Eridu, there is little evidence for settlement, but at Kish both the main mound at
Ingharra and especially Mound P – the location of the plano-convex building and the
three palaces – have evidence for associated residential districts.^17
Southern Mesopotamia was never again as urbanized as it was during the Late Early
Dynastic Period. Adams ( 1981 : 138 , tab. 12 ) estimates that only 10 percent of the sites
dating to this period were non-urban. Until quite recently it was difficult for archae-
ologists to obtain excavation permits to work on small sites and it is for this reason
that we have so few Early Dynastic sites to compare with their urban counterparts.
Nevertheless, the limited data available from high-resolution satellite imagery on the
structure of small sites indicates that they were as crowded as the larger cities.
LATER SUMERIAN CITIES
As we move into the later part of the third millennium–the Akkadian and Ur III
periods–we once again encounter a dearth of information on the overall organization
of settlements. Moreover, the high-resolution satellite imagery does not help here since
the archaeological surveys which provide the dating for unexcavated sites failed to
adequately distinguish between Ur III and later Larsa or even early Old Babylonian
settlements. However, the larger excavated sites evidence continuity of occupation
from at least Ur III through early Old Babylonian times, when many in the south were
largely abandoned and not reoccupied for more than three centuries (Stone 1977 ;
Armstrong and Brandt 1994 ). Data from excavated sites suggest that the surviving
surfaces of southern mounds have broad areas of early second millennium BCdomestic
architecture and, probably, Ur III public buildings (Stone 2002 ). Further north, any
Ur III remains were thoroughly buried beneath the remains of their early second
millennium settlements and have remained inaccessible to archaeologists.
As the center of the Ur III state, the city of that name has the most extensive
preserved public district. Ziggurat, temples,^18 priestly residence–even royal tombs–have
been excavated at Ur and dated to this period (Woolley 1974 ). When kingship moved
first to Larsa and later to Babylon, some of these public buildings were no longer
maintained, but the main ziggurat, temples, the priestly residence, and a probable
treasury built in Ur’s suburbs all testify to the continuity of occupation of this
important city. And it is to this time that we have extensive excavated evidence for
domestic areas–three of which have been sampled (Woolley and Mallowan 1976 ;
Charpin 1986 ; Van De Mieroop 1992 ). Extensive excavated remains also exist at Nippur
(McCown and Haines 1967 ; Stone 1977 ), Tell ed-Der (Gasche and Pons 1989 : Plan 1 ),
Tell Asmar (Delougaz, Hill and Lloyd 1967 ) and Ischali (Hill, Jacobsen and Delougaz,
1990 ). Perhaps the city whose overall organization is best understood is Mashkan-
shapir, where a combination of surface survey, aerial photography (Stone and
Zimansky 2004 ), high-resolution satellite imagery and limited excavations have
allowed the development of an overall plan of this city (Stone forthcoming). The
satellite imagery has also greatly expanded our understanding of the organization of
Tell Asmar and Ischali, as well as outlining the structure of other early second
millennium cities, towns and villages. Moreover, a number of small settlements dated
to this period have been excavated–the most complete being Haradum (Kempinski-
Lecomte 1992 ) and Tell Harmal (Baqir 1946 , 1959 )–and many more have good high-
resolution satellite imagery. A comparison between these later Isin-Larsa to early Old
–– The organisation of a Sumerian town ––