Published online by Cambridge University Press: 07 September 2009
A good supply of water was rightly regarded as one of the essential commodities for the maintenance of urban life in the ancient world. One of the major problems with which city authorities had to deal was the maintenance of adequate supplies of water to satisfy the domestic, public, recreational, and industrial demands of the inhabitants. The Romans were particularly renowned for their hydraulic technology in general and the construction of aqueducts in particular, often bringing water from great distances. The geographer Strabo praised the engineering skills of the Romans, maintaining that veritable rivers of water flowed by means of aqueducts through the city of Rome. Close on a century later the first curator of Rome's water supply and one-time military governor of Britain, Sextus Julius Frontinus stated the same, if a little more pointedly, when he compared the achievements of the Romans in the field of water supply with the ‘idle pyramids of the Egyptians or the glorious but useless monuments of the Greeks’.
1. Aristotle, , Pol. 7. 1330 bGoogle Scholar.
2. E.g. Rome (50–90 m.), Cologne (75 m.), and Carthage (132 m.), see Hodge, A. T., ‘Aqueducts’, in Barton, I. M. (ed.), Roman public buildings (Exeter, 1989), pp. 129–30Google Scholar; Frejus (40 m.) and Nimes (50 m.), see Grenier, A., Manuel d'archeologie gallo-romaine IV (Paris, 1960), pp. 41–45, 88–97Google Scholar.
3. Strabo, 5.3.8.
4. Frontinus, , De Aqueductibus 1.16Google Scholar.
5. Coulton, J. J., ‘Roman aqueducts in Asia Minor’, in Macready, S. and Thompson, F. H. (eds.), Roman architecture in the Greek world (London, 1987), pp. 72–73Google Scholar points out that it was not through lack of the requisite technology that the Greeks did not make extensive use of long aqueducts but the unstable military situation. The siege of Syracuse by the Athenians during the Sicilian expedition illustrates the dangers. One of the actions taken by the Athenians was to cut off the water supply to the city (Thuc. 6.100). In Greece water was brought to Olynthos from springs about 6 miles from the city, see Robinson, D. M. and Graham, J. W., Excavations at Olynthus VIII (Baltimore, 1938), p. 307Google Scholar, Robinson, D. M., Excavations at Olynthus XII, pp. 103–14Google Scholar. Similarly in Sicily Akragas and Syracuse tapped external sources of water, see Burns, A., ‘Ancient Greek water supply and city planning: a study of Syracuse and Acragas’, Technology and Culture 15 (1974), 389–412CrossRefGoogle Scholar.
6. Despite the obvious spectacular structures such as the Pont du Gard, and the aqueducts of Segovia, Lyon, and Rome, in general less than 10° 0 of an aqueduct was usually carried on arched bridges, see Smith, N., ‘Roman hydraulic technology’, SciAm 238.5 (1978), 154Google Scholar; Hodge, , op. cit., p. 129Google Scholar.
7. Pausanias, 10.4.1, includes running water at a fountain head as one of the essential amenities which characterized an ancient city.
8. Philostratus, , Vitae 2.1 (551)Google Scholar.
9. Procopius, , De Aedific. 6.2Google Scholar.
10. See Philostratus, , Vitae 2.1 (548)Google Scholar; Pliny, , Epist. 10.38, 90Google Scholar.
11. Pliny, , Epist. 10.37Google Scholar.
12. Philostratus, , Vitae 2.1 (551)Google Scholar.
13. Stephens, G. R., ‘Civil aqueducts in Britain’, Britannia 16 (1985), 198–9CrossRefGoogle Scholar.
14. Hodge, , op. cit., pp. 130–2Google Scholar.
15. Pliny, , Epist. 10.90–91Google Scholar.
16. Smith, , op. cit., 154–5Google Scholar.
17. See Coulton (n. 5 above).
18. Procopius, , De Bello Gothico 5.9–10Google Scholar.
19. Procopius, , op. cit., 5.8.44Google Scholar.
20. Procopius, , op. cit., 5.19.18Google Scholar.
21. Procopius, , op. cit., 5.19.28–29Google Scholar. Rome relied on wells before the construction of the first aqueduct, see Frontinus, , De Aqueductibus 1.4Google Scholar. Although in general river water was considered unsuitable for drinking purposes (see Palladius, , De Re Rustica 1.17)Google Scholar, during a siege water from any source would be used.
