Hostname: page-component-76dd75c94c-28gj6 Total loading time: 0 Render date: 2024-04-30T08:52:10.778Z Has data issue: false hasContentIssue false

Perachora: the Remains outside the Two Sanctuaries

Published online by Cambridge University Press:  27 September 2013

Extract

In the first chapter of Perachora i, Humfry Payne gave a brief survey of the Perachora peninsula, and of his own excavations. There he distinguished between the area of the town, situated in the plain that lies between Lake Vouliagmeni and the tip of the promontory, and the ‘Heraion Valley’ whose buildings were almost wholly of a public nature. His description of the town envisaged further excavation; but his own activities were concentrated in the area of public buildings, the two sanctuaries of Hera Akraia by the harbour and of Hera Limenia in the Heraion valley itself.

The two volumes of Perachora are concerned with the discoveries Payne made in these two sanctuaries. Omitted from them are the other public buildings in or adjacent to the sanctuaries. These consist of the angled stoa, the so-called ‘agora’, the double-apsidal cistern, and the hestiatorion or dining-hall. Also omitted is the detailed study of the town which he promised. The stoa and ‘agora’ (which is now to be renamed ‘the west court’, since, whatever its actual function, it was certainly not an agora) have now been published separately by Dr. J. J. Coulton. The present account gathers together the remaining public buildings in the vicinity of the sanctuaries, the apsidal cistern and the hestiatorion, together with the ancient remains in the area of the town.

Type
Research Article
Copyright
Copyright © The Council, British School at Athens 1969

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

1 See also Dunbabin's preface to Perachora ii.

2 BSA lix (1964) 100 f. (stoa); BSA lxii (1967) 353 f. (west court).

3 Where I use Dunbabin I refer, unless it is stated other-wise, to his ‘draft chapter’. This is the typescript version, not the preliminary essay in the field notebook. I did not see the draft chapter and the notebook until 1965, that is, after the season devoted to cleaning and studying the houses. The section on the fortifications is based on the work of Dunbabin.

4 See pp. 164 ff.

5 The positions of the double-apsidal cistern, the hestiatorion, and the approximate line of the drain are marked on the general plan of the Heraion promontory, Perachora i, pl. 137.

6 Similar square piers, (but in four rows of five) existed in a rock-cut cistern in the sanctuary of Apollo at Argos (Vollgraff, W., Le Sanctuaire d'Apollon Pythéen à Argos, pl. iii).Google Scholar For the comparison with Perachora, cf. Roux, , L'Architecture 69.Google Scholar The storage chamber behind the ‘fountain-house of Theagenes’ at Megara (probably to be dated to the fifth century B.C.) had a roof supported by octagonal piers, with corresponding capitals. These also taper noticeably towards the top. They are made of several sections joined by empalia, and were originally 5·20 m. high, though the maximum depth of water in the reservoir was only about 1·40 m. (Graben, G., ADelt (1964) A 37 ff.Google Scholar)

7 The lengths of which are normally neither constant nor equal, except in those piers which snpport cross-beams.

8 It is not clear whether in the piers without cross-beams the two upright sections were dowelled to each other.

9 This technique seems to be without parallel. Could it be derived from methods of ship-construction?

10 Stone slabs appear to have covered the great cistern at New Pleuron (personal observation), but they rested on walls constructed across the width of the tank. At Argos (Roux, loc. cit.) remains of a pebble cement floor over the cistern have been found. This was laid on stone slabs. Roux suggests the slabs were supported by stone beams, though none of these has been found. There the span between the supports is only about 1 metre, compared with 2·1 m. at Perachora. At Megara (Gruben, loc. cit.) where the span is about 2·3 m. no trace of the roof or its support above the level of the capitals was found. Gruben suggests beams supporting an earth roof (‘auf einer Balkenlage ein flaches Erddach’) because no tile fragments were found.

11 ‘Cistern north of steps’ in the section, Perachora i, pl. 140.

