Scholia Reviews ns 11 (2002) 31.

A. Trevor Hodge, Roman Aqueducts and Water Supply. London: Duckworth, 2002[2]. Pp. viii + 504, incl. 90 halftones, 135 line illustrations, 8 tables, 7 maps an appendix of facts, figures and formulae, a bibliography and a supplementary bibliography, notes and index. ISBN 0-7156-3171-5. UKú25.00.

Alfonso Burgers
Department of Modern and Classical Languages and Literatures,
University of Cape Town

This seminal book has been brought out in a second edition because the first edition and reprint have probably been sold out. This is not surprising, for it is the most detailed account of Roman aqueducts in the English language. As an engineer I was amazed when I acquired the first edition that such a masterly treatment could be given about a technical subject from a research professor of classics. Trevor Hodge is not alone in this regard: in 1978 J. G. Landels, from the department of classics at the University of Reading, also produced a significant book on engineering in the ancient world.[[1]] Both these authors bring to these subjects supporting insights with details from the ancient historical record. They made use of the technical expertise of engineers and scientists where it was necessary to explain scientific principles. Not being scientists themselves they required that these principles be explained to them in such terms that they could understand them thoroughly, and then translated the ideas into language that is supported by their classical backgrounds suited to their classical readership. The results, two immensely important books relating to ancient engineering.

The second edition differs very little from the first. Only at the end of the original preface and the end of the introduction are there new comments. There is also a supplementary bibliography. In the new comments at the end of the introduction (p. 18), Hodge indicates that he has had a rethink relating to the praise accorded to Frontinus in some modern classical literature,[[2]] though his original comment on Frontinus remains as was stated previously. Frontinus was a very successful soldier and governor of Britain. He was then appointed as curator aquarum to oversee the corrupt department of water affairs in Rome. Although he was not an engineer, he rationalized the water distribution in the city and during his office, wrote an extant book De Aquis Urbis Romae that has become the only ancient source on Rome's water supplies during the Augustan period.[[3]]

After a brief statement about a typical Roman aqueduct, Hodge questions some modern attitudes and interpretations by scholars, both classical and archaeological. He states, 'the aqueducts have generally been studied rather as archaeological monuments than as functioning machines' (p. 4). It was this comment that motivated me to do some research on aqueducts. He says that modern classical students ask 'How did it all work?' (p. 4). Aqueducts were built for several reasons: to provide potable water, to supply water to the high demand Roman baths, and as Hodge says, from the Roman view for 'civic pride' (p. 4). When the modern viewer looks at the visible monumental remains of Roman aqueducts they invariably impress and they certainly filled the contemporary Roman with pride.

The concept of the Roman aqueduct was a development of Hellenistic Greek and other ancient middle-eastern approaches to provide their cities with drinking water and irrigation systems. Nestor's palace had an aqueduct water supply. Hodge briefly also discusses ancient water supply sources such as wells and storage cisterns, stating that aqueducts were not the only sources of water supply to their towns, as these 'also drew copious supplies from a great multiplicity of private (and public) wells and cisterns' (p. 48). Some large Roman cities, for instance in Britain, had no aqueduct water supplies, they relied only on wells.[[4]]

The Romans were very careful from where they obtained 'clean' potable water, which meant that many of their aqueducts were of great length. Some aqueducts were only about three to five kilometres long, but many were in access of eighty kilometres long -- 132 kilometres in the case of the Carthage aqueduct.[[5]] The population of Rome escalated during the fourth and third centuries BC and wells were inadequate to supply the demands for 'clean' water, so the senate appointed the consul Appius to construct the first aqueduct (sixteen kilometers long) to relieve the water shortage. Within the next six centuries Rome was supplied with eleven aqueducts (p. 347). As the empire expanded these structures were built wherever the Roman army established itself or where colonies were provided for retired soldiers and other citizens. Sources of water were often springs situated long distances from the delivery point. Sometimes the Romans resorted to building dams as sources of water supplies and Hodge gives photographic evidence of two Roman earth dams in Spain that survive to this day. Another very old Roman dam that still survives today is on the Orontes at Homs (pp. 87-91). If one considers the technical design implications of modern earth dams it is astonishing that they could construct earth dams that could survive for nearly two thousand years. This is to the great credit of the Roman army engineers; not many modern engineering structures survive that long.

In Chapter 5, 'The Aqueduct' (pp. 93-125), Hodge goes into detail about the construction of the aqueduct without blinding the reader with detailed engineering dimensions. He points out that although there are the magnificent remains of above ground aqueduct bridge structures, they represent a small proportion of the total length of aqueducts, the rest being buried below ground level to various depths. Figure 46 is a very good illustration of the buried channel portion of the 95,4 kilometre long Eifel aqueduct to Cologne. Its source was a very powerful spring, which was still flowing strongly when I saw it in 1996. There are several illustrations of aqueduct channel cross-sections, indicating the variety of closed channel structures that were used by the Romans (pp. 96f.). Wooden and terracotta pipes were also used extensively, with variable efficient ways of sealing the about 0.75 to 1 metre length pipes (pp.106-17).

