¹ C. Anti, Teatri greci arcaici, 55–82, has evolved new theories about the form of the theatre of Dionysus in the sixth and fifth centuries b.c. His fig. 17 gives an attempted reconstruction of what he calls the theatre of Thespis. The result is unconvincing for the following reasons. (1) As pointed out by Pickard-Cambridge, Theatre of Dionysus in Athens, 8, the orientation of the earliest orchestra is almost certainly fixed by the conformation of the hill-side, and must therefore be roughly at right-angles to the older temple of Dionysus (note that the wall SM₃ is parallel to that of the older temple); whereas Anti's reconstruction forms an angle of 64° with the wall of the older temple. (2) The shape of his reconstructed orchestra, rectilinear on three sides and curved on the fourth, is very unlikely. (3) He makes the fragment of wall J3 form the west side of a rectangle in the middle of the orchestra serving as an altar base or the like; but this rectangle would be 13 m. long!

Anti's fig. 18 gives his reconstruction for the theatre of Aeschylus. This includes, as stage buildings, the base T and the long hall to the south of it. As, however, fifth century sherds have been found under the north wall of the latter (cf. Pickard-Cambridge, op. cit., 17), it cannot well date, as Anti would have it, to the ‘primissimi anni” of the fifth century. His cavea for this period is trapezium-shaped, the drainage canal forming the east side of the orchestra. Such a scheme cannot be disproved, but it involves parodos walls converging towards the stage, instead of away from it as do the existing walls.

² See PLATE 1a; D-R 44; Fiechter, A I, 72; Pickard-Cambridge, op. cit., figs. 37, 38.

³ Philios, , PAE 1880, 50Google Scholar; 1885, 62; Dragatsis, , AE 1884, 196Google Scholar; D-R 97 (plan in fig. 34); Bulle 202; Arias 16, and in Dioniso IV, 93. Excavated by the Greek Archaeological Society under Philios in 1880.

^3a Broneer, O., AJA XXXIX, 416Google Scholar, says that the theatre was never completed, and doubts if it was used in its present form. The pavement of the passage between the prohedria and orchestra gutter was left in various stages of completion.

⁴ Cf. Judeich, Topographie von Athen, 442.

⁵ The archaeological evidence does not suffice to determine this, but owing to the size of the Athens cavea, doubling was certainly called for.

⁶ He gives wrong measurements (see fig. 2). Bulle (202) mentions the possibility of a wooden prohedria, but did not notice the incision.

⁷ Dragatsis, loc. cit.

⁸ Up to 27 cms.

⁹ Bulle 204; Arias 18.

¹⁰ Gushing, , Papers of the American School IV, 23Google Scholar; Miller, ibid., 1; D-R 109; Bieber, Denkmäler, 20; R. C. Flickinger, The Greek Theater and its Drama, 227; Bulle 9, 210; Arias, , Historia, 1933, 55Google Scholar (good photographs); Caputo, Dioniso III, 301; IV, 90; Arias 24; Bieber, H.T., 116. Photographs PLATE 2c–d. Excavated by the American School at Athens.

¹¹ Teatri Greci Arcaici, 47 f.; cf. Papers of the American School IV, 1 ff. Broneer, O., AJA XXXIX, 415 f.Google Scholar, suggests that the rock cutting was originally caused by quarrying.

¹² Cushing, , Papers of the American School IV, 30Google Scholar, estimates that in order to reach to the level of H (the first row of seats) it must have been 487 m. (16 ft.) high. This is true, but it is more likely that there was a downward slope from north to south in the orchestra.

¹³ D-R 111. There is also no room for a narrow podium, which Cushing restores on the edge of the orchestra; for the latter is already quite narrow enough without further curtailment.

¹⁴ Cushing, loc. cit., 30; nothing now visible.

¹⁵ Cushing, loc. cit., 31.

¹⁶ Cushing gives 35 cms., Arias 25. The actual measurement of the bottom (first) row proved to be 32·5 cms., of the second row 31 cms.

