^{page 85 note 1} I.G. i². 51 (Tod, , Gr. Hist. Inscr., No. 58)Google Scholar for Rhegion; I.G. i². 52 (Tod, op. cit., No. 57) for Leontinoi. Further references are given by Hiller (in the Corpus) and by Tod. See also Raubitschek, A. E., T.A.P.A. lxxv, 1944, p. 10, note 2Google Scholar. Apparently Hiller copied a mistake of Dittenberger, , Syll. ³, No. 71Google Scholar, in his reference to Foucart's article, which should be corrected to read Rev. arch, xxxiii, 1877, 384Google Scholar. Roberts-Gardner, , Manual, No. 12Google Scholar, also give a mistaken, though different, reference to Foucart.

^{page 85 note 2} The depth of surface removed by the erasure on I.G. i². 51 was c. o.oo1 m., just enough to expunge the letters. Evidently none of the letters had been weathered before 433/2 deeper than they were originally cut. A. H. Smith reported to Bauer, (Klio, xv, 1917/1918, p. 189) that he could see scattered traces of the earlier prescript, but I have myself been unable to see them.Google ScholarBauer, (loc. cit., p. 188)Google Scholar reports from Athens that the depth of surface removed from I.G. i². 52 was c. 0.02 m.

^{page 85 note 3} This is extremely probable; cf. Tod, op. cit., p. 126.

^{page 85 note 4} Cf. Michel, , Recueil, No. 1430Google Scholar, who cites the observation of Wilhelm; cf. also Austin, R. P., Stoichedon Style, pp. 39–40Google Scholar.

^{page 85 note 5} Austin, R. P., Stoichedon Style, p. 33Google Scholar, gives the measurements for the chequer units of I.G. i². 51 as 0.0138 m. across and 0.0176 m. down in lines 1–8, and 0.0143 m- across and 0.022 m. down in lines 9–15; he also gives the measurements for the chequer units in I.G. i². 52 as 0.021 m. both across and down in lines 1–15 and 0.023 m- both across and down in lines 16–32. An excellent photograph of I.G. i². 51 is published by Austin, op. cit., Plate 6.

^{page 85 note 6} The additional letters in line 9, to the end of εịπε.

^{page 85 note 7} The additional letters in line 16, to the end of εịπε.

^{page 85 note 8} Cf. I.G. i². 19, 22, 82, 94, 103, 108, 109, 110, 118, 120. The archon was sometimes named before the epistates, as in I.G. i². 119. For the name of the archon, in I.G. i ². 19 seeGoogle ScholarRaubitschek, , T.A.P.A. lxxv, 1944, p. 10, note 3Google Scholar.Raubitschek, op. cit., p. 12, gives a new text of I.G. i². 20.

^{page 86 note 1} Austin, , Stoichedon Style, p. 40Google Scholar, comments on the engraver's addition of an extra line to care for the longer prescript. Then his observation continues: ‘But he also added an extra stoichos on the right-hand side of the new prescript, so that it had thirty-four letters to the line. This was unfortunate; the result was that, at the end of the new prescript, he had ten vacant spaces dividing the new part from the old; if he had not added the extra stoichos he would have had a gap of only two spaces. Evidently, therefore, this engraver neglected to plan the spacing of the new prescript exactly with the help of the drawn chequer, or else he miscounted the letters’. Austin did not comment on the omission of the word γραμματεύς, for had it been included both the extra line and the extra stoichos would have been necessary. Apparently the engraver planned the chequer pattern correctly, and calculated the number of letters correctly, but became absent-minded after he had begun the cutting.

^{page 86 note 2} Hicks, , Gr. Hist. Inscr., No. 40Google Scholar, thought that πρộτος was accidentally omitted; see also Roberts-Gardner, , Manual, No. 13Google Scholar. Hicks-Hill, , Gr. Hist. Inscr., No. 52Google Scholar, make no mention of it, and do not restore it in their text.

^{page 87 note 1} The remains of this letter are correctly shown by Hicks, , Inscr. British Museum, i, No. VGoogle Scholar. Roberts-Gardner, , Manual, No. 12Google Scholar, thought they might belong to pi, and suggested παντίσθένει, tentatively, as a possible supplement. There is no downward stroke at the right, as of pi, so epsilon is the only letter possible. This may be seen in the photograph published by Austin, , Stoichedon Style, Plate 6Google Scholar.

^{page 87 note 2} Inscr. British Museum, i, No. V.

^{page 87 note 3} Gr. Hist. Inscr., No. 39.

^{page 87 note 4} Rev. Arch., xxxiii, 1877, p. 387Google Scholar.

^{page 87 note 5} This observation was made in the British Museum on 11 January, 1946. The stroke may be seen in Austin, , Stoichedon Style, Plate 6Google Scholar. No trace of the cross-stroke, either sloping or horizontal, was evident to me when I saw the stone, though I reexamined it on 25 February, 1946, to make doubly sure.

