AAJ.—The Architectural Association's Journal.

Beazley, , ABS.—B's Attic Black-figure, a Sketch, London 1928.Google Scholar

Beazley, AV.—B's Attische Vasenmaler des Rotfigurigen Stils.

Fölzer.—F.'s, Die Hydria, Leipzig, 1906.Google Scholar (Numbers refer to her list at the end of the book.)

Judeich.—J.'s, Topographie von Athen, Munich 1936.Google Scholar

Milet.—Wiegand, Milet, Band I, Heft V, Das Nymphaeum, pp. 73 ff.

Orlandos.—O.'s, article in Ephem. 1916, pp. 94 ff.Google Scholar

Pfuhl.—P.'s, Malerei und Zeichnung der Griechen, Munich, 1923.Google Scholar

Rapp.—R.'s unpublished dissertation, in the Archaeological Seminar in Munich, Über Architekturdarstellungen auf griechischen Vasen archaïscher Zeit., 1914.

Reinach.—R.'s Repertoire des Vases grecs et etrusques, 1899.

Robertson.—R.'s, Greek and Roman Architecture, Cambridge, 1929.Google Scholar

Südhoff.—S.'s, Aus dem antiken Badewesen, Berlin, 1910, pt. I.Google Scholar

Weickert.—W.'s Typen der archaischen Architektur, etc., 1929.

Wiegand.—W.'s Die archaische Porosarchitektur der Akropolis zu Athen, 1904.

I should like to express my thanks to Professor Ashmole and Professor Beazley for reading my manuscript and making many criticisms and suggestions, and to the late Mr. Payne for suggestions; to Mr. Bagenal for advice on technical points; to the British Museum for having several photographs taken for me; to Professor Buschor for allowing me to mention the unpublished Laconian kylix-fragment in Samos; to Professor Delia Seta for giving me new photographs of the Lemnian models; to Doctor Diepolder for allowing me to take photographs in the Museum at Munich, and to the Archaeological Seminar in Munich for allowing me to read the dissertation of Franz Rapp; to the German Institute for providing me with a photograph of the lion at Olympia; to Mr. B. H. Hill for allowing me to publish the spouts from the fountain in the Agora at Corinth, and to the American School at Athens for providing me with photographs of them; to Mr. Shear for allowing me to mention an unpublished vase in the Agora; to several students of the British School at Athens for notes and photographs; to the trustees of the Arthur platt and Frida Mond studentships. This article is adapted from a thesis which was written for the degree of Master of Arts at the University of London while I was holding these two studentships.

^{page 143 note 1} The late Minoan painters, according to Buschor Griechische Vasenmalerei, 1914 p. 22, decorated their pots with marine creatures for the same reason.

^{page 143 note 2} BMC Sicily pp. 78 ff.; Gardner Types Pl. VI (2); Ashmole Greek Sculpture in Sicily and South Italy Pl. III (11 and 13).

^{page 143 note 3} BMC Italy p. 388; Head Historia p. 113; Sylloge Lloyd Coll. 737.

^{page 143 note 4} BMC Thessaly and Aetolia Pl. X (1, 2, 9). Head Historia p. 306 says that the spring represented is Hyperia.

^{page 143 note 5} E.g., Furtwängler, Die antiken Gemmen Pl. VIII 39; Pl. XIX 20, 22; Pl. XX 11; B.M. Cat. of Engraved Gems and Cameos, Pl. XI, nos. 646 and 665; Beazley Lewes House Coll., Pls. III and IX 43.

^{page 143 note 6} See below, p. 180.

^{page 144 note 1} Judeich p. 190.

^{page 144 note 2} Dörpfeld, AD II Pl. XXXVIIIGoogle Scholar; Gräber, AM 1905 pp. 7 ff.Google Scholar

^{page 145 note 1} Paus. I, 40:

^{page 145 note 2} AM 1900 pp. 23 ff.

^{page 145 note 3} AJA 1910 pp. 48 f. At first the excavators thought that there was only one wide cistern and basin; the division into two is an inference of Elderkin, from a small piece of the wall which separated the two basins, and from the two drain-pipes through which the water flowed away in front of the fountain, an unnecessary extravagance had there been only one basin.

^{page 146 note 1} Furtwängler Aegina p. 84; Vallois, RA 1908 II 363.Google Scholar

^{page 146 note 2} The earliest dateable columns of this type in Furtwängler's list, from the propylon at Aegina, are as late as the beginning of the fifth century.

^{page 146 note 3} AM 1900 p. 31. They are not described.

^{page 146 note 4} Unfortunately the remains of the fountain lie largely underneath modern houses, so that only certain parts could be excavated; since the time of the excavations the level of the mud in the surrounding streets and courtyards has risen so considerably that almost nothing of the building is now visible and re-excavation is urgently needed.

^{page 146 note 5} AJA 1900 p. 204 and 1902 p. 321. Fowler, H. N.Art and Archaeology 1922 p. 200Google Scholar. Pirene has not yet been published in any detail; B. H. Hill's publication, which will shortly be ready, will supersede all existing literature on the subject.

^{page 148 note 1} Elderkin, AJA 1910 p. 19.CrossRefGoogle Scholar

^{page 148 note 2} Now demolished after a short distance. Probably in Hellenic times the rock was tiot quarried away so close to the fountain.

^{page 149 note 1} Art and Archaeology 1922 p. 223.

^{page 149 note 2} The facts that it is cut out of the same stratum of rock as the temple (AJA 1905 p. 54), and that the same method of dressing blocks is found in both (AJA 1910 p. 24) suggest that Glauke is perhaps contemporary with the temple, which Weickert dates about the middle of the sixth century. Elderkin thinks, however, that this method of dressing blocks may have been a local peculiarity and have lasted a long time.

