4. Curtius Rufus, _Hist. Alex._ V1: Super arcem, vulgatum Graecorum fabulis miraculum, pensiles horti sunt, summam murorum altitudinem aequantes multarumque arborum umbra et proceritate amoeni. Saxo pilae, quae totum opus sustinent, instructae sunt, super pilas lapide quadrato solum stratum est patiens terrae, quam altam iniciunt, et humoris, quo rigant terras: adeoque validas arbores sustinet moles ut stipites earum VIII cubitorum spatium crassitudine aequent, in L pedum altitudinem emineant frugiferaeque sint, ut si terra sua alerentur. Et cum vetustas non opera solum manu facta, sed etiam ipsam naturam paulatim exedendo perimat, haec moles, quae tot arborum radicibus premitur tantique nemoris pondere onerata est, inviolata durat, quippe XX [pedes] lati parietes sustinent, XI pedum intervallo distantes, ut procul visentibus silvae montibus suis inminere videantur. Syriae regem Babylone regnantem hoc opus esse molitum memoriae proditum est, amore conjugis victum, quae desiderio nemorum silvarumque in campestribus locis virum conpulit amoenitatem naturae genere hujus operis imitari. It would lead us too far afield were I to attempt here to emphasise all the points that weigh for or against my contention; I may safely leave the decision to time. According to Berosus, the hanging gardens must have been on the Kasr, as he places them in a formal and detailed manner in the area of the buildings by which Nebuchadnezzar enlarged the palace of Nabopolassar. The Principal Citadel may also be taken into consideration in this connection, and the question can only be settled by excavation. A difficulty that is apparently serious lies in the length of the side of the quadrangle, which is given by Strabo and Diodorus as 4 plethra (about 120 metres). On examining the central building we find this is exactly four times its actual measurement, and any one who holds fast by these figures will naturally reject my hypothesis. I have been too often misled by ancient statements of measurements to treat this information as conclusive, and as in the case of Herodotus’ statements with reference to the town walls, I consider it possible that the length and breadth have been confused with the circumference. The central building rested on the 16 walls that supported the vaults, and on the 4 walls surrounding them, 20 in all. Thus a non-philologist might readily conclude that the text of Curtius quoted above, “Haec moles ... durat, quippe XX lati parietes sustinent,” can be correct without the addition of “pedes” inserted between _XX_ and _lati_. The 10– to 11–foot span of the vaults can be seen to-day in the ruins in approximate accordance with the statement of Diodorus and Curtius. I would attach little importance to any of these details, and lay stress only on the main facts. Our authors here speak of a building, with characteristics completely different from all others, and precisely similar to those of the vaulted building.