12 For this reason (and others; see below p. 171) I cannot accept Dunbabin, 's suggestion (BSA xlvi (1951) 68)Google Scholar that the cistern is a replacement of the archaic sacred pool.

13 In the section Perachora i, pl. 140 the drain overlies the deposit over the Protocorinthian level, and is sunk into the fifth-century level, but this stratification does not do more than indicate a terminus post quem.

14 For this type of construction compare the western out-flow drain of the stadium at Olympia (Kunze, , Olympia Bericht v. 19 f.).Google Scholar The essential purpose of these drains is the rapid removal of heavy rainfall in enclosed collecting areas such as a theatre or stadium, or, as here, the Heraion valley.

15 Payne clearly regarded the cistern as Hellenistic (Perachora i. 26), but he does not give his reasons.

16 Roux, (L'Architecture 69)Google Scholar, apropos the pier cistern at Argos (n. 6 above) says: ‘La cîterne à piliers n'apparaît pas, à ma connaissance, avant l'époque hellénistique; encore la trouve-t-on rarement.’ The ‘fountain of Theagenes’, though earlier (fifth century B.C.?) is not really an exception, since there the piers support a roof considerably higher than the water level (Gruben, loc. cit.).

17 Coulton, , BSA lix (1964) 124 f.Google Scholar; lxii (1967) 368.

18 Martin, R., Manuel d'Architecture grecque i. 260.Google Scholar

19 ‘The sanctuary of Asklepios and Hygieia at Corinth’, AJA xxxvii (1933) 432 n. 1.

20 Perachora i. 14.

21 BSA xlvi (1951) 61 f.

22 See below, pp. 164 ff.

23 On the foundation block at the north-west corner of the west room, there is no trace of the foundations ever having extended further to the north; but this is not conclusive, since it was not necessary that the block here be carefully dressed to take another block to the north, and the existing surface of the stone is badly damaged.

24 Many of the rooms similar in plan and purpose to our main rooms (described below, p. 169), are approached through open colonnades, usually from a peristyle court. The dining-rooms in the Sanctuary of Asklepios on the south side of the Athenian Acropolis (for the identification of these as dining-rooms, see my article JHS lxxxix (1969) 106 ff.) afford the closest parallel. These comprise a line of rooms with a colonnaded vestibule in front, but the colonnade is not part of the original structure, which may have been even closer in arrangement to the hestiatorion at Perachora, with the side walls forming returns across the front.

25 At one point only, at the northern end of the east wall, does the original surface of the surviving blocks seem to have been deliberately trimmed to carry a polygonal block.

26 Hodge, A. T., The Woodwork of Greek Roofs 50.Google Scholar

27 Windows exist to either side of the door and in the side walls of the andron of Idrieus at Labraunda. IG xi. 154, A4, from Delos, refers to ἑστιατόρια (of which the precise location is not mentioned in the surviving part of the inscription) with both doors and windows: τὰς θύρας καὶ τὰς θυρίδας.

28 I cannot accept Dunbabin's suggestion (loc. cit. 68) that this building, together with the cistern, formed part of the μαντεῐον as such, since it is of a type found also (and, indeed, predominantly) in non-oracular shrines.

29 The floor in both doorways is destroyed. The plan which survives from the original excavation marks post-holes in the doorways, but I could not distinguish these.

30 The evidence from other similar rooms, listed below, suggests that the doors opened inwards in the normal way.

31 Equivalent to the πόδες and ἐπίκλιντρα of the couches at the Asklepieion of Delos, , IG xi. 144, 66.Google Scholar See below, p. 170.

32 Roebuck, C., Corinth xivGoogle Scholar, ‘The Asklepieion and Lerna’ 51 f.

33 Roebuck loc. cit. 54.

34 See my article ‘Two buildings in Sanctuaries of Asklepios’, JHS lxxxix (1969) 106 f. The possibility (n. 24) has been mentioned above that the original plan of the building at Athens was similar to that of the hestiatorion at Perachora.

