¹ BSA lii (1957) 152–89, ‘A Survey of the Aigaleos–Parnes Wall’. References to the house: 153–6, 161, pls. 29, 30 (site); 171–2, fig. 6 (description and plan); 184–5 (discussion and date).

² We should like to express our gratitude to the following: Drs. Orlandos, Papademetriou, and Threpsiades of the Greek Archaeological Service for granting permission to work and publish; the Director and Committee of the British School for support and encouragement, and for the use of field equipment; Professor Homer Thompson of the Agora Excavation for encouragement and great kindness in making possible the loan of equipment and the hiring of skilled workmen from the Agora in 1960; Mr. Demetrios Christodoulou of Ano Liosia for negotiations locally in 1958; Mr. S. J. E. Jones for help on the site and with the sherds in both seasons and for the plan in Fig. 1; Miss Lucy Talcott, Miss Virginia Grace, and Professor E. Vanderpool of the American School of Classical Studies and Mr. B. Sparkes of the British School for their guidance on the sherds and small finds.

Fieldwork in 1958, limited by our permit to cleaning operations and aimed at rescuing more of the plan, was carried out with two locally hired men, Aug. 28th–Sept. 4th. The site was completely uncovered, with the aid of four experienced Agora workers and two local men, in August 1960; the house was then left open to view but with earth banked up to cover and protect all walls and features.

^2a A visit in May 1962 revealed that the municipal authorities of Athens have appropriated the area; a new road has been bulldozed to the site itself, and the hollow below and north of the Dema house is very quickly being filled in with city refuse.

³ BSA lii (1957) 172, fig. 6 gives a plan of the then visible remains.

⁴ Karten von Attika; Text ii. 45; BSA lii (1957) 160.

⁵ Levels were taken where the conduit was explored at points along the southernmost 105 m. of its known length of 115 m. There is, to begin with, a total rise northwards of 5·5 cm. in the bed of the channel in the first 29 m. After this there is a fall northwards of 44 cm. in the next 50 m. to the last point where the bed itself was exposed. This fall northwards continues, for at the last level taken, 26 m. farther on, even the top of the channel marked a further drop of 4 cm.

⁶ Curtius, and Kaupert, , Karten von Attika vi.Google Scholar

⁷ Dinsmoor, W. B., The Architecture of Ancient Greece (1950) 43–44, fig. 16Google Scholar; Pritchett, W. K., Hesperia xxv (1956) 281–6, for types and terms.Google Scholar

⁸ Published dimensions of Corinthian rain-tiles show a singular consistency over a long period in Attica: e.g. archaic tiles from Tholos kitchen drain (c. 470 B.C.) 67 × 53 cm. (Hesperia Supp. iv (1940) 76); classical tiles from Stoa of Zeus (c. 430 B.C.) 67 × 58 cm. rain-and cover-tiles in one, 67 × 76 overall (Hesperia vi (1937) 36–37, 45); Roman tiles from the Odeion (c. 15 B.C., and repairs c. A.D. 150) 67 × 56 cm. (Hesperia xix (1950) 49–52, 54, 125–6, pls. 37, 39); its marble tiles 79–83 × 60 cm. Cf. archaic tiles from the Athenian Treasury, Delphi, 68 × 57 cm., and Temple of Aphaia Aegina, 68 × 58 cm. (BSA xlix (1954) 211); classical tiles from Olynthus, 67 × 55 and 65(?) × 53 cm. (Olynthus viii. 232; xii. 242–3, pl. 208 correcting text, and cover-tile suggesting length of rain-tile, but note also a smaller size—50 × 49, 56 × 46, Olynthus viii. 232; xii. 82, 186). Cf. also the tile standard from Assos, 70 × 60 cm. (Clarke, , Bacon, , and Koldewey, , Assos (1902) 71, 73, 167–8)Google Scholar, and tiles from Samos, 81 × 72 cm. (archaic Rhoikos temple, AM 1930, 88), from Priene, 64 × 53 cm. (Wiegand, and Schrader, , Priene 306)Google Scholar, from Delos, 65 × 56 cm. (Délos viii. 319, fig. 196), and Chios, 64 × 52 cm.–70 × 57 cm. (no apparent standard, Roman period: BSA xlix (1954) 170–1).

⁹ As regards material and surface finish, it is interesting that at Athens the Tholos (c. 470 B.C.) (Hesperia Supp. iv (1940) 65 and coloured frontispiece), the Stoa of Zeus (c. 430 B.C.), and the Metroon (150–125 B.C.) (Hesperia vi (1937) 36–37, 191–2, 195) had tiles of buff-yellow clay, with dark grit, and a yellow slip, whereas the later Odeion had buff to brown-red tiles without slip or glaze. Olynthus produced some similar yellow-surfaced Corinthian tiles (Olynthus xii. 242); by contrast, mid-5th-century-B.C. Corinthian tiles at Assos were black-glazed (Clarke, , Assos 167–8).Google Scholar

¹⁰ Stevens, G. P. (Hesperia xix (1950) 174–88, pl. 82)CrossRefGoogle Scholar describes an Athenian Laconian-tile standard of the 1st century B.C. preserving a rain-tile 98 × 49 cm. and a cover-tile 84 × 17–22 cm.; he distinguishes two types of rain-tile, viz. wedge-shaped and parallel-sided, and two sizes for sets of rain- and cover-tiles, a longer of 98 cm. and a shorter of 84 cm. The Assos Laconian-tile standard shows one size only, a rain-tile 93·5 × 41–49 cm. and a cover-tile 93·5 × 20–26 cm. (Clarke, , Assos 71).Google Scholar The third such Laconian-tile standard known, a recent find from Messene, again illustrates the wedge-form (Ergon (1960) 168, fig. 182). Actual examples of tiles from Athens include two rain-tiles and a cover-tile of 5th-century-B.C. context, 87 × 52 cm., 85 × 40 cm., and 91 × 17–21 cm. (Hesperia xx (1951) 174–5; xix (1950) 184–5, 179–80), and a rain-tile of 2nd-century-B.C. context, 98 × 46–50 cm. (ibid. 179–80, 184). Olynthian examples include rain-tiles of 95 × 45–51 cm. and 96 × 47–50 cm. and cover-tiles of 95 × 19–21 cm. (Olynthus viii. 232; xii. 185). A smaller pattern was used in Hellenistic Dura-Europos, 80 × 41–47 cm. and 80 × 19–20 cm. (Excavations of Dura-Europos, Ninth Report, 10) and in late Roman Athens, 80 × 35–41 cm. (c. A.D. 500, Hesperia Supp. iv (1940) 124).

