^{page 152 note 1} The drawings illustrating this article are based almost entirely on the measurements made on the site during the excavations by Mr. P. De Jong, Architect to the British School at Athens, with a few details added from the Excavation Register of Marbles (the work of many hands). Most of the finished drawings are also by Mr. De Jong, with a few additions of my own; but Figs. 1, 2, 3, 15, 16, 17 were drawn out for me by Mr. L. W. Thornton White, A.R.I.B.A., Henry Jarvis Student of the Royal Institute, whilst he was studying at the British School at Rome. I wish to record my deep indebtedness to both the above architects for their helpful and patient collaboration.

^{page 153 note 1} Professor Ernest Fiechter, of Stuttgart, who has most kindly put at my disposal the notes and sketches which he made when he examined this wall, agrees that no other interpretation is possible for it. I am much indebted to him also for his advice and criticism on many other points connected with the study of the stage-buildings at all periods.

^{page 154 note 1} At Oropos there are three triglyphs between those over each pair of columns.

^{page 154 note 2} Omitting the end ones which are 1·22 and 1·215 m.

^{page 156 note 1} Exclusive of the gutter surrounding it; cf. B.S.A. xxvi. p. 139Google Scholar. For the festivals held at the Theatre, ibid. p. 119.

^{page 156 note 2} B.S.A. xxvi. p. 148Google Scholar f.; xxvii. pp. 189 ff.

^{page 157 note 1} B.S.A. xxvii. p. 191Google Scholar, Fig. 5.

^{page 157 note 2} Ibid. xxviii. p. 16.

^{page 157 note 3} Ibid. xxvi. pp. 148 ff.

^{page 159 note 1} B.S.A. xxvii. p. 191Google Scholar.

^{page 160 note 1} Sockets for the support of a wooden stage belonging to a later period are alluded to in B.S.A. xxvii. p. 187Google Scholar, note 2.

^{page 161 note 1} Wiegand-Schrader, , Priene, p. 245, Fig. 242Google Scholar.

^{page 163 note 1} These may very likely have belonged to a structure situated in the centre of the lowest row of seats, similar to that found in the theatre at Miletus, as shewn in Noack, Baukunst des Altertums, Pl. 87; if so, this must have been a quite late addition to the plan of the cavea.

^{page 164 note 1} There are, moreover, distinct differences in the details of the guttae between architrave No. 2 and that used for the restoration, and the triglyphs are much further apart on the former. The scale is in any case too large for the ends of the stage-colonnade with its (probably) smaller intercolumniation than on the long sides.

^{page 165 note 1} A typical example is offered by Santa Costanza, built as a mausoleum perhaps for Constantine the Great, and certainly used for the interment of the ladies of his family. Rivoira, who dates it not later than 327 A.D. (Roman Architecture, E.T. pp. 238 ff.), points out that the capitals, mostly composite, but including a few Corinthian, are re-used materials of dates ranging from the reign of Titus to that of Maxentius, and the columns, with their bases, are of alien origin, and vary in height and circumference.

^{page 166 note 1} P. 203 f.

^{page 167 note 1} Their shape would make them suitable to angle-positions in a colonnade of which the returns were not at right-angles, but we have no foundations to suggest such an edifice at the Theatre. The five sides of the example illustrated measure ·44 m.; ·45 m.; ·28 m.; ·49 m.; ·57 m.; the other one is incomplete, but clearly was not a replica of it.

^{page 170 note 1} Olympia, Tafelband II. Pl. lxxxvi. (lower r. corner).

^{page 170 note 2} Except one piece of an unusually large architrave, No. 10 below (p. 187).

^{page 172 note 1} They may be seen in the photograph reproduced in B.S.A. xxvi. p. 140Google Scholar, Fig. 9.

^{page 172 note 2} B.S.A. xxviii. pp. 6Google Scholar ff.

