^{page 35 note 1} CQ N.S. xvi (1966), 286–90.Google Scholar What I call the front of the stone contains eight inventories, IG i ². 256–63,Google Scholar for the years 434/3–427/6. The reverse contained, on my hypothesis, the inventories for 426/5–423/2, which are now lost.

^{page 35 note 2} These eleven fragments are IG i. 141Google Scholar – 4b–c, IG i. 147–50 a–b, IGi. supp., pp. 26 and 130, three fragments published by Woodward, A. M., JHS xxxi (1911), 35–40,Google Scholar and SEG x. 193–4.Google Scholar

^{page 35 note 3} My colleagues in this examination were Professors Malcolm McGregor, D. W. Bradeen, who had supervised the construction of the stele, and Donald Laing, whose own studies of the re-use of inscriptions first drew my attention to this important facet of epigraphy. They do not necessarily share the views which I finally formed about the nature of this inscription. Professor Eugene Vanderpool examined the mouldings of the stones for me.

^{page 35 note 4} The dimensions of this fragment, IG i. supp., p. 130,Google Scholar are alt. 0.36 m, lat. 0·325 m, cr. 0·154 m. This stone preserves the original top of the stele.

^{page 35 note 5} The position of this fragment in the rear face of the stele extends from 0·65 m from the top to 0·88 m from the top. This fragment is 0·15 m wide and 0·175 m thick.

^{page 35 note 6} What I call the front face contains four inventories, IG i. 121–4Google Scholar = IG i ². 236–9.Google Scholar The reverse face also contains four inventories, IG i. 125–8Google Scholar = IG i ². 244–7.Google Scholar

^{page 35 note 7} The new edges probably came to points at the plane of the rear face, but they have been blunted in an irregular fashion now. The right-angle cuttings have also been damaged and appear as little more than a raised edge toward the bottom of the stele. For a drawing of a very similar moulding cf. Anastasios K. Orlandos, ‘H ii. 377, n. 8.

^{page 36 note 1} Here, too, the right-angle cutting is largely worn away, and the point of the moulding has been completely broken off.

^{page 36 note 2} According to Meritt, AJP lix (1938), 500–1, the stone was in the Cabinet de Médailles of the Bibliothèque Nationale in 1925, but on my visit in 1968 the authorities could not locate it.Google Scholar

^{page 36 note 3} Cf. IG i ². 236.Google Scholar

^{page 36 note 4} The apograph in the Editio Maior is based on that of Raoul-Rochette, which Boeckh published as CIG 137. As in numerous other fifth-century inventories spaces were left uninscribed at the end of several lines in IG i ². 256,Google Scholar but lines 1 and 4 certainly contained sixty-seven spaces (cf. IG i ²., p. 303), and the Paris fragment breaks off after the sixty-second space.Google Scholar

^{page 36 note 5} The moulding appears on a fragment containing parts of lines 24–7, which Broneer published in Hesperia, ii (1933), 375.Google Scholar

^{page 36 note 6} JHS xxxi (1911), 38–9.Google Scholar Unfortunately, there is no record of precisely how far downward the projection extended. Bradeen (per ep.) confirms the join but he would describe the surfaces, not as flat, but as ‘one a little concave, the other convex’.

^{page 36 note 7} According to Woodward, , JHS xxxi (1911), 39,Google Scholar there is ‘a vertical split practically from the top to the bottom of the original stele, which was clearly made before the horizontal split [just below line 27], for it continues in exactly the same line through both halves of the slab’. This reasoning does not seem cogent to me. Since stones tend to fracture along lines of structural weakness, the vertical breakage may be later than the horizontal, just as my interpretation would require.

^{page 36 note 8} The Pronaos stele was probably not much taller than its present height of 1·59 m.

^{page 37 note 1} IG i ². 256, line 1.Google Scholar

^{page 37 note 2} The original top of the Pronaos stele has been cut in such a way that it is no longer parallel to the ground but slopes downward from the original front face to the back face. This new cutting, presumably by the same mason who prepared the other mouldings, would be visible with the stone lying on its original front face.

^{page 37 note 3} The stele containing the Kallias Decrees was used as an altar slab; cf. Wade-Gery, , JMS li (1930), 58.Google Scholar

^{page 37 note 4} In dealing with the re-use of an inscription we must be alert to the possibility that the stone was re-used on several different occasions. At one time the Pronaos stele was certainly used as a door-sill. Its rear face is worn smooth on the right side from top to bottom, while on the left it has two pivot holes and signs of discoloration forming a semicircular pattern. (There is also a rectangular cutting in the upper left-hand corner.) On the front the face of the upper fragment, IG i. 121–4aGoogle Scholar is well preserved, but the face of the lower half, IG i. 121–4b, has been mutilated, perhaps by chisel blows. This mutilation presumably occurred after the stone was broken in two.Google Scholar

^{page 38 note 1} At first it seemed to me that the mason might have sawn the stone in half, but Bradeen says that it was fractured, not sawn, just below line 27.