Published in the United Kingdom in 2014 by

OXBOW BOOKS

10 Hythe Bridge Street, Oxford OX1 2EW

and in the United States by

OXBOW BOOKS

908 Darby Road, Havertown, PA 19083

© Oxbow Books and the individual authors 2014

Reprinted in paperback DATE

Hardcover Edition: ISBN 978-1-78297-420-8

Digital Edition: ISBN 978-1-78297-421-5; Mobi: ISBN 978-1-78297-422-2; PDF: ISBN 978-1-78297-423-9

A CIP record for this book is available from the British Library

Library of Congress Cataloging-in-Publication Data

Brandon, C. J. (Christopher J.)

  Building for eternity : the history and technology of Roman concrete engineering in the sea / by C.J. Brandon, R.L.

Hohlfelder, M.D. Jackson and J.P. Oleson ; with contributions by L. Bottalico, S. Cramer, R. Cucitore, E. Gotti, C.R.

Stern and G. Vola ; edited by J.P. Oleson.

    1 online resource.

  Summary: "This book explains how the Romans built so successfully in the sea with maritime concrete. The story

is a mix of archaeological, geological, historical and chemical research, with relevance to both ancient and modern

technology. It also bridges the gap between science and the humanities by integrating analytical materials science,

history, and archaeology, along with underwater exploration. The book will be of interest to anyone interested in Roman

architecture and engineering, and it will hold special interest for geologists and mineralogists studying the material

characteristics of pyroclastic volcanic rocks and their alteration in seawater brines. The demonstrable durability and

longevity of Roman maritime concrete structures may be of special interest to engineers working on cementing materials

appropriate for the long-term storage of hazardous substances such as radioactive waste"--Provided by publisher.

  Includes bibliographical references and index.

  Description based on print version record and CIP data provided by publisher; resource not viewed.

  ISBN 978-1-78297-421-5 (epub) -- ISBN 978-1-78297-422-2 (mobi (kindle)) -- ISBN 978-1-78297-423-9 ( pdf)

-- ISBN 978-1-78297-420-8 (hardcover) 1. Concrete construction--Rome--History. 2. Concrete construction--

Research--Mediterranean Region. 3. ROMACONS Project. 4. Marine engineering--Rome--History. 5. Technology-

-Rome--History. 6. Architecture, Roman. 7. Rome--Antiquities. 8. Mediterranean Region--Antiquities. 9.

Geology--Mediterranean Region. 10. Volcanic ash, tuff, etc.--Mediterranean Region--Analysis. I. Hohlfelder, Robert L.

II. Jackson, M. D. III. Oleson, John Peter. IV. Bottalico, L. (Luca) V. Title.

  TH16

  627'.702--dc23

2014032468

All rights reserved. No part of this book may be reproduced or transmitted in any form or by any means, electronic or

mechanical including photocopying, recording or by any information storage and retrieval system, without permission

from the publisher in writing.

Printed in the United Kingdom by Berforts Information Press

For a complete list of Oxbow titles, please contact:

Oxbow Books is part of the Casemate Group

Front cover: Reconstruction of workers lowering mortar into formwork at Sebastos (Hohlfelder 1987: 264–65) (National Geographic Society, used with permission).

To the unknown master builders of ancient Rome who challenged and tamed the sea, and to the CTG Italcementi Group for unwavering support from the inception of the Roman Maritime Concrete Study through the publication of this volume. Their embrace of our vision made this book possible.

CONTENTS

Preface and Acknowledgements

List of Abbreviations

List of Figures

List of Tables

List of Contributors

1. THE TECHNOLOGY OF ROMAN MARITIME CONCRETE (J. P. Oleson and M. D. Jackson)

1.1. Introduction

1.2. The unique character of Roman maritime concrete

1.3. Recent research on Roman concrete

1.4. ROMACONS research questions

1.5. Summary of the archaeological and engineering significance of the analyses of the ROMACONS samples

2. ANCIENT LITERARY SOURCES CONCERNED WITH ROMAN CONCRETE TECHNOLOGY (J. P. Oleson)

2.1. Theophrastus

2.2. M. Porcius Cato

2.3. Vitruvius Pollio

2.4. Q. Horatius Flaccus

2.5. P. Virgilius Maro

2.6. Strabo

2.7. L. Annaeus Seneca

2.8. Pliny the Elder

2.9. P. Papinius Statius

2.10. Flavius Josephus

2.11. Pliny the Younger

2.12. C. Suetonius Tranquillus

2.13. Apuleius

2.14. Cassius Dio

2.15. M. Cetius Faventinus

2.16. Procopius of Caesarea

2.17. Inscriptions

3. HISTORY AND PROCEDURES OF THE ROMACONS PROJECT (C. J. Brandon and R. L. Hohlfelder)

3.1. History of the project

3.2. Coring equipment and procedures

4. NARRATIVE OF THE ROMACONS FIELDWORK (R. L. Hohlfelder and C. J. Brandon)

4.1. Portus, Fieldwork July–August 2002

4.2. Antium, Fieldwork August 2002

4.3. Cosa, Fieldwork July–August 2003

4.4. Santa Liberata, Fieldwork June 2003, September 2004, and June 2005

4.5. Caesarea Palaestinae, Fieldwork October 2005

4.6. Baianus Lacus, Baianus Sinus, and Portus Iulius (Bay of Pozzuoli), Fieldwork September 2006

4.7. Alexandria, Fieldwork May 2007

4.8. Chersonesos, Fieldwork September 2007

4.9. Egnatia, Fieldwork May 2009

4.10. Pompeiopolis, Fieldwork August 2009

5. THE BRINDISI PILA REPRODUCTION (J. P. Oleson)

5.1. The reconstruction project: Methods and materials

5.2. Formwork design

5.3. Construction of the formwork

5.4. Preparation of the mortar

5.5. Placement of the mortar and aggregate

5.6. Conclusions from the reconstruction experiment

6. MARITIME CONCRETE IN THE MEDITERRANEAN WORLD (C. J. Brandon)

6.1. Important sites not sampled by ROMACONS

6.2. Catalogue of maritime concrete structures around the Mediterranean and Portugal

7. SEA-WATER CONCRETES AND THEIR MATERIAL CHARACTERISTICS (M. D. Jackson and collaborators)

7.1. Introduction

7.2. Geologic materials of the concretes

7.3. Concrete mix design and preparation

7.4. Pozzolanic cementitious processes in the sea-water mortars

7.5. Material properties of the maritime concretes

7.6. Inferences regarding durability of the ancient sea-water concrete

7.7. Summary of analytical methods

8. ROMAN FORMWORK USED FOR UNDERWATER CONCRETE CONSTRUCTION (C. J. Brandon)

8.1. The Role of formwork in Roman concrete construction

8.2. A Typology of Roman formwork design for marine construction: Fixed forms

8.3. A Typology of formwork design for underwater construction: Prefabricated and floating forms

8.4. Conclusions

9. ROMAN MARITIME CONCRETE TECHNOLOGY IN ITS MEDITERRANEAN CONTEXT (R. L. Hohlfelder and J. P. Oleson)

9.1. Trade in pozzolana, pumiceous ash pozzolan, and caementa

9.2. Mechanisms for the spread of innovation in Roman marine construction

9.3. Conclusions: Society, trade, and technology in the Roman Mediterranean

APPENDIX 1: Glossary of Technical Terms (M. D. Jackson and J. P. Oleson)

APPENDIX 2: Schedule of Samples Collected for Preliminary Study Prior to the ROMACONS Project (C. J. Brandon and M. D. Jackson)

APPENDIX 3: Catalogue and Descriptions of Concretes Drilled from Marine Structures by ROMACONS (J. P. Oleson, M. D. Jackson and G. Vola)

A3.1. Santa Liberata

A3.2. Portus Cosanus

A3.3. Portus

A3.4. Portus Traiani

A3.5. Antium

A3.6. Baiae

A3.7. Secca Fumosa

A3.8. Portus Iulius

A3.9. Egnatia

A3.10. Brindisi

A3.11. Chersonesos

A3.12. Pompeiopolis

A3.13. Caesarea Palaestinae

A3.14. Alexandria

APPENDIX 4: Compositional Analyses of Concretes Drilled from Harbour Structures by ROMACONS (M. D. Jackson and G. Vola)

Bibliography

PREFACE AND ACKNOWLEDGEMENTS

In a project of this nature, in which the four participants all contributed their special skills while at the same time sharing the many collective tasks and interweaving their knowledge into the final publication, the only just way to acknowledge everyone’s contribution is to list the principal authors alphabetically on the title page.

Since this project has lasted for over a decade and has involved numerous countries and archaeological sites, we naturally have been helped by a great number of individuals, institutions, laboratories, and agencies. If we have inadvertently missed anyone here, we express our apologies.