^{page 174 note 1} This would not be surprising, for Hadrian in omnibus paene urbibus et aliquid aedificavit, Spartian, , V. Hadriani, 19Google Scholar, 2. At the same time, we must not regard the Hadrianic date as established beyond all doubt, for there might possibly have been a second, later, Eponymos of the name of G. Julius Antipatros, grandson of the Hadrianic Eponymos.

^{page 174 note 2} I.G. V. 1, 365Google Scholar.

^{page 174 note 3} B.S.A. xxvii. p. 239Google Scholar f., No. 31.

^{page 175 note 1} Cf.B.S.A. xxvi. p. 146Google Scholar, Fig. 11; and xxvii. p. 205 (9) for the filling-up of the stage in its final period.

^{page 175 note 2} There is nothing exactly similar to be found in Chapot, La Colonne Torse, the nearest parallel being an example drawn by Piranesi, Trofei di Augusta, Pl. XI. 5, with a single row of acanthus-leaves at the top of a spiral-fluted shaft, just below the necking. This may be merely a candelabrum, since it is among candelabra that the nearest analogies to our column may be found, though their acanthus-leaves seldom turn outwards at the top. Cf. an example from Lanuvium, , Papers, B.S.R., xi. p. 83Google Scholar, and Fig. 18; and others cited in that connection.

^{page 176 note 1} But we must suppose that they were either replaced or supplemented, at a not very late date, by the ‘acanthus and reed’ type on the spiral columns (type No. 14).

^{page 178 note 1} E.g. at Ephesus, Ionic below, then two Corinthian orders; at Aspendus, Ionic below and Corinthian above.

^{page 178 note 2} The description of the Corinthian capitals owes much to the assistance of Professor Edmund Weigand, of the University of Würzburg, who has kindly allowed me to add notes which he has supplied on parallels and dating. These are distinguished by quotationmarks and printed in round brackets.

^{page 179 note 1} Except where the contrary is stated, all these capitals are of Laconian marble.

^{page 180 note 1} The quality of the workmanship is so definitely inferior to that of No. 1 that it might be permissible to regard this as a later attempt to copy a capital of ‘Hellenistic-Augustan’ type.

^{page 180 note 2} (‘We may compare the capital from the Exedra at Olympia, Olympia, Tafelband II. Pl. xc. 2’). For the evolution of the ‘Eastern’ type of capital in the first century of the Empire, see Weigand, E., Jahrb. xxix. (1914), pp. 58Google Scholar ff.

^{page 181 note 1} (‘May quite well be Byzantine.’) We may compare a similar capital, ·63 m. high over all, standing (in 1929) in front of the Sparta Museum (added since the Catalogue was published); and several, differing in minor details, at Mistra (Millet, Monuments Byzantins de Mistra, Pls. 46 and 55).

^{page 182 note 1} Professor Weigand kindly allows me to add (‘In regard to the distinctive group of “Löffel-akanthus capitals” discussed by Ronczewski, , Arch. Anzeiger, 1928, pp. 46–58Google Scholar, Figs. 5–12, I differ from the author's judgment and dating. I regard a late date as possible, since the type persists almost unchanged down to early Christian times, but do not venture on a definite opinion, since there are no really trustworthy data for comparison.’)

^{page 183 note 1} This type is dealt with in Professor Weigand's article, Jahrb. 1914, p. 89Google Scholar f.; he illustrates examples from Rome (S. Nicola in Carcere) and Pergamon (Gymnasium), op. cit., Figs. 41, 42, and cites numerous Greek examples on p. 90 (notes 1–3), pointing out that the majority belong to the period late-Trajan to Pius. For the (3) Olympia types, Olympia, Tafelband II. Pls. lxxxvi. and xc. 1, 3; two of these have palmettes on the front, of which one is upright, the other inverted, and the third has an acanthus of slender form. None of them has the frontal volutes of our specimen.