^{page 149 note 3} In the National Museum at Athens. They will shortly be published by Dr. Filippo Magi in the Annuario.

^{page 149 note 4} AJA 1934 Pl. XXA.

^{page 150 note 1} AA 1930 p. 146 fig. 22

^{page 150 note 2} AJA 1934 Pl. XXB

^{page 151 note 1} E.g., Berlin 4027.

^{page 151 note 2} See below, p. 175.

^{page 151 note 3} Robertson p. 196 fig. 87.

^{page 151 note 4} AM 1930 pp. 28 flf. The position of the basin suggests that it was built shortly after the South Colonnade (p. 32); this is dated by sherds to the orientalising period (p. 22). The stylistic affinities of the spout confirm this dating. (The basin was covered in again after excavation.)

^{page 152 note 1} See below, p. 194.

^{page 152 note 2} Athenaeus, Deipn XV 672.Google Scholar

^{page 152 note 3} AM 1922 pp. 81 ff. It seems to me that Buschor has proved conclusively that the pediment represents a scene from the Troilos legend and that the building is a fountainhouse, even though his conjectural reconstruction of the whole pediment strikes one as slightly empty and disjointed. Wrede, Attische Mauern p. 44 fig. 10.

^{page 152 note 4} A wall of Peisistratid date at Eleusis shows the same type of construction. Wrede Attische Mauern p. 44 fig. 30.

^{page 152 note 5} Buschor thinks that in the real building, which suggested that in the pediment, the door was in one of the short walls, and the sculptor moved it round so that it might be seen. This would be natural in painting or low relief, but in high relief, where the craftsman is dealing with solid forms, it is surely too naïf even for such an early date; the lack of other examples of doors in this position seems to me unimportant, considering how scanty are the remains of the simpler kinds of Greek buildings. For a door in roughly this position, but in an elliptical house, see AM 1930 p. 17.

^{page 152 note 6} Louvre E. 662. For references to publications of all the vases mentioned, see list at the end. The unpublished dissertation of Franz Rapp, in Munich, on architecture in archaic vase-paintings, which I read after most of this article was written, discusses a large number of the vase-paintings in my list, including a few which I did not already know. Where my independent conclusions are the same as his I make no acknowledgment.

^{page 153 note 1} Athens 277. First quarter of the sixth century (Payne Necrocorinthia pp. 73 etc.).

^{page 153 note 2} Ephem. 1916 p. 96.

^{page 153 note 3} I deliberately omit vase-paintings where water is being drawn from an object like a pithos buried in the ground—e.g., Milan, Castello Sforzesco 06504 (RA 1933 p. 155 fig. 1); Vienna (id. p. 158 fig. 3); Florence 76103 (RA 1935 p. 205 fig. 1); Rome, Museo Artistico Industriale (RM 38–39, Pl. II, 1 and RM 46 Pl. XXI 1) (by Onesimos). Incidentally, these are all red-figure kylikes from the years shortly before or shortly after 480. Also Berlin, Antiquarium inv. 3228 (Pfuhl fig. 276), a late black-figure pelike. In all these the pithos-top must crown quite a deep well, as ropes, and on the pelike a primitive kind of crane, are being used to lower the buckets to water-level.

^{page 153 note 4} The fountain on the temple-terrace at Delphi. See below, p. 175.

^{page 153 note 5} Aus dem antiken Badewesen (Berlin, 1910) I. 64.

^{page 153 note 6} E.g., Madrid 10924; Berlin 1843; Leyden 14 e, 28; Athens Nat. Mus. 1425; Acrop. frag. 2599; Elite Ceramographique IV 17. The oil-flask in B.M. B 333 suggests that someone has just had a bath.

^{page 154 note 1} Milet pp. 73 ff.

^{page 154 note 2} Ephem. 1916 pp. 94 ff.

^{page 154 note 3} New York, G.R. 521.

^{page 154 note 4} B.M. B 325. Berlin 4001 has only a triglyphon and spout, without any columns, and a kalpis in Rhodes from Ialysos, a triglyphon and capital (this does not show clearly in the publication). Spouts without any building, usually with a stand underneath, are of course common; e.g., B.M. B 338.

^{page 154 note 5} New York, Gallatin Collection (here there is a small section of architrave above the column, with a second, frontal spout attached to the bottom of it); Philadelphia, (A.J.A. 1907 p. 429)Google Scholar Athens, Acrop. frag. 563 (r.f.); Madrid 11117 (r.f.); B.M. F 493 (Etruscan).

^{page 155 note 1} E.g., Athens, Acrop. frag. 732; Berlin 1725 and 1908; B.M. B 331, 333, 336; Munich 1716; New York 06.1021.77. Almost all the vases in my list at the end which have no page reference are of this type.

^{page 155 note 2} Berlin 1910; BM. B 337 and 640; Munich 1693; Naples R.C. 205; Paris, Bibl. Nat. 330.

^{page 155 note 3} B.M. B 330.

^{page 156 note 1} Naples S.A. 12; Louvre F 302; Würzburg 345 (without columns).

^{page 157 note 1} Athens, Acrop. frag. 742 (?); B.M. B 334 and 335; Munich 1715; Naples S.A. 157; Louvre F 296 (central wall and part of entablature, but no columns); hydria once on sale at Sotheby's; Taranto; Rome, Torlonia Coll. (r.f.); Berlin 2173 (r.f.). On Thebes, Rhitsonagr. 18, no. 15, there are spouts attached to both sides of a Doric column.

^{page 157 note 2} See Graef pp. 89 ff. on Acrop. Frag. 732. Rapp thinks that this form arose from an attempt to show two spouts in the same wall.

^{page 157 note 3} Boulogne-sur-Mer; Athens 12531.