35 For these, see also Frickenhaus, , JdI xxxii (1917) 114.Google Scholar

36 Waldstein, , The Argive Heraeum i. 131 f.Google Scholar (Tilton); see also Frickenhaus's reinterpretation of this building, loc. cit. 121.

37 Broneer, , Hesperia xxxi (1962) 7.Google Scholar The ‘rock’ is a marl like that found in the underground parts of the waterworks at Perachora: see below, p. 200.

38 Frickenhaus, loc. cit. 114, and Welter, G., Troizen und Kalaureia 31 f.Google Scholar

39 Furtwängler, , Aegina, Das Heiligtum der Aphaia 107 (Fiechter) and pl. 70.Google Scholar

40 Robert, F., Délos xxGoogle Scholar, ‘Trois sanctuaries sur le rivage occidental’ 51 f. and esp. 64.

41 άνδρών is also used for the formal dining-rooms of the Hecatomnid dynasty in the sanctuary of Zeus at Labraunda. Cf. also Aristophanes, Eccl. 676: τὰ δικαστήρια καὶ τὰς στοιὰς ἀνδρῶνας πάντα ποιήσω (quoted by H. A. Thompson, with reference to the South Stoa I of the Agora, Athenian, Hesperia xxiii (1954) 43).Google Scholar

42 For the date, cf. F. Robert, op. cit. 99 f.

43 Several other Delian inscriptions mention ἑστιατόρια: IG xi. 154, A4, several ἑστιατόρια of unknown locality; IG xi. 161, A114, in the ‘Artemision on the island’; and, if the restoration is accepted, IG xi. 165, 43.

44 Payne seems to have regarded it as Hellenistic along with the apsidal cistern, but again without giving any reason (Perachora i. 26).

45 Coulton, , BSA lxii (1967) 359.Google Scholar

46 Roebuck, op. cit., ‘Chronology’ 173.

47 Olynthus viii, ‘The Hellenic House’ 185.

48 This is certainly true of the Asklepieion at Corinth, and, probably, of that at Troizen (where there are careful arrangements to drain off water). It also applies to the cult caves at Isthmia. At the Argive Heraion no mention of water supplies is made by Tilton, but this cannot be regarded as decisive. Draw-basins and bath buildings in the immediate vicinity suggest that water was once available in some quantity there. Epidaurus is a well-watered spot. The peristyle building in the sanctuary of Apollo at Argos has, as noted above (n. 6) an elaborate cistern; Roux (loc. cit.) identifies this building as a sanctuary of Asklepios, and it may have included hestiatoria among its rooms.

49 By Mr. Charles K. Williams, Assistant Director of the American School at Athens.

50 The plain of the Heraion promontory runs westwards from the lake, at first with a marked gradient, then level, then climbing by a steep but low bank to another level area. It is this area which we call the ‘Upper Plain’.

51 In this valley walls, probably of terraces, are marked on the plan Perachora i, pl. 137. Dunbabin in his draft chapter records that the Heraion valley was trenched during the original excavations, but the remains of only one building were found: ‘about 100 yards east of the temenos wall (i.e. of Hera Limenia), on the south side of the valley, is a nort–south wall, returning at each end towards the east; its interior measurement is 5·70 m., and it consists of one course of foundations, of large squared stones carefully laid. Inside the space enclosed was found a sherd from a Droop or Cassel Cup. So this building must be at least as old as the second half of the sixth century.’ Dunbabin believes that the valley was sacred ground, and that ‘the single building in the valley may well have been for the keeper of the sacred herds or the tiller of the sacred soil’. We re-examined this building in 1967. At present it seems that the ‘return’ at the south end is to the west, not the east, while that at the other end could simply be a stone dislodged from its original position (its upper surface has a cutting which suggests that another stone abutted against it, so that originally it must have been a vertical surface). It would be surprising if this wall is any earlier than 400 B.C., and the cup sherd referred to by Dunbabin can hardly be more than a fortuitous find. One must agree that the Heraion valley was most likely to have been undeveloped sacred ground.

