British School at Athens rescue excavations on the shores of the Peneios valley reservoir in the spring and summer of 1969: a summary report.

Some fifty years ago, in the mid-1960s, a new reservoir was created in the Peneios valley in the Province of Elis, in the north-western Peloponnese, to improve the irrigation and agriculture of the region. A great long barrier dam was built across the middle reaches of the Peneios valley, and the reservoir would flood the low-lying land behind it. The Greek Archaeological Service organized ground-surveys of the areas to be flooded, and invited the foreign archaeological schools and institutes to assist, allocating sections to each (Fig. 1). Survey teams would search the shores of the rising lake for hitherto unknown sites, and could test-trench those of interest; excavation permits would be granted for further exploration of the more promising ones. The surveys were carried out in the Autumn months of 1967, and excavations might be undertaken in the next year or two. A summary account of the surveys appeared in AAA I (1968) 46-48, ‘Salvage Archaeology in Elis, 1967’ (by H. S. Robinson), and longer accounts of the surveys of individual schools in Arch. Delt. 23 (1968), Chronika B1, 174-194. More detailed accounts of some of the larger-scale excavations appeared in the next few years.1

Map of the Peneios Valley basin, naming villages, showing the Peneios Dam (top left), marking the survey sectors of the ‘foreign schools’, and also the numbered sites noted in 1967 by the BSA survey team (22a = ‘Kostoureika’; 3b = ‘Keramidia’. Inset map: the NW region of the Peloponnese: Pa = Patras; E = ancient Elis; Py = ancient Pylos; O = Olympia.

Fig. 1    Map of the Peneios Valley basin, naming villages, showing the Peneios Dam (top left), marking the survey sectors of the ‘foreign schools’, and also the numbered sites noted in 1967 by the BSA survey team (22a = ‘Kostoureika’; 3b = ‘Keramidia’. Inset map: the NW region of the Peloponnese: Pa = Patras; E = ancient Elis; Py = ancient Pylos; O = Olympia.


The British School at Athens had been allocated a section of the south-western shores of the planned lake, a section which included part of the tributary valley of the Loukia stream and its minor tributary, the Kleisoura, and included seven villages (from west to east) Agios Elias, Kalatha, Keramidia, Daphniotissa, Ephyra, Ano Loukavitsa (or Kambos), and Kato Loukavitsa (now Augi). Here, much of the low-lying ground was, by report, subject to flooding in wet seasons (and the whole province is one of heavy rainfall). Not surprisingly many of the ‘new sites’ were found on minor ridge-tops or rising hillsides. In November 1967 the BSA ground-survey team, led by Roger Howell and assisted by the Hon. Cressida Ridley and Ken Wardle, surveyed this area, and found some 34 sites, several close-set and so listed as nos. 1-23 by Roger Howell in the BSA’s section of the general survey report (Arch. Deltion 23, 1968, Chronika B1, 178-182); they test-trenched at a number of these sites.

In the Easter vacation of 1969 further excavation work was carried out by a BSA team at two of the sites noted in the published survey: ‘Kostoureika’ and ‘Keramidia’. (Fig, 1, sites nos. 22a and 3b). The first site, ‘Kostoureika’, was on a low-ridge-top on the very edge of the, by then, rising waters of the lake, and was so-called because the land belonged to the Kostouras family of Keramidia village (the base for the 1969 excavation teams). Here a 1967 test-trench (4 x 2m. in extent) had located the corner of a stone-built structure and produced Hellenistic sherds; this gave rise to hopes of finding a substantial structure, perhaps ‘a fairly wealthy rustic villa’. The Easter 1969 team consisted of R.A. Tomlinson, P.A. Rahtz, and Roger Howell of Birmingham University, assisted by the student helpers, also of Birmingham2, and also J. Ellis Jones of Bangor, included no doubt in the hope that the Kostoureika structure might provide a ‘Peloponnesian’ parallel’ to the farmhouses, or ‘rustic villas’, which he had assisted in excavating earlier in Attica, the ‘Dema House’ (dug in 1958 and 1960) and the Vari House (1966).3



