Kekrops then was a dead, divinized hero embodied as a snake; the natural place for his worship was his tomb, probably the earliest sanctuary of the Acropolis. Clement[84] of Alexandria says, ‘the tomb of Kekrops is at Athens on the Acropolis,’ and Theodoretus[85], quoting Antiochos, adds that it is ‘by the Poliouchos herself,’ the goddess of the city. We might safely assume that a hero-tomb was a sanctuary, but we have express evidence: in an honorary decree[86] respecting the ‘ephebi’ of the deme of Kekrops it is ordered that the decree shall be set up ‘in the sanctuary of Kekrops,’ and from another decree[87] we learn the name of a ‘priest of Kekrops.’ But our most definite evidence as to where the tomb of Kekrops lay comes from the famous Chandler inscription[88] now in the British Museum. This inscription is exactly dated by the archonship of Diokles (409-408 B.C.). It is a statement of the exact condition in which the overseers of the unfinished temple took over the work, what part was half finished, what unwrought and unchannelled (_i.e._ columns), and what were completely finished but not set up in their place. The various parts of the temple are described as near or opposite to such and such an ancient shrine, and fortunately among these descriptions occur more than one mention of the Kekropion. The following[89] is decisive: ‘Concerning the porch beside the Kekropion the roof stones above the Korae must be....’ The porch of the Karyatids, or to call it by its ancient[90] name, the porch of the Korae, the Maidens, was beside, close to, the Kekropion. So far all is certain. The tomb of Kekrops was close to the porch of the Maidens; but in which direction? We should expect it to be north-west, because in that direction, as will be immediately (p. 48) shown, lay the precinct of Pandrosos, daughter of Kekrops. Professor Dörpfeld[91] places it conjecturally at D (Fig. 16), and the site is almost certain. It has been already noted that the west wall of the present Erechtheion was set back a short distance within its original plan. It may have been to avoid trenching on the tomb of Kekrops. Moreover, at the south end of this wall there is a great gap in the ancient masonry of about 10 ft. long by 10 high. The gap is evident, though it was filled up by modern masonry. It is spanned by an enormous ancient block of stone, 15 ft. by 5. Here probably was buried the serpent king. [Illustration: FIG. 19.*]