Page - 72 - in The Power of Urban Water - Studies in premodern urbanism

72 PhilippKobusch Fig. 3: Thermos, Sanctuary. The same generally applies to fountains that were positioned in the immediate vicinity of the main altar. For example, the wells in Thermos (Fig.3),16 Magnesia17 and Aulis18 do not have anymonumental structures– the latter follows thesame typeas thewells inTegeaandDelphi. Thisdoesnotmeanthat theycouldnotbepartofadesignedarchitecturalprospectus,especially in Hellenistic times, as is known fromMagnesia, for example. Here the basin is embedded in thepavement in front of themainaltar and is thus integrated in the central view-axis from the propylon via the altar to the temple front. Nevertheless, the source in this ensemble does not receive any independent architectural emphasis, but remains clearly subordinate to the two monumentalarchitecturalunits. InThermos (Fig.3), incomparison, theapproach to the temple was elaborately architecturally designed in theHellenistic period due to the so-called agora, a street which is bordered by long galleries. But the probably older sourcewas not included in this ensemble. Rather, it was located not only on a lower level from this access road, butwas also concealedbyanexedramonument oriented towards the road. In Aulis, the situation is hardly assessable, because the connection to the temple is ob- scured by amodern street. Therefore, it is questionablewhether the small foundation in front of the fountainmustbeaddressedas themainaltar of the sanctuary.However,written sources tell about a plane tree in the direct vicinity of the fountain.19 A natural monument, not an elaborate architectural design, therefore servedas aneye-catcherhere. Evenmore vividly, the low interest in an architectural staging of importantwater installa- tions canbeseen indifferent examples inDelphi.20After the landslide in373BC, awaterbasin (Fig. 2,2) was placed in front of the northeast corner of the temple during the construction of the new retainingwall, the so-called Ischegaon. Awater pipe supplied it with runningwater. The traces in the rockshowthat thebasinwasnot framedarchitecturally. It also laybehind the the thinwallswere possibly reinforcedby a small stonepackage: Courby 1927, 172.However, even the excavators wereuncertainwhether this could indicatea structureaboveground.FernandCourby interpreted the thickblocks that adjoin the slabson thewest sideas foundations for thispackage (Courby 1927, fig. 129).However, theydonot haveaworked surface andare thereforeunsuitable as a foundation for aneatly placedand stablewall. 16 Sotiriades 1899, 61f.; Papapostolou 1984, 127f. fig. 1; pl. 105; Papapostolou 2014, 184 figs. 79. 90. 111. 17 Bingöl 2007,84f.;Hammerschmied2018, 102f. figs. 2. 6. Itwasawidebasin (12×3m)embedded in theground andaccessible by six steps on the east side. 18 Threpsiades1958,48–51 figs.3–4;Drögemüller1961,218f. fig. 12;Glaser1983,16–18figs.30.31.Thereconstruc- tionbyGlaser 1976/77, fig. 1 is in theendhypothetical. Thesituation inXanthoscannotbeassessedwithcertainty. Parts of the ImperialNymphaionseemtodateback to theHellenisticperiodandseemtohave formedanartificial grotto. The source,whichoriginateddirectly in front of the temple,mayhavebeen in ritual usebeforeHellenistic times, as older findings suggest: Balland 1974;Metzger 1979, 10, 14f.; Le Roy 1988, passim, especially 127; Dorl- Klingenschmid 2001, 252f. no. 120 fig. 180; Longfellow2012, 146–151 figs. 13–15. 19 Paus.9, 19, 7;Hom. Il. 2, 305. 20 Glaser 1983, 26–28 figs. 48–50;Bommelaer–Laroche 2015, 222f. fig. 77.