The Mortella III Wreck: a Spotlight on Mediterranean Shipbuilding of the 16th Century 9781407354132, 9781407356228

Citation preview

L E A IN N L IO ON IT D L D IA A ER AT

M

The Mortella III Wreck A Spotlight on Mediterranean Shipbuilding of the 16th Century Arnaud Cazenave de la Roche B A R I N T E R NAT I O NA L S E R I E S 2 9 7 6

2020

The Mortella III Wreck A Spotlight on Mediterranean Shipbuilding of the 16th Century Arnaud Cazenave de la Roche B A R I N T E R NAT I O NA L S E R I E S 2 9 7 6

2020

Published in 2020 by BAR Publishing, Oxford BAR International Series 2976 The Mortella III Wreck: a Spotlight on Mediterranean Shipbuilding of the 16th Century isbn  

978 1 4073 5413 2 paperback isbn   978 1 4073 5622 8 e-format doi  https://doi.org/10.30861/9781407354132

A catalogue record for this book is available from the British Library © Arnaud Cazenave de la Roche 2020 cov er i m age

Framework of the Mortella III wreck at mid-ship (photo: Christoph Gerigk).

The Author’s moral rights under the 1988 UK Copyright, Designs and Patents Act are hereby expressly asserted. All rights reserved. No part of this work may be copied, reproduced, stored, sold, distributed, scanned, saved in any form of digital format or transmitted in any form digitally, without the written permission of the Publisher. Links to third party websites are provided by BAR Publishing in good faith and for information only. BAR Publishing disclaims any responsibility for the materials contained in any third party website referenced in this work.

BAR titles are available from: BAR Publishing 122 Banbury Rd, Oxford, em a i l [email protected] phon e +44 (0)1865 310431 fa x +44 (0)1865 316916 www.barpublishing.com

ox 2

7bp,

uk

Of Related Interest Ships, Boats, Ports, Trade, and War in the Mediterranean and Beyond Proceedings of the Maritime Archaeology Graduate Symposium 2018 Naseem Raad and Carlos Cabrera Tejedor Oxford, BAR Publishing, 2020

BAR International Series 2961

Analysing Maritime Archaeological Archives Collections, Access and Management Julie Satchell with contributions by Jesse Ransley and Julian Whitewright Oxford, BAR Publishing, 2016 The Maritime Archaeology Trust Monograph Series (formerly Hampshire and Wight Trust for Maritime Archaeology Monograph Series), 3

BAR British Series 628

The Gresham Ship Project A 16th-Century Merchantman Wrecked in the Princes Channel, Thames Estuary Volume I: Excavation and Hull Studies Edited by Jens Auer and Thijs J. Maarleveld with contributions by Massimiliano Ditta, Antony Firth, Nigel Nayling, Delia Ní Chíobháin, Christian Thomsen, and Cate Wagstaffe Oxford, BAR Publishing, 2014 Nautical Archaeology Society Monograph Series (NAS), 4

BAR British Series 602

The Archaeology and History of the Flower of Ugie, Wrecked 1852 in the Eastern Solent Edited by Julian Whitewright and Julie Satchell Oxford, BAR Publishing, 2011 Hampshire and Wight Trust for Maritime Archaeology Monograph Series (now The Maritime Archaeology Trust Monograph Series), 1

BAR British Series 551

The Zeebrugge Shipwreck A forgotten early sixteenth-century merchantman discovered off the Belgian coast Hendrik Lettany Oxford, BAR Publishing, 2018 Local Boats

BAR International Series 2898

Fourth International Symposium on the Boat and Ship Archaeology, Porto 1985 Edited by Octavio Lixa Filgueiras Oxford, BAR Publishing, 1988

BAR International Series 438

Post-mediaeval Boat and Ship Archaeology Edited by Carl Olof Cederlund Oxford, BAR Publishing, 1985

For more information, or to purchase these titles, please visit www.barpublishing.com

BAR International Series 256

The Mortella III excavation Scientific and technical contributors Prof. Ana Crespo Solana (CSIC, Spain), scientific advice and collaboration Prof. Nigel Nayling (University of Wales), scientific advice and wood study Fabien Langenegger (OPAN, Switzerland) wood study and dendrochronology Dra. Marta Domínguez, dendrochronology Dr. Franck Allegrini Simonetti (CdC), Prof. Marco Milanese (Università degli Studi di Sassari), Emilie Thomas, ceramics studies Max Guérout (GRAN), artillery study Dr. Jean-Bernard Memet and Philippe de Viviès (A-CORROS), conservation protocol Maite Segura García and Inmaculada Rigo, conservation laboratory Dr. Carole Mathe (University of Avignon), chemical analysis Prof. Antoine-Marie Graziani (University of Corsica Pasquale Paoli) and Dr. Renato Gianni Ridella (University of Genoa), historical research Fabrizio Ciacchella (University of Genoa), anchors study Jesús Guevara (Aingurak), Dr. Samantha Heitzmann (Paris I - Panthéon Sorbonne University), and Bérenger Debrand, drawings Dr. François Gendron (MNHN), lithic studies Dr. Erik Gonthier and Dr. Jean-Marc Valère (MNHN), ropes study Christoph Gerigk, photography and photomosaic Brandon Mason, Christin Heamagi and Garry Momber (Maritime Archaeology Trust), photogrammetry Stéphane Jamme, drone, photography Diving team Barbot, Alexandra