¹⁷ These arches may have been constructed for drainage purposes. Y has an angle turning westwards.

¹⁸ The fortification of Thorikos by the Athenians in 409 B.C., mentioned in Xen. Hell. I, 2, 1, has no immediate bearing upon the extension of the theatre.

¹⁹ W. Wrede, Attische Mauern, 34 suggests that the temple is contemporary with the extension mentioned in (e) above.

²⁰ Wrede, loc. cit., remarks that Bulk's early dating of this extension to the late fifth century is not justified, later sherds having been found below the extension.

²¹ Bibliography in BSA XLIII, 176, n. 2. Photograph PLATE 2a. Excavated mostly superficially in 1879.

²² See PLATE 2a.

²³ Bulle's measurement; it appeared to me to be nearer 5·50 m.

²⁴ On this point Bulle is right, Arias wrong.

²⁵ PAE 1891, 16.

²⁶ IG II 1278.

²⁷ Also recorded in PAE 1891, 16.

²⁸ Bibliography in BSA XLIII, 177 n. 1. Photograph plate ie. Excavated by the American School at Athens, 1888–9.

²⁹ It has naturally been conjectured that the modern name Dhióniso has some connection with the ancient ceremonies in honour of Dionysus; and this view is upheld by D. M. Robinson in Hesperia XVII, 141, n. 1. But the only ancient name for the place was Ikaria, and there is no evidence for the name Dhióniso earlier than Turkish times; one explanation might be to refer it to a landowner or innkeeper of that name!

³⁰ The numerous references to Thespis may be found in Pickard-Cambridge, Dithyramb, Tragedy and Comedy, 97.

³¹ Bibliography in BSA XLIII, 180, n. 1; also Anti, Teatri Greci Arcaici, 110 ff. For Dörpfeld's supposition of an earlier theatre nearby, see p. 33 below. Excavated 1886.

³² See BSA XLIII, 180.

³³ ‘Ro’; ‘Rw’ in Fiechter, Oropos, pl. I.

³⁴ At Priene there is not a throne exactly in the centre, as this is occupied by an altar. See A. von Gerkan, Das Theater von Priene, pl. IX, and BSA XLIII, 180.

³⁵ Op. cit., 66.

³⁶ Dörpfeld, , AM XLVII, 26Google Scholar; Fiechter, Oropos, 26; Arias 74; Caputo, , Dioniso III, 308, n. 20.Google Scholar

³⁷ IG VII 4255, 29; cf. PAE 1887, 62.

³⁸ AM XLVII, 27.

³⁹ The latter was near the modern Markópoulo in north Attica.

⁴⁰ Fiechter, Oropos, 13 f., 26 considers that the whole layout forms an ‘Einheit’, and dates the first period of the theatre, including the cavea, to 300–250 B.c.; Bulle gives 200 B.c.

⁴¹ Anti dates the stone substructures to the fifth century. This is probably too early; but there may have been, in the fourth century, resting on stone supports.

⁴² Photograph PLATE 1b below. Best account by Brownson, AJA 1st ser. VII (1891), 268; Fiechter, Eretria, 26; D-R 112. Excavated 1890–1.

⁴³ Pollux IV, 132.

⁴⁴ Cf. D-R 114.

⁴⁵ Cf. BSA XLIII, 133, 150. As pointed out by Wycherley, R. E., JHS LXVII, 137Google Scholar, Eretria can contribute nothing to Anti's theories.

^45a Bulle (91) Arias (117) date it to about 300 B.C.; cf. Fiechter, Eretria, 7.

⁴⁶ As at Athens, but contrary to Vitruvius' recommendations (cf. BSA XLIII, 133).

⁴⁷ There can only have been twenty-five rows in all.

⁴⁸ AJA 1st ser. VII (1891), 270.

⁴⁹ Ibid., 267.

^49a Oropos, 26 f.