^{page 157 note 4} By this I do not, of course, mean that any vase-painting is a literal portrait of a particular fountain. Details such as the satyr on the fragment in Leipzig or the spouts on B.M. B 329 may be taken from actual fountains, but the buildings are quite general drawings of typical fountain-houses. Kallirrhoe is twice shown by the simplest and most general scheme of all, with the name added to make the meaning clear (Acrop. frag. 732; B.M. B 331), and even these two drawings differ considerably in details.

Two other drawings are perhaps intended to represent individual fountains, though not literally. The fountain on the Hypsis hydria in the Torlonia Collection is labelled and has a satyr's head for a spout. Furtwängler, (FR text II p. 115)Google Scholar thinks that the fountain was near a sanctuary of Dionysos and may even have been Dörpfeld's Enneakrounos (see below, p. 172), since the remains of a temple of Dionysos were found near it.

On a hydria in the B.M. (B 332) the fountain-house is flanked by two colossal figures of Dionysos and Hermes, which Walters (CVA) says are probably statues. On the shoulder of another hydria (B.M. B 334) is a disjointed scene in which Dionysos and Hermes again appear. Beazley suggests that in both drawings the gods are symbols to represent their precincts, between which a Fountain stood. The precincts of the same two gods are perhaps symbolised by their attributes on a double herm which was used as a boundary stone (AA 1915 p. 180 fig. 3), though Beazley regards the pot as a hydria, and the herm as a boundary between a precinct of Hermes and a fountain.

^{page 158 note 1} Agora P 2642; Bari 3083; Florence 4209 (223); Leyden 14 e, 28; B.M. B 332; Madrid 10924; Würzburg 317.

^{page 158 note 2} E.g., B.M. B 329; Würzburg 316.

^{page 158 note 3} B.M. B 332.

^{page 158 note 4} Berlin 1843; Rome, Vatican (Alinari 35777).

^{page 158 note 5} Eleusis; ex-Forman Coll. 282; Vienna 221; Louvre E 876 and F 296; Rome, Conservatori.

^{page 158 note 6} Mr. T. J. Dunbabin has kindly allowed me to publish this photograph.

^{page 158 note 7} See below, p. 173.

^{page 158 note 8} Rapp thinks (p. 20) that these pillars are a compressed version of fountain-houses without columns, like that on the Lakonian dinos, Louvre E 662 (see below, p. 170), and that they were made narrow because of shortage of space. This is, of course, possible; but spouts attached to plain walls must have existed, and there is no reason why they should not appear on vases.

^{page 158 note 9} B.M. inv. 99·;7–21·2; Munich 1436; Berlin 1694; Athens, Acrop. frag. 2146. (Here the central panel is decorated with a rough-scale pattern.)

^{page 158 note 10} If the painter is drawing a natural spring, without architectural additions, so that he cannot omit the rock, he shows only the front surface of the rock, sliced off and drawn in profile, so that it looks like a pillar (e.g., B.M. B 324). A pillar fountain on a Siana-cup in Taranto is probably meant for a natural spring, as there is no masonry pattern and a tree is growing from the top. Cf. the Hellenistic terracotta, Athens no. 886 (Dumont-Chaplain II Pl. XIX p. 237; Priene p. 68 fig. 35), which gives only a thin slice of the wall or rock to which the spout is fixed.

^{page 158 note 11} E.g., Athens, Acrop. frag. 2599; B.M. E 13 (r.f.).

^{page 159 note 1} Louvre E 703 (Gerhard AV Pl. CLXXXV; Ducati Pontische Vasen Pl. IX (other side); Jacobsthal Ornamente Pl. Xc); cf. CVA Villa Giulia IV Bn. Pl. 2.

^{page 159 note 2} Bagenal AAJ Aug. 1936 p. 84. In fountain-houses whose walls were cut out of rock it would be possible to embed the ends of the roof-beams in the rock itself, thus making the whole building more solid and allowing a heavier pediment.

^{page 159 note 3} Except the central walls of ‘double’ fountain-houses, which are often checquered—e.g., Berlin 2173; B.M. B 334; Louvre F 296; Naples S.A. 157. Is this because the painter is a little troubled about how to fill up the space between his two fountains?

^{page 160 note 1} E.g., Madrid 10924; Vatican (Alinari 35777); B.M. B 330; Berlin 1725 and 1843; Athens Acrop. frag. 732, 741; Munich 1690.

^{page 160 note 2} Occasionally there are triglyphs standing actually on the tie-bar; e.g., B.M. B 333 and 336; Vatican, (Mus. Greg. II Pl. IX, 2b)Google Scholar; but this must be due to artistic compression. A hydria in the Vatican (Alinari 35777) has snakes and birds above the bar; Ridder, De (RA 1901 p. 178)Google Scholar thinks that they were cut out in sheet metal and attached. It seems to me that they are much more likely to be either real animals, arranged conventionally to make a better pattern, or, as Payne once suggested to me, a frieze painted on the back wall of the house.

It is interesting that in the house-model from the Heraeum, Argive (AM 1923 p. 52)Google Scholar the columns are joined to the wall by just such struts. Here they are probably the ends of wooden beams the rest of which was imbedded in the wall.

^{page 160 note 3} E.g., Florence 4209 (François-vase); Würzburg 317.

^{page 160 note 4} Louvre F 302.

^{page 160 note 5} Athens, , AD I Pl. XVIIIGoogle Scholar; Perrot and Chipiez VII Pl. LIII 4. Delos id. Pl. LIII 1.

^{page 160 note 6} F de D, I. 3 pp. 32 ff.

^{page 160 note 7} E.g., the two older temples at Paestum, Robertson figs. 30a and 32.

^{page 160 note 8} Durm p. 373.