Filippi, Jean-José

Heamagi, Christin

Martins, Guillaume

Bertoncini, Alain

Gendron, François

Heitzmann, Samantha

Mason, Brandon

Bourdeaud’hui, Cédric

Gerigk, Christoph

Jamme, Stéphane

Momber, Garry

C. de la Roche, Arnaud

Grimond, Jonathan

Joyard, Anne

Nayling, Nigel

Casamarta, Dominique

Coquoz, Xavier

Kühn, Laurent

Philips, Agnès

Catteau, Sidonie

Couppey, Antoine

Langenegger, Fabien

Pinelli, Charles

Clément, Normann

Debrand, Bérenger

Lanleau, Jacques

Sanchez, Didier

Ciacchella, Fabrizio

Guesnon, Joë

Mager, Patrick

Ter-Jung, Marine

Drogue, Gilles

Guevara, Jesús

Marie, Jehan

v

Acknowledgements This book is the culmination of a long journey initiated in 2005 and 2006 with the discovery of the Mortella shipwrecks in the Bay of Saint-Florent and the excavation of one of them, Mortella III, between 2010 and 2019. It is the result of a doctoral thesis defended at Sorbonne University in 2018. I would like to warmly thank the people and institutions that have made this work possible: Professor Ana Crespo Solana (CSIC), Professor Filipe Vieira de Castro (Texas A&M University), Professor Olivier Chaline (University of Paris-Sorbonne), Max Guérout (GRAN), Cayetano Hormaechea, Professor Brad Loewen (University of Montréal), Professor Nigel Nayling (University of Wales), Charles Pinelli (Les Amis des Agriate), Stéphane Orsini (FAGEC). François Pinelli, Vincent Maliet (CdC), Antoine Couppey, Hélène and Philippe Epaminondi. The excavation of the Mortella III wreck has been possible thanks to: • The support of the DRASSM (Department for Underwater Archaeological Research, French Ministry of Culture), the Government of Corsica (Collectivité de Corse, CdC), the city of Saint-Florent. • The collaboration of the University of Paris-Sorbonne, the University of Corsica Pasquale Paoli, the Spanish National Research Council–CSIC-, The ForSEAdiscovery consortium, the Maritime Archaeology Trust (MAT, UK) and the Museum National d’Histoire Naturelle of Paris (MNHN). • The grant of the EU programme Marie Slodowska-Curie Actions – MSCA- Horizon 2020 (MSCA, IF No. 843337). I would also like to express my gratitude to the entire team of people, divers, archaeologists and specialists who have put their knowledge and know-how at the service of this excavation. This work is above all the product of their remarkable and dedicated work. The list of members of this team and the scientific/technical contributors are listed on the previous page. Finally, I would like to express my gratitude to my wife, Maite Segura, for her unwavering support and patience.

vii

Contents List of Figures................................................................................................................................................................... xiii List of Tables..................................................................................................................................................................... xix Lists of Abbreviations ..................................................................................................................................................... xxi Introduction......................................................................................................................................................................... 1 The Mortella wrecks: a valuable source of information.................................................................................................. 1 The issues addressed and the objectives of the study...................................................................................................... 1 The limits of the archaeological study............................................................................................................................. 2 Construction techniques................................................................................................................................................... 3 The architecture of the ship.............................................................................................................................................. 4 Archaeological sources.................................................................................................................................................... 4 Documentary sources....................................................................................................................................................... 4 The sources of the historical perspective......................................................................................................................... 5 1. Shipbuilding in the Mediterranean in the 16th century: State of the art and issues................................................. 7 1.1. The concepts of ‘maritime space’ and ‘technical culture’ at the heart of current issues........................................... 7 1.2. Construction techniques: in search of ‘technical fingerprints’.................................................................................. 8 1.3. The components of the architectural project: in search of an architectural model................................................... 8 1.3.1. Proportions: the relationship between dimensions............................................................................................ 8 1.3.2. The form: the shape of the master-frame........................................................................................................ 13 1.3.3. The design method: the moulding process...................................................................................................... 18 2. Portraits of the Mortella wrecks: The discovery of the sites, their characteristics, their chronology & excavation methods.................................................................................................................................. 23 2.1. The Mortella II and III: discovery of two sites of the same origin. Characteristics and background..................... 23 2.1.1. The discovery of the sites ............................................................................................................................... 23 2.1.2. General description of the Mortella sites at the time of their discovery ........................................................ 23 2.1.2.1. Site characteristics................................................................................................................................... 23 2.1.2.2. Nature of the sites.................................................................................................................................... 25 2.1.2.3. The link between the two sites................................................................................................................ 25 2.1.3. Background and history of interventions........................................................................................................ 25 2.2. Archaeological material and wood of the hull: the artefacts and the material helping the chronology ................. 26 2.2.1. The contribution of the artefacts to the understanding of the wreck and to its dating.................................... 27 2.2.1.1. Artillery.................................................................................................................................................... 27 2.2.1.2. Ceramics ................................................................................................................................................. 27 2.2.2. The wood: its characteristics and contribution to dating using dendrochronology......................................... 28 2.2.2.1. The wood and its origin........................................................................................................................... 28 2.2.2.2. Its use in shipbuilding.............................................................................................................................. 29 2.2.2.3. Woodworking and tool marks.................................................................................................................. 31 2.2.2.4. Dendrochronological study..................................................................................................................... 31 2.3. General layout of the Mortella III site and context of the shipwreck .................................................................... 32 2.3.1. General layout of the remains of the hull (fig.26) .......................................................................................... 32 2.3.2. The context surrounding the sinking............................................................................................................... 32 2.4. Defining the research programme: research axis, methodology and excavation strategy...................................... 36 2.4.1. Methodology and excavation strategy............................................................................................................. 36 2.4.1.1. Objectives and excavation strategy......................................................................................................... 36 2.4.1.2. Means and excavation methodology ...................................................................................................... 36 2.4.2. Conservation and protection of the site .......................................................................................................... 41 2.4.2.1. Preventive conservation of the artefacts ................................................................................................. 41 2.4.2.2. Conservation of the remains on the site.................................................................................................. 42