^{page 183 note 2} C.I.G. 1305. Fourmont gave the finding-place as ‘Spartae, prope ecclesiam D. Nicolai.’

^{page 184 note 1} For a cornice-block with a similar set-back see below, type 2 (0).

^{page 184 note 2} This is the material of which all the types of architrave except No. 2 are made.

^{page 185 note 1} I must accordingly retract my opinion (B.S.A. xxvii. p. 206Google Scholar, note 2) that it could not have belonged to this series.

^{page 186 note 1} The frieze-profile on the two inscribed blocks gave me the impression that there might have been some deepening of the original surface, due to erasing a previous inscription and then dressing the surface to take a new one.

^{page 187 note 1} The profile of the inner face is conjecturally shown as resembling the outer face.

^{page 187 note 2} Antiquities of Athens, III. ch. ix. Pl. III.; Marbres Antiques du Louvre, p. 78, 1391–1404. See also Perdrizet's, P. full publication, in Monuments Piot, xxxiGoogle Scholar.

^{page 187 note 3} Possibly the largest of the column-bases, with an upper diameter of ·71 m., implying a shaft ca. ·64 m. in diameter below, may be part of the Order to which this architrave belongs; if so, we might ascribe to it the largest column-fragment found (No. 11, above).

^{page 189 note 1} The possibility that one or more may have been a lintel from a doorway of the Scenae Frons must not be left unmentioned.

^{page 191 note 1} It is possible that this should be allotted to the upper Order, as it combines cornice and coffered roof.

^{page 192 note 1} Its simple profile would seem to fit it better for a cornice crowning a plain wall than for a free-standing colonnade. If it belonged to the Scenae Frons, this would imply that it may (presumably) have broken forward on to columns standing close against a solid wall (cf. the W. end of the Stoa of Hadrian at Athens). For a structure of this type we lack, however, the necessary architrave blocks with plain backs.

^{page 193 note 1} The egg-and-dart moulding, seen on the photograph, which runs round three sides of the sunk panels, has been omitted, by an oversight, from the drawing.

^{page 194 note 1} Cf. Sisson's admirable publication of the Stoa of Hadrian, in Papers, B.S.R., xi. pp. 50Google Scholar ff.; for the coping, Fig. 1 and Pl. XIX. B. Cf. also Stuart and Revett, op. cit. I. ch. v. Pl. III.

^{page 195 note 1} This would bring it to the 27th or 28th course above the paving-level at the foot of the external staircase (B.S.A. xxvii. p. 183Google Scholar f.), as the diazoma is 9·90 m. above this level. Twenty-seven courses each ·365 m. high would give us 9·855 m., or, if ·37 m. high, 9·99 m. In any case I rather distrust the indication of a string-course in Le Roy's view (Ruines des plus beaux Monuments, Pl. XIII), though I previously accepted it (B.S.A. xxvi. p. 120Google Scholar); he shews a string-course at the S.-W. angle of the retaining-wall, on the W. return, 12 courses (say 4·40 m.) above the ground-level at the time of his visit.

^{page 199 note 1} A capital of type 3 is used in Mr. De Jong's drawing, but this must be of later date.

^{page 199 note 2} See above, p. 179. As we have no Augustan structure to which to allot them, it would simplify matters if we could date these capitals to the Flavian period. There is, however, no evidence to support the view that the Augustan type survived so late, even at Sparta; or indeed that in Roman times Sparta was slower than the rest of Greece in following architectural fashion.

^{page 201 note 1} For the approximate position of the Agora, see B.S.A. xii. p. 433Google Scholar f.; the existence of an external staircase, leading to the diazoma on the E. alone, confirms our view that the normal approach was not from the west, and, moreover, the W. Parodos perhaps gave access merely to the Skenotheke in the original (Augustan) plan.

^{page 202 note 1} Cf. B.S.A. xxvii. p. 193Google Scholar f.

^{page 202 note 2} Op. cit. p. 197 f.