22. Procopius, , op. cit., 5.19.27Google Scholar.
23. Dittenberger, , OGIS 483 lines 203–32Google Scholar.
24. To refresh travellers, see the Dipylon fountain house at Athens, Wycherley, R. E., The stones of Athens (Princeton, 1978), p. 19Google Scholar. See also Olynthos for a fountain house at a gate on the southern hill, Robinson, D. M., Excavations at Olynthus II (Baltimore, 1928), pp. 11–14Google Scholar.
25. Only one instance of the interconnecting pipe was found in situ but the partial collapse of the party walls between tanks, at the point where such pipes were located, indicates that all the internal partition walls were so pierced.
26. Vitruvius, , De Architectura 8.6.5Google Scholar.
27. The Romans were fully conversant with the technology of the syphon and it was probably utilized more extensively than has been supposed, see Hodge, A. T., ‘Syphons in Roman aqueducts’, PBSR 38 (1983), 174–221, esp. 220–1Google Scholar. The Kremna syphon was less than 300 m. in length, compare Aspendos (0.8 km.) and Les Tourillons, supplying Lyon (ca. 6 kms), see Hodge, ibid., 185–9. Hodge, ibid., 220, calculates that the total length of the syphons in the Lyon area comes to 16.6 kms!.
28. Vitruvius, , op. cit., 8.6.5Google Scholar.
29. Vitruvius, , op. cit., 7.6.8–10Google Scholarrecommends the use of clay pipes because they are cheaper, easy to construct and maintain, and because water carried in clay pipes remains pure. Terracotta pressure pipes were used in aqueducts in Spain, Hodge, , op. cit., 175Google Scholar. Lead, stone, and terracotta encased in masonry were also used, e.g. see Hodge, ibid., 181–2, 189 (lead), Coulton in Macready and Thompson (eds.) (1987), p. 74 (stone), Thompson, , ‘The Roman aqueduct at Lincoln’, ArchJ 111 (1954), 119 (terracotta sheathed in concrete)Google Scholar.
30. Vitruvius, , op. cit., 8.6.8Google Scholar recommends a minimum thickness of two digits. As far as the author is aware the Kremna pipes are the thickest yet found.
31. See Hodge, , PBSR 38 (1983), 195–202Google Scholar for discussion of the hydrology of pressure pipes.
32. See Oleson, J., Greek and Roman mechanical water lifting devices (Toronto, 1984), pp. 242–8, 355Google Scholar.
33. Thompson (1954), 120–5. Wacher, J., The towns of Roman Britain (London, 1974), pp. 126–32Google Scholar questions whether the accepted source for the aqueduct is correct.
34. See Hodge, ibid., 192. The original excavator suggested a force pump might have been used.
35. Vitruvius, , De Architectura 10.4.1–7.5Google Scholar. J. Oleson (n. 32) has catalogued the available documentary and archaeological evidence for water lifting devices.
36. Landels, J. G., Engineering in the ancient world (London, 1978), p. 74Google Scholar calculates for a height of 16 m. a theoretical rate of lift of 13.63 litres of water per minute for one man, using a bucket and chain system.
37. Zosimus 1.69–70; Mitchell, S. and Waelkens, M., ‘Cremna and Sagalassus 1987’, AS 38 (1988), 57–58Google Scholar. As already discussed, cutting the water supply would have been one of the first actions of any besieging force.
38. See above n. 31.
39. Twelve interconnecting tanks stored water for the Helena baths at Rome, see Nash, E., A pictorial dictionary of ancient Rome II (London, 1962), pp. 454–7Google Scholar. The baths of Caracalla had sixty-four interconnecting chambers, see Nash, , op. cit., p. 434Google Scholar.
40. Hodge in Barton, (ed.), op. cit., p. 128Google Scholar.
41. Procopius, , De Bello Gothico 5.19.27Google Scholar.
42. Stephens, , op. cit. (n. 13), 197–8, 201, 203Google Scholar.
43. Coulton, in Macready, and Thompson, (eds.), op. cit., p. 82Google Scholar.