Our greatest debt of gratitude is to the CTG Italcementi Group. We remain indebted to Dr. Luigi Cassar of Italcementi who supported our research in so many ways at its inception. Dr. Enrico Borgarello continued this enlightened patronage and provided a generous subvention towards the cost of publishing this book. We also thank Mr. Dario Belotti, Ms. Isabella Mazza, Mr. Piero Gandini, and Mr. Massimo Borsa for their invaluable assistance in surmounting numerous logistical and practical problems at every stage of our project. We especially thank Mr. Bruno Zanga for his very precise and nuanced mineralogical descriptions of the concretes. Dr. Gabriele Vola, Mr. Zanga, and Dr. Emanuele Gotti, the principal CTG Italcementi scientific staff who carried out analytical investigations of the concretes, appear in the book as contributors.

We could not have carried out our field work and laboratory analyses without generous funding from the following foundations, granting agencies, and institutions: The Social Sciences and Humanities Research Council of Canada, Loeb Classical Library Foundation, Taggart Foundation, University of Victoria, and University of Colorado at Boulder. A generous last-minute grant from the Planet Ocean Exploration Foundation topped up the funds needed for the publication subvention for this book.

The staff and associated faculty at the Oxford Centre for Maritime Archaeology provided enormous scholarly and practical advice, particularly Mr. Franck Goddio, Mr. Jonathan Cole, Dr. Damian Robinson, and Prof. Andrew Wilson.

At the University of California at Berkeley, Professor Paulo J. M. Monteiro, in the Department of Civil and Environmental Engineering, and Professor Hans-Rudolf Wenk at the Department of Earth and Planetary Science, and graduate students and colleagues, Ms. Sejung R. Chae, Dr. Abdul-Hamid Emwas, Dr. Penghui Li, Dr. Cagla Meral, Dr. Juhyuk Moon, Dr. Sean Mulcahy, and Dr. Rae Taylor, collaborated with M. D. Jackson in innovative investigations of the fine-scale compositions and material characteristics of the cementitious binding hydrates of the ancient sea-water concretes. Mr. T. Teague provided especially valuable laboratory support. Many of these experiments were conducted at the Advanced Light Source of the Lawrence Berkeley Laboratories. Professor Barry Scheetz, at Pennsylvania State University, also contributed to this research.

At the American Academy in Rome, Dr. Lester Little and Dr. Archer Martin generously facilitated our planning and permit applications in Italy.

At the University of Naples, Professor Vincenzo Morra, Professor Maurizio de’Gennaro, and Professor Piergiulio Cappelletti and their graduate students, Dr. Corrado Stanislao and Ms. Concetta Rispoli, collaborated with Dr. Gabriele Vola, to provide comprehensive mineralogical descriptions of the eastern Mediterranean harbour concretes. We are deeply appreciative of the thoughtful review that Professor Morra and Dr. Lorenzo Fedele gave Chapter 7, which greatly improved the presentation of the analytical results. The conclusions drawn from these analyses are solely our own.

Many individual scholars provided invaluable advice: Alessandra Benini, Dr. Fiona Brock, Mr. Enrico Felici, Prof. Elaine Gazda, Dr. Anna Marguerite McCann, and Robert Yorke. We thank Professor David Blackman for reading the entire manuscript so attentively and providing so many valuable suggestions, and Professor Floyd McCoy for his careful review and thoughtful comments regarding Chapter 7.

The coring at Santa Liberata and Cosa could not have taken place without the generous assistance of Dr. Pamela Gambogi, Direttore, Nucleo Operativo di Archeologia Subacquea, Soprintendenza Archeologica per la Toscana, and her team, Paolo Volpe and Archangelo Alessandiri. The Guardia di Finanza at Porto Santo Stefano generously provided a boat and crew: Paolo Gennaro, Sergio DiMauro, Pero Ronolo, Gianfranco Atzori, Imberto Martini, Enzo Timordidio.

For permits and assistance at Portus and Portus Traiani we thank Dr. Anna Gallina Zevi, Soprintendente, Soprintendenza per i Beni Archeologici di Ostia, Dr. Morelli, Direttore del Museo delle Navi di Ostia and her staff, Dr. Lidia Paroli, Dr. Anna Maria Reggiani, Soprintendente, Soprintendenza per i Beni Archeologici di Lazio. Dr. Cinzia Morelli, Soprintendenza Per I Beni Archeologici di Ostia and Dr. Antonia Arnoldus-Huyzendveld, helped C. J. Brandon survey the levels in the Claudian and Trajanic harbours

We received invaluable assistance at Anzio from Dr. Annalisa Zaratinni, Direttore, Nucleo Operativo di Archeologia Subaquea, Soprintendenza per i Beni Archeologici di Lazio. Our work at Baia was made possible by the kind assistance of Dr. Stefano De Caro, Dr. Paola Miniero, Dr. Maria Luisa Nava, Mr. Jonathan Cole, Dr. Dante Bartoli, and Mr. Derek Klapecki.

For our fieldwork at Caesarea, Prof. Michal Artzy, Director of the Recanati Centre for Underwater Archaeology at Haifa University, generously put the resources of her centre at our disposal and provided invaluable assistance with permits and other practical issues. We were ably assisted on site by Greg Votruba and John Tresman.

Dr. Rita Auriemma of the Università di Lecce graciously assisted with the excavation permit and with many practical aspects of our work at Egnatia.

During the construction of the reproduction pila at Brindisi we were given great assistance by the Lega Navale of Brindisi and its president, Ammiraglio R. Fadda. They provided the venue for the pila experiment and every possible courtesy and consideration while we disrupted the normal activities of the club. We thank Francesco Retta, Fabrizio Orlandino, Peppino Brescia, and Mario Colucci of Italcementi, Brindisi, who volunteered to work with us as their regular schedules permitted. We were assisted for the first core sample at Brindisi by Dante Bartoli, then a graduate student at the Institute of Nautical Archaeology at Texas A&M University, and for the second core sample by Mr. Pietro Brescia and Mr. Antonio Vendeita of Italcementi, Brindisi.

For the fieldwork in the eastern harbour of Alexandria, Franck Goddio, of the Institut Européen d’Archéologie Sous-Marine (IEASM) obtained permits and facilitated our work, supported by Bernard Camier, Jonathan Cole, and Zizi Louxor.

At Chersonesos, Tolis Vougioukas of the Dive Centre Creta Maris kindly supplied the diving tanks and helped with the hire of the boats from which we operated. The Director, Ms. Maria Bredaki, of the 23rd Ephorate of Prehistoric and Classical Antiquities in Heraklion very kindly agreed that her assistant, Eirini Karousou, should supervise our work when no other inspector was available.

Prof. R. Yagçi of Dokuz Eylul University in Izmir made possible our work at Pompeiopolis by facilitating our permit and our coring activities. Mr. Akin Kaymaz gave us invaluable advice and logistical support in the field. Prof. L. Vann, and Prof. Nicholas K. Rauh were also of great assistance in the planning and execution of the work.

In Istanbul, Dr Ismail Karamut and Metin Gökçay very kindly arranged access to the excavations at Yenikapı.

In the text of the book, dates are AD unless labelled BC. Place names in Latin or other foreign languages are not italicized.

LIST OF ABBREVIATIONS

LIST OF FIGURES

Unless otherwise indicated, the photographs were taken by members of the ROMACONS team.

Fig. 1.1. Puteolanus pulvis (pozzolana) from a quarry near Baia.

Fig. 2.1. Reconstruction of the three types of forms mentioned by Vitruvius: a. Inundated form constructed in situ. b. Cofferdam constructed in situ and dewatered. c. Casting pilae sequentially on shoreline (C. J. Brandon).

Fig. 2.2. The Roman pier at Pozzuoli in the mid-eighteenth century (Paoli 1768: pl. XIII) (Courtesy of the Bodleian Library, University of Oxford; Arch. Antiq. B subt. 18, Plate XIII).

Fig. 2.3. Glass bottle with engraving of Puteoli harbour and Baiae (Courtesy of Narodny Museum, Prague).

Fig. 3.1. Drawing of concrete block in Area G, Sebastos (R. L. Vann).

Fig. 3.2. Map of the sites cored by ROMACONS, 2002–2009, along with other sites mentioned in the text (Will Foster Illustration).

Fig. 3.3. Brandon collecting hand cored mortar sample, Chersonesos.

Fig. 3.4. CORMET drill bits with hard metal teeth (Photo: Cordiam).

Fig. 3.5. CORPAX drill bit with PCD (polycrystalline diamond) teeth (Photo: Cordiam).

Fig. 3.6. CORSIN drill bit with a PCD tooth (polycrystalline diamond) surface set (Photo: Cordiam).

Fig. 3.7. CORSET drill bit with natural diamonds set on the cutting surface (Photo: Cordiam.)

Fig. 3.8. CORDIM drill bit with synthetic diamond powder (Photo: Cordiam.)

Fig. 3.9. Corset and Cordim bits and resulting cores (Will Foster Illustration).

Fig. 3.10. Nine centimetre diameter core samples prepared for testing (Photo: Italcementi).

Fig. 3.11. Continuous coring system (Will Foster Illustration).

Fig. 3.12. Santa Liberata core SLI.2004.01 with human scale.

