85 – 87 ). It seems that whatever the motive for the Ubaid travellers had been, the thrust
of the contacts turned north and southeast in the Uruk period. Regular contact with
the Gulf is only resumed after the collapse of the Euphrates route northwards.
There has been speculation about the reasons for the collapse of this wide-ranging
Uruk network and it is still unclear what factors brought it about. Some of the colonies
such as Habuba seem to have been deserted, while Hassek Huyuk was destroyed
(Algaze 1993 : 107 ). We can suggest that many of the traders who had manned the
colonies and outposts of the Uruk system were forced to relocate.
Dramatic changes at much the same time are visible in the north at Arslantepe and
as we have just noted at Susa, and at Godin Tepe in the Zagros. At both Arslantepe and
at Godin, the post-Uruk levels are marked by a new type of pottery, Kura-Araxes ware,
which originated north of the Causcasus (Frangipane 2002 ; Badler 2002 : 83 ). The
appearance of this pottery appears to be related to movements of new people
southwards into Anatolia and northwestern Iran (Kohl 2007 ). Can we suggest that part
of the reason for the end of the colony system was that the production of ore was
disrupted at the Anatolian mines as a result of these incursions? The disruption of
supplies would have made the late Uruk colonies unnecessary and must have had a
major economic impact on the cities of the plain. It is noticeable that there was a sharp
contraction in the urban population at this time and at Uruk there is a break in the
Eanna sequence at the same moment, which we can perhaps attribute to a period of
recession. There is no evidence of an attack or an invasion at Uruk itself and the
recession, if such it was, was followed by another flowering in the third millennium
when the old public buildings were carefully cleared out and a massive terrace built
over them. The so-called Sammelfund, a cache of precious objects probably buried at
the end of the Uruk period, has been used to suggest an attack, but we know that later,
in the mid-third millennium, superfluous items of religious significance were disposed
of by carefully burying them^1 so we can suggest that this, rather than imminent danger,
may be the explanation for the Sammelfund.
In north Mesopotamia, the collapse of the Uruk network had even greater reper-
cussions with urban centres disappearing from the archaeological record for several
centuries. This did not mean that trade ceased completely. In both north and south
there is evidence for some foreign goods, but the scale is hugely reduced and in some
cases the routes too seem to have changed (Quenet 2008 : 36 , 97 ff.). In the north the
distribution of one particular type of seal, the Piedmont, suggests the use of a route
from Elam to Western Syria (Collon 1987 ) that hugged the flanks of the Zagros and
Taurus mountains. The Piedmont seal is typically made of glazed softstone, it is long
and thin, and is decorated with simple geometric motifs and crude animal files. These
seals are relatively rare on the southern plain which does not seem to have been a major
player in this network. The route mapped by these seals is also marked by a string of
circular fortified sites in the Hamrin valley, east of the middle Tigris, which seem to
have acted as staging posts on the route. Some continued in use into the succeeding
Early Dynastic I period but after this the valley is more or less deserted, pointing to
another shift in the direction of the trade.
In spite of the very limited evidence, it seems clear that the supply of foreign goods
in the southern plain was drastically curtailed after the collapse of the Uruk colonies.
For example, the amount of metal brought into the south was apparently sharply
reduced, although as few sites of the period have been excavated, it is difficult to make
–– Trade in the Sumerian world ––