The Symposium-Tent of Ptolemy II: A New Proposal

289 THE SYMPOSlur.,\-TENT OF PTOLEMY II: A NEW PROPOSAL All students of classical antiquity sooner or later make some use of the Deipnosophists of Athenaeus of Naucratis. Those interested especially in Hellenistic architecture have drawn heavily on Athenaeus' excerpts from the book Peri Alexandreias, written (probably in the 2nd c. BC) by Callixenus of Rhodes. Of particular interest have been Callixenus' descriptions of the banquet-tent erected by Ptolemy II Philadelphus in the palace-park at Alexandria, for the celebration of the winter Dionysia, and of the interior design and decoration of the great river-barge, called Thn were eagles facing each other, in gold, fifteen cubits in size. Along the two sides [as seen 7 from the entrance] were set 100 couches ... : the prospect toward the front (i. e. presumably lithe entrance") was left open .... Call ixenus' description of the tent is rather short on specific measurements. Of the ground-plan, he tells us only: that the central area was enclosed by wooden columns, five on the sides, four across the ends (counting the corner columns twice); that this central "nave , II as we may call it, had room for 130 couches KUKA'll; and that the nave was surrounded on three sides by a "peristyle corridor, II for the retinues of the guests. The elevation of the tent is described in greater detail, but with more emphasis on decorative aspects than on physical dimensions. Thus we learn that the columns at the angles were shaped like palm-trees, the others like Dionysiac thyrsoi; the imposing height of the columns (50 cubits) is stressed, but the thickness is not mentioned. Callixenus further records that the architrave above the columns provided the sole support for the covering of the nave, i.e. there was neither a single central pier or column, nor an inner rectangle of supports such as we find in the bouleuteria of Priene and Miletos. Yet we are told nothing of the size and construction of the architrave, wh,ich is described simply as "square" (presumably in section). The covering of the central area of the nave was a "miniature dome of heavenII (ouraniskos); this word, which seems not to occur before the Hellenistic period, presumably indicated originally a roof or ceiling that curved upward from the perimeter toward the centre. In the present instance it must refer to an awning. Since the axial spans of the columns are not recorded, we must try to deduce the size of the awning from the other details that are described. On either side (KClS' £KtlTEPOV ~€poing, 0:" ouranisk_o~; he takes all the decorative details of the orophe to refer to the awning, i.e. in his view the middle section of the awning was scarlet edged with white, while along the longer sides there were painted representations of beams and decorated coffers, rather than real wooden members. This huge canopy was tied down by seven cables on each side, the ends of the cables being fastened to the crossbeams at the top of the surrounding peristyle, which thus had to resist the enormous upward drag of a canopy measuring about 33 x 43.65 m (about 1440 m2 ), and lacking any sort of intermediate support. Studniczka further assumed that the peristyle corridor was two-storeyed, with an upper level of supports; here again, there is no mention of such a feature in Athenaeus' quotation from Callixenus. He also believed that the panel-painting, tunics, and cloaks must have been displayed 011 the outside of the scarlet curtains that hung in the intercolumniations of the peristyle, and so were invisible to the guests ~ining in the nave. He then placed the shields, too, around the exterior of the tent, above the lower architrave of the peristyle, as well as around three sides of the nave, at the same level. He believed that the "caves,,13 must have been mounted in the outer intervals of the (hypothetical) upper storey of the peristyle; the "8-cubit size" of the spaces containing the "caves" he interpreted as indicating height rather than width. 12See the ~c<}le accorr.panying the restored plan and elevations, Studniczka (above, n.3) Plates 3 (plan), 2.1-2 (elevatior::'. here Figs.1,2 and Plate 1. 13A motlt that n:curs not only ill some of the floa~::- :r: the procession celebrating the winter-festival of Dionysus (.l\ th. LOOC; and d. the OK1UC;, hung with grape-vine, ivy, and various CL'ltivated 298 F.E. WINTER AND A. CHRISTIE By means of these assumptions Studniczka obtained a nave certainly large enough to hold 130 couches, with more than ample room left over for circulation of attendants. Almost all decorative elements, however, were relegated to the exterior of the marquee, where, like the eagle acroteria, they would have been visible ~ from the outside. This placing of the decorative elements, though reasonable enough in a temple designed to be seen primarily from the outside, is much less logical in buildings in which people normally gathered in the interior. Thus the paintings and other dedications of the Lesche of the Cnidians at Delphi, and of the Stoa Poikile, Stoa of Zeus, Bouleuterion, and Tholos in the Agora at Athens, like the paintings, mosaics, and carved ornament of Hellenistic houses and palaces, were inside the buildings, not on the exterior. 