212 TANIA PANAgOU
this helps to determine whether production was short-lived or long-lasting.
However, the total number of the stamps found at one place does not deter-
mine the breadth of the distribution by itself. A stamped production, however
small it may be in numbers, may be found at various sites and at a great dis-
tance from their place of origin. Consequently, even such a limited production
can be linked with broader networks of trade.
Obviously, the size of stamp distributions is shaped by three main fac-
tors: (1) the distance of the find spot from the city that produced the
stamped amphoras, (2) the number of different find spots, and (3) the actual
quantities of stamps found at each site and in general. Based on these fac-
tors, after combining and studying the data, one can detect four patterns in
amphora stamp distribution: (1) limited, (2) medium, (3) wide and (4) wid-
est distribution.
Category 1: Cities with Limited Distribution of their Amphora Stamps
There are twenty-two cities in the category of limited distribution. These
cities do not offer strong evidence for a consistent or far-reaching export
trade. However, it is possible to identify various tendencies in this group
of cities. Because of considerable differences in the distribution of their
stamps, these productions can be further divided into the following three
subgroups.
Subgroup 1a: Regional Finds
First of all, we find production centers showing an extremely narrow circu-
lation of their products. Their stamps remain undated and are found either
locally or in the wider region of the city in question. In this pattern we can
place nine cities: Dyme (one stamp in Aetolian Chalcis), Hephaistia (one stamp
at the Cabeirion on Lemnos, plus one uncertain in Odessos), Iasos (one stamp
in Iasos and one of unknown origin in the National Archaeological Museum
in Athens), Meliboia (two or three stamps in Pella), Mytilene (two stamps
on Lesbos, one on Thasos), Pydna (one stamp in Pella), Pylos (one stamp in
Corinth), and the uncertain attributions of productions to Antandros (one
stamp in Pergamon) and to Lamponia (one stamp).
It is interesting that all of these cases are known from between one and three
stamps. They have all been found near their city of origin, but not in it, with
the exception of Iasos. It is quite possible that this picture of such a narrow
distribution is misleading; future finds or future identifications may increase
the number of stamped amphoras produced at any of these sites. Nevertheless,
the combination of restricted distribution with limited volume of production
probably reflects an extremely low production of stamped amphoras. Based on