ix

The Mortella III Wreck 3. The hull of the Mortella III wreck and its construction method.............................................................................. 43 3.1. The transversal framework: organization, dimensions, morphology and scarfs methods...................................... 43 3.1.1. General layout of the transverse framing ....................................................................................................... 43 3.1.1.1. Definition of the frames groups............................................................................................................... 43 3.1.1.2. Sequencing of the frame components, room and space ......................................................................... 45 3.1.1.3. The frames of the aft part of the hull....................................................................................................... 47 3.1.1.4. Frames located between the tailframes.................................................................................................... 49 3.1.1.5. Frames of the fore part of the hull........................................................................................................... 52 3.1.2. Components of the frames (see Table 10 for a summary of dimensions)....................................................... 57 3.1.2.1. The floor-timbers..................................................................................................................................... 57 3.1.2.2. Scarfs of the floor-timbers to the first-futtocks........................................................................................ 60 3.1.2.3. The first-futtocks ..................................................................................................................................... 63 3.1.2.4. Scarfs from first to second-futtocks ........................................................................................................ 64 3.1.2.5. The second-futtocks................................................................................................................................. 66 3.1.2.6. Summary tables of the frames measurements......................................................................................... 66 3.1.3. Scarfs typology of the frame components ...................................................................................................... 66 3.1.3.1. A technical choice ................................................................................................................................... 66 3.1.3.2. A possible architectural design function.................................................................................................. 69 3.2. The longitudinal framework: layout, dimensions, morphology and scarfs methods.............................................. 71 3.2.1. The keel, the heel of the keel and the keelson................................................................................................. 71 3.2.1.1. Search for keel length.............................................................................................................................. 71 3.2.1.2. Morphology of the keel and types of scarfs............................................................................................ 73 3.2.1.3. The heel of the keel, its scarf to the keel and the stern............................................................................ 75 3.2.1.4. The keelson.............................................................................................................................................. 79 3.2.2. Clamps, sill and ceiling .................................................................................................................................. 80 3.2.2.1. Clamps..................................................................................................................................................... 80 3.2.2.2. The ceiling and the sill. ........................................................................................................................... 84 3.2.3. The planking and its fastening method to the frames...................................................................................... 84 3.2.3.1. The wood ................................................................................................................................................ 84 3.2.3.2. Morphology and dimensions................................................................................................................... 85 3.2.3.3. Fastening method of the planking to the frames..................................................................................... 86 3.2.3.4. The forward end of the planking............................................................................................................. 88 3.2.3.5. Caulking and sealing products................................................................................................................. 90 3.3. Fastening system for the timbers............................................................................................................................ 92 3.3.1. Iron nails and absence of treenails ................................................................................................................. 92 3.3.2. Circular section of the nails............................................................................................................................. 93 3.3.3. Folded down nails tips .................................................................................................................................... 94 4. The attributes of the hull: The mainmast step, the pump, and the rudder............................................................. 95 4.1. The main mast-step................................................................................................................................................. 95 4.1.1. The longitudinal timbers of the mast-step arrangement (or sister-keelsons).................................................. 95 4.1.2. The buttresses.................................................................................................................................................. 95 4.1.3. The keys.......................................................................................................................................................... 96 4.1.4. A major Mediterranean technical ‘fingerprint’ ............................................................................................... 96 4.2. The pump.............................................................................................................................................................. 100 4.2.1. The foot-valve............................................................................................................................................... 102 4.2.2. The pump-tube.............................................................................................................................................. 102 4.3. The rudder............................................................................................................................................................. 107 4.3.1. The rudder device: definitions and vocabulary ............................................................................................ 108 4.3.2. Morphology of the rudder blade.................................................................................................................... 108 4.3.3. The nailing .................................................................................................................................................... 109 4.3.4. The bolts........................................................................................................................................................ 109 4.3.5. The iron-works...............................................................................................................................................111 4.4. Analysis and interpretation of the remains uncovered.......................................................................................... 114 5. The architectural profile of the Mortella III wreck: Shapes and proportions.......................................................117 5.1. Attempt to restore the shape of the master-frame M27......................................................................................... 117 5.1.1. Position of the master-frame on the keel....................................................................................................... 117 5.1.2. Characteristics and shape of the frame.......................................................................................................... 118 5.1.2.1. The rising and narrowing of the master floor-timber ........................................................................... 118 x