^{page 161 note 1} On the François-vase the capitals are painted white, like the bases which are pre sumably stone. Sulze, therefore (AA 1936 pp. 14 ff.)Google Scholar thinks that the capitals were also of stone. He is discussing two archaic Doric capitals from Agrigentum and Tiryns, the first of which probably, and the second possibly, had a wooden shaft. He claims that stone capitals would be more practical than wood because they would protect the top of the shaft from damp, and because wooden Doric capitals are structurally unsound; and he suggests that stone capitals were fairly common. It seems to me that the shafts would usually be quite adequately protected by the overhang of the roof; and Bagenal tells me that there is no reason why Doric capitals should not be made of wood. Stone capitals above light wooden shafts seem a very awkward kind of construction. However, the capital from Agrigentum seems to show that such a combination did sometimes occur, and the capitals on the François-vase may be a case in point. On a few other vases there are white capitals or abaci—e.g., B.M. B 335 (Doric) and Munich 1716 (Ionic). Sulze regards these also as stone. But of course they may be due simply to love of poly chromy.

^{page 161 note 2} E.g., B.M. B 330 and 333; Munich 1690.

^{page 161 note 3} E.g., Treasury of Syracuse at Olympia, Olympia, Pl. XXXIV, Weickert p. 138; Basilica at Paestum, K. and P. pp. 13 ff. figs. 8, 9, 10, Perrot and Chipiez VII Pl. XXVI 1–8; Temple of Ceres at Paestum, K. and P. pp. 18 ff. figs. 17, 20, 21; Perrot and Chipiez id. Pl. XXVI 9–12. The capital from the Tomb of Xenvares at Corfu has instead of leaves a thin vertical strip with the outlines of an egg-and-dart pattern engraved on it and its lower edge cut out to follow the pattern. The pattern was originally coloured. At first sight the egg and dart looks very like the tips of leaves.

^{page 161 note 4} Munich 1693. Rapp had already noticed this. On a very fragmentary r.f. kylix lent to the Ashmolean there is a capital with the same kind of projecting strip under the echinus and painted on the strip a pattern which might represent either the tips of leaves or a simplified egg-and-dart pattern. (CVA Oxford fasc. 1 Pl. 14, 34.)

^{page 161 note 5} E.g., B.M. B 334; Munich 1715 and 1716.

^{page 161 note 6} Munich 1715; Louvre F 302 (here the volutes are suggested only by dots).

^{page 161 note 7} See above, p. 160, note 5.

^{page 162 note 1} B.M. B 332, Würzburg 317 and Bari 3083 have a line joining the top of the inner curve of the volutes; in the last two the curve of this part of the volute-member is suggested by a curved line close up against the abacus. The two capitals on Würzburg 317 are drawn quite differently from each other, which shows how unsafe it is to rely too much on vase-paintings. Similarities in the shape and decoration of the vase, and in the drawing of figures and of architectural details show that B.M. B 329, Würzburg 316 and 317, Bari 3083 and the fragment in Florence are by the same hand. Perhaps also B.M. B 332.

^{page 162 note 2} B.M. B 332.

^{page 162 note 3} Munich 1716.

^{page 163 note 1} Taranto, cup of Siana shape.

^{page 163 note 2} AD. I. Pl. XXIX, i; Perrot and Chipiez VII, p. 573, Fig. 5.

^{page 163 note 3} B.M. B 672.

^{page 163 note 4} The painter of the Taranto cup was interested in capitals; at the other end of the same picture is an isolated column with a very exactly-drawn Aeolic capital.

^{page 163 note 5} Leyden, 14 e, 48. See Beazley, ABS p. 26 note 4.

^{page 163 note 6} E.g., B.M. B 333; Madrid 10924.

^{page 163 note 7} E.g., B.M. B 334; Berlin 1908.

^{page 163 note 8} E.g., B.M. B 331; Berlin 1910.

^{page 163 note 9} I cannot believe that these are really separate rectangular pilasters (FR text I pp. 56 f.); I know of no example of such pilasters used in the same row as columns. Buschor's, explanation (AM 1922 p. 84)Google Scholar seems to me much more plausible: that the object, platform or basin-rim, on which Rhodia is standing, was really inside the house, close to the spouts, but that Klitias moved it partially outside because there was no room for a figure inside, regardless of the fact that in an actual building this would involve a gap in the wall.

^{page 164 note 1} The only known early example of this in Greece, Treasury X at Delphi (Weickert p. 130), is of imported stone, and may have been made by craftsmen imported from Magna Graecia. Engaged columns are found in temple D at Selinus.

^{page 164 note 2} Berlin 1843. On two vases, Vatican (Alinari 35777) and Louvre F 296, there are half-columns attached on each side of the central wall; such an unlikely arrangement is surely another indication that walls with spouts on both sides did not really exist, and that the painter was forced to invent their decorative details for himself.

^{page 164 note 3} E.g., B.M. B 330 and 335.

^{page 164 note 4} B.M. B 330 and 335; Taranto, cup of Siana shape (see above); New York G.R. 521.

^{page 164 note 5} RA 1908 p. 384. If triglyphs were still the ends of real beams, they would, of course, occur only on opposite sides of the building.

^{page 164 note 6} Weickert p. 39.

^{page 164 note 7} Id. p. 45.

^{page 164 note 8} Id. p. 42.

^{page 164 note 9} Id. p. 50.