52 The existence of this older road had already been postulated by Dunbabin in his draft chapter. There are cuttings in the rock above the western end of the lake, higher than those described by Payne, (Perachora i. 10Google Scholar;, which may be part of an earlier road.

53 See below p. 212.

54 See p. 203.

55 Broneer, O., Hesperia xxviii (1959) 306.Google Scholar

56 Personal observation; cf. Cassels, , BSR xxiii (1955) 14 and pl. viiiaGoogle Scholar, though the blocks used at Perachora are by no means as regular as those of this tomb.

57 Similar grooves appear to be represented on the spit carried by the two obeliaphoroi on a polychrome oinochoe from the Athenian Agora; see Crosby, Margaret, ‘Five Comic Scenes from Athens’, Hesperia xxiv (1955) 76CrossRefGoogle Scholar, pl. 35b and c.

58 If they had formed an outer wall they would have had to face north.

59 Cf. Robinson, and Graham, , Olynthus viiiGoogle Scholar, ‘The Hellenic House’ 171 f. This example at Perachora is considerably older than any other example known. Robinson and Graham, on the evidence then available, considered that the type probably developed in the fifth century B.C.

60 Dunbabin, Field Notebook.

61 The identity of A I with this building depends on Dunbabin's Field Notes. The other building mentioned in JHS lix (1939) 194 is house XIV.

62 Cf. Perachora i, fig. 5.

63 Perachora i. 11.

64 See below p. 218.

65 Recorded in his Field Notebook. The part where he had excavated was still plainly visible in 1966.

66 The west end has fallen away, and is in a dangerous condition.

67 It is also possible that the powdered rock once extended up to the wooden beams which surrounded the top of the shaft, and has either fallen or been dug away.

68 The position of these niches is shown on the plan (Fig. 16) and on the west side of the staircase in the section (Fig. 20).

69 The exact levels are impossible to determine without clearance of the tunnels. It does not appear that the tunnels are absolutely level; they seem to decline away from the junction. The ceiling of the section from the west shaft towards the well-shaft, which we could compare with the water level, definitely does slope down; though, as can be seen in the photograph, this slope is not very pronounced.

70 A similar runnel exists in the temenos of Poseidon at Isthmia. An earlier runnel there appears to have been simply cut in the clay and lined with plaster; rounded edges indicate that it was not covered. A later loop was constructed with fragments of roof tiles taken from the classical temple (damaged by fire in 394 B.C.), and also covered with them. See Broneer, O., Hesperia xxviii (1959) 306.Google Scholar

71 For these, see below p. 207.

72 For the technique of removing a soft layer from beneath a harder cap which is left to form a roof, cf. the fountain of Lerna at Corinth (Roebuck, , Corinth xiv, 100Google Scholar, the corridor).

73 The remains of this road which were still visible in the 1930s are shown on Payne, 's general plan, Perachora i, pl. 137.Google Scholar West of the fountain house this line has been incorporated into the modern road. In 1966 we were able to draw and photograph some of the ancient ruts then still visible before they were damaged by a bulldozer re-levelling the road. In 1968 the modern road was remade, and the ruts completely buried.

74 Compare the similar parapet in front of the drawbasins of the ‘fountain of Theagenes’ at Megara (Gruben, , ADelt 19 (1964) A 39 and plan 1Google Scholar). There also, the floor of the fountain house façade was higher than that of the drawbasins and storage chamber behind.

75 Cf. the similar technique in the temple of Asklepios, Corinth: Roebuck, , Corinth xiv, 30.Google Scholar

76 Cf. Roux, , L'Architecture, chapter xi, 333.Google Scholar

77 Broneer, O., Corinth i. ivGoogle Scholar, ‘The South Stoa’ 46.

78 Similarly the very slender proportions of the north porch columns of the Erechtheion (9·52 diameters) are caused by special circumstances, the positioning and level of the porch.