Fig. 3.13. Use of wrenches to separate core tubes.

Fig. 3.14. Cordiam Hydro 25 rotary drill motor unit.

Fig. 3.15. Drill motor unit mounted on Cordiam M60 rack clamped to scaffold frame.

Fig. 3.16. Frame constructed of iron scaffold tubes.

Fig. 3.17. Cordiam M60 rack clamped to purpose-made aluminium frame.

Fig. 3.18. Feet clamped to frame and fixed to concrete with an expansion bolt.

Fig. 3.19. Diagram ofexpansion bolt assembly (Will Foster Illustration).

Fig. 3.20. Hydraulic fluid pump (foreground), electric generator (background).

Fig. 3.21. Twin hydraulic hoses connected to the drill motor unit.

Fig. 3.22. MAG 15 hydraulic rotary hand drill.

Fig. 3.23. Submersible water pump hanging off tender boat.

Fig. 3.24. Drilling equipment being loaded into a rental van.

Fig. 3.25. Drilling equipment packed into a crate for air freight transportation.

Fig. 3.26. Guardia di Finanza patrol boat with support vessel at Santa Liberata.

Fig. 3.27. Small inflatable dinghy used to transport equipment at Santa Liberata.

Fig. 3.28. Fishing boat and zodiac used to transport equipment at Caesarea.

Fig. 3.29. Two small rental boats used for offshore coring at Baia.

Fig. 3.30. Small caique and rowing boat at site of CHR.2007.02, Chersonesos.

Fig. 3.31. Princess Douda in the Eastern Harbour of Alexandria.

Fig. 3.32. Dive boat at Alexandria.

Fig. 3.33. Coring on the Molo Sinistro at the Claudian harbour of Portus.

Fig. 3.34. Dive boat and equipment boat positioned over off-shore coring site at Sebastos.

Fig. 3.35. Selecting the core location at Santa Liberata.

Fig. 3.36. Drilling frame positioned on a pila at Caesarea.

Fig. 3.37. Drilling holes for the feet anchor bolts at Caesarea.

Fig. 3.38. Frame feet being fixed to an uneven concrete surface at Caesarea.

Fig. 3.39. Clamping the feet to the frame at Caesarea.

Fig. 3.40. Cordiam M60 rack being clamped to the frame at Santa Liberata.

Fig. 3.41. Cordiam Hydro 25 rotary drill mounted on the M60 rack.

Fig. 3.42. Drilling into concrete with a plume of flushed debris.

Fig. 3.43. Concrete core removed from the drill tube at Portus Iulius.

Fig. 3.44. Weak mortar mix being trowelled into core hole.

Fig. 3.45. Core location labelled at Secca Fumosa.

Fig. 3.46. Oleson recording a core at Caesarea before shipment to Italcementi.

Fig. 4.1. Plan of Portus with location of the coring sites (Will Foster Illustration).

Fig. 4.2. Portus, coring at site of POR 2002.02 (Testaguzza’s lighthouse).

Fig. 4.3. View of eastern end of Molo Sinistro, with sockets left by formwork. Site of POR.2002.01 in foreground, POR.2002.03 in background.

Fig. 4.4. Taking core POR.2002.03.

Fig. 4.5. Reverse of Nero’s Portus issue (Courtesy of the British Museum; CM BMC132, AN31942001).

Fig. 4.6. Taking core PTR.2002.01.

Fig. 4.7. Taking core PTR.2002.02.

Fig. 4.8. Plan of Antium harbour with location of core (Will Foster Illustration).

Fig. 4.9. Taking core ANZ.2002.01.

Fig. 4.10. Plan of Portus Cosanus (Will Foster Illustration).

Fig. 4.11. Portus Cosanus, view of Piers 1–3.

Fig. 4.12. Portus Cosanus, impressions of wooden formwork shuttering on Pier 1 (after McCann et al. 1987: fig. III-13). (Courtesy of A. M. McCann)

Fig. 4.13. Portus Cosanus, coring of Pier 2.

Fig. 4.14. Portus Cosanus, seaward portion of Pier 2, showing upper and lower concrete mixtures and hole left by catena.

Fig. 4.15. Portus Cosanus, view of Pier 1.5.

Fig. 4.16. Portus Cosanus, taking core PCO.2003.05 from Pier 5.

Fig. 4.17. Plan of Santa Liberata (Will Foster Illustration).

Fig. 4.18. Santa Liberata, aerial photograph of villa, piscina, and pilae (Courtesy of P. Gambogi).

Fig. 4.19. Santa Liberata, view of piscina and adjacent pila during coring of STL.2003.01.

Fig. 4.20. Santa Liberata, taking core SLI.2004.01 on the villa pila.

Fig. 4.21. Santa Liberata, the side of the fallen villa pila.

Fig. 4.22. Sebastos, aerial view of submerged harbour installations (R. L. Hohlfelder).

Fig. 4.23. Sebastos, plan of harbour remains, with indication of coring locations (Will Foster Illustrations).

Fig. 4.24. Sebastos, reconstruction of harbour remains (C. J. Brandon).

Fig. 4.25. Sebastos, photo of concrete block and formwork in Area G.

Fig. 4.26. Sebastos, sonar image and plan of blocks in Area K (C. J. Brandon).

Fig. 4.27. Sebastos, coring ofpila at CAE.2005.05.

Fig. 4.28. Sebastos, view of block from which core CAE.2005.02 was taken.

Fig. 4.29. Sebastos, drawing of block from which core CAE.2005.02 was taken (C. J. Brandon after P. Dessauer and L. Reynafarje).

Fig. 4.30. Sebastos, drawing of group of blocks from which core CAE.2005.02 was taken (C. J. Brandon after P. Dessauer and L. Reynafarje).

Fig. 4.31. Sebastos, reconstruction drawing of the harbour in Late Antiquity, following renovation with dumped rubble (S. Giannetti, in Holum et al. 1988: 159, fig. 110).

Fig. 4.32. Map of Baiae area, with indication of structures mentioned and coring locations (Will Foster Illustration).

Fig. 4.33. Plan of structures at entrance to Baianus Lacus, with indication of coring location (Will Foster Illustration).

Fig. 4.34. Plan of structures at entrance to Portus Iulius, with indication of coring locations (Will Foster Illustration).

Fig. 4.35. Plan of pilae at Secca Fumosa, with indication of coring location (Will Foster Illustration).

Fig. 4.36. Eastern Harbour of Alexandria, with indication of coring locations (Will Foster Illustration).

Fig. 4.37. Oleson examining core ALE.2007.02 on the Princess Douda in Alexandria harbour.

Fig. 4.38. Chersonesos, plan of harbour and breakwater, with indication of coring locations (Will Foster Illustration).

Fig. 4.39. Chersonesos, coring in progress at location of CHR.2007.02.

Fig. 4.40. Egnatia, plan of harbour with indication of coring site (Will Foster Illustration).

Fig. 4.41. Egnatia, view of pila with coring in progress for EGN.2008.01.

Fig. 4.42. Egnatia, opus reticulatum facing on outer pila.

Fig. 4.43. Map of harbour sites along the southern coast of Turkey (Will Foster Illustration).

Fig. 4.44. Pompeiopolis, plan of harbour with indication of coring sites (Will Foster Illustration).

Fig. 4.45. Pompeiopolis, aerial photograph of harbour and adjacent portion of city (Courtesy of R. Yagçı).

Fig. 4.46. Pompeiopolis, view of west breakwater, looking south.

Fig. 4.47. Pompeiopolis, west breakwater, reconstruction of cell containing concrete (C. J. Brandon).

Fig. 4.48. Pompeiopolis, detail of outer wall of west breakwater.

Fig. 4.49. Pompeiopolis, detail of clamp recesses on west breakwater.

Fig. 4.50. Coin of Antoninus Pius representing the harbour of Pompeiopolis (Courtesy of the American Numismatic Society, Newell Collection).

Fig. 4.51. Pompeiopolis, taking core POM.2009.01.

Fig. 4.52. Pompeiopolis, taking core POM.2009.02.

Fig. 4.53. Reconstruction of Pompeiopolis harbour in the second century AD (C. J. Brandon).

Fig. 4.54. Pompeiopolis, map of the harbour by Beaufort in 1811–12 (from Beaufort 1817: fig. 3).

Fig. 5.1. Location of the completed pila reproduction in the marina (to the left).

Fig. 5.2. Tuff blocks and bags of pozzolana assembled for the reconstruction project.

Fig. 5.3. Reconstituted lumber ready for use in the formwork.

Fig. 5.4. Grassello di calce provided for the pila reconstruction.

Fig. 5.5. Tuff blocks being reduced to caementa for the pila reconstruction.

Fig. 5.6. First two plank walls in position; bolt hole being drilled in horizontal collar beam joint.

Fig. 5.7. Trimming the formwork planks after installation of shuttering.

Fig. 5.8. Pozzolana and lime putty in the mixing trough.

Fig. 5.9. Detail of the mortar after mixing.