Excavation at ‘Kostoureika’ sadly disappointed all high hopes of finding the full plan of an Elean parallel to those Attic houses! The 1967 test-trench had located the one and only surviving corner of the stone-built structure – it became clear that all the rest had been ploughed out or robbed away (perhaps to build the modern hut and sheep-fold nearby on the same ridge). Probing around the test-trench with a resistivity meter to decide how to extend the original test-trench gave surprisingly negative results.4 Opening up a wider trench (Kostoureika I KTR I; 15m. EW x 7.50-8.0m NS) proved that the only surviving part of the structure was the right-angled corner formed by the two short lengths of wall uncovered in 1967: all the rest seems to have been ploughed out, and ridge and furrow marks across the ridge was a further hint. Indeed the iron point of a primitive wooden plough,5 found by the broken end of the walling, just outside the limits of the 1967 test-trench suggested that the destruction was not very recent. Furthermore, the nature of the surviving stonework, which incorporated a reused shaped sandstone block and another stone with white plaster stuck to it, also suggested that the building located in 1967 may itself have been, not the original Hellenistic structure, but some later farm building or hovel incorporating reused material salvaged from a much earlier one. The remains represented the foundations, some 0.40m. wide, for, probably, mud-brick walling, too narrow to support safely an upper storey. So the building located in 1967 may just have been an earlier precursor of the abandoned and ruinous ‘modern’ farm hut still standing a little further E on the same ridge.6 However, that there had definitely been some roofed structure of Hellenistic date on this ridge was clear enough, and that one intended for human occupation too, as the quantity and range of domestic pottery collected in 1969 indicated well enough. Some typically glazed ancient roof-tile fragments were also found (the local modern roof-tiles were unglazed). However, pitifully few small finds came to light: an illegible bronze coin, two ancient terracotta loom-weights, four terracotta spools, and a small lead weight (the report of a ‘golden ram’ – part of a necklace perhaps – found by the owner’s grandfather on this site suggested that any more interesting and valuable chance finds turned up by the plough in earlier years would have disappeared long ago).

When the ground-soil over this first trench, Kostoureika I (KTR I) was cleared away, and the immediate surrounds of the two walls excavated to a deeper level, an interesting feature was noted: the subsoil along the outer edges of the corner formed by the two walls had been worn away into a shallow hollow, following the line of the corner, and then veering away slightly at both ends (Fig. 2). This feature was comparable to similar hollows noted that Eastertide at other nearby abandoned farm-buildings around the lakeside, hollows along and partly under the outer edges of their foundations (Fig. 3). These had been formed by sheep huddling in the lee of the walls, gradually wearing the subsoil away with their feet, to form a ‘shelter-trench’, which in some cases slightly under-mined the stone foundations; indeed at some places we noted sheep actually sheltering in such hollows, their backs touching the wall foundations above them. This feature at KTR I might again suggest that the building found there had been used at some stage, if not all along, as a farm-hut rather than a permanent home.

Fig 2

Fig. 2    ‘Kostoureika’: Plan of the excavated area (KTR I and II).


Fig. 3    A view (basically SE to NW), across Trench KTR I. showing the only surviving corner of stone foundations (marked X), and across the rising waters of the lake, the Peneios Dam (arrowed), and ‘Farmhouse Ridge’ (with A marking one of the ‘outholdings’).


To test the possibility that the ‘Hellenistic house’ (or ‘rustic villa’), which the quantity of pottery had promised, may have existed just a little further away from the stone structure located in 1967 and again in 1969 in KTR I, another wide trench was opened a little further W on the ridge where some more sherds had been picked up. This new trench, Kostoureika II (KTR II) extended N of the base-line adopted for the Easter 1969 excavations. KTR II measured 11.25m EW x 7.0m. NS, but was divided into a wider and a narrower part (KTR II a and b) by a strip-baulk (0.25m. wide) left between them for the convenience of drawing trench sections. KTR II revealed various hollows, some producing yet more Hellenistic sherds; perhaps they had been dug to deposit household rubbish. Fortunately, the wet conditions of Easter 1969 emphasised colour-differences in the soil and made the identification of such hollows all the more easy. Much the most important feature in KTR II, however, was the deeper hollow in the NE corner of Trench KTR II b which was followed down to a depth of 2.0m. This produced evidence of very early occupation hereabouts, flecks of charcoal, an obsidian blade, flint chips, and over a thousand sherds of Early Bronze Age pottery. This pottery (of Early Helladic II date) was earlier than any material found elsewhere in the BSA survey area, indeed than any found in all the Peneios Valley surveys of 1967, and so was of some significance.

