Hogarth, David G.; Smith, Cecil Harcourt [Contr.]: Excavations at Ephesus: the archaic Artemisia: Text (London, 1908)

Excavations of 1904-5. 29

painted red-figured ware, but no black-figured or other painted ware. The
terracottas included interesting statuettes of " Leto and child," and a seated
Goddess. Nothing of a date necessarily earlier than the beginning of the
fifth century B.C. was found in these spaces, which had evidently been filled
with the dt'bris most readily obtained from recently formed refuse-heaps.
Every object mixed with the earth, except the solid " loom-weights,"
was broken, and most objects were mere disjecta membra, thrown in pell-
mell. The test of keeping apart the sherds, etc., from successive levels was
applied, while certain of the least disturbed spaces were being excavated ; but
it proved conclusively that there was no stratification. The filling and ramming
had all been done at one time. Lying, as they do, at a very low level, these
spaces were the first to be flooded, and in the second season could only be
investigated by the aid of a hand pump, supplementing the steam pump ; and,
where much denudation had taken place, as, e.g., in the east and south, were not
found worth cleaning out completely. It seems probable that Wood was right
in his belief {Discoveries, p. 187), that a stream or flood-torrent passed along
the south side of the site after its abandonment, and washed out, amontr other
things, the contents of the spaces between the Hellenistic piers.

Reference has already been made to explorations carried out round the
limits of the latest platform by broad trenches dug at various points along the
line of the lowest step. These were carried along almost the whole length or
the northern and eastern limits. But along the southern limit pits only were
sunk at intervals ; and along the western limit, owing to the immense depth of
deposit, and the extreme saturation of the bottom soil, only three or four pits
were made. Lengths of step not planned by Wood were thus found on the
north and at the south-east angle (see plan, Atlas, I.); and the fragment used
by him as an eastern limit was verified. Sections of the courtyard pave-
ment, which was of uniform character wherever opened out, were obtained ;
and a broad gutter, on the north, east, and south, dug probably by the
Hellenistic builders, at a uniform distance from the lowest steps on all those
sides, was discovered. This was paved with tiles, and had been roofed with
slabs, some of which were still in position, and at the north-east and south-east
corners it was entered by tributary gutters descending from the higher ground
of the eastern precinct. The north and south gutters discharged to westward
below the site. The discovery of this gutter on the south, where no step has
survived, supplied a criterion whereby to check Wood's measurement of the
breadth of the latest platform.

The objects found in these marginal trenches were, without exception, ol