¹¹ Red, brown, or black glazing on Laconian-type tiles was common in Attica in the 5th and 4th centuries B.C.; cf. the Tholos kitchen, 2nd period (4th century B.C.), brownglazed tiles, and red-black glazed tiles at the Leipsydrion fort and the Dema wall towers (BSA lii (1957) 185).

¹² Find places were as follows: I—1 frag., III—2 frags., VIII—1 frag., IX—9 frags., X—1 frag., XI—1 frag.

¹³ Such basins, originally supported on a separate central pedestal at waist height and used for washing, are commonly found in ancient domestic and public buildings, and also as dedications. At Athens houses in the Agora area have produced few marble basins, more often terracotta ones; at Olynthus marble and terracotta bowls, were found in roughly equal numbers; and at Delos they were common. Illustrated Acropolis examples, having the shallow profile and picked floor of our louter, vary in size (65–120 cm. diameter), but are all decorated with reeded or moulded rims and/or carinated bowls and all but four are dated 530–480 B.C.; the few with plain rims have deep bowls and polished interiors and are again pre-480 B.C. (Raubitschek, A. E., Dedications from the Akropolis (1949) 370–413.)Google Scholar Three illustrated examples from the public buildings of the Agora (Tholos area) have the shallow profile (64–76 cm. diameter × 9·8–11·5 cm. height) and roughened floor, but again moulded rims; two are pre-500 B.C. and one probably c. 400 B.C. in date (Hesperia Supp. iv (1940) 142–3). Olynthian marble basins (Olynthus viii. 317–20) described and illustrated include (A) four with shallow profile, but moulded rims and/or carinated bowls: (1) 85 cm. diameter – 12 cm. int. depth (Olynthus ii. 67, figs. 174, 175), (2) 62 cm. diameter (Olynthus xii. 218, pls. 186, 2; 187, 2), (3) 84 cm. diameter (ibid. 242, pl. 211, 1), and (4) 90 cm. diameter × 7 cm. int. depth × 10 cm. height (ibid. 246–7, pls. 218, 219; 4th century B.C.); of these, three, (1), (2), and (4), had fluted marble pedestals: 43, 52·5, and 68 cm. high respectively, and one, (3), had a terracotta pedestal on a 43·5 cm. square base; and (B) one basin (5) apparently with uninterrupted curve from bowl to lip, but also deeper, 90 cm. diameter × 15 cm. int. depth (Olynthus viii. 319, pl. 78, 6); all five had square tenons below (8–10 cm. square) to fit mortises in their pedestals, and two display a roughened circle around the tenon, (1) (32 cm. diameter) and (5). Examples at Delos illustrate both the low moulded profile (66, 63, and 71 cm. diameter; Délos xviii, Le Mobilier Délien 48, pls. 148, 149, 177) and also the deeper, unmoulded bowl (in diameter and height: 60 × 15, 53 × 18, 50 × 30 cm., ibid. pls. 223–4, 243, 244).

¹⁴ Ancient repairs with lead clamps of marble and terracotta louters, bathtubs, pithoi, and fine pots are frequently found: e.g. Olynthian marble louters, nos. (1), (2), and (3) in previous note, and complete Olynthian terracotta basin and stand in next note. They indicate the value of these objects and ancient frugality, cf. Cato, De Re Rustica xxxix.

¹⁵ The reconstruction depends, as regards height and basin diameter, on Olynthian parallels. At Olynthus terracotta basins were c. 85–90 cm. in diameter and shallower than the marble basins; some had tubular tenons to fit into separate terracotta pedestals (Olynthus viii. 320), while at least one example was made in one piece with its pedestal. This (Olynthus xii. 229, pls. 191, 2; 219, 2) was 73·5 cm. high overall, having a basin 86 cm. diameter × 7·5 cm. deep int. × 11·5 cm. high ext. with a triple-moulded rim, 4 cm. wide × 8 cm. high, and a pedestal 62 cm. high, comprising a square abacus, 37·7 cm. square × 2·5 cm. high, a bowlshaped echinus, 35 cm. diameter tapering down to 20 cm., a fluted column, 40 cm. high, and a spreading circular base, 60 cm. diameter × 10·5 cm. high. Other terracotta louter stands include a fragmentary hollow fluted shaft (ibid. 205, pl. 173, 3), and four with columnar shafts on square bases: (1) 66 cm. high with flaring square base, 38–39 cm. sq. × 11·5 cm. high, rounded torus, c. 21 cm. diameter × c. 5 cm. high, and fluted column, c. 47 cm. high (Olynthus viii. 320, no. 17, pl. 78, 4); (2) 45 cm. preserved height, with flaring square base, 45 cm. sq. × 14 cm. high, with torus, 1·8 cm. high, and fluted shaft, 30 cm. high (Olynthus ii. 92, figs. 200, 204; viii. 320, no. 18); (3) square base, 45 cm. sq. with fluted shaft (ibid. 320, no. 19); (4) square base, 43 cm. sq. × 13 cm. high, with round shaft (Olynthus xii. 242).