Two more trenches were opened (4-5 April 1969) on a hillside some 500m. further NE, but still on ‘Kostouras’ land, near Agios Athanasios (KTR III and KTR IV), on sites not identified in the 1967 ground-survey but where tile-fragments had been found.7 Trench KTR III (12.00 x 5.00m. in extent) produced evidence of local occupation – disturbed stone foundations, some Laconian-type rain and cover-tile fragments, and sherds of Hellenistic – Roman date, mainly cooking-ware and coarse-ware fragments. The trial-trench, dug down to 0.60 – 0.70m. depth, produced no traceable built foundations and was abandoned. Uphill of this trench and some 85m. NE of it, another test-trench, KTR IV, was opened and extended to some 4.50 x 3.0m. in extent. This actually disclosed about half the foundations of a round pottery kiln and its flue, abandoned and ruinous. It was the more easily recognized from the mixture of bright red-brown and grey-black lumps of burnt clay from its collapsed roof. Also found here were the fragments of a large coarse pot left inside it, at the junction of the flue and the kiln.

At Easter 1969 the writer also explored, measured and recorded some abandoned farmhouse buildings along the south shores of the reservoir in the BSA’s sector. These, locally known as exochai (‘out-holdings’), were ancillary structures built in part with mud-brick walling, in part with wood and reeds, and roofed with tiles – and generally with attached sheep-folds, lightly fenced with branches and reeds. They were occupied and used for temporary summer or seasonal periods at busy sheep-rearing and farming times. Four examples were measured and planned: a ruined hut (‘Kostoureika Hut’) on the E crest of the main Kostoureika ridge site (i.e. KTR I and II) and three more elaborate structures on a nearby ridge, (Farmhouse Ridge’ Buildings A, B, and C) just across a stretch of the rising lake. Two are illustrated here (Fig. 4). It was unlikely that these buildings would ever be used again, for farming practices hereabouts might well alter, once the reservoir was fully formed. It seemed proper to record a few examples of these before they disintegrated.8

Fig. 4    ‘Farmhouse Ridge’: farm buildings B (top) and C (bottom), plans with side and end views, showing also the lightly fenced sheepfolds at their N ends.



The other main excavation site of Easter 1969 was ‘Keramidia’ (KER), about 1km N of Keramidia village. Hereabouts the 1967 survey team had located a number of sites fairly close together, perhaps parts of a ‘loosely knit settlement’. This was an area where local report spoke of chance finds made over the years, and of traces of buildings and wall foundations and quantities of tile fragments. A 1967 test-trench bore those reports out, and produced many tile fragments and ‘late Hellenistic or early Roman pottery’.

At Easter 1969 the writer led off a few workmen to extend that test-trench and explore further over four days (8–12 April 1969). At very little depth below the surface (indeed just under the ploughsoil), solid wall foundations and plaster floors were found, considerably disturbed, but the signs were that these had been successive phases of occupation over a long period.

In Summer 1969 the writer returned to explore this site further, for four weeks (18 August – 12 September 1969).9 At first two shallow trenches were opened (KER I, KER II) but these were joined up and the exposed area enlarged by successive extensions in various directions, to an irregular area of 37.50m. EW x 26.50m. NS. Lines of stone walls were revealed, laid floors, in one place of stone slabs, elsewhere of laid plaster, elsewhere of earth, and all fairly close to the surface (and so found in damaged condition); also sunken ‘vats’ or rectangular tanks, also damaged. Some wall foundations overlapped or overlaid others, which suggested successive phases of repair or rebuilding and occupation over a considerable period (Figs 5, 6). Parts of some walls had been almost completely grubbed out. Attempts were made to define the ‘outer limits’ of this structural complex, and two or three small test-trenches were opened further away in the same field, but no clear boundary line or limit was located. Finds included some 25 bronze coins, the legible ones mostly of late Roman imperial date, fragments of glass bottles, of green or bluish glass, some small lead objects such as detached ‘pot mends’, iron nails, lumps of corroded metal or slag, a few terracotta loomweights, and a mass of pottery (one complete hand-made pithos in the fill of one ‘vat’), ranging from some earlier sherds to late Roman pottery, including some tableware items, but mostly fragments of cooking and household wares.