Fig. 5.10. Reconstruction of workers lowering mortar into formwork at Sebastos (Hohlfelder 1987: 264–65) (National Geographic Society, used with permission).

Fig. 5.11. Trajan’s column, Cichorius Scene XII. Soldiers using baskets to shift earth (P. Rockwell, used with permission).

Fig. 5.12. Bardo Museum, third-century mosaic showing construction scene (J. P. Oleson).

Fig. 5.13. Tomb of Trebius Iustus, on the Via Latina, Rome. Fresco depicting construction of a brick-faced concrete wall (Marucchi 1911: fig. 5).

Fig. 5.14. Reproduction basket with load of mortar and ropes for lowering and dumping.

Fig. 5.15. Basket of mortar floating in the inundated form.

Fig. 5.16. Empty basket returning to the surface. Note tip rope attached to base.

Fig. 5.17. Basket loads of mortar visible in shallow water in formwork.

Fig. 5.18. Trough filled with measured volume of tuff caementa.

Fig. 5.19. Surface of the concrete after settling overnight.

Fig. 5.20. Final upper surface of the concrete within the formwork.

Fig. 5.21. Completed pila with paved upper surface.

Fig. 5.22. Condition of the formwork planks at low tide, November 2005.

Fig. 5.23. Coring the completed pila in March 2005.

All maps and plans in Chapter 6 are by Will Foster Illustration.

Fig. 6.1. Map of coring sites in Italy, and sites with marine concrete.

Fig. 6.2. Location map of Quarteira.

Fig. 6.3. Two fragments of concrete wall from the Quarteira piscina, now in the Loulé Archaeological Museum. The near fragment is a cross-section showing the use of an embedded amphora body as a nesting pot.

Fig. 6.4. City area of Cadiz, with location of concrete.

Fig. 6.5. City area of Ampurias, with location of concrete wall.

Fig. 6.6. Marseilles, Place Jules Verne area, with location of concrete wall.

Fig. 6.7. Harbour area of Forum Iulii, with location of East Quay.

Fig. 6.8. Harbour area of Forum Iulii, with location of South Quay.

Fig. 6.9. Pianosa, circular fish-ponds.

Fig. 6.10. Santa Liberata, fish-pond and pilae.

Fig. 6.11. Cosa, harbour with pilae.

Fig. 6.12. Torre Valdaliga, fish-pond with concrete walls.

Fig. 6.13. Torre Mattonara, fish-pond with concrete walls.

Fig. 6.14. Civitavecchia, plan of ancient harbour with conjectured position of breakwaters.

Fig. 6.15. Punta della Vipera, fish-pond with thick concrete outer walls.

Fig. 6.16. Santa Marinella, semicircular concrete fish-pond.

Fig. 6.17. Santa Severa, concrete fish-pond.

Fig. 6.18. Palo, concrete fish-pond.

Fig. 6.19. Portus, location of Molo Sinistro.

Fig. 6.20. Portus, long concrete mole on the western side of the Darsena.

Fig. 6.21. Portus, concrete mole along northern side of the canal leading to Trajan’s Basin.

Fig. 6.22. Portus, concrete embankment near the Severan Warehouses.

Fig. 6.23. Anzio, plan of harbour.

Fig. 6.24. Astura, La Saracca, semicircular concrete fish-pond.

Fig. 6.25. Astura, La Banca, rectangular concrete fish-pond.

Fig. 6.26. Torre Astura, Punta di Astura, rectangular concrete fishpond.

Fig. 6.27. Torre Astura, Porto di Astura, two long concrete moles enclosing a harbour on the east side of Torre Astura.

Fig. 6.28. Ponza, Porto di Ponza, modern harbour mole overlying a Roman concrete mole.

Fig. 6.29. Ventotene, concrete structures within a rock cut fish-pond.

Fig. 6.30. Circeo, Lake Paola Canal, large concrete mass on one of a pair of concrete jetties at entrance to the canal.

Fig. 6.31. Piscina di Lucullo, circular concrete fish-pond.

Fig. 6.32. Terracina, concrete mole forming outer edge of Roman port.

Fig. 6.33. Sperlonga, Grotta di Tiberio, concrete and rock cut fishpond.

Fig. 6.34. Gaeta, La Catena or La Nave, row of five concrete pilae.

Fig. 6.35. Porto di Caposele, concrete foundation for a quay.

Fig. 6.36. Formia, rectangular concrete fish-pond.

Fig. 6.37. Miseno, Punta Sarparella, concrete jetty.

Fig. 6.38. Miseno, Punta Terone, row of eight concrete pilae.

Fig. 6.39. Miseno, Punta di Pennata, concrete pilae.

Fig. 6.40. Miseno, Punta di Pennata, concrete quay.

Fig. 6.41. Miseno, Punta di Pennata, concrete pilae.

Fig. 6.42. Baia, Castello di Baia, 14 concrete pilae offshore.

Fig. 6.43. Baia, Cantieri di Baia, 3 concrete pilae offshore.

Fig. 6.44. Baia, two concrete moles forming entrance channel to Baianus Lacus.

Fig. 6.45. Baia, Villa dei Pisoni, cluster of concrete pilae.

Fig. 6.46. Baia, Secca Fumosa, 30 concrete pilae offshore.

Fig. 6.47. Baia, row of concrete pilae at entrance to Portus Iulius.

Fig. 6.48. Baia, two concrete moles forming entrance to Portus Iulius.

Fig. 6.49. Baia, Portus Iulius, five concrete pilae on the east side of the eastern mole.

Fig. 6.50. Pozzuoli, 13 concrete pilae forming foundation for main harbour breakwater.

Fig. 6.51. Pozzuoli, concrete pilae along the coastline, to the south of the headland of Pozzuoli.

Fig. 6.52. Nisida, four concrete pilae.

Fig. 6.53. Pausilypon. Left side: Gaiola (Palaepolis), concrete wall, mass, and pila. Right side: small harbour protected by a row of concrete pilae.

Fig. 6.54. Pausilypon. Right side: Marechiano harbour with concrete mole. Left side: Regio Marechiano, three concrete pilae and irregular pier.

Fig. 6.55. Pausilypon, Regio Rosebery, row of concrete pilae.

Fig. 6.56. Capri, Palazzo a Mare East, concrete pilae and landing stages.

Fig. 6.57. Capri, Palazzo a Mare West, concrete pilae and landing stages.

Fig. 6.58. Island of Gallo Lungo, row of concrete pilae.

Fig. 6.59. San Marco di Castellabate, concrete mole.

Fig. 6.60. Sapri, concrete arched pier.

Fig. 6.61. Lecce, San Cataldo, Porto Adriano, Hadrianic or early modern mole (D. Klapecki).

Fig. 6.62. Egnazia, two concrete moles and several pilae.

Fig. 6.63. Mavra Litharia, natural beachrock formation with raised concrete wall in background.

Fig. 6.64. Anthedon, naturally concreted rubble behind clamped ashlar marginal walls.

Fig. 6.65. Anthedon, naturally concreted rubble, detail.

Fig. 6.66. Chersonesos, concrete moles.

Fig. 6.67. Kyme, concrete mole.

Fig. 6.68. Side, long concrete mole in four sections.

Fig. 6.69. Pompeiopolis, two concrete moles.

Fig. 6.70. Sebastos, Area K, row of five concrete blocks and two isolated pilae.

Fig. 6.71. Sebastos. southwest mole, between Areas K and E/F, large concrete blocks.

Fig. 6.72. Sebastos, Area G, large concrete block.

Fig. 6.73. Alexandria, Antirhodos Island, block of concrete.

Fig. 6.74. Alexandria, Dock at Ball Trap, 16 concrete pilae.

Fig. 6.75. Alexandria, concrete jetty.

Fig. 6.76. Thapsus, long concrete mole.

Fig. 6.77. Carthage, Quadrilateral of Falbe, concrete walls forming the entrance channel to the inner harbours.

Fig. 6.78. Carthage, Circular Harbour, Roman concrete extension to the piers supporting the causeway.

Fig. 6.79. Carthage, Neptune block, large, isolated concrete block, plan.

Fig. 6.80. Carthage, Neptune block, view of holes left by catenae.

Fig. 6.81. Carthage, Neptune block, detail of concrete.

Fig. 6.82. Tipasa, large block of concrete.

Fig. 6.83. Cherchel, Seven large concrete pilae in a line.

All figures in Chapter 7 are by M. D. Jackson and Bronze Black Design, unless otherwise noted.

Fig. 7.1. Characteristic fabric of ancient Roman maritime concrete, shown in the drill core of the Baianus Sinus pila (BAI.2006.03). a. Photograph of a core sample showing pumiceous tuff caementa, and pumiceous volcanic ash and relict lime clasts in the pozzolanic mortar. b. Partially dissolved relict lime clasts (1), poorly crystalline, calcium-aluminium-silicate-hydrate (C-A-S-H) in the cementitious matrix (2), relict pumice clasts with associated cementitious hydrates (3), and chloride and sulphate microstructures (4, 5) (SEM-BSE image). c. Detail of (b) showing the pozzolanic reaction rim around a pumice clast (SEM-BSE image). d. Al-tobermorite crystals in a relict lime clast (SEM-SE image).

