1. Map the site
2. Do a surface survey of pottery
3. Do a series of soundings

We concentrated our work on what we termed "the central area," north of the tarmack road going into Rissani, south of the convergence of the Oued Ziz and the Oued Shurfa, and between the two riverbeds.

The topographers produced a detailed planometric map of:

1. The central area on a scale of 1:2500, showing all standing walls and structures, cisterns, mounds and other important features of the landscape, and marking elevations of certain key locations.

1. The mosque area (scale 1:400). The mosque area was defined as such because local tradition refers to the best preserved of the structures as "the mosque." Yet, our topographers determined that the qibla wall of the so-called mosque was several degrees off the correct orientation toward Mecca. Local inhabitants assured us that that was the case with many mosques built in Morocco during the middle ages, a fact corroborated by a study done in by Michael Bonine of Arizona State University.
2. the citadel area (scale 1:1000),
3. the industrial (?) structure in the Oued Ziz (scale 1:100), and the area around the water tower.

We conducted a surface pottery survey in the area mapped (figure 1), dividing the area into four quadrants. Shards were collected with defining shapes and/or painted designs. By comparing the shards collected with those of nearby sites, with those in the Archaeological Museum in Rabat, and with the collection of a local ceramics expert, Mr. Lahcen Taouchickt, we were able to establish a preliminary typology containing twenty-five different types, based on glaze, color, and texture. This typology was used to classify pottery found in the soundings. The twenty-five types are:

The category labeled "Bhayr," types I, III, IV, and V, is similar to pottery that was manufactured in a nearby village, Bhayr al-Ansar, between the 17th and l9th centuries. It is a monochromatic glazed ware (figure 8). The types that are labeled "Sijilmasian," types VI, XIII, XV, XVI, XVII, XIX, XX, and XXI are believed to have been manufactured locally between the 11th and 14th centuries. Except for type VI they are unglazed (figure 9).

A 5 x 5 meter trench was excavated approximately forty meters west of the mosque, at NS 2, EW 0.7 on the Sijilmasa Central Area map. A large fragment of red mortar rubble, probably a fallen wall made of material similar to the mosque, was removed at a depth of 65 cm to 140 cm. At 220 cm beneath the surface, a stone lined drainage canal was exposed. This canal was superimposed by a lime plaster floor with a reddish mortar base. Emanating from the east balk, the canal empties into a catch basin. Three pits, possibly cesspools, were dug into virgin soil. Of these, only one was completely excavated to 140 cm below virgin soil, which contained the most important finds of the season. These include a blue-green vase with the phrase "al-baraka" twice inscribed on one side and "Allah" twice on the other. The vase is similar to a XIth century vase discovered at the Qala at Bani Hammad in Algeria and described by George Marcais. Other finds in this pit include fragments of a molded glass bowl with floral motifs and common utility wares.

A 2 x 2 meter trench was dug beneath a lime plaster floor exposed by Benchemsi in 1974. This trench is located approximately midway between the mosque and trench 1, at NS 2, EW 1 on the Sijilmasa Central Area map. Sixty cm beneath the lime plaster floor, which probably dates from the eighteenth century, we discovered a stone pavement. In the 25cm of fill between this pavement and virgin soil, Sijilmasian pottery shards were found. The stone pavement constitutes a level of occupation that predates the one exposed in the 1974 excavation.

This 5 x 5 meter trench is located along the southern border of our area of concentration, approximately mid way between the water tower and the stadium, at NS -0.3, EW -0.9 on the Sijilmasa Central Area map. Two mud walls, perhaps the entrance to a house, rest on a hard mortar foundation, which in turn descends to virgin soil at a depth of 360 cm. Ceramics from this trench were almost entirely common utility wares, suggesting residential occupation. The Bhayr pottery was found only in the upper levels of the fill, whereas Sijilmasian pottery was distributed among the various levels.

Trench 4A and 4B consist of the two small soundings done in the industrial (?) complex in the Oued Ziz. The industrial complex consists of several chambers, less than 2 x 2 meters, some of which appear to be basins connected by a series of drains. At first it appeared that the structure was a tarmery or a dying plant. However, similar structures were excavated in Greece, at Laurion and Thorikos. They were described as facilities to wash mineral ores, facilities that would have been very useful in Sijilmasa. Soundings of an inconclusive nature were excavated in two of these basins, to depths of between two to three meters. A small number of Sijilmasian shards were found; none of the Bhayr type was found. We are hesitant to suggest a date for the structure at this time

Trench 5 is a 5 x 5 meter trench located between trenches 1 and 4, at NS 2.2 EW -0.2 on the Sijilmasa Central Area map. Virgin soil was reached at depths varying from 217 to 237 cm. Two new types of pottery were found in this sounding: type XXII, black geometric design under glaze; and type XXV, crude redware with painted and incised exterior, which was tentatively identified by Susan McIntosh as Middle Niger pottery. The amount of Bhayr pottery in this trench is very small and restricted to the surface deposits.

We knew at the end of the first season that at least two distinct stratigraphic levels could be discerned, a level dating to early modern times, i.e. seventeenth to nineteenth centuries, and an earlier medieval level. The 2 x 2 meter trench within the Ben Shemsi excavation revealed two levels of occupation beneath the early modern one. Pottery from stratified contexts in our five soundings helped define our types. At this point, there were two broad categories of ceramics that exhibit characteristics that were time sensitive: the category that we called "Bhayr" that was thick and heavy and glazed, dating from the seventeenth to the nineteenth centuries; and the category that we called "Sijilmasian" that was thin and light and predominantly unglazed (except for type VI), that was pre-fifteenth century.

The results of the 1988 survey indicate that a major excavation of the site is warranted. A geomorphological survey of the site and surrounding area should be done. This will allow us to begin to assess the impact of changing water resources on the city's history. Ultimately, further excavation at Sijilmasa will produce more information about the various institutions in the city: the mosque, the citadel, the industries, the market place, and the water resources. As the relationship between these becomes clearer, we will come closer to a definition of "city" in the context of Sijilmasa.