⁵⁰ Leake, , Travels in Northern Greece, ii, 112Google Scholar; Ulrichs, , Reisen in Griechenland, i, 159Google Scholar; Bursian, , Geog. Gr., i, 205Google Scholar; Vischer, Erinnerungen, 590; Baedeker, Greece 194; PAE 1907, 108; Guide Bleu 259 f.; Arias 64. An experimental excavation, in the orchestra and below, by the Greek Archaeological Society in 1907 revealed nothing. Photograph PLATE 2b.

⁵¹ IG VII 3409.

⁵² This peculiarity, which has not been reported in any account, is shown in PLATE 2b.

⁵³ The latter is visible in the photograph, which was taken from a point immediately above it.

⁵⁴ The first step is completely rectilinear; the second more curved.

⁵⁵ IG VII 3407, where we are told: ‘Lollingius bis frustra quaesivit’. The Guide Bleu says some letters are visible, but nothing could be seen on our visit.

⁵⁶ Quaest. Symp. 717.

⁵⁷ Pomtow in Delphica, III, 77; Arias 58; N. Valmin, Fouilles de Delphes, III, vi, i ff. Plan in RE Suppl. Bd. IV, 1199. Photograph PLATE 5a.

⁵⁸ Not six, as Arias says.

⁵⁹ The inscriptions on them are, however, of the Roman period (cf. BSA XLIII, 184).

⁶⁰ Arvanítopoulos, ADelt 1915, 88; F. Stählin, Das Hellenische Thessalien, 97; Arias 39. Photograph PLATE 3a.

⁶¹ F. Stählin (and others), Pagasai und Demetrias, 23 and 119; id., Das Hellenische Thessalien, 73; Bulle 237; PAE 1912, 154.

⁶² Boëthius, , BSA XXV, 418Google Scholar; Arias 86; Bulle 259; A. J. B. Wace, Mycenae, 38, figs. 54a, b. Excavated by the British School at Athens, and further cleared in 1939. Photograph by F. H. Stubbings, PLATE 2ƒ.

⁶³ Curtius, Peloponnesos, ii, 352. Excavation report in AM XVI, 363 (Kophiniotis). A notice in ADelt 1891, 86 mentions excavations by him in 1890, and promises further reports, but none exist. Other refs.: Gardner, E. A. (and others), The Excavations at Megalopolis, 35, 42Google Scholar; Kophiniotis, 86; Vollgraff, , Nieuwe Opgravingen te Argos (pamphlet, 1931)Google Scholar; Id., BCH XLV, 223; best plan in Guide Bleu 387. Photographs PLATE 3b–c.

⁶⁴ Travels in the Marea, ii, 396.

⁶⁵ ii, 90, fig. 58.

⁶⁶ See PLATE 3b; full description and measurements in McDonald, W. A., The Political Meeting Places of the Greeks (Baltimore, 1943), 80.Google Scholar

⁶⁷ E.g. Bleu, Guide, loc. cit.; Expédition de la Morée, ii, 91Google Scholar, which goes so far as to identify some foundations with its ‘proscenium’. Anti, in Dioniso XII, 70 (review of my article in BSA XLIII), adduces it as supporting his theory of the trapezoidal origin of the cavea; but if, as argued below, it was not a theatre, it is of no value as a parallel.

⁶⁸ McDonald, op. cit. 84, considers that they were either an earlier theatre or a bouleuterion or used for both purposes in the fifth century.

⁶⁹ II, 20, 7. See Pausanias, ed. Hitzig and Blümner, I, 582. Vischer and others do not think this likely from the text. McDonald does not discuss the point.

⁷⁰ Griechische Theatergebäude (1874), Pl. I, 22.

⁷¹ Vollgraff excavated the lower eighteen rows.

⁷² See PLATE 3c. The enumeration of kerkides and stairways is always from left to right, looking from the orchestra.

⁷³ H. = height; W. = width. Dimensions in centimetres.

⁷⁴ Op. cit., 10.

⁷⁵ Kavvadias, PAE 1881–3, 1900, 1903 (full refs. in Arias and Bulle); Id., Fouilles d'Epidaure, i, 10; Id., 75; Defrasse and Lechat, Epidaure, 193; D-R 120; Bieber, H.T., 132; Bulle 167; Arias 88. Excavated by the Greek Archaeological Society in 1881. Photographs, plates 5b, 4a.