Fig. 7.2. Geologic sketch map of Bay of Naples showing the Flegrean Fields and Somma-Vesuvius volcanic districts and the limestone bedrock of the Sorrento peninsula (after Orsi et al. 1996).

Fig. 7.3. Cementitious hydrates in the ancient maritime mortars. a. C-A-S-H in the cementitious matrix and the perimeters of relict lime clasts, Baianus Sinus (SEM-BSE image). b. Platy crystals of Al-tobermorite, relict lime clast, Baianus Sinus (SEM-SE image). c. Platy Al-tobermorite crystals and C-A-S-H, cementitious matrix, Baianus Sinus (SEM-SE image). d. Cementitious hydrates in a tubular pumice clast, Caesarea (G.Vola). e. Hydrocalumite crystals in a relict void surrounded by thread-like Al-tobermorite crystals, Baianus Sinus (SEM-BSE image). f. Ettringite, sub-spherical microstructure at the perimeter of a relict lime clast, Baianus Sinus (SEM-BSE image).

Fig. 7.4. Zeolite mineral microstructures in components of the ancient maritime concretes. a. Pumice clast with relict geological phillipsite and C-A-S-H in a large vesicle, Baianus Sinus mortar (SEM-SE image). b. In situ phillipsite in a pore of the cementitious matrix, Portus Cosa mortar (SEM-BSE image). c. Possible in situ phillipsite in Tufo Lionato caementa, Portus Traiani concrete (petrographic image, plane polarized light). d. Tufo Lionato from the Salone Quarry, northeast of Rome (petrographic image, plane polarized light) (see Jackson et al. 2005).

Fig. 7.5. Macroscale map of a segment of the Portus Traiani concrete, showing the principal components of the ancient Roman maritime concrete fabric.

Fig. 7.6. Slices of drill cores of the ancient maritime concretes and the young Brindisi concrete reproduction, showing macroscale fabrics of the mortars and glassy volcanic tuff caementa. All drill cores are 9 cm in diameter. a. Santa Liberata (SLI.2004.01) concrete, possible Flegrean tuff caementa (specimen SLI.2004.01. T1) and pumice clast (specimen SLI.04.01.P1). b. Portus Claudius (POR.2002.01) mortar, relict lime fabrics (C. Hagen) (see also Fig. 7.14). c. Portus Traiani (PTR.2002.01) concrete (see also Fig. 7.5), Tufo Lionato caementa from Alban Hills volcano (specimen PTR.02.02.T1, see also Fig. 7.4) and possible Flegrean pumice clasts (specimen PTR.02.01.P1). d. Portus Neronis (ANZ.2002.01) mortar, with possible Flegrean pumice and pumiceous ash (C. Hagen). e. Sebastos, Caesarea (CAE.2005.05) concrete, showing diverse mortar components (C. Hagen), with pumiceous tuff caementa and pumice clasts (specimen CAE.05.05.P1). f. Brindisi (BRI.2009.01) concrete, Bacoli Tuff (specimens 05.BRI.02.T1, 06.BRI.01.T1) and lime-ash putty (See Fig. 7.14).

Fig. 7.7. Maps of the central Italian coast and the Bay of Pozzuoli, showing harbour drill sites and volcanic districts (after Jackson et al. 2013b).

Fig. 7.8. Photographs of the ancient maritime concretes of eastern Mediterranean harbour structures. a. Egnázia (EGN.2008.02), mortar with pale orangish-gray pumiceous pozzolan (C. Brandon). b. Chersonesos (CHR.2007.02), concrete with fossiliferous limestone caementa and porous mortar with pale yellowish-gray pumice (C. Hagen). c. Caesarea (CAE.2005.05), mortar with various calcareous clasts and grayish-green pumice, and fragments of calcarenite (1), a hardened clot of lime putty (2), and relict lime clasts (3). d. Caesarea (CAE.2005.01), concrete with calcareous sandstone caementa and voids at the interface with a mortar that has pale yellowish gray pumice (C. Brandon). e. Caesarea (CAE.2005.02), mortar with pale yellowish-gray pumice (C. Brandon). f. (Caesarea (CAE.2005.05), mortar with greenish-gray pumice. g. Pompeiopolis (POM.2009.02), concrete with pumiceous tuff and rounded river cobble caementa (G. Vola). h. Pompeiopolis (POM.2009.02), concrete with river cobble caementa and mortar with yellowish-gray pumice (G. Vola). i. Alexandria (ALE.07.03), mortar with pale yellowish-gray pumice (C. Brandon).

Fig. 7.9. Map showing locations of Mediterranean volcanic districts, pumice deposits and possible limestone sources described in the text.

Fig. 7.10. Trace element studies of volcanic tuff caementa in the maritime concretes and the Bacoli Tuff caementa in the 2004 Brindisi concrete reproduction (Table A4.2). a. Zr/Y and Nb/Y ratios, compared with glassy tuff deposits of the Vulsini, Vico, and Monti Sabatini volcanic districts (for generalized compositional fields, see Marra et al. 2013b), Tufo Lionato (this study), and the Neapolitan Yellow Tuff. b. Eu/Th and Ba/Th ratios. c. La/Yb and Ta/Th ratios of all specimens of volcanic tuff caementa and pumice clasts. (1) Parker 1989; (2) Turbeville 1993; (3) Lancaster et al. 2011; (4) Orsi et al. 1992; (5) Steinhauser et al. 2010.

Fig. 7.11. Trace element studies, Zr/Y and Nb/Y, of pumice clasts from the volcanic ash pozzolan of the ancient maritime mortars compared with Mediterranean pumice deposits beyond the Bay of Naples (Fig. 7.9; Table A4.2). Monti Sabatini deposits: Lancaster et al. 2011; Marra et al. 2013b; Aeolian Islands deposits: (1) Lipari pumice, Gioncada et al. 2003 (Pomiciazzo pumice); Daví et al. 2011 (Monte Pilato pumice); (2) Vulcano pumice, De Astis et al. 1997; (3) Stromboli pumice, Bertagnini and Landi 1996; Aegean Islands deposits: (4) Kyparissiakos Gulf pumice, Ionian Sea, Bathrellos et al. 2009; (5) Santorini, Thera pumice, Pre-Minoan and Minoan eruptions, Druitt et al. 1999; (6) Minoan pumice, Vinci et al. 1984; (7) Knossos pumice, Warren and Pulchelt 1990; (8) Milos pumice, Fytikas et al. 1996; (9) Yali pumice, Margari et al. 2007 (see also Allen and McPhie 2000); (9) Nisyros pumice, Margari et al. 2007 (see also Francalanci et al. 1995); (10) Santorini air fall deposits at Gölhisar Gölü, Turkey, pumice glass, Eastwood et al. 1999.

Fig. 7.12. Trace element studies of pumice clasts from the volcanic ash pozzolan of the ancient maritime mortars compared with Campi Flegrei pumice deposits (Fig. 7.2; Table A4.2). Post Neapolitan Yellow Tuff volcanic chronostratigraphy from Fedele et al. 2011. a. Zr/Y and Nb/Y. b. Nb/Zr and La/Yb. (1) Fedele et al. 2011; (2) Tonarini et al. 2009; (3) Di Vito et al. 2011; 4) di Vita et al. 1999; (5) Lustrino et al. 2002; (6) Orsi et al. 1992; (7) Scarpati et al. 1993; (8) Pabst et al. 2008; (9) De Astis et al. 2004; (10) Civetta et al. 1991; (11) Lancaster et al. 2011, Marra et al. 2013b.

Fig. 7.13. Trace element studies of pumice clasts from the volcanic ash pozzolan of the ancient maritime mortars compared with Somma-Vesuvius pumice deposits (Fig. 7.2; Table A4.2). Volcanic chronostratigraphy from Di Renzo 2007. a. Zr/Y and Nb/Y. b. Nb/Zr and La/Yb. AD 79 pumice: (1) Cioni et al. 1995, (2) Ayuso et al. 1998, (3) Paone et al. 2006, Piochi et al. 2006; Protohistoric pumices: (3) Paone et al. 2006, (4) Somma et al. 2001; Avellino pumice: (2) Ayuso et al. 1998, (3) Paone et al. 2006, (5) Sulpizio et al. 2010; Ottaviano pumice: (3) Paone et al. 2006, (6) Piochi et al. 2006; (7) Aulinas et al. 2008; Novelle-Seggiano-Bosco pumice: (1) Cioni et al. 2003; (3) Paone et al. 2006; Sarno pumice: (2) Ayuso et al. 1998, (3) Paone et al. 2006; Codola pumice: (2) Ayuso et al. 1998, (3) Paone et al. 2006; Camaldoli della Torre pumice (CI): (8) Di Renzo et al. 2007; Camaldoli della Torre pumice (Post-CI): (8) Di Renzo et al. 2007; Camaldoli della Torre pumice (Pre-CI): (8) Di Renzo et al. 2007; Pumices, Roman architectural mortars: (9) Marra et al. 2013, (10) Lancaster et al. 2011.