¹⁶ For a brief discussion of ancient Greek bathtubs, Greece and Rome ² vi (1959) 31–41. Olynthian examples are numerous; generally they are 100–125 cm. long, 70–75 cm wide, and c. 40 cm. high, flat-bottomed with a shallow basin, c. 35 cm. diameter at the foot, and steep-sided, higher at the back than the front, with out-turned rims (Olynthus viii. 200–1, pls. 53–54; xii, pls. 200, 236); many were found in situ (Olynthus viii. 46–50; viii. 199–204), frequently in small compartments, some with cement floors and drains, clearly bathroom cubicles within a kitchen or living unit (e.g. Olynthus viii, pls. 49, 106, 87, 89; xii, pls. 14, 16, 202, 207), occasionally open in a utility room (e.g. Olynthus viii, pls. 28,95; xii .55–56, pls. 44, 234), and, rarely, in small isolated rooms (Olynthus viii, pls. 36, 97, house A vi 7).

¹⁷ Cf. Olynthus viii. 312 ff., pl. 77; Hesperia xx (1951) 180, fig. 6.

¹⁸ For similar loomweights, cf. Hesperia xviii (1949) 340, no. 123, pl. 101; Hesperia Suppl. vii 80 no. 4 (Type B); Olynthus ii. 118–28.

¹⁹ A common classical type; cf. Olynthus ii, fig. 295.

²⁰ The type extends from neolithic times onwards, alongside other and more complex forms of mill. Archaic and classical mill-stones, all of saddle-quern type, were found in the domestic parts of the Agora public buildings (Tholos area); see Hesperia Supp. iv (1940) 143–4, illustrating a late 5th-century-B.C. example. Similar examples were found at Olynthus (Olynthus ii. 69–71, figs. 180, 186; viii. 326, ff. pl. 79) and Delos (Délos xviii, Le Mobilier Délien 123 ff., pl. xlviii f.).

²¹ Especial thanks are due to Miss Christine Sapieha, who did the drawings and profiles in Figs. 4–11.

^21a It is likely that a small proportion of coarse sherds also is from the later period.

^21b We are most grateful to Sir John Beazley for confirming this attribution, as also for the references under no. 3, and for a number of other suggestions and corrections. We are indebted also to Dr. R. Lullies of Munich who examined the first tracings of the RF frs. and gave us preliminary advice.

²² See, however, under nos. 10 and 94.

²³ We are grateful to Miss Lucy Talcott for this suggestion; and also for the reference under no. 100 below.

²⁴ BSA lii (1957) 184.

²⁵ Thuc. ii. 19. 2.

²⁶ Perhaps, as noted earlier, the paucity of the finds at the site is evidence of the wholesale plundering of property by the Thebans, extending even to structural elements such as beams and tiles (Hellenica Oxyrhynchia xii. 4, cf. CP xxi (1926) 346–55). It is calculated (p. 113 below) that 720 Corinthian roof-tiles would be required for the restored pastas-block of the house, but only 600 fragments, including the smallest scraps, were found, and these gave on the ground an impression of a very thin scatter. The majority of the tiles have plainly disappeared, together, doubtless, with other re-usable structural features and furnishings— perhaps looted by the men who destroyed the house, or salvaged later.

²⁷ BSA lii (1957) 186, 189.

²⁸ For general discussion of domestic architecture see Rider, B. C., The Greek House (1916)Google Scholar; RE vii (1912) cols. 2523 ff.; RE Supplb. vii (1938) cols. 224 ff.; Robertson, D. S., Greek and Roman Architecture (1945) 297–302Google Scholar; Wycherley, R. E., How the Greeks built Cities (1949) 175–97Google Scholar; Dinsmoor, W. B., The Architecture of Ancient Greece (1950) 211, 252–3, 262–3, 322–5Google Scholar; Martin, R., L'Urbanisme dans la Grèce antique (1956) 220–52Google Scholar; Plommer, H., History of Architectural Development iGoogle Scholar, Ancient and Classical Architecture (1956) 114, 123, 202–5; and Lawrence, A. W., Greek Architecture (1957) 240 ff.Google Scholar

²⁹ Several authors imply the existence of fine country houses: Thucydides, ii. 65. 2 (describing popular reaction to the Peloponnesian war in 430 B.C.): Isocrates, Areopagiticus 52 (in the role of laudator temporis acti): And the Oxyrhynchus Historian (describing conditions in Attica before Theban and Spartan raids from Decelea in 413 B.C.; Ox. Pap. v (1908), no. 842, xii 5, with Bury's restored readings ibid. 230–1; cf. Hellenica Oxyrhynchia, ed. V. Bartoletti (Teubner, 1959) 25 on xviii (xii) 5)observes:

But no such house has yet been clearly identified. Known domestic buildings of classical date in Attica include the following: (a) some homesteads in South Attica characterized by sturdy round or square towers, probably humbler farming or industrial establishments, as they are without trace of large luxurious houses (Hesperia xxv (1956) 122–46); (b) a deme settlement at Draphi in the south foothills of Mt. Pentelikon, having houses terraced into the hillside, built of adobe on stone socles, and preserving both large fitted facing blocks with short picked striations and quadrangular blocks alternating with stackwork (cf. Hesperia xxviii (1959) pl. 20; Olynthus xii, pls. 12, 243); especially noteworthy is house A of the 4th-1st centuries B.C., measuring 7·5 × 12·5 m. (? an eastern wing of a larger complex), and comprising paved anteroom (6·3 × 2·1 m.), main room (6·3 × 4·0 m.) containing two stone bases for wooden columns and an oval hearth (c. 1·0 × 0·6 m.) ringed with a thin wall and provided with adjacent pithos-reservoir, and behind this room at a higher level, two smaller ones, one plain (4·4 × 2·5 m.), the other containing a corridor and bathroom with mosaic floor of tile-fragments, plastered cylindrical basin against the rear wall, and tile drains; adjacent, to the west, further 5th-century-B.C. structures with a wine-press and vat (BCH lxxx (1956) 246–8; lxxxi (1957) 515–17; lxxxii (1958) 681); (c) an isolated house in the Mesogaia, at Karellas to the right of the Liopesi–Markopoulo highway, and itself now covered by a branch road, still shows razed remains (socles, plaster floors, and stone column base) of its six northern rooms, 32 paces wide over all, lying parallel but not aligned at the rear, rather falling into three blocks set back progressively to the east and measuring (east–west, width and depth as far as the superimposed roadway, in paces): 6 × 9; 6 × 6, and 5 × 6; 3 × 3, 7 × 3, and 3·5 × 3 (excavated by A. D. Keramopoulos); mentioned, but not described, along with the partial publication of another country-house, of Roman and early Christian date, on the coast at Punta, in PAE 1919, 32: (i.e. ) (d) a large homestead (?) rt. of the Koropi–Vari road, showing surface traces of a large rectangular enclosure (c. 40 × 70 paces) with several enclosures and buildings on one side, concentrated in one corner, and a cistern in the opposite corner (cistern reported as Hellenistic; site unexcavated). We owe the references to sites (b)–(d) to the kindness of Professor Homer Thompson and Professor E. Vanderpool; (e) the ‘priests' house’ of 6th-century-B.C. date associated with the temple at Cape Zoster, Vouliagmeni, has a large enclosed court, with colonnades on two sides, and its main room on the south, fronted by a colonnade (AE 1938, 1–31). None of these sites parallels the Dema house fully.

³⁰ Pseudo-Dicaearchus, De Graeciae Urbibus frag. 1. 1; Demosthenes iii. 25–26, xiii. 29, xxiii. 207 (stressing the essentially modest nature of 5th-century-B.C. town houses, even those of eminent Athenians). Archaeological evidence comprises (a) small irregular houses on the fringe of the Agora, e.g. Simon the cobbler's shop, of late-5th-century B.C. date (Hesperia xxiii (1954) 51–55; Archaeology xiii (1960) 234 ff.); (b) a residential and industrial quarter south-west of the Agora, preserving house plans of irregular shapes and sizes, predetermined by previous occupation, in particular the complete plans of two houses (C, D) of pre-450-B.C. date; these, built of adobe, with earth floors throughout and tile roofs of several pitches, had a corridor-entrance, a small plain inner court, off-centre to the south, and various rooms of undifferentiated use; despite extensions c. 400 B.C. both remained plain, gaining as new amenities, the one only a single-columned verandah and pebble flooring in its court and the other pebble terrazzo floors in court and two rooms (? bathroom and shop) (Hesperia xx (1951) 135 ff. esp. 187–252); (c) comparable houses on the Areopagus slopes, also laid out pre-450 B.C., but more regularly as a block, c. 25 × 22 m., of four modest rectangular houses in pairs against a party wall, the eastern pair being square (11 × 11·10–11·40 m.); all had a small court, generally cobbled, and earth-floored rooms of undifferentiated functions, apart from one store-room with embedded pithoi, a possible andron with pitching as for a cement floor, and one certain example of a rudimentary portico, with a single central stone-based column (Hesperia xxviii (1959) 98–103); (d) irregular streets and houses near the Pnyx, preserving notably two rooms identifiable as dining-rooms (andron), one with a raised cement border, pebble-mosaic centre, and anteroom, the other similarly with a raised cement border, and dated to the 4th century B.C. (Judeich, W., Topographie von Athen (1931) 290, 299Google Scholar; Olynthus viii. 180–1); cf. another mid-4th-century-B.C. bordered example on the Pnyx, hill (Hesperia xii (1943) 312, 333)Google Scholar; (e) houses of the 5th and 4th centuries B.C. on the south slope of the Acropolis, the plans of which, on account of later disturbances, were not fully recoverable; one had rooms grouped round a court (Ergon 1957. 5–7; 1959. 157; BCH lxxx. 232; lxxxii. 657, 660–1; lxxxiv. 622–4); (ƒ) a classical house (4th century B.C.) lying north of the Olympieion; the incomplete plan suggests a large oblong with rooms along its north side (AJA lxiv (1960) 267–8; BCH lxxxiv (1960) 632–5); (g) a 5th-century-B.C. house recently located not far within the ancient walls, revealing part of its west side-wall and a peristyle with a column in situ (BCH lxxxiii (1959) 574). Doubtless there were more luxurious late-5th-century-B.C. town houses: cf. the house of the rich Callias, large enough for a school of sophists, with outer porch, inner court, a peristyle or at least two facing colonnades, one accommodating five strollers abreast (3–4 m. wide or more), and ground-floor store-rooms (Plato, , Protagoras 314–315 D).Google ScholarXenophon, (Symposium 1. 2)Google Scholar locates Callias' house in the Peiraeus. Certainly the new 5th-century-B.C. Hippodamian grid-planning of Peiraeus would have encouraged the building of spacious houses: ancient streets traced on Akte, perhaps belonging to a 4th-century extension, imply the existence of some insulae 40 m. across (Judeich, op. cit. 430) and some 60 m. across, accommodating an average of two houses in this width; two such houses, part planned, are large oblongs, wider than deep from front to rear, with courts on the sunny east-south-east side backed by a range of rooms next to the street (Milchhoefer, , Karten von Attika, Text i. 55–56Google Scholar (plan but no scale)). Probably closely related also to Athenian town-housing is the building block (c. 27 × 30 m.) within the Eleusis sanctuary walls, which includes a 5th-century-B.C. house and, next to it, a ‘state residence’, the Prytaneion—an oblong structure (c. 15·5 × 10·5 m.) with high socles and pebbled floors, inner vestibule, and a small southerly court surrounded on three sides by small rooms and probably roofed over in its northern half by a portico fronting two larger dining-rooms (4·5 m. square), each for seven couches (PAE 1955, 62 ff.; BCH lxxxi (1957) 509–11; AJA lx (1956) 267–8).