^{page 202 note 3} Zonaras tells us (xi. 17) that Vespasian did not put his own name on buildings which he restored. For his building-activities in general see Suet., Vespasian, 8, 9Google Scholarinit., 17; and the evidence collected by Weinand in Pauly-Wissowa, vi. 2689 f. I have not seen Newton's, H. C.Epigraphical Evidence for the Reigns of Vespasian and Titus (Cornell Studies in Class. Philology, XVI.)Google Scholar.

^{page 202 note 4} He is unlikely to have visited Sparta when in Greece with Nero, in 66 A.D., owing to the Emperor's refusal to go there (Dio Cassius, lxiii. 14, 3). Where he retired to, when dismissed for his social shortcomings, is not known, but he apparently sailed direct from Greece to Judaea on appointment to the Jewish command, early in 67. Though he passed through Greece on his way to Rome in 70, it must have been quite a brief visit (to Corinth only?), since, after leaving Alexandria in August and touching at the Lycian coast and at Rhodes he was back in Rome in late September or early October. Apart from our inscription his only recorded connection with Sparta is the rescript concerning liberti mentioned in Pliny's letter to Trajan, , Epp. x. 65Google Scholar, 3. We can hardly presuppose personal interest in the demarcation of the Laconian-Messenian frontier recorded in I.G. V. 1, 1431Google Scholar.

The evidence for Vespasian's visits to Greece is usefully collected in Braithwaite's edition. of Suetonius, Vespasian, pp. 30 and 44 f. For other Peloponnesian inscriptions to the Flavian House cf. I.G. IV. 584Google Scholar, a dedication to Titus at Argos, before 79 A.D., and I.G. IV. 670Google Scholar, a statue base to Vespasian copied at Nauplia by Fourmont (C.I.G. 1162) but no longer extant, which bears no indication of date.

^{page 203 note 1} Cf. p. 166.

^{page 203 note 2} The only cornice of white marble is that described above as type 3, of which we found only four pieces in all, two of them quite small. The remains of the inscription suggest a date not later than the Antonine era, whereas the numerous returns on the architrave of white marble indicate a broken front which cannot be dated so early.

^{page 204 note 1} See below, pp. 212 ff.

^{page 205 note 1} B.S.A. xxvii, p. 193Google Scholar f.

^{page 206 note 1} Though, as is pointed out above, this type of column goes back to the first century.

^{page 208 note 1} There seemed to be enough pieces to form at least two shafts each two metres in height.

^{page 208 note 2} No fragment gave us a complete section through the shaft at either end, and many of the pieces were only a few inches long.

^{page 209 note 1} This appearance is presumably accidental, though this fantastic creature might well have had a snake as its tail.

^{page 209 note 2} Wulff, , Altchrist. und Byzant. Kunst, p. 509Google Scholar, Fig. 439. I am much indebted to Professor Weigand for this and the following parallel.

^{page 209 note 3} Wulff, op. cit. p. 505, Fig. 435: Σωτηπὶον, Ὁδηγὸς Βυζαντ Μουσεὶοu, 29 f., Fig. 12.

^{page 210 note 1} He was trib. pot. IX. from July 1, 77-June 30, 78; imp. XIX. in the spring of 78; and cos. VIII. in 77 (cos. IX. in 79).

^{page 211 note 1} iii. 16, 6; and cf. the authorities collected by Ziehen, in Pauly-Wissowa, s.v. Sparta, col. 1496 f.

^{page 211 note 2} At least fourteen different holders of this priesthood are known. Cf. I.G. V. i, Index vi. 2, s.vv. ὰρχιερεὺς and Σεβαστοί; and add to this list B.S.A. xxix. p. 35Google Scholar, No. 58.

^{page 212 note 1} This is proved by an inscription at Mantineia, , I.G. V. 2, 281Google Scholar; see also Dittenberger's, note on it in Syll. ³841Google Scholar. For Eurykles' descent see below, Appendix No. 1.