^{page 87 note 6} Sylioge ¹, 24; Sylioge ², 25; Sylioge ³, 71.

^{page 87 note 7} Though Hicks and Hill in 1901 set the standard, incorrectly, for the accepted text by reading the rough breathing before ⋯μοσάντον. Cf. Hicks-Hill, , Gr. Hist. Inscr., No. 51Google Scholar.

^{page 87 note 8} Roberts-Gardner, , Manual, No. 12Google Scholar.

^{page 87 note 9} Michel, , Recueil, No. 1430Google Scholar, reads ⋯ομοσάντον.

^{page 87 note 10} Inscr. British Museum, i, p. 15: ‘The Λ at the beginning of line 12 is clearly a blunder for A’.

^{page 87 note 11} Rev. Arch, xxxiii, 1877, p.388Google Scholar.

^{page 87 note 12} C.I.G. i. 74. Boeckh too had ΛION in majuscule.

^{page 88 note 1} Silenos died in Athens, and was buried in the Kerameikos, where his metrical epitaph has been discovered. Cf. Arch. Anz. xlvi, 1931, p. 216Google Scholar; R.E.G. xlv, 1932, p. 214Google Scholar.

^{page 88 note 2} Kilo, xv, 1917/1918, pp. 190–1Google Scholar. Cf. also Accame, S., Riv. di Fil. xiii, 1935, p. 75Google Scholar.

^{page 88 note 3} Michel, , Recueil, No. 1430Google Scholar, restored [⋯ς τòν αἰεì χρόνον καί] ⋯φελέσομεν… (too long by two letters); Kirchner, , in Dittenberger, , Sytt. ³, 71 restored [⋯ς τòν αἰεì χρε⋯έσομεν… (too short by one letter)Google Scholar.

^{page 88 note 4} The restoration here adopted in line 15 was suggested by John S. Galbraith of Merton College, who rightly observes that such a supplement is necessary in the terms of the oath to substantiate the motivation as it appears in line 12.

^{page 89 note 1} Gr. Hist. Inscr., No. 52: ‘several small fragments ofthis inscription, hardly allowing of restoration, are still unpublished’. These are probably the same as those that Bauer (Klio, xv, 1917/1918, p. 191, note 2)Google Scholar says had been attributed by Wilhelm.

^{page 89 note 2} Klio, xv, 1917/1918, p. 188Google Scholar.

^{page 89 note 3} Austin, , Stoichedon Style, p. 51Google Scholar, did not fail o t note the stoichedon order. It is misleading to speak here of an ‘ordo turbatus’, as in the notes preceding I.G. i². 52.

^{page 89 note 4} Bauer, , Klio, xv, 1917/1918, p. 188, note 2Google Scholar, recorded this putting of OI in one space and observed: ‘Die Ursache dieser Schreibung von Diphthongen scheint eine lautliche zu sein’. This is too pedantic a solution. There are no laws of phonetics involved, for the whole aberration may be charged solely to the absent-mindedness of the engraver.

^{page 89 note 5} There can hardly be any doubt about the restoration of the letters IN in line 22, spaced as they are, as part of the word Λεοντίνοις.

^{page 90 note 1} The utilization of the old stones was thus not merely a matter of economy; cf. Tod, , Gr. Hist. Inscr., p. 126Google Scholar. Bauer, , Klio, xv, 1917/1918, p. 189Google Scholar, records tha t I.G. i². 52 has anathyrosis (Anschlussflache) on top, a fact which might point to a superimposed sculptural adornment and hence to a willingness to tolerate a longerasure rather than discard a valuable relief, However this may be, the stone in London has no anathyrosis; the top is delicately and uniformly pebbled and was not prepared to have anything placed upon it. This stele has also a bevel along the right side where the lateral and reverse faces join, cutting about 0.01 m. off each adjacent face. The back of the stone is rough-picked, but fairly even, and along the side and top of the reverse is a band c. 0.03 m. wide dressed down with a drove chisel. At present the stone measures 0.318 m. high, 0.228 m. wide, and 0.092m. thick.

^{page 91 note 1} Tod, , Gr. Hist. Jnscr., p. 126, suggests 446 or later; but sigma with four bars was in fact generally used as early as 448Google Scholar.

^{page 91 note 2} Riv. di Fil. xiii, 1935, p. 74Google Scholar.

^{page 91 note 3} Wade-Gery, , J.H.S. lii, 1932, p. 216, note 43Google Scholar, has restated the time-relationship correctly.

^{page 91 note 4} Earlier, that is, than the alliance with Segesta, but this must now be dated in 458/7. See Raubitschek, , T.A.P.A. Ixxv, 1944, p. 10Google Scholar.