^{page 165 note 1} E.g., B.M. B 329; cf. the white ground lekythos in the Louvre, Rayet-Collignon Pl. XI. Probably, as Vallois says, the taenia at the top is here made to do duty for the whole architrave. Bagenal, (AAJ Aug. 1936 p. 89)Google Scholar thinks that this may represent a real arrangement; the reglets would be the ends of thin planks, fastened with pegs, the guttae, to the under-side of the architrave, and perhaps used to hang things from (as in the Apulian volute-crater, Buschor fig. 160). But, even so, one must assume that the architrave has shrunk to negligible proportions, so that it is hardly less drastic to imagine that all but the taenia has vanished. In the second place, this would, as he says, involve regarding the capitals as brackets, not true Doric capitals, since the reglets project in front of them. And in the third place, on the François-vase the reglets are in the position in which they are found in normal Doric; so that it seems that the arrangement found in B.M. B 329 can only be an artistic abbreviation. On the other hand, the function of reglets in wooden Doric architecture is hard to explain. Bagenal thinks that there is another possible solution in an idea of H. W. Richmond, that the architrave was originally formed by two beams lying side by side, that the taenia was the front edge of a plank which lay above them, and the reglets small pieces of wood notched into the plank and pegged to the beams, so that these were firmly fixed together. If this is the true solution, the spacing of the reglets would of course originally be quite independent of that of the triglyphs, as it some times is on vases.

^{page 165 note 2} E.g., B.M. B 334.

^{page 165 note 3} E.g., B.M. B 329 and 332; Würzburg 316, 317; probably they often were plain.

^{page 165 note 4} E.g., B.M. B 334.

^{page 165 note 5} E.g., Vatican (Alinari 35777); Berlin 1843 and 1908.

^{page 165 note 6} Agora P 2642.

^{page 165 note 7} E.g., Vatican (Alinari 35777); B.M. B 329; Würzburg 316, 317; probably only because there is more room for them and because they make a more balanced pattern under the metopes which are white.

^{page 166 note 1} These do not occur in any other vase-painting of a fountain.

^{page 166 note 2} E.g., the Hecatompedon (Weickert p. 98; Wiegand Pl. I); the earlier temple of Aphaia at Aegina (Weickert p. 95, Furt. Aegina Pls. LX, LXII).

^{page 167 note 1} E.g., B.M. B 333. 336; Vatican, (Mus. Greg. II Pl. IX, 2B).Google Scholar

^{page 167 note 2} B.M. B 333.

^{page 167 note 3} Graef 2549, Heft IV Pl. CV p. 248.

^{page 167 note 4} See above, p. 160, note 2. So is the frieze on the other vases quoted in note 1.

^{page 168 note 1} The painter is not greatly troubled about which way the pediment really faces, and even adds pieces of pediment to the ‘shorthand’ type of fountain (see above, p. 22), e.g., B.M. B 330; Munich 1690; and to ‘double’ fountains, e.g., B.M. B 334, 335.

^{page 168 note 2} E.g., B.M. B 334, where the Nike and the seated woman look at first sight as if they belonged to the same scene.

^{page 168 note 3} I therefore regard the curious domed roof of the François vase fountain as an abbreviation of a pediment, as Furtwängler does; though Orlandos, (Ephem. 1916 p. 102)Google Scholar thinks it is a stamped-earth roof, slightly domed to let the rain drain off.

^{page 168 note 4} B.M. B 334 and 335; Agora P 2642.

^{page 168 note 5} AAJ 1936 pp. 110 ff.

^{page 168 note 6} Perhaps also the house model in Samos, AM 1930 p. 17.Google Scholar Thatching does not seem to be used on any permanent buildings in Greece to-day, but in Epirus I have seen round Vlach huts built entirely of a very regular thatching of leaves.

^{page 168 note 7} E.g., Leipzig frag.; Madrid 10924; Munich 1690; B.M. B 330, 334, 335; Würzburg 317.

^{page 168 note 8} Jahresh. 1899 pp. 16 ff., esp. p. 35.

^{page 168 note 9} Wiegand pp. 38 ff. Pl. I; Schuchhardt, AM 1935 pp. 1 ff.Google Scholar

^{page 168 note 10} It might, of course, equally well be the facing of the ridge-beam.

^{page 169 note 1} AM 1900 p. 204 and 1933 p. 41. Here the volute is a separate addition, not the end of the sima or of the band at the top of the sima.

^{page 169 note 2} Mon. Linc. 1935 col. 150 Pls. I, III, IV, V. Gabrici suggests no absolute date for this building, but says it is ‘antichissimo.’ The way in which the roof-tiles fit under the back of the raking-cornice is, as he points out, paralleled in the Megaron of Demeter at Selinus. This is dated by Koldewey and Puchstein in the early sixth century (pp. 80, 231 ff.).

^{page 169 note 3} Milet I 8 pp. 16 ff.

^{page 169 note 4} B.M. B 330.

^{page 169 note 5} Würzburg 317.

^{page 169 note 6} Schrader Archaische Marmorskulpturen im Akropolismuseum zu Athen, pp. 8 ff.; Jahrb. 1928 pp. 54 ff.; Schuchhardt op. cit. pp. 31 ff.

^{page 169 note 7} Agora P 2642.

^{page 169 note 8} Koldewey, Neandria p. 46.Google Scholar

^{page 169 note 9} E.g., the Athenian Treasury at Delphi, F de D, IV 1 Pl. XLVI.

^{page 169 note 10} B.M. B 334. In the B.M. catalogue and the Corpus the figure is called Eris, and is said to belong to the picture on the shoulder of the hydria; but as this picture is disjointed and unexplained, and as the leg of the figure actually touches the pediment, she is much more likely to be an acroterion; cf. a figure as acroterion on Acrop. frag. 2549, Heft IV. Pl. CV p. 248. For real Nike acroteria cf. Acrop 690, 691, 693, 694, and the Nikes from Delos and the Alcmaeonid temple at Delphi. See Eduard Schmidt Jahresh. 1920 pp. 97 ff.