Contents 5.1.2.2. The shape of the master-frame M27...................................................................................................... 121 5.1.2.3. The design of the master-frame............................................................................................................. 122 5.1.2. Restoration of the master-frame shape ......................................................................................................... 125 5.2. The deck and depth of hold................................................................................................................................... 129 5.2.1. The number of decks..................................................................................................................................... 129 5.2.2. Decks height.................................................................................................................................................. 129 5.2.3. The depth of hold.......................................................................................................................................... 132 5.3. Rake of the sternpost and shape of the stem......................................................................................................... 133 5.3.1. The rake of the sternpost and the overhang................................................................................................... 133 5.3.2. The stern and its overhang............................................................................................................................. 135 5.4. Proportions of the ship, the attempt to restore her shape and longitudinal dimensions ....................................... 137 5.4.1. The main proportion ratios............................................................................................................................ 137 5.4.1.1. The keel width/length ratio.................................................................................................................... 138 5.4.1.2. Summary of proportions........................................................................................................................ 139 5.4.2. Attempt to restore the longitudinal shape and the dimensions of the ship.................................................... 139 5.5. The gauging........................................................................................................................................................... 139 5.5.1. An essential unit of volume in 16th century shipbuilding.............................................................................. 139 5.5.2. Calculation of tonnage in the 16th century..................................................................................................... 141 5.5.2.1. The Venetian method............................................................................................................................. 141 5.5.2.2. The Spanish method.............................................................................................................................. 141 6. Historical study: Issues, references, and identification attempt of the Mortella wrecks ..................................... 143 6.1. Hypothesis and historical issues........................................................................................................................... 143 6.2. The naval combat and shipwrecks of 1555........................................................................................................... 143 6.2.1. The historical context of the period............................................................................................................... 143 6.2.2. Related Texts ................................................................................................................................................ 143 6.2.3. Analysis of the hypothesis of the shipwrecks of 1555.................................................................................. 145 6.3. The shipwrecks of 1526........................................................................................................................................ 146 6.3.1. The historical context of the period............................................................................................................... 146 6.3.2. Related texts ................................................................................................................................................. 146 6.3.3. Analysis of the texts on shipwrecks of 1526 ................................................................................................ 147 6.4. The shipwrecks of 1527........................................................................................................................................ 148 6.4.1. The historical context of the period............................................................................................................... 148 6.4.2. Related texts.................................................................................................................................................. 148 6.4.3. Analysis of the texts on shipwrecks of 1527................................................................................................. 149 6.5. Provisional lessons drawn from historical research.............................................................................................. 150 6.5.1. The architectural portrait of a Genoese nave................................................................................................. 150 6.5.2. Historical landmarks on the development of the Genoese nave................................................................... 151 Conclusion....................................................................................................................................................................... 153 The architectural portrait of the Mortella III ship........................................................................................................ 153 The historical portrait of the Mortella III ship............................................................................................................. 153 The contribution of the wreck to a Mediterranean technical model............................................................................ 154 The contribution of the wreck to the definition of a Mediterranean architectural model............................................ 155 The exploration of the Venetian treatises in relation to the Mortella III wreck. ......................................................... 156 Future prospects for knowledge of Mediterranean shipbuilding. ............................................................................... 157 Sources snd Bibliography............................................................................................................................................... 159 Glossary........................................................................................................................................................................... 169 Annex 1. General layout and planimetry of the Mortella III wreck.......................................................................... 173 Annex 2. Photomosaic of the Mortella III wreck......................................................................................................... 175 Annex 3. Dendrochronological study........................................................................................................................... 177 Annex 4. Study of the transverse compression of the Mortella III timbers.............................................................. 191 Annex 5. Chemical analysis of caulking and sealing materials.................................................................................. 197 xi

The Mortella III Wreck Annex 6. Main original texts resulting from literature research................................................................................ 203 Annex 7. Selection of texts mentioning or inducing the As-Dos-Tres rule................................................................. 209 Annex 8. Study of the West Anchor of the Mortella III wreck....................................................................................211 Annex 9. Divers team list................................................................................................................................................ 213 Annex 10. Photos of the excavation .............................................................................................................................. 215 Annexes 1 and 2 are also available for download from barpublishing.com/additional-downloads.html

xii

List of Figures Figure 1. Geographical location of the Mortella wrecks...................................................................................................... 6 Figure 2. Master-frame shape of the 16 codos galleon as prescribed by the Spanish Ordinances of 1613........................ 14 Figure 3. Shape of the master-frame of a 48 palmos ship, according to F. Oliveira, 1570, fº112...................................... 14 Figure 4. Shape of the master-frame of a 500-ton ship, according to Manoel Fernandes, 1616........................................ 15 Figure 5. Top (a): Shape of the master-frame recommended by Matthew Baker Bottom (b): Representation of the master-frame of the Red-Bay wreck......................................................................................................................... 15 Figure 6. Shape of a Venetian ship master-frame obtained by four tangent arcs according to Matthew Baker................. 16 Figure 7. Zorzi Trombetta da Modon’s method for tracing the shape of the master-frame of a nave................................ 16 Figure 8. Restoration of the shape of the master-frame of the nave of 700 botte by Zorzi Trombetta da Modon............. 17 Figure 9. Superposition of fig. 6 and 8: similarity of Zorzi Trombetta’s shapes and Mathew Baker’s Venetian model.... 17 Figure 10. Superposition of fig. 5a. and 8: forms of Zorzi Trombetta and M. Baker’s English model.............................. 18 Figure 11. Shape of the master-frame of the nave cuadra of ‘Fabrica di galere’ obtained from the measurements provided in folio 37............................................................................................................................................................. 19 Figure 12. Superposition of fig. 5a and 11: coincidence of shapes of the ‘Fabrica di galere’ and Matthew Baker’s English models.................................................................................................................................................................... 19 Figure 13. The ‘graminho’, an instrument for the narrowing (OLIVEIRA, 1570, Fº93)................................................... 20 Figure 14. Fernando Oliveira illustrates the rising of the ‘madeiros da conta’: (OLIVEIRA, 1570, Fº103)..................... 21 Figure 15. The ‘partisone’ system according to Zorzi Trombetta de Modon (TROMBETTA, 1445, Fº45) The ‘Mezzaluna’......................................................................................................................................................................... 21 Figure 16. SONAR image of the site of Mortella III that led to its discovery.................................................................... 23 Figure 17. General organization of the site at the time of its discovery............................................................................. 24 Figure 18. Site of Mortella III. Tumulus B at the time of its discovery.............................................................................. 25 Figure 19. Wrought iron bombard (Cn3)............................................................................................................................ 26 Figure 20. 225 mm caliber shot. Note the inscription carved in stone............................................................................... 27 Figure 21. Medium cup. Edge with incised decoration, the colour of the paste is buff to pink......................................... 27 Figure 22. European distribution of sessile oak (Quercus petrae)..................................................................................... 28 Figure 23. 300-year-old sessile oak from Burgundy........................................................................................................... 29 Figure 24. The adze, an essential cutting tool in the shaping of the timbers...................................................................... 31 Figure 25. Cross-section of the M20 futtock that gave an absolute dating......................................................................... 31 Figure 26. General planimetry of the Mortella III site........................................................................................................ 33 Figure 27. Remains uncovered under tumulus A. Extract from the photomosaic.............................................................. 34 Figure 28. Top: tumulus B (extract from the mosaic photo). Bottom: front part of tumulus A (crutches and fore end of the keel).......................................................................................................................................................................... 35 Figure 29. Excavation areas of the Mortella III site from 2007 to 2015. In green: survey areas (2007)............................ 37