^{page 212 note 2} B.S.A. xxvii. p. 206Google Scholar.

^{page 212 note 3} Pp.222 ff.

^{page 212 note 4} See especially his note on V. 1, 89, in which Gorgippos appears again.

^{page 214 note 1} Cf. B.S.A. xxvii. p. 205Google Scholar, and note 2.

^{page 214 note 2} I was clearly wrong in suggesting in a previous report (B.S.A. xxvii. p. 205Google Scholar) that these epithets could have been applied equally to the reigning Emperors (Diocletian and Maximian) as well as the Caesars. For ἐπιφανέστατος as almost a sollemne epitheton of the Caesar(s) see the chronological list of imperial titles in Preisigke, Wörterbuch, III. Abschn. 2, and cf. Abschn. 9, s.v. It appears to be given first to Valerian the younger, and becomes a normal epithet before the end of the third century; and is in use frequently, if not habitually, throughout the fourth. It is applied to Honorius as Caesar, in O.G.I. 723Google Scholar (= Dessau, I.L.S. 8809). For an altogether exceptional usage, in reference to Hadrian, as Emperor, see P. Berlin, 19, 1, 21.

^{page 214 note 3} Maximinus appears as ᾀνδρειὸτατος Καὶσαρ in P. Oyr. 1318; Galerius as επιφανὲστατος καὶ ᾀνδρειὸτατος, at Gortyn, I.G. Rom. i. 973Google Scholar (= 1512).

^{page 214 note 4} Cf. B.S.A. xxvii. p. 208Google Scholar.

^{page 215 note 1} Cf. B.S.A. xxvii. p. 206Google Scholar. For a still later inscription with the names of Theodosius, Arcadius and Honorius, from Megara (dated to 401/2 A.D.) see Syll. ³ 908.

^{page 215 note 2} Cf. op. cit. p. 208.

^{page 215 note 3} I.G. V. 1, 571Google Scholar.

^{page 215 note 4} Ibid. 537.

^{page 215 note 5} B.S.A. xxvi. p. 166 f., 1, B9, l. 6Google Scholar.

^{page 215 note 6} I.G. V. 1, 32B, l. 10Google Scholar; 101, l. 3.

^{page 216 note 1} I.G. V. 1, 32A, l. 3Google Scholar f.; B.S.A. xxvi., l.c. 1, B9, 1. 1Google Scholar.

^{page 216 note 2} Repeated in part, with some amplifications, in I.G. IV. i², p. xxxiGoogle Scholar.

^{page 216 note 3} I.G. IV. i², 86Google Scholar.

^{page 216 note 4} Ibid. 80, 665.

^{page 216 note 5} I.G. V. 1, 581Google Scholar.

^{page 216 note 6} Ibid. 470; cf. stemma on p. 117, and B.S.A. xxvi. p. 205Google Scholar.

^{page 217 note 1} This is not, by itself, suitable to a damaged building, and certainly not to its πὲτασος; it may have followed some other participle, e.g. from διαφορεῖσθαι, καταστπὲφεοθαι, or συμπὶπτειν. No more likely verb, ending in -οω ΟΓ-wνw occurs to me.

^{page 217 note 2} B.S.A. xxvi. p. 228Google Scholar f.

^{page 217 note 3} Forschungen in Ephesos, ii. pp. 162Google Scholar ff., 39, l. 4; 40, l. 6 f.; 41, 1. 2 f.

^{page 217 note 4} C.I.G. 3422, l. 17 ( = I.G. Rom. iv. 1632Google Scholar).

^{page 219 note 1} Note also the rough breathing over the first omikron in the name of Honorius, on No. 5. For these signs see Larfeld, Gr. Epigraphik (von Müller, I., Handbuch, I. 5, 1914), p. 301 f.Google Scholar, and Wilhelm, , Beiträge, pp. 160 f., 312.Google Scholar

^{page 219 note 2} The finding-place, near the E. end of the stage confirms this view (v. supra, p. 191).