⁷⁶ II, 27, 5.

⁷⁷ Bulle 167.

⁷⁸ Cf. BSA XLIV, 328.

⁷⁹ See PLATE 5b. The proscenium, as in some other theatres, occupies exactly the width of the orchestra circle.

⁸⁰ As does Arias (88).

⁸¹ V, 9, 1.

⁸² Vallois, L'Architecture hellénique et hellénistique à Délos, 221.

⁸³ See PLATE 5b.

⁸⁴ The bend is visible in PLATE 5b.

⁸⁵ The natural slope was semicircular, but its flanks did not extend very high. This was the real reason for the absence of lateral kerkides above the diazoma. The reasons given by Kawadias, 78, are (a) a desire to avoid blocking the upper exits; (b) the restricted view of the stage which would be obtained from the upper seats. Upper exits have been deduced, at the wings of the diazoma, by the discovery of certain coping-stones (cf. Defrasse 198, 210, pis. III, IA); but exits would have been equally possible if an artificial embankment had permitted an extension of the epitheatron. As to the restriction of the spectators' view, the angle relative to the stage would not have been different from that in the lower seats close to the walls.

⁸⁶ Bruyn, Corneille Le, Voyage au Levant, 1725 ed., vol. V, quoted by Defrasse 6 ffGoogle Scholar. (cf. ibid., 198, n. 2). The accumulation of earth over the cavea, removed in 1881, did not exist before the eighteenth century.

⁸⁷ See FIG. 1 above. Dimensions in centimetres.

⁸⁸ Including Fiechter, A III, 54.

⁸⁹ The numbering of the rows includes rows of prohedria benches. Row 33 is that below the central benches.

⁹⁰ AJA XXX, 74.

⁹¹ Or 3 cms. more iri each case, including foot-space.

⁹² The fig. in Defrasse 199 (a) is not really to scale, as it implies a width of only 1·80 m.

⁹³ See plate 4a.

⁹⁴ See BSA XLIII, 173, fig. 28.

⁹⁵ See PLATE 4a. For examples of lion-legged thrones see Seltman, G., JHS LXVII, 24.Google Scholar

⁹⁶ References in Bulle, 174.

⁹⁷ Bulle, 168.

⁹⁸ Fougères, Mantinée et l'Arcadie Orientale, 165; Bulle, 248. Excavated by the French School. Photograph plate 5c.

⁹⁹ On the north side the old parados wall must have had a bend in order to include the last kerkis.

¹⁰⁰ Op. cit., 169.

¹⁰¹ See PLATE 5c.

¹⁰² Bulle 248.

¹⁰³ E. A. Gardner (and others), The Excavations at Megalopolis; D-R 133; Pausanias, ed. Frazer, iv, 330; Bulle 97; Fiechter, Megalopolis; Arias 100 (good photographs). Excavated by the British School, 1890–1. During my visit (Dec. 1938) much of the theatre was under water.

¹⁰⁴ For this reason it was not included in the note on theatres and town-planning, BSA XLIII, 138 f.; it should, however, be taken into account for the reasons given below.

¹⁰⁵ Op. cit., ch. IV.

¹⁰⁶ Probably the invasion by the Spartans in 352 b.c. would not have caused the destruction of the cavea.

¹⁰⁷ Its front faced towards the river.

¹⁰⁸ Woodward, , BSA XXVI, 123, pls. XIV, XVGoogle Scholar; XXVII, 175; XXVIII 3; Bulle, Dos Theater zu Sparta; Arias 99. Excavated by the British School in 1924. Photograph plate 3d.

¹⁰⁹ Woodward gives 179°, but from the point of view of the cavea this means 181°, as an obtuse, angle is formed.

¹¹⁰ See fig. 11.

^110a BSA XXVII, 192, 204; XXX, 152. For some problems raised by the altered orientation see the next paragraph.