xiii

The Mortella III Wreck Figure 30. Sediment clearance on the remains of the hull.................................................................................................. 38 Figure 31. Transverse cross-section survey at the beam..................................................................................................... 38 Figure 32. The grid square in place on the remains of tumulus A...................................................................................... 39 Figure 33. Scales placed on the bottom to allow calibration of the photomosaic.............................................................. 39 Figure 34. 3D survey of the heel of the keel...................................................................................................................... 40 Figure 35. Photography, drawing and information processing in the excavation laboratory............................................. 40 Figure 36. Video recording by a R.O.V. and manual recording on the bottom.................................................................. 41 Figure 37. Covering the excavation areas with a geotextile fabric (a.) in turn covered by a layer of sediment (b.).......... 41 Figure 38. The frame of the starboard side at the level of the beam................................................................................... 44 Figure 39. Framing at mid-ship.......................................................................................................................................... 44 Figure 40. The floor-timber V30......................................................................................................................................... 45 Figure 41. Planimetry of the central and fore areas............................................................................................................ 46 Figure 42. Room and space and intervals at mid-ship........................................................................................................ 47 Figure 43. Reverse assembly sequence from master-frame M.27...................................................................................... 49 Figure 44. The master-frame on the starboard side............................................................................................................ 50 Figure 45. The fracture of the floor-timbers V16 and V17................................................................................................. 50 Figure 46. The fracture of the floors-timbers V20 and V21............................................................................................... 51 Figure 47. Top: Broken head of floor-timber V27. Bottom: Breakpoint indication at the level of the transverse profile of the M27 frame measured on the bottom............................................................................................................. 51 Figure 48. Cross-section of frames M20 and M22............................................................................................................. 52 Figure 49. Cross section of the frame M24......................................................................................................................... 53 Figure 50. Finding of the port side frame under the tumulus B in 2012............................................................................. 54 Figure 51. Planimetry of the excavation area AF12/B (tumulus B).................................................................................... 54 Figure 52. Cross-sections of frames MB2, MB5 and MB (tumulus B).............................................................................. 55 Figure 53. Location plan. Excavation area AF13/12.......................................................................................................... 56 Figure 54. Planimetry of the fore end of the wreck (Excavation area AF 13/1)................................................................. 56 Figure 55. Cross section of the M2P, M4P and M5P frames (Excavation area AF13/1).................................................... 57 Figure 56. Planimetry of the excavation area AF12/B (tumulus B).................................................................................... 58 Figure 57. Photograph of the master floor-timber V27....................................................................................................... 59 Figure 58. Master floor-timber record V27......................................................................................................................... 59 Figure 59. Scarf of the floor-timber V18 with its first-futtock G18.................................................................................... 61 Figure 60. Oblique nailing of the second nail joining the floor-timbers to the first futtocks.............................................. 62 Figure 61. Assembly of floor-timber V27 with first first-futtocks G27A and G27B.......................................................... 62 Figure 62. Photographic representation of the assembly of the floor-timber V27 with the First-futtock G27A................ 62 Figure 63. Photomontage: Transparent overlay of the two timbers V27 and G27............................................................. 63 Figure 64. Transparent representation of the V27 - G27A assembly and nailing............................................................... 63 Figure 65. G27B : face attached to V27 (Left) and fore face of the floating end (right).................................................... 63 Figure 66. V1P to V8P frames of the aft of the ship........................................................................................................... 63