14 In the case of Ptolemy's tent, too, we believe that the decorative elements were all, or almost all, in the interior; and we note that the same is true of the vast majority of the decorative details of Philopator's river-barge, as described by Callixenus. We further believe that Studniczka gave insufficient attention to the structural problems raised by his philological interpretations. In fact, when we began to work out the details of our own restoration, we concluded that his version of the tent could in all likelihood never have been erected. In our view, it would have been virtually impossible, without intermediate supports, to stretch above the nave an awning measuring more than 1400 m2 ; this objection remains valid despite Studniczka's assumption that there was a large central oculus. If Callixenus correctly states that the nave architraves formed the sole support of the awning, the nave itself must have been much fruits, that sheltered the colossus of Dionysus at 198d), but also in the IIDionysiac saloon ll of Philopator's river-barge (Ath. 205f). 14Delphi: Paus. 10.25-32; Athenian monuments: Paus. 1.3.2-3, 3.4, 5.1, 15.2-4. Houses: cf. the description of the decorative elements of the block of the House of the Comedians, in Delos 27 (Paris 1970). SYfviPCSIUM-TENT OF PTOLEMY" 299 smaller than 32 x 42 m. Ever. in our rroposed nave, with an axial rectangle of not more than 36 x 48 cubits (about 18.75 x 25 m) the "drag" of the great canopy would have been such that the lofty columns would have collapsed, unless they were sunk at least 0.65-0.75 m into the ground. Thus if they had bases at all (and such may well have been the case) these must have been for "S how" only. The peristyle supports must also have been sunk into the ground, to help "buttress" the lofty nave. Even so, the freestanding exterior supports of the peristyle, if they were to serve as anchors for the cables of the awn ing, must themselves have been guyed down, lest the "drag" of the guy-cables of the awning uproot them. In any event, the awning would have been easier to deal with, and also most like a miniature ouranos, if it could be made to curve upward from the sides toward the middle. We have therefore based our restoration of the pavilion on quite a different set of assumptions from those made by Studniczka (Figs. 3-5). We have used thicker columns, assuming a lower diameter of not less than 3, and possibly as much as 3.5-4 cubi ts; the columns, we bel ieve, must have been sunk into the ground, so that the visible bases, if any, were really false bases. 15 We further bel ieve that virtually all decorative elements, and particularly the "caves, II were intended to be visible to the diners gathered in the tent. We therefore suggest that the "caves" were in the upper part of the nave intercolulT1niations. Taking the 8-cubit "s ize" of the caves recorded by Callixenus to refer to the width of the spaces occupied by the "caves" (rather than to the height, as Studniczka assumed), we arrive at axial spans for the nave colonnades of (8+3) to (8+4) = 11-12 cubits; the axial rectangle would then have measured 33 x 44 to 36 x 48 cubits, or some 17.25 to 18.75 rn in width, by 23 15 Like the existing false bases fltted around the columns at the corners of the bays of crossvaulting in the nave of S. Maria degl i Angeli at Rome. 300 F.E. WINTER AND A. CHRISTIE • • • • • • • • • • • • • • • • .. • • • .. • • • • .. • • • • • .. • • • • • • .. • • • • • • 'f FIGURE 3 Proposed new plan of the tent, drawn by A. Christie (scale in cubits) T OF PTOLEMY II SYMPOS IUM-TEN FIGLJRE 4 from above: Structure of tent, roposal . new p Christie drawn by A. 301 302 F.E. WINTER AND A. CHRISTIE FIGURE 5 Proposed new elevation of front of tent: drawn by A. Christie SYMFCSIUM-TENT OF PTOLEMY II 303 to 25 m in length. Whi Ie not insisting on the correctness of interpretation of the "size" of the "caves," we nevertheless maintain that the axial width of the nave colonnade cannot have been significantly larger than the figure that we propose, if it was spanned by real wooden crossbeams. We assume that the beams "on either sidell of the orophe were in fact real beams, not painted representations; these beams would have been laid above the second and fourth columns on either side of the nave, and would have supported light wooden coffers over the bays between the beams and the architraves of the short ends of the nave. The free span of the beams, in our restoration, would have been just over 16 m, i.e. well within the capability of Hellenistic builders to handle. 16 In addition to the two beams above, we bel ieve that there was a third crossbeam, supported by the third column on each long side of the nave. Sunk into the ends of all three beams were bow-shaped metal rods, or diatonaia toxoeide, as Callixenus calls them in his account of the awning over the upper deck of the great river-barge of Ptolemy IV; 17 our diatonaia span about 17 m, as against only 8 m on the river-barge. 18 We assume that the diatonaion over the central beam curved upward more than those over the two end beams, thus imparting a flattened domical shape to the awning stretched over the beams. All diatonaia may have been joined to the wooden beams themselves by braces, to form a sort of triangular truss-system. 19 16Cf. the free spans of the roofed council-chamber in the bouleuterion complex at Miletus (about 15.75 m) and of the original roof-structure of the <.Jssembly-hall at Priene (about 14.25 m). 17Cf. Ath. 205f-206a. 18The "bow"-shape of the diatonaia used on the river-barge must surely indicate that they were r.1etal, not wooden; and the technology that produced 8-meter metal rods could also handle 17-meter rods. 19The time and place at which genuir:c trusses were first used in C reek architecture remains uncert