Imported Building Materials of Sebastos Harbour, Israel

Abstract

Archaeological research of the remains of Herod the Great's Sebastos Harbour has identified foreign building materials, corroborating Josephus’ description of its construction. The three materials are 17,000 m³ of volcanic ash (pozzolana) from Puteoli, 11–14,000 m³ of timber from Seleucia Pieria, and cobblestones of unidentified origin. These materials and their importation are discussed, using past provenance analysis, new data, and historical and logistical considerations. The pozzolana and timber importation demanded dozens of transport ships, and is evidence of the sophistication of the construction of Sebastos Harbour and the efficient organization of the Mediterranean under Augustus’ control.

© 2007 The Author

Key words:

• Caesarea Maritima
• Roman harbour
• provenance analysis
• Puteoli
• Antioch
• Seleucia Pieria

Acknowledgements

This article is derived from the author's University of Haifa Department of Maritime Civilizations MA thesis, and was supported by a Maurice Hatter Award. The author wishes to thank those associated with the Department for their support, particularly his advisors Professor Yossi Mart, Professor Michael Artzy, and the late Professor Avner Raban. He also greatly appreciates the efforts of the members of the 2002 Caesarea marine excavation team and many others who assisted with this research.

Notes

1 Samples of pozzolana collected in 2005 during ROMACONS research were also determined to have come from Puteoli (pers. comm. J. P. Oleson, 29 July 2006).

2 A figure proposed by Boyce et al. (2004: 135) based on a conceptual model of the mole structures, in which the harbour is built primarily of Area K type concrete walls. This figure equates to some 200 forms, each containing a volume of about 390 m³ of concrete.

3 About 17,300 tons, considering a pozzolana bulk density of 1000 kg/m³. This density figure is an estimate derived from the specific gravity of 1.58 for the pozzolana isolated from a concrete sample from Area E2 (Oleson, 1989b: 202), which does not account for the interstitial pore space and the fact that there is considerable variability of pozzolana even within a single quarry.

4 See Oleson (1985) and Brandon (1997) for analysis of the architecture of Areas G and K formworks respectively (Fig. 1).

5 These samples were placed in deionized water rather than sea‐water during their transport to the botanical laboratory, causing deformation, limiting analysis to the genus level (Liphschitz et al., 1989: 191). Nevertheless the researchers were able to determine that the Pinus, Populus and Quercus sp. samples were distinct from the local varieties (Liphschitz et al., 1989: 194). A report of Picea europea (Oleson, 1985: 168) is not accurate as the species does not exist.

6 Pinus silvestris, if indeed it was the species (rather than Pinus nigra), does not grow as far south as Cyprus. Pinus brutia and Pinus nigra are also found in large numbers elsewhere such as Turkey.

7 The sampling was performed with the guidance of Chris Brandon and Avner Hillman, and the species identification was performed by Nili Liphschitz of the Institute of Archaeology at Tel Aviv University, to all of whom I am grateful. In Dr Liphschitz's report it is mentioned that the only country in which these species grow together is Turkey.

8 Determinations of the tree‐growing regions are based on their positive reference for particular locations in one or more of these texts: G. E. Post and J. E. Dinsmore (1932); P. H. Davis, M. J. E. Coode, and J. Cullen (1965); M. Zohary (1973); R. D. Meikle (1977); K. Browicz and J. Zieliânski (1982); G. Krüssmann (1991); M. Rival (1991); M. Vidakovic (1991); T. G. Tutin (1993); K. Browicz (1996); T. Skroppa (2003). The sample identified as Pinus silvestris or Pinus nigra mentioned above is not considered further as both are only possibilities.

9 Quercus coccifera ssp. calliprinos, such as that documented by Meikle (1977: 1484–1486), is distinguished by Dr Nili Liphschitz and this author differently, rather as Quercus calliprinos. For the same reason Quercus coccifera ssp. coccifera, not identified on Cyprus, is considered here as Quercus coccifera.

10 Quercus coccifera is the only species not keyed to the Amanus mountains, specifically growing only west of the Taurus mountains (Zohary, 1973: 356). Anatolia's coast west of the Amanos is however deficient in Picea orientalis and Fagus sylvatica. The Quercus coccifera was unlikely to have been found in the region, and may be considered to have been trans‐shipped from a location further west. The Abies sp. found in the Amanos region, is of the genus cilicica, and that of the Picea sp., orientalis.

11 In contrast, the six river‐stones collected from Carmel coastal streams displayed Mohs ratings of 3–3.5. Thanks are due to the General Directorate of Highways, Technical Research Department, Geological Services Branch in Ankara, Turkey for performing these Mohs hardness tests. For the stone‐type identification of the samples I thank the petrographic analysis of Dr Eşref Atabey of the Mineral Research and Exploration General Directorate, Department of Geology, Balgat‐Ankara, Turkey.