xiv

List of Figures Figure 67. First futtocks G27 A and G27 B. Note the tapering of the ends........................................................................ 64 Figure 68. First futtock recording G27A............................................................................................................................. 64 Figure 69. First futtock recording G27B............................................................................................................................. 65 Figure 70. View of the lower face of the assembly of G27A to A27.................................................................................. 65 Figure 71. Photographic representation of the assembly G27A to A2................................................................................ 66 Figure 72. Frame M27 and timbers assembly layout.......................................................................................................... 67 Figure 73. Record of the first futtock G27A....................................................................................................................... 68 Figure 74. Record of the second futtock A27..................................................................................................................... 68 Figure 75. Hook scarfs of the frames of the wreck of Yassiada I....................................................................................... 70 Figure 76. Port side of the keel end.................................................................................................................................... 71 Figure 77. End of the keel seen on its upper face............................................................................................................... 72 Figure 78. Keel scarf as revealed under floor-timber V26.................................................................................................. 72 Figure 79. Butt scarf discovered under floor-timber V25................................................................................................... 73 Figure 80. Two overlapped keel timbers............................................................................................................................. 75 Figure 81. Schematic representation of the keel................................................................................................................. 75 Figure 82. Diagram of the keel morphology....................................................................................................................... 76 Figure 83. The heel of the keel........................................................................................................................................... 77 Figure 84. Heels of the keel of the wrecks of Villefranche and Mortella III...................................................................... 77 Figure 85. The heel of the keel of the Red-Bay wreck....................................................................................................... 78 Figure 86. The heel of the keel according to Lavanha, fº43............................................................................................... 78 Figure 87. The heel of the keel of the Aveiro A wreck....................................................................................................... 78 Figure 88. The heel of the keel of the Corpo Santo wreck................................................................................................. 79 Figure 89. The heel of the keel of the 29 M wreck (Red-Bay)........................................................................................... 79 Figure 90. The heel of the keel of the Calvi 1 wreck.......................................................................................................... 79 Figure 91. Assembly of the heel of the keel to the stern on the Red-Bay wreck................................................................ 79 Figure 92. Scarf keel / heel of the keel of the Villefranche-sur-mer wreck Hypothesis a. and b....................................... 80 Figure 93. Scarf of timbers of the keel of Nossa Senhora dos Mártires............................................................................. 80 Figure 94. The scarf recommended by J. B. Lavanha, fº43................................................................................................ 80 Figure 95. The keelson of the fore part of the wreck.......................................................................................................... 81 Figure 96. The foot wale S1 and the bilge clamp S2.......................................................................................................... 82 Figure 97. The clamps S4 and S5 (starboard side)............................................................................................................. 82 Figure 98. Longitudinal frame layout................................................................................................................................. 83 Figure 99. The sill: the inboard side next to the bilge clamp S2........................................................................................ 84 Figure 100. Carpenters’ mark............................................................................................................................................. 85 Figure 101. Oblique cutting line from the fore end of the strakes of planks...................................................................... 85 Figure 102. Nailing traces on the planking......................................................................................................................... 86 Figure 103. Right, Concretion formed by the tip of a folded plank nail Left, Impression of the nail left in the concretion............................................................................................................................................................................ 86

xv

The Mortella III Wreck Figure 104. Section of a hole left by nailing the planks (First futtock G 20)..................................................................... 87 Figure 105. G27B: Moulding of a plank nail (polyurethane mastic moulding) (1)............................................................ 89 Figure 106. G27B: Moulding of a plank nail (polyurethane mastic moulding) (2)............................................................ 89 Figure 107. AF15/2 – Diagram of the fore end planking slicing........................................................................................ 89 Figure 108. AF15/2 - Photomosaic of the port side fore planking panel............................................................................ 90 Figure 109. Hypothesis of fitting of the fore end planks into the stem rabbet.................................................................... 90 Figure 110. Fore end of the planking of a 16th century ship.............................................................................................. 91 Figure 111. Edge of a plank with remains of caulking products........................................................................................ 91 Figure 112. Outside face of a plank covered by sealing material....................................................................................... 91 Figure 113. Sketch of the attachment of the Red-Bay wreck planking to the frames........................................................ 92 Figure 114. Impression of the nail tips on the inner side of the frames of the wreck of Nossa Senhora dos Martires...... 93 Figure 115. Diagram of the main mast step........................................................................................................................ 95 Figure 116. Buttresses T1 to T3 arrange to wedge the port sister keelson......................................................................... 96 Figure 117. Buttress T2 fit into the groove of the port sister keelson................................................................................. 97 Figure 118. Sketch of the keys joining the two sister keelsons and grooves...................................................................... 97 Figure 119. The keys joining the two sister keelsons......................................................................................................... 98 Figure 120. Drawing of the main mast step of the Mortella III wreck............................................................................... 98 Figure 121. Diagram of the main mast step of the wreck of Villefranche-sur-Mer............................................................ 99 Figure 122. Diagram of the main mast step of the wreck of Red Bay................................................................................ 99 Figure 123. Diagram of the main mast step of the wreck of the Mary-Rose.................................................................... 100 Figure 124. Remains of the pump well where the debris of the upper deck lay............................................................... 101 Figure 125. Wooden pieces joining the end of the planks of the pump well.................................................................... 101 Figure 126. Foot pump between V21 et V22 (port).......................................................................................................... 102 Figure 127. Lower end of the pump tube in situ............................................................................................................... 103 Figure 128. Location of the pump..................................................................................................................................... 104 Figure 129. Drawing of the foot-valve............................................................................................................................. 105 Figure 130. Photography of the foot-valve....................................................................................................................... 106 Figure 131. Location of the foot valve upon the floor-timbers......................................................................................... 106 Figure 132. a. Foot valve of the Villefranche wreck b. Foot valve of the Red Bay wreck............................................... 106 Figure 133. Inner part of the pump................................................................................................................................... 107 Figure 135. Drawing of the pump tube............................................................................................................................. 107 Figure 134. External part of the pump.............................................................................................................................. 107 Figure 136. Excavation area AF 15/1 located at the fore end of the wreck...................................................................... 108 Figure 137. Rudder parts (BONNEFOUX, 1848)............................................................................................................ 108 Figure 138. Photomosaic of the remains of the rudder..................................................................................................... 109 Figure 139. Edge of the end of the rudder........................................................................................................................ 109 Figure 140. Organization of the remains of the rudder..................................................................................................... 110 Figure 141. Bottom view: A diagonal cut plank holds the two timbers together...............................................................111 xvi