^{page 220 note 1} E.g. τὴν οτοὴν σὺν ταῖς ὲν αὺτῆ ὲξὲδραιs erected by C. Julius Eurykles Herklanos at Mantineia, I.G. V. 2, 281Google Scholar.

^{page 220 note 2} No. e in my list above, p. 162.

^{page 223 note 1} Cf. his full stemma of the Euryklids, V. 1, p. 307.

^{page 223 note 2} In V. 1, 288, from the Orthia Sanctuary, we find a Julius Eurykles, belonging to a later generation, who was victor there as a boy ca. 134 A.D. (see my restoration of the name in The Sanctuary of Artemis Orthia, p. 231, No. 45) and συνέϕηβος to another victor in the same year (V. I, 287, = op. cit. No. 44). He may possibly have been a grandson of Eurykles Herklanos, but there is no convincing reason for postulating any connection between them.

^{page 223 note 3} I follow Kolbe's dating, though it is to be noted that the consul M. Peducaeus Stloga Priscinus, from whose nomen the lady's name is derived (originally by Le Bas, , Rev. Arch. ii. 1845, p. 211Google Scholar; Voyage, ii. 245Google Scholar) was not the first consul of this nomen. Thus she or a parent of hers may have derived it from Peducaeus Saenianus, cos. 89, or Peducaeus Priscinus, cos. 110.

^{page 223 note 4} This alternation of the names Lakon and Spartiatikos, if my suggestion is correct, seems to point to some connection, at an earlier stage, and perhaps through the female line, with the family of the Euryklids.

^{page 226 note 1} I am indebted to Messrs. R. P. Austin, H. Box and W. L. Cuttle for their careful copies in the Inventory of Stamped Bricks kept during the excavations, and to Miss G. Parkin for her patient help in the sorting and classification of those found at the Nymphaeum in 1927.

^{page 226 note 2} B.S.A. xiii. pp. 191–196Google Scholar.

^{page 226 note 3} All the types here published (I. II. IV. V.) which comprise properly a different stamp for each of the four edges of the brick, are apt to show variations in the placing of the stamps, as well as duplications and omissions. These are not taken into account where the type is recognisable.

^{page 227 note 1} This presumed type has the further peculiarity, which would distinguish it from other brick-stamps, of having no contractor's name.

^{page 228 note 1} For the ‘Ἀθάνα’ tile-works see Wace, , B.S.A. xii. p. 347Google Scholar. Kolbe's alternative explanation, I.G. V. 1, p. 168Google Scholar, note on No. 887, does not carry conviction, and Hiller's (quoted by Kolbe, ibid.) seems most improbable.

^{page 232 note 1} B.S.A. xii. pp. 344–350Google Scholar; xiii. pp. 17–43.

^{page 232 note 2} I regret that I was unable to go over the obscurer fragments in hopes of obtaining further clues to aid in their identification. If any of my suggestions are confirmed or condemned by subsequent study I shall be equally contented.

^{page 235 note 1} Cf. B.S.A. xiii. p. 24 fGoogle Scholar.

^{page 235 note 2} I.G. V. 1, 270Google Scholar.

^{page 236 note 1} B.S.A. xiii. p. 42Google Scholar; I.G. V. 1, 884Google Scholar.

^{page 236 note 2} B.S.A. xiii. p. 41Google Scholar.

^{page 237 note 1} B.S.A. xxvi. p. 165Google Scholar, 1 B 1 (γ).

^{page 237 note 2} Mr. Wace, who has kindly read the proofs of this section, suggests that the end of l. 1 might perhaps be restored as δα[μόσιοι]; but this is an unusual position for the epithet on a tile-stamp, and we should thus have to regard -τιμο as the end of the contractor's name, which is not normally put at the beginning of the stamp.