¹¹¹ Summary in Bulle, Das Theater zu Sparta, 15, 49; Woodward, locc. citt. See, however, note 114 below.

¹¹² BSA XXVI, 153.

^112a BSA XXVI, pl. XV; cf. fig. II above.

¹¹³ It is uncertain whether the absence ot hollowing above row 6 has any chronological significance.

¹¹⁴ Mr. Woodward suggests in a letter to me that the earlier cavea may have been much smaller, so that the massive retaining walls at its outer ends would not have been necessary; and wonders if the altered layout was due to the decision to build a skenotheke, for which there would have been no room in the earlier west parados. Bulle, Das Theater zu Sparta, 35, postulatesi small cavea with earthen or wooden seats.

¹¹⁵ Until at least the Hellenistic age there ban have been no stage at Sparta. The peculiar character of the Sparta theatre is recognised in several passages from ancient authors, for which see Bulle 108.

^115a See BSA XXVII, pl. XXVII. In this case the west water-channel may also have been moved.

¹¹⁶ O. Walter, ÖJh 1915, Beiblatt, 68; Bulle 247; Arias 84. Excavated by the Austrian School in 1914.

¹¹⁷ VI, 26, 1.

^117a Bulle 247.

¹¹⁸ See Müller, A.. Phihlogus, Suppl. Bd. VII, 65.Google Scholar

¹¹⁹ O. Walter, ÖJh 1919, Beiblatt, 20; short notice in Bulle 259; Arias 82. Excavated by the Austrian School in 1914.

¹²⁰ AJA ist ser. V (1889), 267, pl. q (Trowbridge); IX (1893), 388; 2nd ser. IX (1905), 263, pls. 8–9 (Fossum); Fiechter, Sikyon, 27; D-R 117; Bulle 192; Arias 75. Excavated 1886–7 and later. Photograph PLATE 3e.

¹²¹ The indication of north given in AJA ist ser. V (1889), pl. 9 is very different from that of H. Wirsing's plan in Bulle. Neither is quite correct, but the American plan is to be preferred.

¹²² As do A. Müller, Lehrbuch der griechischen Bühnenaltertümer, 30 n. 3, Curtius, , Peloponnesos, ii, 490, and others.Google Scholar

¹²³ Fiechter calls them vomitoria, but this word is confusing, as it makes one think of the arched exits in the Roman theatre, which were an architectural necessity due to its totally different method of con struction.

¹²⁴ See BSA XLIII, 169; they are not mentioned elsewhere, even in Fiechter's comprehensive survey.

¹²⁵ Trowbridge, in AJA 1st ser. V (1889), 267Google Scholar speaks of ‘masonry covered with earth’. This appears to be a mistake.

¹²⁶ These can clearly be seen in PLATE 3e.

¹²⁷ All authorities now concede the existence of Greek arches; but there is no parallel in this position in a theatre.

¹²⁸ Modern Greek villages often have flights of steps with a considerable upward slope on each step.

¹²⁹ AJA ist ser. V (1889), 278. Perhaps he means 3–5 cms.

¹³⁰ Collated from AJA ist ser. V (1889), Fiechter, Sikyon, and from observation.

¹³¹ The measurement at the rear; Fiechter gives 82 cms., but this is incorrect.

^131a Bulle (199) suggests the first half of the third century; Fiechter gives a date shortly after 303 B.C. (cf. p. 53 above).

¹³² AJA 2nd ser. IX (1905), 272.

¹³³ Chamonard, J., BCH XX (1896), 256, pls. 19–23CrossRefGoogle Scholar; Vallois, R., Nouvelles Archives des Missions Scientifiques, 1921, 213Google Scholar; Fossum, , AJA XXX, 74Google Scholar; Béquignon, Y. and Replat, J., BCH LI (1927), 401Google Scholar; Bulle 174; Vallois, R., L'Architecture hellénique et hellénistique à Délos, I, 220, 231Google Scholar; Dilke, O. A. W., Mus. Helveticum, V, 60.Google Scholar Excavated 1882, 1892, 1912. See FIG. 16, and photograph PLATE 4b below.

