long duration since a sequence of internal features was found in Room 28. Here two built cooking hearths were found, against the south and east walls. Their bases were made of small quartzite pebbles, covered over with a layer of clay, which was repeatedly replastered. The wall behind each was reinforced with stones and mud-bricks. The room was later converted to storage, with three pithoi added at different times. One was set into the centre of the eastern cooking hearth, while a mud-brick bench appeared to have been built against the north wall when the last pithos was added, perhaps to help support it. Another hearth was found in an adjacent room (24). In Room 27, traces were found of a pit cut in the floor, which had been lined with wicker-work and was presumably for storage.

The cause of the rebuilding and raising of the floors to the higher level of Phase 6 is not clear, unless damage had been caused by one of the earthquakes which regularly affect the Langadas basin. The Phase 7 floors were not easy to detect and did not seem to have been specially prepared. They were most often observed as the change from relatively clean mud-brick fill, within the walls, to the greyish, charcoal flecked levels of Phase 8 underlying them in this area. Several irregular pits were cut into this stratum from the level of the Phase 7 floors. The mortar used in the Phase 7 walls on either side was grey with charcoal flecks, just like the Phase 8 level below them, whereas in most cases the Late Bronze Age walls are built with fresh, clean clay, yellow or reddish in colour. The simplest explanation for the nature of the mortar and the irregular pits, is that the builders were too idle to go further than the area around their feet for fresh mortar, and created large 'borrow' pits which were then filled with any available refuse lying around, including heaps of pithos sherds.

Only two of the walls in this part of the settlement can be shown to rest on walls of Phase 8 and it seems likely that this area was used in a different way in the preceding phase. Walls in the long trench in the south-east of the Toumba probably belong to Phase 7.

PHASE 8. LATE BRONZE AGE? C.I 250-I 200 BC (FIG. 6, PLATE 51d)

Only a limited amount of work was carried out in the levels of this phase, particularly in the north-east part of the mound, below Rooms 6-8, and I I, of Phase 7. Even so, it is apparent that some major changes of plan took place between the collapse of the Phase 8 buildings, and those of Phase 7 which followed them. Room 932 is much larger than those which succeed

32 For simplicity the room numbers of Phase 6 have been kept for the rooms of earlier phases where these have basically the same wall alignments. Where an earlier room was replaced by several smaller ones, only one of the room numbers from the upper level will be used and the others will be ignored.

Thus the room below Rooms 6-9 of Phase 7 will be identified as Room 9 in Phase 8. All finds are assigned to phase and room with a five digit code in the study records, e.g. 00708 refers to Phase 7, Room 8, etc.



324 K. A. WARDLE

16

14

'

ps 12

11

20

F21 10

9i 8 7

6 15

5;

4= 2

3 I1l

23

7321\

22

:24

27 29

28

25

/

10

\ I Early Iro Age Pt

/

/

/

e

0 5 fOm


Posts -a Pithoi 0 Yards Mudbrick walls

FIG. 5- Plan of Phase 7.




ASSIROS TOUMBA 1986 PHASE 8 c.1250B.C. 14

12

11

10

9


it, measuring 8.5 x 4 m. The other rooms dug to this level so far are I I, 12, and 14 while io has not yet been excavated.

The fill attributable to Phase 8 in Room 9 was fairly clean, cut only by the post pits of the Phase 7 walls built over it. A slight trace of a floor level was found only just below the Phase 7 walls, which would suggest either a rebuild during the phase, or a surprising depth of underfloor fill between walls which are the best constructed we have found at Assiros. In Room I I, a sequence of floors in relatively clean fill suggests that it may not have been a roofed area even if not a typical yard. A series of superimposed hearth bowls, found against the north wall, were not as elaborate as those found in Phase 7 Room 28, but in one case broken sherds were used in the base.

Among the finds from this area was a small fine Mycenaean piriform jar shattered and scattered but restorable apart from its missing foot (PLATE 5 Id). The decoration of neat zigzags is typical of LH IIIB. The stemmed form is a feature of other examples known from northern Greece.33 Its fabric is Macedonian, but not that shown to be from the Langadas basin. At Assiros, Mycenaean can be divided visually into three categories: local-from the Langadas area; provincial-? from coastal Macedonia; imported-from further south.34

In Room 9 the Phase 8 walls were an exact rebuild of those of Phase 9, except that the doorway in the north wall was now at the east end. The burnt Phase 9 walls were carefully

33 Cf. Stemmed piriform jars from Saratse and Kastanas. W. A. Heurtley, Prehistoric Macedonia 223 no. 448, pl. xix. JRGZM 26 (1979) 186, Abb. 14: 4 and 5.