Fig. 7.14. Petrographic photomicrographs of the lime putty mortar fabric of the Brindisi concrete reproduction and the complex Portus Claudius mortar. Optical micrographs, plane polarized light. a. Brindisi mortar (BRI.2005), 12 months hydration. b. Brindisi mortar (BRI.2006), 24 months hydration. c, d. Portus Claudius mortar (POR.2002.02C). The opaque selvages seem to follow the relict surfaces of partially hydrated lime-volcanic ash clumps.

Fig. 7.15. Petrographic photomicrographs of relict lime clast microstructures, and the composition of C-A-S-H and Al-tobermorite in the Baianus Sinus mortar. Optical micrographs, plane polarized light. a. Portus Cosanus mortar, possible relict quicklime clast, with cracks produced by in situ hydration of quicklime in sea-water, followed by dissolution during pozzolanic reaction. b. Portus Neronis mortar, possible matured slaked lime clast, showing gradual dissolution during pozzolanic reaction. c, d. Typical, partially dissolved, relict lime clast in the Baianus Sinus mortar, and compositions as Ca/Si/Al=100 cation atomic ratios from SEM-EDS analyses. The dotted line shows the approximate gradational boundary between Al-tobermorite crystal clusters in the core and poorly crystalline C-A-S-H phase in the perimetral rim (after Jackson et al. 2013b).

Fig. 7.16. Results of bulk chemical analyses of the ancient maritime mortars, as weight % oxides, determined from powdered specimens (Tables A4.2, A4.3). a. CaO-Al2O3-SiO2. b. MgO vs CaO+MgO, compositions below the dotted line may represent mortar specimens with high calcium lime, and little dolomitic (magnesium) component. The wide range of values for mortars of specific harbour concretes is due, in part, to heterogeneous proportions of volcanic ash (or limestone particles) in the centimetre-sized specimens.

Fig. 7.17. Results of bulk chemical analyses, as weight % oxides, of powdered bulk mortar specimens, centimetre-sized specimens of the Brindisi mortar reproduction and the ancient maritime mortars (see Fig. 7.16; Table A4.3). a. CaO+MgO and Al2O3+ SiO2. b. Na2O and SiO2. c. K2O and SiO2 of the mortars. The outlying compositions of specimens from the 48- and 60-month core samples may reflect the influence of influxes of sea-water and atmospheric gases during repeated drilling episodes.

Fig. 7.18. Determinations of the material characteristics of the ancient concretes and pozzolanic mortars, measured in drill core specimens. a. Measurements of the relative proportions of mortar and decimeter-sized caementa (Table 7.1). b. Unit weight and uniaxial compressive strength of drill core segments of the ancient maritime concretes and the young Brindisi concrete reproduction (Table 7.3). Bacoli Tuff unit weight is about 1300 kg/m³ (this study); Neapolitan Yellow Tuff unit weight is 1200 to 1400 kg/m³, compressive strength is 0.7 to 12 MPa (Colella et al. 2001); calcareous sandstone (calcarenite) unit weight is about 2020 (Scicchitano et al. 2007), compressive strength is 2 to18 MPa, and limestones are similar (Marcari et al. 2010); Tufo Lionato from the Salone quarry has unit weight about 1520 kg/m³ and compressive strength about 26 MPa (Jackson et al. 2005). c. Young’s modulus (elastic modulus) and uniaxial compressive strength of drill core segments of the ancient maritime concretes and the young Brindisi concrete reproduction (Table 7.3). d. Porosity, as volume %/total volume, of mortars of the ancient maritime concretes (dark gray bars) and young Brindisi reproduction (light gray bars), and of the Neapolitan Yellow Tuff and the Bacoli Tuff (medium gray bars) (Pellegrino 1967, in Ottaviano 1988). Each bar represents a single sample measurement (see Table 7.4); Alexandria mortar determinations by Rispoli (2011). e. Mortar porosity and uniaxial compressive strength of drill core segments of the ancient maritime concretes, the young Brindisi concrete reproduction, the Bacoli Tuff, and the Neapolitan Yellow Tuff.

Fig. 7.19. Determination of the pore structures of the young and ancient sea-water mortars through mercury intrusion porosity experiments (after Gotti et al. 2008). a. Relative pore size distribution in a typical maritime mortar specimen with Flegrean pumiceous ash pozzolan from Santa Liberata (SLI.2004.01A), compared with modern Portland cement mortars. b. Cumulative porosity of the Santa Liberata (SLI.2004.01A) mortar specimen compared with modern Portland cement mortars. c. Pore size distribution of the Santa Liberata (SLI.2004.01A) mortar specimen compared with the Brindisi mortar reproduction at 6 months hydration (BRI.2005.01) and at 12 months hydration (BRI.2005.02).

Fig. 7.20. Images showing the pore structure ofFlegrean tuffpozzolan and the pumiceous sea-water mortar fabric from a Portus Neronis core sample. a. Vesicular fabric of the Bacoli Tuff, showing the porous coarse pumice clasts and the altered vitric matrix, composed of fine pumiceous ash (petrographic image, plane polarized light). b. Cementitious matrix of the Portus Neronis mortar. Vesicles of pumice clasts (1, 2, 3) are lined with cementitious hydrates; vesicles of fine pumiceous ash particles (4) are filled with cementitious hydrates, mainly C-A-S-H and Al-tobermorite (SEM-BSE image).

Fig. 7.21. Results of magic-angle nuclear magnetic resonance (MASNMR) analysis, showing aluminium and silicon bonding environments in Al-tobermorite from relict lime clasts, Baianus Sinus mortar, and crystallization conditions based on temperatures computed in an adiabatic thermal model of the Bainanus Sinus pila (after Jackson et al. 2013b). a. ²⁹Si NMR study; Q¹ dimers or chain terminations, Q² chain middle groups, and Q³ branching sites describe the connectivity of SiO2 tetrahedra. b. ²⁷Al NMR study. c. Schematic diagram showing types of measured linkages of tetrahedral SiO4–⁴ or AlO4–⁵ units (triangles). Light and dark gray triangles indicate examples of linkages of silicate tetrahedra and green triangles indicate linkages of silicate and aluminium tetrahedra. d. Maximum temperatures (Θ) at the specimen site and the body centre of the 5.7 m thick Baianus Sinus block. The model configuration calculates heat evolved through formation of C-A-S-H cementitious binder. Exothermic hydration of lime produced an initial temperature of about +5 °C above ambient sea-water temperatures (Tw). The model block attained 14–26 °C sea-water temperatures about two years after installation.

Fig. 8.1. Formwork impressions on a concrete foundation on the Palatine Hill, Rome.

Fig. 8.2. Opus reticulatum facing on the sides of a concrete pila at Secca Fumosa near Baiae.

Fig. 8.3. Reconstruction of a Category 1 inundated form constructed in situ (C. J. Brandon).

Fig. 8.4. Reconstruction of piles (destinae) installed in the first phase of building a Category 1 form (C. J. Brandon).

Fig. 8.5. Plan of the eastern mole at Anzio (Felici 1993: fig. 8; used with permission).

Fig. 8.6. Misenum, Punta Terone, details of a pila with vertical and horizontal pile and beam impressions (Gianfrotta 1996: fig. 8; used with permission).

Fig. 8.7. Plan of the entrance channel moles to the harbour of Baianus Lacus (Scognamiglio 2002: pl. 1; photo E. Scognamiglio).

Fig. 8.8. Portus Iulius, outer pila on the western side of the entrance channel with positions of vertical pile impressions (C. J. Brandon).

Fig. 8.9. Portus Iulius, top of a 30 cm diameter pile on the outer pila on the western side of the entrance channel.

Fig. 8.10. Plan of a concrete pila at Bagni di Tiberio on Capri (after Scognamiglio 2010: 123).

Fig. 8.11. Plan of the southeast mole at Egnázia (Auriemma 2004: 48, fig. 29; used with permission).

Fig. 8.12. Reconstruction sketch of the southeast mole at Egnázia (Auriemma 2004: 52, fig. 35; used with permission).

Fig. 8.13. Reconstruction of the outer piles (stipites) installed in the second phase of building a Category 1 form (C. J. Brandon).

Fig. 8.14. Reconstruction of the fish-pond formwork at Santa Severa (after Pellandra 1997: pl. II a–b).

Fig. 8.15. Sketch of concrete formwork on the Molo Sinistro of the Claudian harbour of Portus (O. Testaguzza, used with permission).

Fig. 8.16. Reconstruction of the inner harbour concrete pier at Anzio (Felici 2002: fig. 8; used with permission).

Fig. 8.17. Inner harbour concrete pier at Anzio, details of lower portion of two vertical timber planks from the shuttering (Felici 2002: fig. 14; used with permission).

Fig. 8.18. Base of shuttering on the side of a concrete dock south of the Baianus Lacus entrance channel (Scognamiglio 2002: fig. 2; photo E. Scognamiglio).