Construction of aqueducts, whether it was the monumental bridge structures that carried the water channels or the buried pipes and channels, required great skill from the Roman engineers, surveyors[[6]] and skilled labourers - often the soldiers themselves. Hodge illustrates some of the better-known bridge structures such as the Pont du Gard, Taronga, Segovia, Ephesos, Metz, Marcia, and Beaunant at Lyon (pp. 132-44). He mentions some of the statistics of these structures, which indicate their sizes, heights and slenderness. The fact that many of the high aqueducts had no bonding material between the layers of dressed stone points to the meticulous care that was taken by the engineers to ensure their stability. There no doubt were many failures, but they would have contributed to the expertise that the engineers accumulated and passed on for future constructions.

A very important aspect of aqueduct construction was to bring water across valleys that were too long and deep to span for all of its length with stone bridges. They borrowed the concept of the so-called 'inverted' siphon (pp. 147-60),[[7]] to bring water from one high point of the valley to a slightly lower high point on the other side of the valley. The Romans developed this concept to a high level of efficiency. It presented the engineers with great problems, because it required a closed system of piping that was watertight. For this purpose lead piping was encased in concrete so that the system could withstand high hydraulic pressures (pp. 307-15). By this time the Romans had developed a reasonably strong concrete with which they encased the various types of piping and which they used for improved construction of other structures (Chapter 6, pp. 315-20).

Although this new edition of Roman Aqueducts does not use the new available archaeological evidence of the last ten years, this does not detract from the book's importance and pedagogical usefulness. Hodge provides an excellent introduction to the mechanisms of Roman water supplies, illustrated with many examples of their workmanship. The author has tackled a large subject with many technical complications and presented it in a very readable and interesting style, pitched at classicists and archaeologists rather than at scientists. His extensive bibliography of most of the available literature pre-1992 is a great foundation source for further research. However, it must be mentioned that a considerable amount of new evidence has become available during the last ten years that is not reflected here. A very important part of the book is the extensive endnotes for each chapter that appear at the end of the book. The appendix (pp. 346-55) provides valuable statistical information and a guide on the formulae used to determine flow in channels and 'inverted siphons'. The overall impression I have of the book is that a classicist has tackled the demanding task of presenting technical matter in such lucid detail so that non-scientist scholars and students can understand it. This book on Roman aqueducts can be compared to the simple and lucid style of Euclid's thirteen books on geometry and other forms of mathematics. Like Euclid's combined book, Hodge's book will remain a basic source on Roman aqueducts and water supplies for a long while yet.


[[1]] J. G. Landels, Engineering in the Ancient World. (Berkeley and London 1978).

[[2]] D. R. Blackman and A. Trevor Hodge, Frontinus' Legacy (Ann Arbor 2000).

[[3]] See Hodge 2002[2] 16-18. I have not seen the work of Blackman and Hodge ([2] above). Hodge is critical of Frontinus because he was fuzzy in his thinking on engineering matters and that he seemed to have written in a subtle way to enhance his own ego. However, as he points out Frontinus 'remains our prime source for Roman aqueducts' (p. 18). Not only was his book a prime source of aqueducts, but he is also a prime source on Roman pipe sizes and the quinariae capacities of pipes. His methods of calculation may not have been accurate or correct, but it has provided some insight for modern scholars on how the Roman system of arithmetic was used and applied to practical problems. Sextus Iulius Fontinus (tr. Charles E. Bennett, ed. Mary B. McElwain), The Stratagems and The Aqueducts of Rome (Cambridge, Mass. and London 1925) 365-403.

[[4]] A. Burgers, The Water Supplies and Related Structures of Roman Britain (Oxford 2001) 100-3.

[[5]] F. Rakob, 'L'aqueduc de Carthage', Doss. Arch. 38 (1979); cf. Hodge [3] 348.

[[6]] Frontinus and several other ancient authors compiled a compendium on Roman land surveying practice Referred to as the Corpus Agrimensorum Romanorum, Recently edited by B. Campbell and published by the JRS Monograph No.9 (Oxford 2000). Roman surveying was amazingly accurate considering the very crude instruments available to them. For their aqueducts, bridge structures, roads and town layouts they used specially trained soldiers to survey the routes and detailed measurements of the structures and land. A prime example of the degree of accuracy they achieved is illustrated by the Nimes aqueduct that had a difference in height between the source and delivery point of only 17 metres in 51 kilometres (Hodge 2002[2] 371). This achievement was repeated on many aqueduct routes, although there were also some interesting failures.

[[7]] The term 'inverted siphon' was a misnomer introduced by some engineer to describe the 'pressure' system that forced water along the up-slope of a pipe system in a valley as illustrated in Hodge's figure 102 (p. 148). Water flows down the slope from the source (left side) and then must be forced under pressure up the slope on the right side. It is purely a pressure system - neither a siphon nor an inverted siphon. I think it is a pity that this incorrect terminology continues to be used in the literature.