34 The local fabric is very distinctive and its manufacture

from local clay has been confirmed by spectrographic analysis carried out by Dr Jones at the Fitch Laboratory. R. E. Jones, Greek and Cypriot Pottery, (1986) 108-12.



326 K. A. WARDLE

ASSIROS TOUMBA 1986 PHASE 9 c.1250 B.C. 14

S18 12

10

doo i

Adoor 9


rebuilt as had also been observed in Room 14 in I980. In the west wall reused mud-bricks had the burnt faces turned sideways so that they could be seen as reddened lines running the width of the wall. Room II occupied only part of the area of the presumed Phase 9 room below (to be known as Room 12) while Room I12 seems to share only its north wall with the preceding phase.

PHASE 9. LATE BRONZE AGE. C.I 300 BC (FIGs. 7 and 8a, b, PLATE 52C, d) The destruction level which distinguishes this phase, the most marked of all those found at

Assiros, was first discovered in 1977 below the floor of the Phase 7 Room 14.35 Unlike the other burnt destructions, the debris here was rich with charred cereals and other seeds suggesting that it was an agricultural storage area. In later seasons the rest of the room was cleared, together with an area to the south as far as the eastern end of Room 9. All the rooms discovered produced seeds, in greater and greater quantity, as excavation proceeded southwards. By 1981 it was already clear that this area represented a concentration of storage unique on this site and only rarely found in other Aegean prehistoric sites. A pattern of storage could be observed and it was possible, after careful study, to suggest not only what crops were stored in each area (as opposed to being scattered during salvage operations after the destruction) but also how many containers of each crop there had been.36

13 BSA 75 (1980) 239 pl. 19b. 36 Scientific American (March 1986) Ioi. See also G. E. M.

Jones, The Use of Ethnographic and Ecological Models in Interpreta-

tion of Plant Remains (unpublished Ph.D. dissertation for the University of Cambridge 1983) chapter 5.2.1.




Door

Pithos

Clay bin

Basket

Door

FIG. 8a. Plan of storeroom, Phase 9-.

Pit

FIG. 8b. Reconstruction of storeroom, Phase 9.



328 K. A. WARDLE

One of the major aims of the renewed excavation in 1986, was to discover how far this concentration of storerooms spread. It was, however, particularly surprising to discover the size of the storerooms in this phase. Both areas excavated, below the Phase 7 Rooms 6, 7, and 8, and below Room I I, proved to be the continuation of rooms already found.

The southern room (9) was the larger. It measured nearly 4 m north-south, and must have been well over 9 m from west-east. The walls have been discovered on north, west, and south, but that to the east lies outside the area of excavation. Excavation in the narrow trench to the east earlier did not reveal this wall, but it may have been quarried away or collapsed outwards down the slope of the mound. We presume that it was more or less on the line of the eastern wall of the Phases 8 and 7 rooms above. The walls survive only to a height of 30 cm but they seem to have been levelled everywhere to provide firm foundations for the Phase 8 walls built on them. The north wall was cut by a doorway 1.5 m wide at the western end and appears to have had another in the north-east corner.

The roof was supported by a number of irregularly spaced oak posts. Two of these survived as large masses of charcoal projecting above the floor level while others were indicated by empty sockets. It appears that three posts were placed at intervals against the southern wall, three down the centre of the room, and three more close to the northern wall. Not all have been discovered, perhaps because of the Phase 8 post pits which cut down from above. The burnt debris contained large lumps of burnt clay with the impressions of beams and branches, perhaps representing the roof.

The floor itself was packed with storage containers, so close that there can hardly have been room to move around. Along the south wall were preserved three large wicker-lined containers, oval in shape and 1.5 m across. These were set slightly into the floor itself. A fourth could easily have been present in the south-east corner of the room, but this area had been dug away by a later pit, perhaps as a result of building operations during Phase 8. Along the centre of the room, and in the gaps against the south wall, were smaller containers, made of clay mixed with chaff and straw, which were 40-50 cm in diameter. The northern half of the room was less crowded. Here were set at least six pithoi, with, probably, two more in the part of the room excavated previously. Two bases were still in position in the floor while the remainder were indicated by the neat holes from which they had been removed, after the destruction, during salvage operations. In places the pits cut in Phases 8 and 7 had removed parts of the containers, but the damage was relatively light.

In parts of the debris the charred cereals were heaped up to 40 cm in depth. Often they were packed so thickly that there was hardly any earth or other debris to remove. The containers in the southern half of the room seemed to have held einkorn wheat, while the pithoi contained a wider variety of crops in smaller quantities. These included hulled barley, bitter vetch, and broomcorn millet as well as a possible mixture of emmer and spelt. The clay bins against the north wall held einkorn. In places whole spikelets of wheat were found showing that the crop was stored in that form to help prevent pest or fungal infestation.37

The crops proved to have been preserved in three different ways. Some were apparently unburnt, some charred, and others converted to ash in the fire, providing the opportunity to study the phytoliths.