Fig. 5    ‘Keramidia’:  a view basically N to S across the excavated area, showing disturbed walls and floors, and the thin cover of field soil over them.


Fig. 6   ‘Keramidia’:  a schematic plan of the main excavation area (with letters marking the main features, such as ‘rooms’ and floors).



A low ridge stood c . 300m. E of the main Keramidia (KER.) site. At Easter 1969 it was spotted as another possible site from the scatter of tile fragments and stones on its summit. In September 1969, three men were detached here, to ‘Keramidia Hill’, to open a test-trench which soon struck built remains. The trench was widened (to 10 x 9m. extent) and revealed very close to the surface (0.15m. down) the mere foundations of a small rectangular building set E-W, with an apsed E end, with the edges of ‘raised platforms’ inside along the two long sides, and a low square platform on the base line of the apse. Local report spoke of an ‘altar stone’ found here and tumbled downhill years before (but, oddly, no name was associated with this low hill, hence the term ‘Keramidia Hill’ adopted here). Most of the N wall had been ploughed away. Outside the building, traces of graves were found. Beyond the remnants of the N wall, and set parallel to it, were traces of three graves (A, B, C), each containing a single skeleton, quite close below the surface. In grave A the body had been laid in a ‘trough’ of reused roof-tiles and covered with similar tiles laid as a curving ‘lid’ (but cracked and broken from pressure from above); an open-ended thimble was found, suggesting this had been a woman’s grave. Grave B also contained a skeleton, similarly sheathed inside a ‘tile coffin’, but nothing else. Grave C, further out from the line of the N wall, was only partially explored; it held a third skeleton, less well preserved and without any coffin or cover of tiles. Two more graves (D, E) were partially exposed outside the apsed E end of the building, but only one end of each; a skull was found in both, showing again that the bodies had been laid with the head at the W end and the feet at the E; and two bronze wire ‘ear-rings’ were found by the skull in Grave E. The skeletal remains were cleaned and recorded, but not further disturbed, and were decently re-covered with earth, when this trench was back-filled on the last day of excavation. This hillock must have been the site of a very small church, perhaps almost a family’s private ‘burial chapel’, long abandoned and long forgotten.10



Post Excavation Studies and Publications

The finds from the summer excavations of 1969 were studied first at Keramidia village by the writer and then, until the end of September, at Olympia, in the Museum storehouse where both the Easter and summer 1969 excavation finds were taken for safe keeping. Ellis Jones resumed study and drawing of the pottery sherds as opportunities allowed, over varying periods, in the next few years: at Easter 1971 (when he was joined by Roger Howell, who was originally ‘booked’ to study the Early Helladic pottery from ‘Kostoureika’), and again on his own in September 1972 and August 1973. Many years went by; other duties intervened; other excavations both in Crete and the Laurion district of Attica occupied much time and effort.11 Finally, in retirement, Ellis Jones returned to Olympia in summer 2002 for two periods of intensive study, 8 May – 7 June, and 8 August – 4 September (in the second period with the most valuable help of his younger brother, Hywel Wyn Jones, who took digital photographs of the sherds and helped in the general rearrangement of the material. The problem was to find someone familiar with Helladic pottery to study those extremely interesting Early Helladic sherds from ‘Kostoureika’. Most fortunately and most kindly, Dr Ourania Kouka of the University of Cyprus agreed to do so; she joined Ellis Jones at Olympia for some days in June 2002 to review the Early Helladic material together, and then returned in December 2002 to complete her own studies and prepare the Early Helladic catalogue of finds for the full final report.

Brief early reports on the 1969 excavations were published in Archaeological Reports 16 (1969-1970), 14-15, under the title ‘Peneios Valley Rescue Operations’, and in Arch. Deltion 25 (1970), Chronika B1, 197, ‘Excavations in Peneios Valley’. A full and detailed report was eventually composed, and appeared in the Archaeopress series in early 2016, under the title of Elis 1969: The Peneios Valley Rescue Excavation Project. British School at Athens Survey 1967 and Rescue Excavations at Kostoureika and Keramidia 1969. In that report Ourania Kouka was responsible for the detailed account and catalogue of the Early Helladic II pottery from ‘Keramidia’ (KER II); Ellis Jones for the rest. The account given here is a brief summary only, to acknowledge, duly and gratefully, that the rescue excavations of 1969 were undertaken in the School’s name and with the School’s most generous support.

