248

DUNCAN MACKENZIE

some time in the early Neolithic Age, had already reached large propor-
tions in Melos by the period to which belong the earliest elements of the
First City at Phylakopi. By this time accordingly trade-intercourse
between Crete and Melos must have been on a considerable scale. The
comparative evidence fumished by the ceramic finds points in the same
direction. We found at Phylakopi that a series of potsherds overlapping
the pottery of the earliest Cycladic period in the test-pit in J 1, but occur-
ring also elsewhere on the site, could not on the other hand without further
ado be eqnated with what was elsewhere found more or less intact on floors
of houses of this general period. The indications were that this frag-
mentary pottery belonged to deposit of a more or less dispersed char-
acter referable to the earliest elements of the First City. This earliest
painted geometric wäre—that with lustrous glaze design, especially the
variety with glaze slip—is so like the painted geometric Cretan wäre with
similar lustrous glaze design or slip found in similar deposit at Cnossos, that
most if im! all of ii gives the impression ofbeingan importation from Crete.]

With these indications, however, we come to the limit of our data
referable to the earliest period of the First City at Phylakopi. The
outlines of the historic picture can only be filled out by the aid of finds of
the same period from other sites such as those already available for com-
parison from Paros, Amorgos and Syros.

2. With the deposit found on the floors of houses of the First City
we are on surer ground.

Pottery found in fioor-deposit has to be associated with the remains of
wall-constructions that belong to a maturer period. Thus of the wall-remains
marked yellow on the plan we are not able defmitely to assign any to the
earliest period of the settlement; some of the pottery however may belong
to an earlier period in the life of those same houses, while some of it again
probably b.elongs to houses that have disappeared before the period to which
the earliest floor-cleposit belongs. On the other hand, we cannot defmitely
assign all these constructions to the final period of that settlement.
There are indications which enable us to distinguish earlier from later
constructions.

Falling back on the architectural illustration and description of the
wall-remains of the settlement as a whole given on pages 21-23, it will in
this conncction be sufficient from the point of view of development if we
bear in mind the following facts of a general character:—The First City
is not like the later ones a fortified city, but an open town. On the
other hand like the later Settlements it has its walls in a general way
orientated east-west and north-south. Like the later towns, its constructions
consist of earlier and later elements.

The data referable to the earlier elements of the First City, repre-
sented by pottery in a fragmentary condition from stray deposit, have
been considered above.

1 This view is also expressed by Mr. Edgar, p. ST.