^{page 169 note 11} E.g., B.M. B 334 and 335; Leyden 14c, 28; Torlonia Coll (r.f.)

^{page 169 note 12} E.g., B.M. B 330; Leyden 14e 28; Agora P 2642; cf. later (r.f.) vase p. 58 (Bibl. Nat. 940).

^{page 170 note 1} Wiegand Pls. IV, V, VIII fig. 213.

^{page 170 note 2} Leipzig frag.

^{page 170 note 3} Athens 277; first quarter of the sixth century (Payne Necrocor. p. 73).

^{page 170 note 4} Würzburg 354.

^{page 170 note 5} Reggio, inv. 1169.

^{page 170 note 6} Louvre E 662. Lane, BSA 1933–1934 pp. 146 and 164.Google Scholar

^{page 170 note 7} See below, p. 182.

^{page 170 note 8} Weickert, p. 30; BSA, Artemis Orthia pp. 118 ff.

^{page 170 note 9} Unpublished fragment found by the German Institute, at Samos. Inst. Phot., Samos 1600.

^{page 170 note 10} That the story has been so ‘ruthlessly cut’ that the most important person in it is relegated to an exergue; and that the space behind Polyxena is filled with a quite irrelevant cock, is typical of the mythological scenes on Lakonian pottery. JHS 1932 p. 26.

^{page 171 note 1} The lower part of the walls and of the column are broken away.

^{page 171 note 2} Louvre E 669.

^{page 171 note 3} A fragmentary kylix in London with part of a kneeling warrior, and, in the exergue, the hindquarters of a horse; and a fragment from the exergue of a similar kylix, by the same hand, in Samos, led Lane, (BSA 1933–1934 P. 164)Google Scholar to the conclusion that Louvre E 669 was probably an abridged Troilos scene. The new Samos kylix, which is very like Louvre 669 in composition and details (in both there are birds on the roof of the house, a snake in front of it and a bird behind the warrior) confirms his view.

Lane considers that Louvre 669 is shortly after the middle of the century; and that the London and the first Samos fragment are about thirty years earlier; the new Samos kylix would be between the two.

^{page 172 note 1} About 510 b.c.

^{page 172 note 2} B.M. E 13.

^{page 172 note 3} There is a plain profile fountain spout on a hydria-kalpis of about this date in the British Museum (E 159), painted by Phintias. On a fragment of an early red-figure krater in Delos, are preserved a fine profile spout and some distance in front of it, the top of a Doric column. A very fragmentary r.f. hydria in Taranto has on the shoulder girls drawing water from a tubular spout; of the scene on the body of the vase nothing is preserved but a lion's-head fountain-spout.

^{page 172 note 4} Pausanias I 14, 1: (to the Odeion)

^{page 172 note 5} Perrot and Chipiez VIII p. 33.

^{page 172 note 6} Hesperia 1935 p. 360, Pl. III.

^{page 172 note 7} Practically all traces of the Ilissos fountain were swept away by a flood in 1896. The American Enneakrounos has not yet been published. Wachsmuth (Paully-Wissowa, Supplement I pp. 211 ff.) opposes Dörpfeld's view, Judeich (pp. 193 fF.) supports it.

^{page 172 note 8} Gräber, AM 1905 pp. 33 ff.Google Scholar

^{page 173 note 1} Cf. Elderkin, AJA 1910 pp. 46 f.Google Scholar

^{page 173 note 2} Op. cit. fig. 32.

^{page 173 note 3} Op. cit. fig. 30.

^{page 173 note 4} Op. cit. fig. 31.

^{page 173 note 8} Wiegand p. 50.

^{page 173 note 6} In vase-paintings, hydriae of various shapes are almost always used for fetching water; pointed amphorae, however, do sometimes occur, always, I think, carried by men—e.g., Vatican, (Mus. Greg. II. Pl. XI 2a)Google Scholar, Berlin 2173 and 4027; see Fölzer pp. 6 ff. For drawing water from deep wells, amphora-shaped, metal pots with moveable baskethandles are used (cf. Milan, Castello Sforzesco B 6504, RA. 1933 pp. 154 ff., and the other vases illustrated in this article) and once (Berlin Antiquarium, inv. 3228, Pfuhl, fig. 276) pelikes with extra basket-handles, probably detachable and made of rope or metal.

^{page 173 note 7} Weickert p. 171.

^{page 173 note 8} See below, p. 195.

^{page 173 note 9} AJA 1902 pp. 306 ff. (Richardson). Fowler's account in Art and Archaeology 1922 pp. 213 ff. is more confident, but naturally gives no references; a summary of the evidence for his chronology is much to be desired.

^{page 175 note 1} ILN Feb. 28, 1931 p. 330.

^{page 175 note 2} F de D II 1 pp. 171 ff. Figs. 128–39.

^{page 175 note 3} Probably soon after 548. Courby, p. no, says that the temple was begun soon after the burning of the old temple, and the Alcmaeonids only finished it. Naturally the foundations would be the earliest part to be built. The steps and walls of the fountain may not be so early, but must be no later than the polygonal wall, in the last third of the century.

^{page 176 note 1} AM 1902 p. 200 (Rubensohn).

^{page 177 note 1} Louvre. About 490–480 b.c.

^{page 177 note 2} Leningrad, 1588. About 490.

^{page 178 note 1} Madrid 11117.

^{page 178 note 2} Athens, Acrop. frag. 563.

^{page 178 note 3} B.M. E 772.

^{page 178 note 4} On a roughly painted pelike of about 480, in Athens (1425) (in the style of the Nikoxenos Painter), is an abbreviation consisting of two columns with an architrave and frieze above, and a spout close under the capital of one of the columns. On the shaft of each capital is drawn a thick black line which tapers towards the top and ends in what looks like a small capital; this must be fantasy.