79 Cook, and Plommer, , The Sanctuary of Hemithea at Kastabos 89.Google Scholar Though the heyday of this sanctuary, i.e. the Hellenistic age, is later than that of the Perachora Heraion, it forms in many ways a useful comparison.

80 Roux op. cit. 345.

81 Cf. the fifth-century capital from the Athenian Agora, now on display in the Agora Museum.

82 See below.

83 Our impression is that pent roofs are normal in fountain houses set against vertical surfaces, whether natural rock or wall, as so many of them are. This point is not adequately discussed by Dunkley, B. (BSA xxxvi, 142).Google Scholar

84 BSA lix (1964) 100 f. See also Llinas, C., BCH lxxxix (1965) 484 f.CrossRefGoogle Scholar

85 Coulton, op. cit. fig. 7.

86 For general comment on Peloponnesian Ionic capitals see Roux, , L'Architecture 339 f.Google Scholar He defines this form of the echinus as one of the peculiarities of his Peloponnesian type.

87 There is also a bottle cistern on the acropolis summit by the lighthouse (see below p. 241).

88 Mr. Megaw tells me that he was able to enter the storage chamber. The section to the left (south) of the entrance tunnel was short. That to the right (north) was much longer, and curved away to the west, giving an impression that it was once a tunnel. He suggests that it had been blocked off subsequently, and plastered to serve as a cistern.

89 This suggestion was made by Mr. Megaw. Qanats are described by Wulff, H. E. in his article ‘The Qanats of Iran’ in Scientific American, April 1968, 94 f.CrossRefGoogle Scholar

90 Roebuck, Corinth xiv, loc. cit.

91 Hill, B.Hodge, Corinth i. vi.Google Scholar Compare also the fountain of Lerna at Corinth (Roebuck, op. cit.), where Roebuck suggests that water was collected on the surface, and entered the underground supply tunnels through vertical shafts. But this system differs from ours, since at Corinth the supply tunnels are deep underground, whereas the Perachora runnel is close to the surface, and was obviously closed for its entire length, so that water could only enter it at its beginning, near the deep shafts.

92 Vitruvius x. 4–7. A description of wheels and Archimedean screws from Spain is given by Palmer, R. E., Transactions of the Institution of Mining and Metallurgy xxxvi (19261927) 299Google Scholar (TIMM). This forms the basis of the account in Forbes, R. J., Studies in Ancient Technology vii. 211Google Scholar (henceforward Studies vii; cf. also Studies ii) which, however, includes some serious errors (see below). There is a brief account of a series of Archimedean screws, wheels, a bucket chain, and a force-pump in an article by Gossé, G., Ampurias iv (1942) 50 f.Google Scholar See also A. G. Drachmann. The Mechanical Technology of Greek and Roman Antiquity.

93 Studies vii. 214.

94 The text is that of Fensterbusch (Darmstadt, 1964 ).

95 Boon, G. C. and Williams, C., JRS lvi (1966) 122 f.Google Scholar

96 This photograph was reproduced by Forbes, who erroneously describes it as a wheel ‘used by the Romans’, not a modern copy.

97 Forbes, , Studies vii. 216Google Scholar, misquotes this as 13½ lb., through 12 feet per minute, a most misleading error.

98 The meaning of ‘duplex catena’ is uncertain. Fensterbusch: ‘Unklar bleibt, ob die Kette duplex genannt wird, weil sie bis zum Wasserniveau hinunter und wieder bis zur Welle emporreicht, oder ob damit 2 nebeneinander laufende Ketten gemeint sind. Oder ‘doppelt herumgeschlungen?’ It is preferable to think of two separate chains, attached to either side of the rectangular buckets or containers, corresponding to the two sides of the drum. ‘A pair of’ seems a legitimate translation of ‘duplex’.