³¹ Sites reasonably claimed as farmhouses or the nuclei of rural estates include, for example, (a) the rural sites in South Attica, having enclosures, store-towers, grain-mills, and threshing-floors (Hesperia xxv (1956) 122–5, 133–43); (b) some comparable sites on Siphnos, again with round towers, enclosures, outbuildings, olive-presses (AJA lx (1956) 51–54); and (c) an isolated rural site at Delphinion, Chios, with house and outbuildings (BSA li (1956) 49–51).

³² For invaluable information about Olynthian and other Greek houses, cf. Robinson, D. M. and Graham, J. W., Excavations at Olynthus viii, The Hellenic House (1938)Google Scholar, and Robinson, D. M., Excavations at Olynthus xii, Domestic and Public Architecture (1946)Google Scholar, referred to in this article as Olynthus viii and Olynthus xii. Supplementary discussions in Hesperia xxii (1953) 196–207; xxiii (1954) 320–46; xxvii (1958) 318–23 BCH lxxx. (1956) 483–506.

³³ Wiegand, T. and Schrader, M., Priene (1904)Google Scholar; Schede, M., Die Ruinen von Priene (1934).Google Scholar

³⁴ Hesperia xiii (1944) 91 ff.

³⁵ Chamonard, J., Exploration archéologique de Délos viii (1922–1924)Google Scholar; BCH lvii (1933) 98–169; lxxvii (1953) 444–96.

³⁶ AM xxiv (1899) 458–67.

³⁷ ADelt i (1915) 124–31; Olynthus viii. 148–50.

³⁸ PAE 1914, 133–48; 1915, 237–44.

³⁹ Antiquity xxxv (1961) 91–102.

⁴⁰ Memorabilia iii. 8. 8–9 and Oeconomicus ix. 4. Cf. Aristotle Oec. 1. vi. 7 (1345a); Aeschylus PV 450–3; Eupolis 378 (Kock).

⁴¹ Olynthus viii. 142–6; fig. 4 on p. 99 shows a reconstruction model of a terrace of houses illustrating the advantage of this southerly aspect.

⁴² Wiegand, and Schrader, , Priene, figs. 298, 316, pl. 21.Google Scholar

⁴³ Olynthus viii. 150, fig. 7.

⁴⁴ Ibid. 149, fig. 6.

⁴⁵ Hesperia xiii (1944) pls. 9, 10.

⁴⁶ Antiquity xxxv (1961) fig. 2, pl. 10.

⁴⁷ Hesperia xx (1951) 204, fig. 11.

⁴⁸ Hesperia xxviii (1959) pl. 17.

⁴⁹ The Dema house (22·05 × 16·10 m.) would fall midway in a series of the oblong Olynthian houses arranged in order of size: Row A 1–5, 11–13 (to city wall 20·5–21·0 × 16·50–16·70 m.), South Villa (21·0 × 16·80 m.), House of the Comedian (?22 × 16·20 m.), Row A6 (20·50 × 22·0 m.), House of the Twin Erotes (22·45 × 17·15 m.), House Av6 (25·50 × 16·85 m.), Villa of Good Fortune (25·70 × 16·90 m.), House ESH1 (27·80 × 16·10 m.).

⁵⁰ The old Bouleuterion (late 6th century B.C.) measured 23·80 × 23·30 m. over its foundations and c. 23·30 × 22–76 m. over its walls, with the meeting hall proper measuring c. 22·76 × 16·40 m. externally and 21·26 × 15·70 m. internally. The New Bouleuterion (last quarter of the 5th century B.C.) measured 22·50 × 17·50 m. at foundation level, c. 21·50 × 16·90 over its walls, and 20·10 × 15·60 m. internally. Cf. Hesperia vi (1937) 128, 133, 142, pls. 6, 8; Hesperia Supp. iv (1940) 82, fig. 62.

⁵¹ The ideal plot, 60 (Ionic) feet square (17·70 × 17·70 m.) was reduced, owing to the requirements of streets and alleys, to actual building plots 58 feet square (17·05 × 17·05 m.) which would have an internal area (plot less wall thickness) of about 55 feet square (16·05 × 16·05 m.), with sides equivalent to the depth of the Dema house; the use of party walls instead of separate side walls increased the individual frontage very slightly. Cf. Olynthus viii. 29–36, esp. 33–34, 47. Variations in ground measurements (ibid. pls. 94, 95, 99) merely show that minor inexactitudes could occur without invalidating the general scheme or the architect's original conception.

⁵² The Olynthian houses exceeding the standard 60-foot plot were not necessarily of accidental or unconsidered dimensions. Measurements in Greek feet can be given to the houses listed in n. 49 above in accordance with the evidence for the use at Olynthus first of the shorter Ionic and then the longer Doric foot. Thus, Row A houses would measure 70 × 56 (Ionic) feet, but no. 6 be exceptional at 70 × 75 feet. The South Villa had walls centred at 70 × 55 (Ionic) feet. Terrace house Av6 was extended during construction (employing the longer foot) to an internal width of 75 (Doric), feet. Finally, two free-standing houses, the Villa of Good Fortune and ESH1 measured respectively 80 × 52 (50 internally) and 85 × 50 (Doric) feet. Cf. Olynthus viii. 15–17, 20, 27–28, 34–35, 45–51, 55, 69, 92, pls. 84, 89, 96, 105; Olynthus xii. 259–60.

⁵³ Priene insulae measured 47·2 × 35·4 m. or 160 × 120 Ionic feet. Most held varying numbers of houses of different sizes, but some were quartered, having four houses built up against party walls. These houses occupied therefore plots of 80 × 60 Ionic feet (23·6·7 m.); they measured internally about 75 × 55 feet (22·2 × 16·2 m.).