^{page 178 note 5} Pind. Paean VI. 7.

^{page 178 note 6} Art and Archaeology 1922 p. 201.

^{page 178 note 7} See p. 143, note 2.

^{page 178 note 8} See p. 143, note 3.

^{page 178 note 9} B.M. E 204 and Constantinople, inv. 2179.

^{page 179 note 1} Bibl. Nat. 422.

^{page 180 note 1} See below p. 182.

^{page 180 note 2} Furtwängler Die antiken Gemmen Pl. XXXIX 26 and 27; and text p. 188; B.M. Catalogue of Engraved Gems and Cameos Pl. X no. 561; Beazley Lewes House Collection fig. 5 no. 134.

^{page 180 note 3} Délos V 103.

^{page 181 note 1} See below, p. 192, note 2.

^{page 181 note 2} AJA 1933 pp. 423 ff.

^{page 181 note 3} Dugas Le Sanctuaire d'Aléa Athéna à Tégée, p. 69, Figs. 26–27 Pl. LXXXI.

^{page 181 note 4} Perhaps when the metal of which clamps are made has been analysed, and their size and proportions studied more carefully, a more exact dating will be possible.

^{page 182 note 1} A.M. 1902 pp. 205 f. and 236.

^{page 182 note 2} Leonardos Ephem. 1918 p. 110.

^{page 182 note 3} Bibl. Nat. 940.

^{page 182 note 4} Its style is approximately that of the later of Miss Moon's early S. Italian vases (BSR XI pp. 30ff.).

^{page 182 note 5} Karlsruhe 388; Naples 690.

^{page 183 note 1} American School at Athens, Corinth, III 40.

^{page 183 note 2} There are no proofs of any roofing earlier than the vault. The publication states that the tiles found above the vault were of ‘good classic pattern,’ and so earlier than the vault; but I understand that there are now being found at Corinth tiles of this pattern and of undoubtedly Hellenistic date.

^{page 183 note 3} Clara Rhodos I 79. The fountain has not yet been published in any detail, but will shortly be so, by Dr. Maiuri; so far there is no published plan. I am grateful to Mr. T. J. Dunbabin and Mr. G. Deeley for the photographs, Figs. 15 and 16.

^{page 184 note 1} AM 1879 p. 38; Judeich p. 136 fig. 10 ff.

^{page 185 note 1} The water has worn a narrow groove in the floor.

^{page 187 note 1} Arch. Delt. 1927 1–7.

^{page 187 note 2} Orlandos, AJA 1934 p. 153CrossRefGoogle Scholar; 1935 p. 408; 1937 p. 336. Prakt. 1934, p. 116.

^{page 188 note 1} Jahresh. 1901 p. 15.

^{page 188 note 2} Orlandos, Prakt. 1925 p. 52.Google Scholar

^{page 188 note 3} Kourouniotis, , Prakt. 1909 pp. 198 ff.Google Scholar A very large third- or second-century basin at Lykosoura (Orlandos, Ephem. 1911 p. 200)Google Scholar, with spouts along its parapet, is interesting in that it very probably had wooden columns and entablature.

^{page 188 note 4} Humann, , Magnesia pp. 113 and 135.Google Scholar

^{page 188 note 5} Dörpfeld, , AM 1902 p. 36.Google Scholar

^{page 189 note 1} AM 1884 pp. 165 ff.

^{page 189 note 2} AM 1905 pp. 23 f.

^{page 189 note 3} E.g., Samos, AM 1884 p. 175 Pl. VIIIGoogle Scholar; Athens, AM 1905 pp. 24 f.Google Scholar

^{page 189 note 4} It has often not been possible to discover much about the outlets of a fountain because of difficulties of excavation and later alterations. Often details are not published.

^{page 190 note 1} Sylloge Lloyd Coll. vol. II no. 737.

^{page 190 note 2} Ashmole Greek Sculpture in Sicily and S. Italy Pl. III 13, 15.

^{page 190 note 3} E.g., Athens 277; Eleusis; Munich 1436; Taranto, cup of Siana shape.

^{page 190 note 4} In red-figure such basins are very common in bathing scenes, presumably inside houses, and spouts for filling them are sometimes shown. E.g., Naples 2848 (Südhoff, Fig. 35).

^{page 190 note 5} E.g., B.M. B 329; Würz. 316; Madrid 10924.

^{page 190 note 6} E.g., Ath. Acrop. frag. 732; Berlin 2173 and 4027; Vatican (Alinari 35777). There are occasionally patterns on the sides of the trough or stand: Acrop. frag. 732 (hooks, perhaps meant to suggest a meander, with a masonry pattern, breaking joint, underneath); B.M. B 640 (row of small circles); Oxford, Ashmolean 568.

^{page 190 note 7} Leyden 14e, 28.

^{page 190 note 8} Beazley, ABS p. 26 n. 4.

^{page 190 note 9} Berlin 1843.

^{page 191 note 1} At Phigaleia there are stone troughs with semicircular grooves in their top edges, to let the water flow in; two seem to have fitted against similar grooves in the front rim of the basin, so that water filled them and animals could drink from them. Troughs of this kind beside fountains are fairly common in Greece to-day, especially in the islands.

^{page 191 note 2} E.g., Ialysos; Dipylon Gate; third-century fountain at Lykosoura.

^{page 191 note 3} Though the cisterns could not be emptied separately from the draw-basins with which they connect, and the two easternmost cisterns, which share a draw-basin, would have to be emptied at the same time.

^{page 192 note 1} In the country to-day the right to use up the overflow of a fountain is nearly always legally secured to some individual.