99 Riad, H., Archaeology xviiGoogle Scholar, no. 3 (Autumn, 1964). I am most deeply indebted to my colleague Mr. R. F. Willetts, who saw the original in the Graeco-Roman Museum at Alexandria, for drawing my attention to the existence of this wall-painting.

100 Hero, , Dioptra 34.Google Scholar This point is made by Drachmann, op. cit. 202, who also refers to a figure (in the manuscript) illustrating Hero, , Pneumatics ii. 32.Google Scholar

101 Op. cit. 200 f.

102 Studies ii. 32.

103 The tentative restoration (Fig. 26) gives essentially a simplified and schematic version of what the arrangements may have been.

104 In the modern, but still primitive version, (i.e. the sakia or Persian wheel), the water containers—earthenware pots—are attached to ropes suspended from the wheels, but there are no wheels at the bottom. ‘The only supply of water was a muddy fluid brought up from the river by Persian wheels with hanging earthernware jars that were worked by circling oxen’ (Sir Samuel Baker, quoted by Alan Moorehead in The White Nile, of Khartoum in the 1860s). But the height to which these wheels raise water is modest, compared with Perachora. There the dimensions of the shaft suggest that the upward and downward sections of the chains were constantly spaced an even distance from each other, and this suggests that they were kept apart at the bottom as well as at the top.

105 In late spring 1933 when Kenny investigated the underground structures there was no water in the system at all. In late summer 1966, when the investigations here published were carried out, the western tunnel held water, and while we were working in it underground we could hear the occasional and intermittent dripping of water from the roof. The spring of that year had been particularly wet. By Feb. 1967, after a summer when there was some rainfall, and a wet winter, the water level appeared to have risen.

106 TIMM loc. cit.

107 A figure based on the table in Forbes, , Studies ii. 83.Google Scholar

108 Coulton, BSA lxii (1967) 369.Google Scholar

109 Demetrius’ raid on Babylon: Diodorus xix. 100. The Hanging Gardens: Strabo xvi. i. 5; Diodorus ii. 10. Strabo calls the water-lifting machines κοχλίας Diodorus ὄργανα. Both authors say they raised water from the Euphrates, but Diodorus says the machines were concealed μηδενὸς τῶν ἔξωθεν τὸ γινόμενον συνιδεῐν δυναμένου. A structure identified by R. Koldewey as the Hanging Gardens contained wells in one room (Koldewey, R., Excavations at Babylon 91 f.).Google Scholar These wells comprise three shafts ‘placed close to each other, a square one in the centre, and oblong ones on each side’. Koldewey supposes this to have been for a chain pump. Unfortunately he gives no further details, neither the size of the openings, nor the depth of the shafts. As the shafts are contained within a rectangular room it may be doubted whether the machine could be turned by animals. Oldfather, in his note on Diodorus, loc. cit. (Loeb edition) supposes the central square shaft to have served as an inspection chamber.

110 It is now clear that the west court by the harbour, once called the ‘agora’ is nothing of the sort (Coulton, , BSA lxii (1967) 353Google Scholar). In any case it is too far from the upper plain to be considered part of the settlement there.

111 This perhaps requires some modification, now that an Early Helladic cemetery has been found at the western end of the lake; but this was revealed only by the construction of a new road. Archaic or Classical graves would have been more easily recognized: see Payne's account of the grave-robbing in the area of the modern village.

112 Perachora i. 23. The argument depends largely on Payne's contention that the Heraion was the only town on the Perachora promontory, and therefore responsible also for the Peiraion. There is good evidence for an ancient town, including a Doric building, perhaps a temple, about 1½ kilometres from the modern village of Perachora, along the road to Bissia.

113 This would be a different matter if the water in the lake, prior to the cutting of the canal which links it to the sea, had been drinkable. As it seems unlikely that a considerable body of fresh water would have been deliberately contaminated by admitting the sea, one imagines that the lake has always been brackish and undrinkable. The construction of the cisterns, ancient and modern, by the lake would not make sense if drinking water could have been taken straight from the lake.