There were also smaller pottery vessels in the room: these included two fine examples of the hand-made burnished globular jars with incised, paste-filled decoration and a number of

31 A much larger quantity of intact spikelets was found in this room than in the other areas dug earlier, confirming the hypothesis put forward earlier on the basis of a few intact

spikelets and the quantities of chaff. Scientific American (March 1986) 1oo.




four-handled burnished jars. One of these may have contained lentils. Of particular interest was a rough clay stopper, of a size to fit one of these smaller jars, which carried the impression of vine leaves on its lower surface. The mouth of the jar it was used on was first sealed with leaves (perhaps first oiled, a method in use in Macedonia until the Second World War) and then plastered over with wet clay to make the seal permanent.

The quantity of cereal stored in this one room, on first estimates, could have been as much as 18 cubic metres. Allowing for the removal of the chaff, since the wheat was presumably all in spikelet form, this would provide some io tons of foodstuff,38 enough to feed ten families for a whole year.39

The room to the north (12) was apparently smaller. Only the west and east walls have been located with certainty. The north wall presumably follows the line of the south wall of Room 14. In the south, it is not yet clear whether there will be a wall under the north wall of Room io, or whether the room will extend the whole width of Rooms II and Io above (Room 10 has not yet been excavated below the Phase 7 floor level). No internal posts have been located so far and the floor was not covered in the same way with pithoi and clay containers. Two small baskets c. 6o cm in diameter had been found set against the north wall at the eastern end in 1979, and a hole in the floor there suggested at least one pithos. Only small quantities of seeds were found on the floor of the new area cleared. So far einkorn, emmer, hulled barley, and broomcorn millet have been identified.

So far we have discovered two large rooms devoted to storage, with a third that was at least partly used in the same way. The total area of the three is 75 m2, with a capacity of 45 m2 of crops stored in pithoi and other containers. This concentrated pattern is very different from that which we have found in the later Bronze Age levels at Assiros and the quantity is unique in prehistoric Greece outside the palaces of the south. Until the total extent of this area of storage has been discovered and its relationship to other buildings on the site as a whole is understood, it will be difficult to assess their significance. It seems likely, however, that we are dealing with an unusually complex social organization at this period. Individual families may have collaborated in building communal storerooms for their crops or a local leader may have collected this quantity under his own control. Either way, the settlement at Assiros and its agricultural storage have already provided exceptional evidence about the nature of a small rural community in late bronze age Greece.

K. A. WARDLE

38 According to J. Percival, The Wheat Plant (Duckworth 1974) grain accounts for 50 per cent of the bulk of wheat in spikelet form. From their field work G. Jones and Paul Halstead estimate that the grains of barley represent 70 per cent of the unthreshed bulk (pers. comm.). The actual total of foodstuff in this storeroom will thus depend on the proportions

of wheat to barley in the room which cannot yet be precisely estimated, and the figure given here is a guide only.

39 C. Clark and M. Haswell estimate in The Economics of Subsistence Agriculture (4th edn., Macmillan 1979) that each adult member of a family needs about 2oo kg. of cereal per annum.



PLATE 50 B.S.A. 82

(a)

(b)

(c)

(d)

EXCAVATIONS

AT

ASSIROS

TOUMBA

1986

(a)

Phase

I: cobbled

alley

with

apsidal

buildings

either

side.

From

east;

(b)

Phase

i: west

end

of apsidal

buildings.

From

west;

(c)

Phase

2: loomweights

from

room

io;

(d)

Phase

2: room

I from

south

Doorway

in north

wall



B.S.A. 82 PLATE 51

(a) (b)

(c) (d)

EXCAVATIONS AT ASSIROS TOUMBA 1986

(a) Phase 2/3: stone circuit wall at west of Toumba. From south; (b) Phase 2/3: iron double axe; (c) Incised spindle whorl from

phase 2/3 pit; (d) Phase 8: Mycenaean piriform jaw, ht. ext. ii..5 cm.



PLATE 52 B.S.A. 82

(a)

(b)

(c)

(d)

EXCAVATIONS

AT

ASSIROS

TOUMBA

1986

(a)

Phase

6, room

23:

burnt

wicker

basket

in floor.

From

south;

(b)

Phase

7, rooms

22

9. From

west;

(c)

Phase

9: store-room

with

wicker

baskets,

clay

bins,

and

pithoi.

From

east;

(d)

Phase

9: store-room

from

west



Excavation at Assiros Toumba 1986: A preliminary report, Annual of the British School at Athens, 82, 1987, 313-329

Documents