^{page 192 note 2} The inscription from Minoe (Delos V 113), which forbids washing anything or bathing in the fountain, suggests that public basins for such purposes were not uncommon. The inscription is fragmentary; the first few lines are decipherable:—

^{page 192 note 3} 117 D., trans. Bury (Loeb).

^{page 192 note 4} B.M. inv. 99.7–1.2; Munich 1436 & 1690; ex-Forman Coll.; Philadelphia, (AJA 1907 p. 429)Google Scholar; Taranto, hydria (r.f.), Athens 1698, B.M. E 772; B.M. F 493; Karlsruhe 388 (Italiote). On a series of r.f. and white-ground lekythoi, by the Bowdoinpainter, tubular spouts are shown from the front, by a ring. (Girgenti, Baron Giudice's Coll.; Athens 1791; Syracuse; Berlin, inv. 3338–9; Oxford, Beazley's Coll. cf. B.M. B 329 (b.f.), E 204; Constantinople inv. 2179.)

^{page 192 note 5} See above, p. 188, note 3.

^{page 192 note 6} Tubular spouts are also used as rainwater-outlets on the simas of many early buildings—e.g., the Geloan treasury at Olympia and the pre-Peisistratid Hecatompedon; on the Peisistratid pteron of the Hecatompedon and the sixth-century temple of Apollo at Delphi they are mixed with lions' heads.

^{page 192 note 7} Munich 1690. Most of the horse is restoration.

^{page 192 note 8} B.M. B 329.

^{page 193 note 1} Olympia III p. 26 fig. 23 Pl. V; Crome Mnemosynon Theodor Wiegand pp. 47 ff. Pls. VII, IX, X.

^{page 193 note 2} Olympia III p. 26; Crome op. cit. p. 49 note 4.

^{page 193 note 3} There is no evidence of pressure in Greek water-systems before that of Pergamon in Hellenistic times (Gerkan Städteanlagen p. 89). But several trick vases, including one which is as early as the first quarter of the sixth century (Louvre, Corinthian comastvase BCH1895 pp. 225 ff. Pls. XIX, XX; Payne Necrocorinthia p. 176; Amsterdam, Allard Pierson Museum, Snijder, Mnemosyne, 1937; Boston, Fine Arts Museum, id. and 25th Annual Report of the F.A. Mus. 1900 p. 71. These are both Campanian fourth century) show that the Greeks knew how to utilise atmospheric pressure.

^{page 193 note 4} Crome points out that the lions which form the fountain in the fresco of the Tomba dei Tori at Corneto are in much the same position. (AD II Pl. XLI.)

^{page 193 note 5} Op. cit.

^{page 193 note 6} Payne Necrocorinthia, Pl. IV 3 and V 1–4.

^{page 193 note 7} Though scales as a rendering for fur do not otherwise occur in Protocorinthian, the scales on the hedgehog vase (Payne Necrocorinthia p. 171 fig. 71) may be meant to suggest prickles.

^{page 194 note 1} Crome op. cit. p. 49 note 4. Phot. Germ. Inst. Ath. Smyrna 20.

^{page 194 note 2} AM 1906 p. 155 fig. 4; Phot. Germ. Inst. Ath. Sam. 130, 131.

^{page 194 note 3} AM 1930 p. 30 and Pl. I; Altsamische Standbilder III figs. 213, 216, 217; Crome op. cit. pp. 49, 51 f.

^{page 194 note 4} Hall, Babylonian and Assyrian Sculpture in the B.M., Pls. XVIII, XLVIII, LIX.

^{page 194 note 5} Inv. 118144, unpublished. Mr. Barnett tells me that it must have been made between about 850 and 700.

^{page 195 note 1} See Necrocor. p. 70.

^{page 195 note 2} Except in Berlin 1843, where a spotted neck, rather awkwardly added, seems definitely meant to show that the creature is a panther or a leopard.

^{page 195 note 3} B.M. B 329; Würz. 316–17; Bari 3083; Florence frag.

^{page 195 note 4} Würz. 354.

^{page 195 note 5} E.g., Torlonia Coll.;. B.M. E 159.

^{page 195 note 6} Louvre.

^{page 195 note 7} Madrid 11167, Leningrad 1588.

^{page 195 note 8} E.g., Naples 690; B.M. F 236.

^{page 195 note 9} There is an inadequate photograph of the later one in AJA 1902 p. 319. The other is not published at all.

^{page 195 note 10} Payne and Young Archaic Marble Sculptures from the Acropolis. Pl. 132.

^{page 195 note 11} F de D IV 1, Pls. XIII, XXXIII.

^{page 196 note 1} Payne Necrocor fig. 200 (Attic) and fig. 71.

^{page 196 note 2} F de D V p. 56 Pl. XV.

^{page 197 note 1} See above, p. 178.

^{page 197 note 2} Priene (p. 296), in the gymnasium; Lykosoura, unpublished Hellenistic basin, with one very rough lion's head.

^{page 197 note 3} Louvre F 296; Vatican, Alinari 35777; Berlin 4027; B.M. E 13 (r.f.).

^{page 197 note 4} E.g., Berlin 1843; ex Sotheby's sale, London.

^{page 197 note 5} Torlonia Coll.; Bibl. Nat. 422 (de Ridder describes them as gorgoneia; perhaps they are).

^{page 197 note 6} Eleusis, Noack Pl. XXIX and p. 64 (a false spout; without any hole), and the Peisistratid Hecatompedon, Wiegand p. 125 fig. 121. I agree with Noack that the ram's head at Kaisariane, now used as a fountain-spout, probably belonged originally to a sima, not to a fountain; the stone on which it is carved, though too broken for one to judge of its dimensions, ends a few centimetres above the animals' head in a finished surface. Here it is very thin, and much more like the edge of a sima than part of a slab from a wall or parapet.