Hence we must figure to ourselves two great steles standing up, side by side, on the east of the tomb; and this is exactly in accord with the next period that we know, in which, at Medum, Snofrui had two great steles and an altar between them on the east of his tomb; and Rahotep had two great steles, one on either side of the offering-niche, east of his tomb.  Probably the pair of obelisks of the tomb of Antef V., at Thebes, were a later form of this system.  Around the royal tomb stood the little private steles of the domestics, placed in rows, thus forming an enclosure about the king.

Some of Professor Petrie’s most interesting work at Abydos was commenced in November, 1902.  In the previous season a part of the early town of Abydos had been excavated, and it was found that its period began at the close of the prehistoric age, and extended over the first few dynasties; the connection between the prehistoric scale and historic reigns was thus settled.  The position of this town was close behind the site of the old temples of Abydos, and within the great girdle-wall enclosure of the twelfth dynasty, which stands about half a mile north of the well-known later temples of Seti I. and Ramses II.  This early town, being behind the temples, or more into the sandy edge of the desert, was higher up; the ground gently sloping from the cultivated land upward as a sandy plain, until it reaches the foot of the hills, a couple of miles back.

The broad result of these new excavations is that ten different temples can be traced on the same ground, though of about twenty feet difference of level; each temple built on the ruins of that which preceded it, quite regardless of the work of the earlier kings.

In such a clearance it was impossible to preserve all the structures.  Had Petrie and his companions avoided moving the foundations of the twenty-sixth dynasty, they could never have seen much of the earlier work; had they left the paving of the twelfth dynasty in place, they must have sacrificed the objects of the Old Kingdom.

[Illustration:  393.jpg GENERAL PLAN OF BUILDINGS AT ABYDOS]

Also, had they only worked the higher levels, and left the rest, the inflow of high Nile would have formed a pond, which would have so rotted the ground that deeper work could not have been carried on in the future.  The only course, therefore, was to plan everything fully, and remove whatever stood in the way of more complete exploration.  All striking pieces of construction, such as the stone gateways of Papi, were left untouched, and work carried on to deep levels around them; in this way, at the end of the season, the site was bristling with pieces of wall and blocks of stonework, rising ten or fifteen feet above the low level clearances.  As the excavations progressed, there was an incessant need of planning and recording all the constructions.  Professor Petrie always went about with a large dinner-knife and a trowel in his pocket, and spent much time in cutting innumerable sections and tracing out the lines of the bricks.  The top and base level of each piece of wall had to be marked on it; and the levels could then be measured off to fixed points.