Fig. 8.19. Reconstruction of a section of the harbour mole at San Marco di Castellabate (after Benini 2002: pl. 3; used with permission).

Fig. 8.20. Remains of a timber pile driven into the rock seabed at San Marco di Castellabate (after Benini 2002: fig. 10; used with permission).

Fig. 8.21. Post hole drilled into the rock seafloor at San Marco di Castellabate (After Benini 2002: fig. 11; used with permission).

Fig. 8.22. Reconstruction of the upper and lower horizontal rails fixed to the outer piles (stipites) installed in the third phase of building a Category 1 form (C. J. Brandon).

Fig. 8.23. Remains of the lower section of formwork on the northern mole at Portus (O. Testaguzza, used with permission).

Fig. 8.24. Detail of the fixing bracket securing the lower rail to a stipes on the formwork on the side of a concrete dock to the south of the entrance channel into the harbour of Baianus Lacus (C. J. Brandon after Miniero 2001: fig. 5).

Fig. 8.25. Detail of the fixing bracket securing the lower rail to a stipes on the formwork on the side of a concrete quay to the south of Punta Pennata at Misenum (C. J. Brandon after Benini and Lanteri 2010: 114–15).

Fig. 8.26. Fixing bracket securing the lower rail to a stipes on the formwork on the side of a concrete quay to the south of Punta Pennata at Misenum (Benini and Lanteri 2010: fig. 14; used with permission).

Fig. 8.27. Detail of the fixing bracket securing the lower rail to a stipes on the formwork on the side of a concrete quay to the South of Punta Pennata at Misenum (Benini and Lanteri 2010: fig. 14; used with permission).

Fig. 8.28. Reconstruction of vertical timber board shuttering fixed to the upper horizontal rail installed in the fourth phase in building a Category 1 form (C. J. Brandon).

Fig. 8.29. Horizontal planked formwork around the platform in front of the Spring House at Cosa (McCann et al. 1987: 100–1, fig. V.4–V.12) (Photo: A. M. McCann, used with permission).

Fig. 8.30. Vertical timber retaining wall around the concrete platform in front of the Spring House at Cosa (McCann et al. 1987: 101, fig. V.6–V.7) (Photo: A. M. McCann, used with permission).

Fig. 8.31. Log cut into slabs (C. J. Brandon).

Fig. 8.32. Staggered and lapped vertical board shuttering (C. J. Brandon).

Fig. 8.33. Impressions of lapped vertical boarding on Block III of the western mole at Anzio (after Felici 1998: 307, fig. 38; used with permission).

Fig. 8.34. Vertical planked formwork around the concrete dock at Santa Severa (A. M. McCann; used with permission).

Fig. 8.35. Remains of vertical timber planking shuttering at the southwest corner of the Darsena basin at Portus (Verduchi 2005: 257; used with permission).

Fig. 8.36. Miseno, formwork comprising untrimmed timber slab cut planks with vertical battens sealing the joints between boards (C. J. Brandon after Benini and Lanteri 2010: 114–15).

Fig. 8.37. Reconstruction of horizontal tie beams (catenae) fixed to the upper horizontal rail installed in the fifth and final phase of building a Category 1 form (C. J. Brandon).

Fig. 8.38. Impression of horizontal tie beams on Molo Sinistro of the Claudian basin at Portus.

Fig. 8.39. Fragment of the northern mole of the harbour at Astura.

Fig. 8.40. Axonometric sketch of the large concrete block on the shoreline north of Carthage (C. J. Brandon after R. A. Yorke; used with permission).

Fig. 8.41. Thapsus; horizontal beam impressions in the upper section of the concrete mole (R. A. Yorke; used with permission).

Fig. 8.42. Plan of the mole at Kyme (Esposito et al. 2002: pl. II; used with permission).

Fig. 8.43. Reconstruction of a Category 2 cofferdam constructed in situ and dewatered (C. J. Brandon).

Fig. 8.44. Grooved piles into which horizontal boards are slotted (C. J. Brandon).

Fig. 8.45. Offset shuttering used to form the quasi-opus reticulatum faced concrete pier in the harbour at Ponza (Gianfrotta 1996: 71; used with permission).

Fig. 8.46. Category 3 form, stage 1, after Vitruvius 5.12.3–4: A platform is to be built out with a level upper surface over less than half its area. The shoreward section is to have one side sloping (C. J. Brandon).

Fig. 8.47. Category 3 form, stage 2, after Vitruvius 5.12.4: Retaining walls one and one half feet thick are to be built at the end facing the sea and on either side of the platform, equal in height to the level surface described above (C. J. Brandon).

Fig. 8.48. Category 3 form, stage 3, after Vitruvius 5.12.4: Then the sloping section is to be filled in with sand and brought up to the level of the retaining walls and platform surface (C. J. Brandon).

Fig. 8.49. Category 3 form, stage 4, after Vitruvius 5.12.4: "Next, a concrete block (pila) of the appointed size must be built there, on this levelled surface, and when it has been formed it is left at least two months to cure" (C. J. Brandon).

Fig. 8.50. Category 3 form, stage 5, after Vitruvius 5.12.4: "Then the retaining wall that holds in the sand is cut away, and in this manner erosion of the sand by the waves causes the block (pila) to fall into the sea. By this procedure, repeated as often as necessary, the breakwater can be carried seaward" (C. J. Brandon).

Fig. 8.51. Reconstruction of a prefabricated floating caisson being launched into the sea near Sebastos (C. J. Brandon).

Fig. 8.52. Reconstruction of prefabricated caissons being positioned and loaded with concrete at Area K at the northern end of the main enclosing mole at Caesarea (C. J. Brandon).

Fig. 8.53. Axonometric sketch of the concrete block and caisson at Antirhodos Island in the Eastern Harbour of Alexandria (C. J. Brandon after Goddio et al. 1998: 32–37; used with permission).

Fig. 8.54. Base of caisson, Antirhodos Island, within the eastern harbour of Alexandria (C. J. Brandon).

Fig. 8.55. Axonometric details of the caisson used to build a long rubble jetty in the harbour of Laurons (Ximenes and Moerman 1988: 229–52; used with permission).

Fig. 8.56. Model of the bow section of the caisson used to construct the bridge pier at Chalon-sur-Saône (C. J. Brandon).

Fig. 8.57. Detail of the floor timbers and keel on the model of the bow section of the caisson used to construct the bridge pier at Chalon-sur-Saône (C. J. Brandon).

Fig. 8.58. Concrete block with remains of double-walled floating caisson in Area G at Sebastos (Courtesy of CAHEP).

Fig. 8.59. Reconstruction of double-walled floating caisson in Area G at Sebastos (C. J. Brandon after S. Talaat).

Fig. 8.60. Reconstruction of the framing of a double-walled, prefabricated floating caisson in Area G at Sebastos (Courtesy of CAHEP).

Fig. 8.61. Reconstruction of floating formwork being manoeuvred into position over Area G at Sebastos (Hohlfelder 1987: 264–65) (National Geographic Society, used with permission).

Fig. 8.62. Reconstruction of a prefabricated rectangular caisson as found in Area K at Caesarea (C. J. Brandon).

Fig. 8.63. Detail of Area K caisson, junction of chine with side wall and flooring (C. J. Brandon).

Fig. 8.64. Detail of Area K caisson, junction of chine with end wall and floor planking, caisson K-2, Caesarea (C. J. Brandon).

Fig. 8.65. Detail of Area K caisson, junction of chine with end wall and floor planking, caisson K-3, Caesarea (C. J. Brandon).

Fig. 8.66. Detail of a floor frame, Area K caisson, Caesarea (A. Raban).

Fig. 8.67. Detail of an external corner stanchion with side wall planking let into it, Area K caisson, Caesarea (C. J. Brandon).

Fig. 8.68. External corner detail with cover-piece over junction of chine beams, Area A caisson, Caesarea (C. J. Brandon).

Fig. 8.69. Corner of caisson K-3, Caesarea (A. Raban).

Fig. 8.70. Cross section throughArea K caisson, Caesarea (C. J. Brandon).

Fig. 8.71. Detail of stringer let into chine beam at bow and stern, Area K caisson, Caesarea (C. J. Brandon).

Fig. 8.72. The inner cell within caisson K-2, Caesarea (A. Raban).

Fig. 8.73. Detail of the corner of the inner cell within caisson K-2, Caesarea (C. J. Brandon).

Fig. 8.74. Detail of the corner of the inner cell within caisson K-2, Caesarea (A. Raban).

Fig. 8.75. Section of the side of caisson K-2, Area K, Caesarea (A. Raban).

Appendix 3. In these figures the upper part of the core is always at the left side of the photograph. Images of individual cores progress from more general to more detailed views.

Fig. A3.1: SLI.2003.01, Overview of core. Pumiceous mortar with abundant relict lime clasts and sea-water saturated pumiceous tuff caementa.

Fig. A3.2: SLI.2003.01, central section, -0.40 to -1.02 m. Pumiceous mortar with abundant relict lime clasts and sea-water saturated pumiceous tuff caementa.

