suspected that there was a second complex around a courtyard to the east
of the residential block, in which case the cellar with strengthened lateral
walls (probably used for storing cereals, pulses, olives, and grape prod-
ucts, judging by the residues under collapsed beams and tiles) may well
have supported a tower, which would then have been located in the
centre of the estate. This wine cellar was evidently slightly earlier than the
farmhouse itself. It seems that a vineyard already existed in the vicinity
before the farmhouse was constructed. An earlier period of domestic use
is suggested by a hearth, whose structural associations were evidently
destroyed in the construction of the new complex.
A roofed corridor linked the store-room to the courtyard and house.
Further store-rooms extended to the north and south of the investigated
area, and the suspected second courtyard is to the east. On the west side
of the residential block, there was a storage building, with a stone-paved
courtyard, which seems to have been the main storage and processing
area of the farm. In a regular space south of the courtyard, with a lowered
floor covered with a lime plaster (5.30 m x 2.30 m), there were two
pointedpithoifor storing wine, and a large quantity of grape pips.
The farm estate was the centre of a huge concentration of activity.
Outside the main residential block, traces of roofing materials show that
there werestoas, or covered corridors, used as workshop areas, with
evidence of tools and equipment—a kiln and a hearth on the south side;
numerous household ceramic vessels; iron tools, daggers, nails, loom-
weights, and coins. The overall number of coins discovered in the farm
complex was 232, dating from the reign of Philip II to thefirst quarter of
the third centurybc.Thisisquiteanimpressivenumberforasingle
property, but many of the recently excavated rural sites also produced
coins. Both the estate at Tria Platania and at Komboloi show the same
essential organizational template, consisting of a rectangular walled residen-
tial complex, with suites of rooms along each side; a scheme that could be
repeated to provide subsidiary structures, catering for agriculturally related
activities. Although there may in both cases have been other, more remote
structures, including separate accommodation for retainers, the rationale
seems to have been for the residents and their domestic assistants—
whether employed retainers, relatives, or servants—to live in the main
residential complex. The order of magnitude for these spaces implies that
the inhabitants would have numbered in tens rather than hundreds.
Along the course of the Egnatia Odos farther along the coastline, in
Mygdonia, a larger number of sites has been disclosed on the west bank
of the River Strymon. Twenty-three separate sites, from various periods,
have been registered. Among the most interesting for our purposes are
the fortified complex at Vrasna, the property with an underground
Thelongue duréein the north Aegean 145