⁵⁴ Olynthus had blocks of 300 × 120, 240 × 120, and 120 × 120 (Ionic) feet; Priene, blocks of 160 × 120 (Ionic) feet; and Miletus, with two main but variable sizes, blocks of 60 × 70, 80 × 100, and 120 × 140 (Ionic) feet. The Peiraieus grid-plan is not known, but the ancient streets (Judeich, , Topographie von Athen 430, 451Google Scholar) suggest blocks with at least one side measuring 120 (Doric) feet (c. 40 m.).

⁵⁵ It is interesting, for example, (a) that the Maison de la Colline at Delos, though larger over all (18·89 × 18·56 m.), had an internal area equivalent to the ideal house-plot at Olynthus (17·7 m. square); (b) that the Pella house resembled the Dema house and Olynthian villas in size; (c) that at Seuthopolis the pastas house was of the same class (20 × 16 m., equivalent to the interior of the Olynthian South Villa) and the peristyle house was even larger (22 × 28 m.); and (d) that the early Athenian house-block on the Areopagus measured 22 × 25 m., so that its two quarter-block houses, at 11 m. square, had half the length and two-thirds the depth of the Dema house.

⁵⁶ Memorabilia iii. 1. 7.

⁵⁷ Some examples of adobe building in Attica are: the geometric period ‘Sacred House’ in the Academy (Ergon 1958, 5–7), the archaic houses and public offices on the west side of the Agora (Hesperia vi (1937) 17–18, 121; Hesperia Supp. iv (1940) 15–19, 34–36), the classical houses by the Agora, in the industrial quarter, on the slopes of the Areopagus, and in the deme centre at Draphi (described in nn. 29 and 30 above), and classical public buildings such as the law-courts (Hesperia xxiii (1954) 58–61) and the South Stoa I (ibid. 39–45).

⁵⁸ Hesperia vi (1937) 18 (42 cm.), 121 (50–53 cm.), 122 (40 cm.); Hesperia xx (1951) 171–2, 191, 193, 197, 202, 207, 239 (45–50 cm.); Olynthus viii 223–9 (45–50 cm.).

⁵⁹ Vitruvius ii, 8. 17; Olynthus viii. 214 ff., 227–8.

⁶⁰ Cf. Hesperia xx (1951) pls. 63b, 67a, 68d, 75c, d; xxviii (1959) 101, pl. 20; also, for earlier work, Hesperia vi (1937) 17, 119, 125; Hesperia Supp. iv (1940) 9, 19, 36.

⁶¹ Hesperia xx (1951) 240–1 draws attention to walls in a contemporary Athenian house, showing stylistic differences in work of the one period, caused less by structural needs than the whim of the masons.

⁶² Hesperia xx (1951) 177; xxviii (1959) 101.

⁶³ Both stucco (previous note) and clay or mud plastering are known to have been applied, at least sometimes, to adobe at Athens (Hesperia, Supp. iv. 9, 19; Hesperia vi (1937) 18; xxiii (1954) 58). Cf. Olynthus viii. 226–7, 291 ff.

⁶⁴ Thuc. iii. 20.

⁶⁵ Demosthenes iii. 29; xxiii. 208; IG ii². 1672, ll. 83–84, 203.

⁶⁶ Olynthus viii. 158, 281 ff.

⁶⁷ e.g. Draphi and the Karellas houses, n. 29 above, sites (b) and (c).

⁶⁸ For the houses, n. 30 (a)–(d) above and Hesperia xx (1951) 191, 205, 208, 215–16, 218, 229; xxviii (1959) 101; for earth-floored archaic public buildings, Hesperia Supp. iv (1940) 5, 8, 12, 15, 17, 19–20, 24, 34, 36, 39; and for 5th-century public buildings ibid. 59 (Tholos), Hesperia xxiii (1954) 39–45 (South Stoa, with androns all clay floored), 58 (law courts); xx (1951) 172 (Poros Building).

⁶⁹ Olynthus viii. 165–6, 239–48.

⁷⁰ Column bases occur at the rural Attic sites Draphi, Karellas, and Vouliagmeni (n. 29 above) and, in the city, in the archaic Prytaneion complex (Hesperia Supp iv (1940) 15–17, 20–21, 23, 36) and in houses of the industrial quarter Hesperia xx (1951) 205, 282, pl. 71a) and on the Areopagus (Hesperia xxviii (1959) 100).

⁷¹ The published Athenian column bases (previous note) had diameters of 30, 35, 36, and 40 cm. in various parts of the Prytaneion complexand 30 cm. in the houses. The few columns noted in Olynthian houses have base diameters of up to 35 cm. with two of 40 cm., and capital diameters of 13–35 cm.; these sizes agree well with the average pillar sizes (40 × 25 cm. maximum). The wooden shafts may have been c. 2 m. high (cf. two surviving monolith pillars, each 1·82 m. high). The occurrence of columns along with stair-bases in some houses showed that these, no less than pillars, might support galleries overhead. Cf. Olynthus viii. 71, 74, 80–82, 109–10, 119, 166, 240–1, 247–8, pls. 51, 58, 66–68, 88, 89, 97, 100; xii. 40–44, 236, 241, 244–5, 293, pls. 26, 34, 38, 202, 209–12, 243.

