The bare mountain of Stavrokoraki, lying at the northern side of the plain of Marathon, terminates on the east in a rounded hill above the village of Kato-Souli. Today the remains indicate that there was a ring-wall 343.5 m. long around the peak, which, where measurable, is 2.40-2.60 m. thick. From the southern corner a rather narrower (ca. 2.00 m. wide) wall runs for some 95 m. downhill to the southeast. From a point just outside the circuit at its northern edge, a second outer wall runs first down the hill to the northeast, then turns toward the south to encircle the hill at a level about half way up the hill; it finally turns toward the plain again and vanishes. This wall, where measurable, is 1.90 m. thick, relatively close in size to the southern long wall. How these two arms of the outer enclosure were linked at the west side of the central enclosure is not entirely clear. It does not, however, appear that there was merely a second wall parallel to the inner circuit on this side; for at least two other lines of wall roughly parallel to the two noticed by Lolling can now be followed, so that the following walls are seen to lie on this side of the hill:
1. The inner circuit, 2.60 m. thick.
2. A southern long wall forming part of the outer defense. 2.00 m. thick. Runs southeast from the south corner of (1).
3. A northern long wall, 1.90 m. thick.
4. A western wall, running parallel to the inner circuit (1) for most of its length and ca. 6 m. from it. Ca. 1.60 m. thick. Runs into the intersection of (1) and (2) and connects at the north with (3).
5. A second western wall, outside (4) and 2 to 3 m. from it. 1.30-1.50m. thick. Appears to merge at the south with (1), (2), and (4).
6. Begins from the northern long wall (3) and runs counter-clockwise around the hill 25-30 m. from the inner circuit (1). Roughly concentric with the other walls on western side of hill. Ca. 0.90 m. thick. Poorly preserved. Becomes impossible to trace.
All these walls are clearly distinguishable on the western side of the hill, and (6) is clearly distinguishable on the north, but the course of the other walls on this side is questionable. The whole area is covered with fallen rubble, and it is only with the greatest difficulty that a wall face can be distinguished. The basic plan seems to have been that of a fortified acropolis protected by the inner circuit and a lower area protected by the north and south long walls. There seems to have been no system of towers or other elaboration of the fortifications, and only possible traces of one entrance were found; on the southeastern side of the inner circuit there is a 2.80 m. gap in the wall which may have been a gateway. A wall runs inward from its north side, and immediately north of this, on the circuit wall and behind it, is a pile of rubble both larger and higher than is generally associated with the wall. This could be the remnants of a tower or platform belonging to the entrance. At the southwest corner (presumably of the "outer circuit") both Lolling and Milchhofer noted a gateway ca. 1.30 m. wide, of which the stones were larger and more regularly cut than those of the rest of the wall; Lolling compares it with the postern at Phyle. We were, however, unable to locate this gate or the towers also mentioned by Milchhofer on this side. The remains on this hill have almost invariably been connected with the Attic deme of Trikorynthos.60 Not only does the site fit well with the topographical information in the authors, but Trikorysian gravestones have been found near Kato-Souli. The deme was a member of the Attic tetrapolis, and the walls have therefore been dated in Mycenaean times. There does not, however, seem to be much archaeological evidence to support this date, though it cannot currently be disproved. Mycenaean sherds do not appear to be very numerous on the site. There is some obsidian, and the pottery ranges from later Geometric times (latter part of the eighth century B.C.) to Classical. The evidence is not conclusive. The dating is based on a handful of sherds from various periods, and the connection of the site with the deme-center of Trikorynthos assumes that the deme-center should be identical with the most prominent remains in the area of the deme. If the fortification is to be identified with the deme-center, it is an almost unique example of the "fortified deme," that is, an acropolis which served a single deme rather than a garrison of the Athenian army. (MCCREDIE, FORTIFIED MILITARY CAMPS OF ATTICA, 37-40.
Info: McCredie, Fortified Military Camps
James R. McCredie, Fortified Military Camps in Attica, Hesperia Supplement 11, 1966,