List of Figures Figure 142. Outboard edge of the lower timber of the rudder (P3)...................................................................................111 Figure 143. Concretions located on the West edge of the rudder..................................................................................... 112 Figure 144. Concretion enclosing a rudder pintle (pintle a. of fig.140)............................................................................ 112 Figure 145. Broken concretion enclosing a second rudder pintle (pintle b.).................................................................... 113 Figure 146. 18th c. rudder pintle....................................................................................................................................... 113 Figure 147. Rudder of the wreck of Villefranche-sur-mer................................................................................................ 114 Figure 148. 1 (Left). One-piece construction of the Red Bay wreck 2 (Right) The rudder of the Mary Rose................. 115 Figure 149. Rudder pintle of Red Bay.............................................................................................................................. 115 Figure 150. Rudder pintle hook of the Mary Rose........................................................................................................... 116 Figure 151. Forward side face of the master-frame M27 after reassembling the parts on land....................................... 117 Figure 152. Drawing of the frame M27............................................................................................................................ 118 Figure 153. The ideal hull of a ship according to Mathew Baker..................................................................................... 118 Figure 154. Drawing of the floor-timber V27 and representation of its rising................................................................. 119 Figure 155. Comparison of the master floor-timbers of The Mortella III and Red-Bay wrecks...................................... 120 Figure 156. Morphology and rising of the floor-timber W54 of the Villefranche wreck................................................. 120 Figure 157. Morphology and rising of the floor-timber C22 of the Calvi 1 wreck.......................................................... 120 Figure 158. Profile of the master-frame according to the Ordinances of 1613................................................................. 121 Figure 159. Floor surface of the Mortella III wreck......................................................................................................... 121 Figure 160. Reconstitution of M27, essential cohesion and adjustment of the timbers................................................... 121 Figure 161. Survey of the master-frame on land.............................................................................................................. 122 Figure 162. Record of the master-frame on land.............................................................................................................. 123 Figure 163. Adaptation of the frame M27 underwater record to the shape of land record............................................... 123 Figure 164. Superposition of the underwater and land records of M27........................................................................... 124 Figure 165. Superposition of the profile of a nave de 700 botte, according to Z. Trombetta (1445) and that of Mortella III wreck............................................................................................................................................................. 124 Figure 166. Design of the shape of the master-frame of a merchant ‘nave’, according to the prescriptions of Pré Theodoro de Nicoló (1550)............................................................................................................................................... 125 Figure 167. Shape of the frames at midship of the Villefranche-sur-Mer wreck.............................................................. 126 Figure 168. Shape of the frame C20 of the Calvi I wreck at midship.............................................................................. 127 Figure 169. Shape of the master-frame and of the ‘almogamas’ or tails frame, according to Fernando Oliveira (fº114)................................................................................................................................................................. 127 Figure 170. Hypothesis of the shape of the master-frame of the Mortella III wreck designed with a single arc (hypothesis 1).................................................................................................................................................................... 128 Figure 171. Evolution of the shape of the frame W54 of the wreck of Villefranche-sur-Mer with two arcs................... 128 Figure 172. Hypothesis of the design of the master frame master-frame of the Mortella III wreck based on the model of the Villefranche wreck (hypothesis 2)............................................................................................................... 129 Figure 173. Height of the decks of the Villefranche-sur-Mer wreck................................................................................ 130 Figure 174. The position of the bridges of the 500 toneladas ship according to Manoel Fernandes............................... 130 Figure 175. Hypothesis of the height of the Mortella III wreck (h1)............................................................................... 131 Figure 176. Hypothesis of the height of the Mortella III wreck (h2)............................................................................... 131 xvii

The Mortella III Wreck Figure 177. 3D image of the upper part of the keel heel.................................................................................................. 133 Figure 178. The rake of stern posts of the wrecks of Mortella III, Calvi 1 and Red-Bay................................................ 134 Figure 179. Hypothesis of the rake of the stern of the Mortella III wreck....................................................................... 134 Figure 180. Rake and height of the stem post of Z. Trombetta nave................................................................................ 134 Figure 181. Fore end of the strakes of the planking......................................................................................................... 135 Figure 182. The stem post of Zorzi Trombetta’s 700-botte nave...................................................................................... 136 Figure 183. The stern and stem posts of the Mortella III wreck, according of the method of F. Oliveira....................... 137 Figure 184. Longitudinal shape of the Mortella III wreck and main measurements........................................................ 140 Figure 185. Captain Paulin de la Garde............................................................................................................................ 144 Figure 186. Nave during the expedition led by Charles V against Algiers (1541)........................................................... 150 Annexes Figure A1.1. General layout of the site and excavation areas........................................................................................... 173 Figure A1.2. General planimetry of the site...................................................................................................................... 174 Figure A3.1. Liste des échantillons prélevés sur l’épave de la Mortella 3....................................................................... 177 Figure A3.2. Courbes moyennes en position relative....................................................................................................... 178 Figure A3.3. Courbe moyenne du l’épave de la Mortella 3.............................................................................................. 178 Figure A3.4. Coefficients de corrélation des échantillons avec la courbe moyenne de la Mortella 3.............................. 179 Figure A3.5. blocs-diagramme des différents éléments d’architecture en datation absolue............................................. 179 Figure A3.6. Datation absolue des échantillons prélevés sur la Mortella 3...................................................................... 179 Figure A5.1. Chromatogrammes obtenus par CPG-SM................................................................................................... 199 Figure A5.2. Agrandissement des chromatogrammes entre 22,5 et 33 min..................................................................... 199 Figure A5.3. Schéma de l’oxydation et de la dégradation thermique conduisant à la formation des composés caractéristiques de la poix de pin (Colombini, et al., 2009). ........................................................................................... 201