Fig. A3.3: SLI.2004.01. Overview of core, the longest recovered by ROMACONS, 5.8 m.

Fig. A3.4: SLI.2004.01, detail, -1.80 to -2.40 m. Pumiceous tuff caementa and well-consolidated pumiceous mortar with abundant relict lime clasts. A possible trace of a relict lime putty-volcanic ash mixture occurs at the break in the core.

Fig. A3.5: SLI.2004.01, detail -3.38 to -3.65 m. Pumiceous tuff caementa and well-consolidated pumiceous mortar with abundant relict lime clasts,. An in situ reaction rim occurs in the interfacial zone of the tuff caementa on left.

Fig. A3.6: SLI.2004.01, detail, -0.35 to -0.55 m. Mortar with abundant relict lime, particles of gravel-sized, one ceramic fragment, and sea-water saturated volcanic glass and pumice clasts. The caementa is sea-water saturated pumiceous tuff on the right, and ceramic on the left.

Fig. A3.7: PCO.2003.01. Overview of core, pumiceous tuff caementa and one limestone fragment, and pumiceous mortar with abundant lime clasts.

Fig. A3.8: PCO.2003.01, detail, 0 to -0.58 m. Pumiceous tuff and gray limestone caementa, and mortar with relict lime clasts and possible clots of relict lime putty.

Fig. A3.9: PCO.2003.01, detail, -0.50 to -1.10 m. Pumiceous tuff caementa and pumiceous mortar with abundant lime clasts.

Fig. A3.10: PCO.2003.02. Overview of core, with diverse caementa, ceramics, lava, and pumiceous tuff, and a relatively low proportion of mortar with complex relict lime clasts. The red tint is caused by rust in core tube.

Fig. A3.11: PCO.2003.02, detail 0 to -0.46 m. Ceramic and lava caementa in iron stained concrete, an artefact of the drilling process, and mortar with pale orange pumiceous pozzolan and complex relict lime clasts.

Fig. A3.12: PCO.2003.02, detail -1.25 to -1.35 m. Concrete with pyroclastic rock caementa, containing large lithic fragments of carbonate bedrock. The origin of these rocks is not known.

Fig. A3.13: PCO.2003.03, detail -0.78 to -1.31 m. Pumiceous tuff caementa and pumiceous mortar with small relict lime clasts and fragments of poorly calcined limestone.

Fig. A3.14: PCO.2003.03, detail -0.98 to -1.10 m. Mortar with pale orange-gray pumice and relict lime clasts.

Fig. A3.15: PCO.2003.04. Overview of core, with pumiceous tuff caementa, and mortar with abundant relict lime clasts.

Fig. A3.16: PCO.2003.04, detail -0.20 to -0.35 m. Mortar with complex relict lime fabrics and precipitation of carbonate textures on the surfaces of compaction flaws.

Fig. A3.17: PCO.2003.04, detail -0.60 to -0.70 m. Stratified mortar with fine ash pozzolan at base (right) and coarse ash with gray pumice clasts coarsening upwards.

Fig. A3.18: PCO.2003.05. Overview of highly weathered concrete core, with pumiceous tuff caementa and mortar altered to dark green earthy fabrics. Iron staining is an artefact of the drilling process.

Fig. A3.19: POR.2002.01, surviving fragment, ca. -1.0 m. Concrete with Tufo Lionato caementa from the Alban Hills volcanic district, and porous mortar with scoriaceous ash from the Alban Hills volcanic district, yellow-gray pumiceous ash from the Campi Flegrei volcanic district, and ceramic fragments.

Fig. A3.20: POR.2002.02, lower sections, -1.58 to -1.98 m, and ca. -2.32 to 3.14 m. Tufo Lionato caementa and mortar with relict lime clasts.

Fig. A3.21: POR.2002.02, detail -1.06 to -1.38 m. Mortar with abundant lime, partly as relict putty clasts.

Fig. A3.22: POR.2002.02, detail -2.42 to -2.48 m. Tufo Lionato caementa with palagonitic glass and natural zeolite cements, and mortar with abundant relict lime.

Fig. A3.23: POR.2002.03. Surviving fragment of sea-water saturated concrete with pumiceous mortar and Tufo Lionato caementa.

Fig. A3.24: PTR.2002.01, detail -0.42 to -1.16 m. Tufo Lionato and ceramic caementa, and a well-consolidated mortar with gray pumiceous ash pozzolan.

Fig. A3.25: PTR.2002.01, detail -1.10 to -1.81 m. Well-consolidated mortar with pale orangish-gray pumiceous ash pozzolan, lava lithic fragments and scoriaceous ash, and relatively few coarse relict lime clasts. Ceramic and sea-water saturated Tufo Lionato caementa.

Fig. A3.26: PTR.2002.02, detail -0.60 to -0.80 m. Well-consolidated mortar with abundant sand- to gravel-sized particles of scoriaceous volcanic ash and Tufo Lionato caementa.

Fig. A3.27: PTR.2002.02, detail -1.35 to -1.50 m. A layer of white to light greenish-grey lime (to left) overlies light greenish grey volcanic pozzolan, and Tufo Lionato caementa on right.

Fig. A3.28: ANZ.2002.01, detail -1.20 to -1.70 m. Pumiceous tuff caementa and mortar with pumice and relict lime clasts.

Fig. A3.29: ANZ.2002.01, detail -1.65 to -1.80 m. Sea-water saturated pumiceous tuff caementa and mortar with abundant relict lime and lava lithic fragments.

Fig. A3.30: ANZ.2002.01, detail -1.76 to -2.01 m. Stratified deposit of fine relict light greyish white lime and fine pumiceous ash.

Fig. A3.31: BAI.2006.01. Overview of core, with pumiceous tuff caementa and mortar with relict lime clasts.

Fig. A3.32: BAI.2006.01, detail -1.60 to -1.75 m. Mortar with pale yellowish-gray glass and pumice clasts, and a large clot of relict lime.

Fig. A3.33: BAI.2006.01, detail -1.85 to -2.0 m. Stratified mortar with fine ash pozzolan and relict lime at base (right), and overlain by mortar with coarse yellowish-gray pumice clasts.

Fig. A3.34: BAI.2006.03. Overview of sea-water saturated core, with pumiceous tuff caementa and mortar with relict lime clasts.

Fig. A3.35: BAI.2006.03. detail -0.50 to -1.02 m. Pumiceous tuff, and mortar with relict lime clasts.

Fig. A3.36: BAI.2006.03. detail -1.90 to -1.98 m. Mortar with yellowish-gray and moderate gray pumice and glass fragments, and relict lime clasts.

Fig. A3.37: BAI.2006.02. Overview of sea-water saturated core with abundant weathered pumiceous tuff caementa.

Fig. A3.38: BAI.2006.04. Overview of sea-water saturated core with abundant weathered pumiceous tuff caementa.

Fig. A3.39: BAI.2006.04, detail -1.20 to -1.35 m. Sea-water saturated mortar and pumiceous tuff caementa.

Fig. A3.40: BAI.2006.05, detail -0.40 to -0.50 m. Sea-water saturated mortar and pumiceous tuff caementa with blue-green alteration in interfacial zones.

Fig. A3.41: EGN.2008.01. Overview of sea-water saturated core, with calcareous sandstone caementa.

Fig. A3.42: EGN.2008.01, -1.91 to -2.18 m. Sea-water saturated calcareous sandstone caementa and pumiceous mortar with relict lime clasts.

Fig. A3.43: BRI.2005.01. Overview of core, experimental concrete reproduction after six months hydration in sea-water.

Fig. A3.44: BRI.2005.01, detail -0.80 to -1.15 m. Pumiceous Bacoli Tuff and experimental mortar reproduction after six months hydration in sea-water.

Fig. A3.45: BRI.2005.02. Overview of core, experimental concrete reproduction after twelve months hydration in sea-water.

Fig. A3.46: BRI.2005.02, detail -1.27 to -1.53 m. Experimental concrete reproduction at 12 months hydration in sea-water, showing compaction flaws and a single relict lime clast in the hydrated lime putty – Bacoli volcanic ash mortar.

Fig. A3.47: BRI.2006.01. Overview of core, sea-water-soaked experimental concrete reproduction after twenty-four months hydration in sea-water.

Fig. A3.48: BRI.2008.01. Overview, experimental concrete reproduction after forty-eight months hydration in sea-water.

Fig. A3.49: BRI.2008.01, detail -0.65 to -0.83 m. Pumiceous Bacoli Tuff and experimental mortar reproduction after forty-eight months hydration in sea-water.

Fig. A3.50: BRI.2008.01, detail -1.48 to -1.53 m. Experimental concrete reproduction at forty-eight months hydration in sea-water shows yellowish-gray Bacoli Ash pumiceous pozzolan, and a clot of lime putty mixed with pumiceous ash.

Fig. A3.51: CHR.2007.02. Overview of core, limestone caementa and porous, earthy mortar with abundant pale orange pumiceous ash and tuff