John Ellis Jones


1. E.g. the detailed publication of American excavations at Armatova (ancient Pylos) reported in Hesperia Supplement XXI (1986), John E. Coleman Excavations at Pylos in Elis.

2. Travel out to Greece and back during the Easter vacation of 1969 was by means of a large Birmingham University Land Rover which carried the whole team. The group occupied an empty house in Keramidia village. Excavations lasted for two weeks (31 March – 12 April 1969), and were carried out with several local workmen. Work was hampered at times by wet weather. All sherds collected were washed and sorted in the evenings by the team at the house occupied in Keramidia village.

3. ABSA 57 (1962), 75-114, J. E. Jones, L. H. Sackett, and A. J. Graham, ‘The Dema House in Attica’; ABSA 68 (1973), 355-452, J. E. Jones, A. J. Graham, and L. H. Sackett, ‘An Attic Country House below the Cave of Pan at Vari’ (also produced as an offprint monograph); also J. E. Jones, The Greeks (1971; in the Young Archaeologists Series, Rupert Hart Davies), 33-52, Ch. 2, ‘Hard at Work – on site in Attica’, for a brief narrative account of the Vari House dig and subsequent studies.

4. The resistivity meter, housed in a convenient wooden case, had been specially made at the University College, Bangor, for J. E. Jones by his old school friend and college colleague, Alwyn R. Owens, of the College’s Electronic Engineering Department.

5. Wooden ploughs with iron tips were in common use once in many parts of Europe. Examples of two types of wooden iron-tipped ploughs (one with a top cross-bar, to be pushed by a man’s sheer human weight and strength, another a more ‘developed’ horse-drawn plough) were included in the collection of agricultural ‘bye-gones’ formed in the 1920’s-1930’s by the late Prof. R. Alun Roberts, Professor of Agriculture at the University College of North Wales, Bangor. The writer had seen this collection early in his student years (1947-48), and thirty years later, as Hon. Curator of the College’s Museum of Welsh Antiquities (1976-1991) had the pleasure of re-forming that collection and arranging for its refurbishment (through a government ‘job-creation scheme’).

6. This farm hut was one of the various ancillary farm buildings (exochai) noted and studied at Easter 1969. See further below.

7. At Easter 1969 Roger Howell and the writer explored a number of likely places in the BSA’s survey area and spotted a number of sites not identified in 1967.

8. These exochai were of particular interest to the writer. They reminded him of similar seasonal migratory habits for stock-rearing purposes in Wales, many generations ago, from lowland to mountain pastures; farmers had ‘summer houses’ (Welsh haf = summer; hafod / hafodty = ‘summer house, pl. hafotai) that would be occupied in summer months in upland grazing areas, and the occupants moved down to more permanent lowland homes (‘hendre’: ‘old house’) in winter. However the growth of large estates in the 18th century and land enclosures led to the creation of many small tenant farms, some centred on old hafotai (‘summer houses’, exochai), and some of these were barely self-sufficient units. In the 20th century farmers in North Wales tried to rent extra fields so as to be able to move stock from upland to lowland grazing areas at need.

9. No Birmingham staff came out in Summer 1969. Travel to and from Greece was by train, and from Athens to Keramidia by bus. The writer had the help of Mrs Stephanie Gee and Miss Olivia Johnson, both of London, for the first two weeks, to the end of August. Thereafter he conducted the excavation on his own, washed and sorted the pottery and started studying the finds, first at Keramidia and then at Olympia Museum, where they were deposited. He worked till the end of September, and returned to Bangor in early October to start a new academic session.

10. It was interesting that these skeletons by this forgotten chapel had been left undisturbed, ‘in everlasting peace’ (at least until September 1969), and not exhumed to free the ground for further new burials. In the village cemetery of Keramidia village in the 1960’s the local practice had been that bodies (i.e. bones) were exhumed after three years, the bones disposed of somewhere else (perhaps in ossuary pits) and the ground made available again for new burials. But by 2002 it was noticed that a seeming prosperity allowed some families to build impressive marble faced tombs, perhaps now for permanent occupation.

11. In Crete, excavations and studies at Knossos on the Unexplored Mansion site; in Attica work on the Agrileza site in the Laureotike, between 1977 and 1997.