¹³⁴ AJA XXX, 73.

¹³⁵ BCH XX (1896), 259.

¹³⁶ Classified by R. L. Scranton, Greek Walls, 173 as trapezoidal isodomic tooled work. See below, PLATE 4b, and BCH LI (1927), 404, fig. 1.

¹³⁷ See Chamonard, , BCH XX (1896), 264Google Scholar; Béquignon and Replat, BCH LI (1927), 401, work out a complete geometric layout for the theatre; but Vallois, L'Architecture hellénique etc., 223, points out that originally the proscenium was shorter than its present length, which cannot therefore be used as a unit of measurement. Dörpfeld's plan reproduced in FIG. 15 above over-emphasises the difference between the two centres, making them appear to be about 12 m. apart.

¹³⁸ Vallois, , Nouvelles Archives des Missions Scientifiques, 1921, 212Google Scholar; cf. Béquignon, and Replat, , BCH LI (1927), 407.Google Scholar

¹³⁹ See Vallois, L'Architecture hellénique etc., 221, n. 1; also FIG, 17.

¹⁴⁰ For the doubling in the epitheatron, see below. Chamonard estimates the number of spectators at our thousand three hundred for the lower part and one thousand two hundred for the epitheatron, a total of five thousand five hundred. But this is calculated on the over-generous allowance of 50 cms. width per person. If we allow 41 cms., we arrive at about six thousand seven hundred in all.

¹⁴¹ Bulle 174; Vallois, , L'Architecture hellénique etc., I, 231Google Scholar; Dilke, , Mus. Helveticum V, 60Google Scholar (where my support of Bulle's postulate of an earlier wooden cavea now seems less warrantable, as I had omitted consideration of two inscriptions overlooked by Bulle, , IG XI 2, 142 and 150 A).Google Scholar

¹⁴² Vallois, L'Architecture hellénique etc., 234, n. 4, gives 305 B.C.

¹⁴³ AJA XLI, 109 (short summary of unpublished paper).

¹⁴⁴ For bibliography, see BSA XLIII, 172, n. 2. Cavea excavated in 1899. Photographs PLATES 3ƒ, 5d.

¹⁴⁵ Height ratio is 3·45 cms. to 66, i.e., 1 in 1·91.

¹⁴⁶ See Thera III (Hiller), fig. 237.

¹⁴⁷ Nothing need be added to Dörpfeld's discussion on the function of this narrow space (UVWX in Thera III, fig. 237), owing to its poor state of pre servation. It is possible that the foundations, which are not much above rock level, have become displaced by an earthquake; deductions are therefore difficult. See Thera III, 256; Arias, , Dioniso, IV, 96 ff.Google Scholar

¹⁴⁸ See PLATE 3ƒ; Thera III, 253.

¹⁴⁹ Thera III, 256.

¹⁵⁰ One thousand two hundred to one thousand three hundred seems a more reasonable figure, considering the shortness of the front rows and the cutting off of others.

¹⁵¹ Where these overlap, it is noticeable that they are trimmed at right angles with their front edge rather than exactly parallel with the stairs. This perhaps indicates rough workmanship.

¹⁵² Thera III, 260.

¹⁵³ Anti, Dioniso, XII, 71, challenges my inter pretation of all similar seating as an economical device; but it must be remembered that stone was more precious than labour where transport was involved.

¹⁵⁴ Magnesia, however, was probably copied from Priene after a long interval; see Dörpfeld, , AM XIX, 70.Google Scholar

¹⁵⁵ Not mentioned in works on the Greek Theatre. Bursian, , Geogr. Gr., ii, 543Google Scholar; Pashley, , Travels in Crete, i, 37.Google Scholar Not excavated.

¹⁵⁶ This is the same monastery as is now there, and on visiting it I was told about its mother church at Patmps in almost the same way as Pashley was.

¹⁵⁷ Cf. Admiralty Chart, Suda Bay and Canea.