The Fayum Project

Season 2009

Introduction

Work in the Fayum started on September 26 and ended December 16, 2009. The purpose of the Fayum project is to gain an understanding of the landscape north and east of Lake Qarun and specifically of the land and water use in the Fayum over time in relation to the development of agriculture.

Excavation and Field School in Karanis

Work on the topographical map of the site was continued, and the area previously excavated by the University of Michigan (1924-1934) was mapped out (Figure 1). The comparison of the present state of the site to the maps published by the University of Michigan represent a shocking truth: 80 years of exposition to the weather has eroded most of the buildings to ground level, has caused collapse of walls that were standing two stories high in 1934. An example is Granary C123, excavated by Michigan, which consisted of vaulted rooms, and open bins (Figure 2), of which only a vague outline of the lower part of the outer wall has survived the 80 years of exposure to rain and wind erosion (Figure 3).

Karanis map

Figure 1: Plan of Karanis by the URU team (75% of the site finished). Drawing by H. Barnard.

Granary C123 in 1920s

Figure 2: Granary C123 after excavation in the 1920s

Granary C123 in 2009

Figure 3: Granary C123 in 2009. Wind and water erosion have destroyed the vulnerable mud brick building within 80 years.

Summary of the trenches:

Work continued in the Northeast area of the site, where no excavations have taken place before 2008. Eight trenches were laid out and partially excavated (Figure 4). Trenches 11, 15 (started in 2008), and 18 (started in 2009) were all part of the same house (Figure 5).

2009 Trenches

Figure 4: Trenches excavated in 2009

House N1

Figure 5: House N1 with trenches KA11, 15 and 18

Trench KA11: 3 x 4 meter trench in building with painted wall plaster in the Northeastern part of the site. Work was started in 2008, but halted after we noticed that the plaster deteriorated due to salt formation when exposed to the air. In 2009 a conservator was present and stabilized the plaster, by adding lime plaster with glass micro balloon fillers. Study revealed the remains of two paint layers, with geometrical panels and stripes in red and green. This room was connected with a narrow hallway to a larger room to the west (Trench KA15). A stairwell was built into the structure at a later date, and the space under the stairwell was well-plastered and probably used for storage. The house was build on a number of first century CE garbage layers, which continues under the walls in the area of the street. The entrance of the house was to the south, and opened into a large room, in which excavation was not finalized (Trench KA18).

Trench KA13, started in 2008 and expanded to the west in 2009, was a 7 x 5 m. trench excavated in the street between the house and the granary (Figure 6). The lowest street levels, overlying bedrock, were dark brown sebakh layers. The entrance of the granary was protected from street level rises by an L-shaped wall, built on top of the sebakh layer. A rebuilding of the wall, related to a street level that was more than 50 cm above the original street level, revealed that sebakh cutting had been going on in antiquity, probably during the third century CE.

Street

Figure 6: The street between house N1 and the granary

Excavations in the granary, started in 2008, were expanded considerably. Along the north side of the granary the tops of the walls were articulated (Trenches KA17 and 19), while four further storage bins and a vaulted hallway were excavated in Trench KA12, which was expanded towards the west (Figure 7). Trench KA20 had four hatches, leading to four vaulted underground chambers, with ventilation ducts lined with the neck of a broken amphora. The structure of this granary is quite similar to C65, excavated by Michigan. Work will continue next year to determine the exact construction order, and explore the living quarters of the granary manager.

Trench KA19

Figure 7a: Trench KA19

Trench KA12

Figure 7b: Trench KA12

Trench KA20

Figure 7c: Trench KA20

Damage Report and Excavation at Kom W

In December 2008 it was decided in consultation with the Chief Inspector of the Fayum, Mr. Ahmed Abd el-Aal, that the unique Neolithic settlement site of Kom W needed immediate protection with a barb wire fence because the agricultural development had come very near, and because the Kom had a large number of very recent tire tracks. This site is of world importance, because it is one of the very few well stratified Neolithic sites in Egypt. Upon inspection in November 2009 it appeared that the fence had been destroyed, probably by a large backhoe, or loader, which had been driven over the metal poles. Many of the poles are bent, and the concrete base unearthed, so that the barbed wire is sagging. In addition two enormous holes were dug with a large machine. One was approximately 4 x 4 meters and 15 meters deep, the other one was approximately 6 x 20 meters, and four meters deep (Figure 8). This was probably done quite soon after the expedition had left the site, because the tracks that this large machinery must have left behind, were no longer visible. This destruction follows the same pattern as previous ones, and seems to be the work of a group of ignorant people who think they can find treasure in places where archaeologists work.

Kom W damage 1

Figure 8a: Inspecting damage at the Neolithic site of Kom W.

Kom W damage 2

Figure 8b: Inspecting damage at the Neolithic site of Kom W.

We have recorded the sections of the deep robber hole. The long, shallow robber hole cut through an area with at least 1.5 meters of Neolithic deposits. It showed that, just like we discovered in Kom K in 2007, Gertrude Caton-Thompson who excavated Kom W in 1924 stopped excavation when she reached a thick hard sand layer, thinking she had reached bed rock. The section of the robber hole showed that there are traces of human occupation which go at least two meters deeper. To capture this information we started Trench KW06, which comprises the section and the three meters behind it. The excavations revealed three round, regularly cut marks, which seem to be post holes, and several hearths. Work in this trench will continue in 2010.

Survey on the North Shore of Lake Qarun

To better understand the density of the surface remains along the north shore of Lake Qarun, a survey strategy was designed and implemented for the first time this year, to lay out large areas in the form of a plus-sign consisting of two 10 meter wide strips of 100 meter long, by 100 meter wide. In these plus-signs every piece of flint, bone, pottery, or ostrich egg shell, was measured in with a total station. By distributing the survey areas along four parallel lines from the ancient lake shore to the north, a large area could be mapped out (Figure 9).

Qarun survey map

Figure 9: Dense clusters of hearths, grinding stones and artefact scatters recorded in 2009.

Based on these first results, other measuring areas were added to augment the picture. It appears that most human activity concentrated in a band of about 200 meters wide along the lake shore, but this is not a continuous stretch. Further analysis and next year's survey should clarify what caused these areas of concentration. In the last week of survey a third stratified site was discovered between Kom K and Kom W, which has been named Kom L (Figure 10). The team will return here next year to investigate the depth of deposit and the nature of this site through excavations.

Neolithic sites map

Figure 10: Location of the Neolithic sites of Kom K, Kom L and Kom W

Finds Analysis

The finds are typically modest and fragmentary, as is typical for most settlement sites. The work of the specialists concentrated mostly on the organic find groups. The study of wood, charcoal, plant remains, basketry and zoological materials is ongoing. Glass, ceramics and terracotta specialists provide not only dating information, but contribute to map out a range of economic activities, which complement the agricultural emphasis of the project.