⁷² Corinthian-type terracotta tiles roofed the imposing secular buildings at Athens in classical and hellenistic times, e.g. the Stoa of Zeus, Philo's Arsenal, the Metroon, the Odeion, the Middle Stoa, and the Stoa of Attalus (IG ii². 1668, l. 58; Hesperia vi (1937) 36–37, 54, 191–2; xix (1950) 50–55). Laconian tiles, general in Athenian houses, were used also in some utilitarian public constructions, e.g. the Parabyston law courts, the Poros Building, the Tholos kitchen (2nd period), and the coping of the city walls (Hesperia xxiii (1954) 60–61; ibid. xx (1951) 172–4; Hesperia Supp. iv (1940) 75, 78–79, 81; IG ii.² 463, 1. 69). Contemporaneously at Olynthus most houses used the narrow Laconian rain-tile exclusively, the more prosperous used the more regular Laconian system of broad and narrow tiles, but only the public buildings and a very few houses used Corinthian ones, and the latter only to edge their eaves (Olynthus viii. 232–4; xii. 90, 186, 242–3). The more common use of Corinthian tiles for houses at Delos and Priene is later in date. The occasional use of Corinthian tiles in classical Attic houses is indicated by their inclusion in the list of the confiscated properties of the Hermocopids (Pollux, , Onomasticon x. 157, 182Google Scholar; Hesperia xxii (1953) 282; xxv (1956) 283–4, 318–19) and their cost, two or three times greater than that of Laconian tiles, is clear from the Eleusinian state contracts (IG ii². 1672, ll. 71–72, 188).

⁷³ The open area VIII, at 11·80 × 10 m., is 118 sq. m. in extent or 33 per cent. of the total area of the house (355 sq. m.). This is far larger than the proportion noted elsewhere (Olynthus viii. 157, with percentages of 5–12 per cent.; cf. Priene house 33 with 18 per cent.). A pastas with the depth of room VI would be c. 3·80 deep within the colonnade (e.g. the norm at Olynthus was 3–4 m., ibid. 163) and would cut down the open court to c. 65 sq. m. or 18·2 per cent. of the total area.

⁷⁴ If there had existed the one column alone, it could only have been a prop for (1) a small portico, 2·5 × 3 m., in the south-west angle of area X, (2) a portico, 2·5 × 4·5 m., in the north-west quarter of X, or (3) a portico, 2·5 × 7 m., along the west wall of X. But the first would be a ridiculously small portico in an already shaded corner on the north side of a room, having no apparent special function to justify it, such as roofing an altar or hearth (Olynthus viii. 321 ff., Hesperia xxii (1953) 196–8; xx (1951) 222); the second would be, besides ungainly, structurally unsound, as the column is in line with the doorway of room I and the weight of the architrave would bear down on the door lintel, not on solid wall; finally, the third combines this structural weakness with an improbable off-centre position for its sole support (cf. Hesperia xxviii (1959) 100, pls. 16, 17). As the one column is exactly half-way between the west wall of the house and the projection of the east walls of rooms I and IX, it is not overbold as a second step to restore another column at 2·5 m. east of the first. These two columns allow of (1) a portico, 3 × 5 m., against the north wall of room IX, (2) a portico, 4·5 × 5 m., in front of room I, or (3) a portico, 5 × 7 m., covering the whole of area X. The objection to these arrangements are that the first, while not structurally impossible, is improbable for the same reasons as (1) above, being a small unnecessary north-side penthouse protecting nothing; the second is an unbalanced arrangement, having an interval column on the south and none on the east; and the third is unsatisfactory because a portico planned merely to cover X would have its two columns set centrally between rooms I and IX or, better, side by side on the east front of X to form a distyle portico. The position of the one base is, however, perfectly intelligible if it is regarded as the survivor of a row of columns extending to the corner of room VI. Measured from its centre to the centres of the walls, the base is distant 2·5 m. from the west house wall and 15 m, from the west wall of VI, or 6 × 2·5 m. Five bases may be restored to the east of it, giving six columns in all with inter-axial measurements of 2·5 m. (cf. Olynthus viii. 165 for the spacing of pastas pillars at Olynthus, which averaged 2·20–2·75 m.). This east–west colonnade would be logically completed by a return southwards from the second column to the north-east corner of room IX, a distance of just over an intercolumniation; this would cover an otherwise uselessly small open area and draw an otherwise isolated corner room into an organic relation to the northern rooms, suggested in the first place by the common alignment of its east wall with that of room I.

⁷⁵ Olynthus viii. 227–8.

⁷⁶ Olynthus viii. 151; Hesperia xx (1951) 226–8; ibid. xxviii (1959) 103.

⁷⁷ Olynthus viii. 167–9, 171–85; Hesperia xxii (1953) 199–203.

⁷⁸ Olynthus viii. 185–204; xii. 369–98; Hesperia xxiii (1954) 328–46; xxvii (1958) 318–23.

⁷⁹ The significant points are (1) the door off-centre to the left, (2) the length of the right-hand wall, 180 cm., just sufficient to fit a couch against it, and (3) the size and shape of the room. Diners reclined on their left sides on couches set end to end along the walls with the foot end of the first couch in each row set into a corner and the head of the last couch in the adjacent row abutting against its right side. In the average andron, 4·5–5 m. square (such as Dema House room I, most of the Olynthian androns, and the rooms in South Stoa I and the Eleusis Prytaneion), the side and rear walls would accommodate two couches in a row and the front wall could take one couch only, extending from the right-hand corner up towards the door, but only if the doorway were set to the left to allow 180 cm. at least to the right-hand wall. For illustrations, see Hesperia xxiii (1954) 39–45, fig. 4; Travlos, J., Πολεοδομικὴ Ἐξέλιξις τῶν Ἀθηνῶν (1960) 63, figs. 29, 30 (South Stoa I)Google Scholar; PAE 1955, 63, fig. 1 (Prytaneion at Eleusis). Some few Olynthian androns (e.g. the Villa of Good Fortune) have doors off-centre to the right, but these were large enough to take two couches against the front walls, the right-hand couch running up from the corner to the door, the left-hand one running from the door to abut against the left-hand side row, i.e. the right-hand wall would have to be at least 180 cm. long and the left-hand wall at least 180+80 cm. or 260 cm.

