Very large merchant vessels (500 tons burden or more) and elaborate port facilities certainly existed in Roman world, and they have attracted much attention in recent years. It is important, however, to provide an accurate context for such spectacular technical achievements: do they represent norm, or are they uncommon, or even highly anomalous? In absence of documentary evidence, we can make good use of comparative materials from other preindustrial contexts. Post-Roman merchant fleets regularly consist overwhelmingly of smaller vessels, often carrying small quantities of a large variety of items. Port facilities in these contexts are generally minimal, and such practices as beaching and offloading into lighters are common. It is argued here that most Roman merchant vessels were small (i.e., below 100 tons), and that man-made port facilities were both small and unusual in Roman Empire. Evidence from ancient world, including wrecks known from underwater excavations and numerous literary, legal, and epigraphical texts, supports precisely picture of ancient shipping we would predict on basis of comparative material. As we move ever more deeply into an age of technical marvels, it becomes increasingly difficult to comprehend nature, and especially scale, of ancient world. It is most spectacular technical achievements of ancient world which draw our attention, and this interest is encouraged by nature of our evidence. In literature, for example, we are more likely to find descriptions of gigantic ships such as Isis or Syracusia built by Hiero than of ordinary merchant vessels, and so far our archaeological evidence is too slender to permit inferences about Roman merchant fleet in general. Interest has therefore centered on larger, and in particular largest, merchant vessels. Thirty-five years ago, Lionel Casson proved that ancient merchant vessels could be much larger than previous scholars had thought,' and his evidence and conclusions have been used, reconsidered, and developed further by scholars such as Wallinga2 and Pomey and Tchernia.3 These writers have justifiably noted that our evidence does not permit us to establish an average size for ancient merchant vessels,4 and they have therefore concentrated on establishing probable sizes of larger and/or largest vessels. These are matters of great interest for history of technology and of large-scale trade in ancient Mediterranean, but such an emphasis on larger ships can cause problems: first, there is danger that scholars using these materials will accept figures on larger vessels as if they were norm;5 and second, without a realistic context it is difficult for us to appre* This is a revised and expanded version of a paper originally delivered at 1985 AIA Annual Meeting in Washington, D.C. as part of a Colloquium on Ports, Technology, and Trade in Ancient World, AJA 90 (1986) 201-202. My thanks to Anna Marguerite McCann for organizing colloquium and for her encouraging words as I revised paper. Thanks also to Lionel Casson and anonymous referees for AJA, all of whom made helpful and instructive suggestions. SL. Casson, Size of Ancient Merchant Vessels, Studi in onore di Aristide Calderini e Roberto Paribeni I (Milan 1956) 231-38; summarized and updated in L. Casson, Ships and Seamanship in Ancient World (Princeton 1971) 171-72, 183-200. While I disagree with some of Casson's emphases and general conclusions, my debt to these fundamental studies will be manifest throughout this paper. 2 H.T. Wallinga, Nautika (I). The Unit of Capacity for Ancient Ships, Mnemosyne 17 (1964) 1-40. P3 . Pomey and A. Tchernia, tonnage maximum des navires de commerce romains, Archaeonautica 2 (1978) 233-51. Note, however, that evidence from literary sources concerning unusually large merchant vessels must be handled with care. The figures Lucian gives for such a ship, e.g., are not to be trusted: see G.W. Houston, Lucian's Navigium and Dimensions of Isis, AJP 108 (1987) 444-50. 4 The statement of Pomey and Tchernia (supra n. 3) 233 is characteristic: ne cherchons pas 'i determiner les tonnages moyens.... les sources disponibles ne nous paraissent pas permettre de s'en faire une id6e suffisamment justifie. Cf. Wallinga (supra n. 2) 26. 5 Thus both J.G. Landels, Engineering in Ancient World (Berkeley 1978) 164, and G. Husson, Lucien. Le navire ou les souhaits 2 (Paris 1970) 13, regard ships of 300 tons and over as normal, even though very few actual vessels this large are known. Similarly, O. Vallespin G6mez, C pper Wreck (Pecio del Cobre), IJNA 15 (1986) 322, argues that we have discovered only a small part of this wreck because we have well under 200 tons of cargo and the 553 American Journal of Archaeology 92 (1988) This content downloaded from 157.55.39.211 on Tue, 27 Sep 2016 05:50:20 UTC All use subject to http://about.jstor.org/terms 554 GEORGE W. HOUSTON [AJA 92 Table 1. Ships Docking in London, 1567 and 1568 Tonnage English Ships Foreign Ships Total