114 For the fortifications at the Heraion, see below, p. 240.

115 Dunbabin, (Perachora ii, Appendix 1, 528)Google Scholar describes the slight evidence of manufactures at Perachora. This evidence suggests some activity connected with the sanctuary. but cannot possibly be considered as evidence of the carrying out of manufacturing processes on a commercial scale, and it is not necessarily to be connected with the existence in the area of a town.

116 For the ‘Harbour’ see now Blackman, D. J., BSA lxi (1966) 192.Google Scholar

117 The site of an ancient town near Perachora has been noted above. Perachora itself at present draws most of its water from the vicinity of Bissia, higher up under Geraneia. There is also an independent and old fountain behind the upper part of Perachora village.

118 Perachora i. 21.

119 Strabo ix. 390.

120 Each of these routes is marked by an ancient made road.

121 For the oracle, see Dunbabin, , BSA xlvi (1951) 61.Google Scholar

122 Some—but how many?—could have come by boat.

123 Ancient Greek festivals were, of course, by no means limited to single days.

124 There is a wall, shown on the plan Perachora i, pl. 137, which runs along the ridge to the north of the Heraion valley, south of the line of the modern (and, here, the ancient) road. Like the other continuous walls of the promontory, this seems too slight to have had a military function, and appears rather to mark the limit of the sacred area. It is noticeable that the ordinary houses, etc., are all outside this line.

125 Jones, J. E., Sackett, L. H., and Graham, A. J.: ‘The Dema House in Attica’, BSA lvii (1962) 75 ff.Google Scholar

126 It is possible that A I is also basically of this type.

127 Robinson, and Graham, , Olynthus viii. 171 f.Google Scholar

128 Cf. IG ii2. 334.

129 Roussel, P., Les Cultes égyptiens à Délos, no. 216 (p. 204).Google Scholar

130 See above, pp. 169 f.

131 iv. 138 f., περὶ τῶν Λακωνικῶν συμποσίων.

132 It is worth noting that the dining-rooms of the Asklepieion at Corinth (see above p. 169) like those at Troizen, which, as healing centres, would presumably have been used all the year round, had provision for heating them in winter (Roebuck, , Corinth xiv. 54Google Scholar; Welter, , Troizen und Kalaureia 32).Google Scholar That there was nothing of this sort at the Heraion suggests that the rooms there were used only in the summer. This slight evidence points to a season when the majority of the pilgrims could camp out in the open without undue hardship.

133 It is in this respect, particularly, that the sanctuary of Hemithea at Kastabos affords a useful comparison with the Heraion of Perachora. See The Sanctuary of Hemithea at Kastabos (J. M. Cook and W. H. Plommer).

134 In the projected chapter on the fortifications. This section is based largely on Dunbabin's work, with my own observations.

135 See above.

136 Dunbabin thought this was of re-used blocks. The sherds found in it were, with one exception, all Hellenistic, and so Dunbabin dates it to the third century B.C.

137 Presumably not one of the half-drum fragments listed by Coulton, , BSA lix (1964) 112.Google Scholar

138 Until Coulton proved otherwise, the stoa was considered to be of the early fourth century, and its destruction by early Hellenistic times was then possible.

139 Dunbabin described this as a ‘scrambling way’. It seems to me to be a carefully constructed, wide path.

140 Cf. Perachora i. 14.

141 Xen. Hellenica v1. iv. 3.

142 McCredie, J. R. and Steinberg, A., Two Boeotian Dedications, Hesperia xxix (1960) 123.CrossRefGoogle Scholar

143 This wall, which would not have survived once it was exposed to the elements, and which seriously obstructed the fountain house floor, was removed.

144 Broneer, O., Corinth iv. iiGoogle Scholar, ‘Terracotta Lamps’ 56 f., 149 f.