⁸⁰ The Villa of Good Fortune (Fig. 12) has two rooms, one large and one small, next to the andron, earth-floored but without trace of fires and cooking. But if the kitchen identification is rejected for this reason, the Villa is robbed of its only claimant, and surely it had a kitchen.

⁸¹ Olynthus viii. 186–8.

⁸² See n. 29 above; BCH lxxxi (1957) 515, fig. 12.

⁸³ Antiquity xxxv (1961) 98, pl. 11a.

⁸⁴ Cf. PAE 1914, 114–15, fig. 11; Olynthus viii. 186, pl. 52.

⁸⁵ Cf. Olynthus xii. 377.

⁸⁶ It might be argued that a continuous partition between rooms V and VI would require slighter or no foundations, being an interior wall not even indirectly exposed to the damp, and that the requirements of the roof were no counter-argument because the eaves of the gable roof would rest both to north and south on solid walls rather than on wall to north and colonnade to south. But Olynthian and Athenian practice, while admitting rather slighter socles, does not omit foundations altogether.

⁸⁷ Olynthus viii. 152–6, 249–63. At Olynthus, Row A1, 2, 3, 6, 8, 9, and 10 have rather similar rooms inside the entrance, regarded as roofed forecourts rather than vestibules proper; ibid. 78, 153, pls. 88–90.

⁸⁸ Olynthus viii, pl. 97, house A vi 7.

⁸⁹ Cf. p. 102, n. 29 above, site (d).

⁹⁰ Olynthus viii. 214–19, 267–80; Hesperia xxiii (1954) 321–8.

⁹¹ See Olynthus viii. 227–8. The pastas-block, 10 m. wide or about 35 ft., could by that formula attain a height of 25 ft. with walls less than 0·45–0·50 m. thick.

⁹² Plato Prt 315d describes the use in an emergency of a ground-floor store-room as a bedroom, implying clearly a makeshift arrangement. Aristophanes Th. 478–89 and Lysias i. 9–10 indicate that bedrooms upstairs were normal.

⁹³ A possible interpretation of the kerbing and the displaced slab in room VI is that they are part of a staircase, the former representing perhaps the limit of an adobe-built landing for a two-flight staircase and the slab a stairbase, slighter than the Olynthian average but, at 0·85 m., long enough; it lacks those cuttings which, being found on a few examples, identified all stairbases at Olynthus. This identification, however, lacks conclusive evidence. Olynthian staircases were usually located in the pastas or in the court (Olynthus viii. 271), but sometimes within a kitchen or utility room (e.g. houses A v 9, A vi 5; ibid. 96, pls. 95, 97; Hesperia xxiii (1954) 327–8, fig. 1). Menander describes what would be a parallel to this tentative suggestion of a staircase in room V–VI, namely an Athenian house with a double room, the antechamber containing the staircase and women's looms and the inner chamber being a store-room, Samia (Loeb, 1921) ll. 17–21:

⁹⁴ Cf. Antiphon, i. 14; Aristophanes, Eccl. 607–9; frag. 133 (K); Aristotle, , Ath. Pol. 50, 2Google Scholar; Oec. ii. 2, 4 (1347a); Ponticus, Heraclides, De rebus publicis i. 10Google Scholar (FHG).

⁹⁵ Olynthus, viii. 278–9 notes two stone pillars 1·82 m. high and suggests columns c. 2 m. high on bases of 0·35 m. diameter, or a total ceiling height of 2·25–2·50 m. Cf. the slimmer stone columns in Delian houses, e.g. the house of Hermes with proportions of 1:8 in both ground-floor and upper-floor columns, BCH lxxvii (1953) 457, fig. 6, 462 ff., 480–1, fig. 13.

⁹⁶ Lysias, i. 9 and 23 mentions first this duplication of room area, and then in particular a case of dining upstairs. Here the master of the house had temporarily moved his quarters upstairs, interchanging with the women; probably there was no regular dining-room but a room usable as an andron, perhaps over the downstairs one.

⁹⁷ The roof pitch of the Stoa of Zeus was calculated to be c. 12° (Hesperia vi (1937) 36); this is very close to that of the south-east angle tower of Aegosthena, which has its gableends preserved (cf. Plommer, , Ancient and Classical Architecture 197Google Scholar). The Olynthian roofs were restored with a pitch of c. 18° (Olynthus viii. 236–8). The length of the Dema house roof from east to west would be c. 22·40 m., and the length of its slope from eaves to gable at least 5·50 m. The effective size of each rain-tile was 56 cm. in width × 62 cm. in length (allowing for 6 cm. overlaps), so that 40 tiles would be needed to cover the length and 9 the slope, or 360 tiles for the north rooms and 720 for the whole pastas-block. if it was a plain rectangular roof. These figures represent the minimum needed to roof to the edge of the house walls. Over hanging eaves (protegismata), common in Greek houses and very likely a feature here also, might necessitate extra rows of tiles front and rear, according to the projection. The horizontal extent of pastas eaves at one Olynthian house (A 10), calculated from sheltered uncobbled areas below, would be c. 1.25 m. in front and perhaps up to 2 m. at the rear (Olynthus viii. 238, pls. 88, 90). Possibly at least one extra row may be restored in the Dema house, thus making the total for the north slope 10 rows of 40, or 400 tiles.

⁹⁸ BSA lii (1957) 161.

⁹⁹ Statius, , Theb. xii. 618–21.Google Scholar

¹⁰⁰ See BSA lii (1957) pl. 30 for terrace lines visible on Aigaleos, c. 500 m. south and south-east of the Dema house site; cf. the evidence of intensive cultivation and denser population on the slopes of Hymettus, Bradford, J., Ancient Landscapes (1957) 29–34, pls. 7–10Google Scholar; AntJ (1956) 172–81. Cf. also for an ancient description of erosion in Greece, Plato, Cri. 111 b–d.