xviii

List of Tables Table 1. ‘As-Dos-Tres’ proportions rule expressed by different authors between the 16th and 17th centuries.................... 10 Table 2. ‘Fabrica di galere’, beginning of the 16th c.Biblioteca Nazionale Centrale di Firenze, codex Magliabecchiano, XIX.7..................................................................................................................................................... 11 Table 3. Zorzi Trombetta ‘da Modon’ c. 1444, ‘Libro’, British Library, Cotton MS Titus A XXVI , fº12a to 16a and 37a to 60b..................................................................................................................................................................... 11 Table 4. Pre Theodoro de Nicoló, c.1550, ‘Instructione sul modo di fabricare galere’, Biblioteca Nazionale Marciana di Venezia, manoscritti italiani, cl. IV cod. XXVI (5131).................................................................................. 12 Table 5. Species used in some wrecks of Mediterranean building tradition....................................................................... 30 Table 6. Mortella III – Room and space of the framing (tumulus A -Starboard)................................................................ 48 Table 7. Comparison of the moulding sizes of the frame-timbers between Mortella III and other wrecks....................... 57 Table 8. Overlaps of the timber framing components......................................................................................................... 60 Table 9. Mortella III – Main measurements of the frames (tumulus A - Starboard)........................................................... 69 Table 10. Average of the main transverse framework measurements................................................................................. 70 Table 11. Summary of the dimensions of the longitudinal framework............................................................................... 92 Table 12. Measurement of the depth of hold in various authors of the 16th and early 17th centuries................................ 132 Table 13. Evaluation of the proportions of the Mortella III ship in comparison with those of the Villefranche-sur-Mer and Calvi I. .................................................................................................................................... 139 Table 14. Comparison of technical ‘fingerprints’ on some Mediterranean wrecks of Mediterranean building tradition............................................................................................................................................................... 155 Annexes Tableau A5.1. Description des échantillons étudiés.......................................................................................................... 197 Tableau A5.2. Gradient de température pour l’analyse en CPG/SM................................................................................ 198 Tableau A5.3. Terpènes identifiés par CPG-SM............................................................................................................... 200 Tableau A5.4. Acides gras identifiés par CPG-SM........................................................................................................... 200 Tableau A5.5. Résumé des résultats obtenus par CPG-SM.............................................................................................. 202 Tableau A8.1. Dimensions et mesures.............................................................................................................................. 211 Tableau A8.2. Angles et rapports...................................................................................................................................... 212 Tableau A8.3. Comparaison entre les ancres Gn-A1 et M III-AW................................................................................... 212

xix

Lists of Abbreviations AGS

Archivo General de Simancas (Spain)

ASG

Archivio di Stato di Genova (Italy)

AGI

Archivo General de Indias (Spain)

ASV

Archivio di Stato di Venezia (Italy)

BNF

Bibliothèque Nationale de France (France)

MNM

Archivo del Museo Nacional de Madrid (Spain)

SHM

Service Historique de la Marine (France)

xxi

Introduction In the 16th century, Mediterranean shipbuilding was renowned throughout Europe. Master carpenters, particularly those from the Italian States, were sought after for the quality of their work and their high degree of technical knowledge. This technical culture inherited from centuries-old tradition passed down orally from generation to generation. But today it is largely unknown because in addition to the scarcity of written documentation, there is also a lack of archaeological documentation.1 This observation has led Renaissanceperiod Mediterranean shipbuilding to become a research priority over the last thirty years.

a study that has been punctuated by doubts, questions, sometimes false leads, but also discoveries and answers to essential questions. It gives a perspective and maturity to the research work that justifies the writing of this book even if the excavation of the wreck of Mortella III is not yet fully completed. During the initial five years of excavations, the work objectives were defined in such a way as to try to provide answers to the major problems raised by the archaeology of Renaissance ships in the Mediterranean maritime space. The issues addressed and the objectives of the study

The Mortella wrecks: a valuable source of information

Several issues related to shipbuilding processes and naval architecture are addressed in this book but it should be pointed out at the outset that all are linked to underlying spatial questions. From the end of the 1980s, influenced by Th. J. Oertling (OERTLING 1989 and 2001), archaeologists working on the Renaissance period, Brad Loewen among others (LOEWEN, 2001), sketched the features of the ‘Atlantic’ tradition. This was identified using common ‘traits’, in contrast to a ‘Mediterranean’ tradition. The work, carried out on the wreck of Villefranche-sur-Mer dated 1516 (GUEROUT, RIETH and GASSEND, 1989) has helped to identify certain specific ‘Mediterranean’ features that have been described as ‘architectural signatures’ (RIETH, 1998). As a result, the constructive and architectural study of the Mortella III wreck seeks both to situate it within the ‘Mediterranean nautical space’ to which it belongs, and also to define, as far as possible, the limits of this large and still somewhat loose concept, whose contours should be tightened.

In this context, the discovery of the Mortella wrecks (Bay of Saint-Florent, Upper Corsica) in 2005 and 2006 provides a valuable source of information that may significantly enrich this documentation on 16th century Mediterranean naval architecture.2 This aim led the Centre d’Etudes en Archéologie Nautique (CEAN) to undertake an excavation programme on one of them—the Mortella III—between 2010 and 2015, which mainly focused on the study of the particularly well-preserved remains of the hull. This project, financially supported by the French Ministry of Culture and its Département des Recherches Archéolog­ iques Subaquatiques et Sous-Marines (DRASSM) as well as by the Territorial Communities of Corsica (CTC), has been the subject of a scientific collaboration agreement with the University of Paris-Sorbonne and the University of Corsica Pasquale Paoli. It has been recently awarded by the European Marie-Curie Programme (MSCA, IF No. 843337) and supported by the Spanish National Research Council, CSIC.

Even if it is still too early to be able to link the construction of the Mortella III wreck to a sub-area of the Mediterranean maritime space, it shall initially be addressed from a spatial perspective. The first step is to identify and list the processes used, specifically the assembly of timbers, fastening methods and all the technical solutions applied to the building of the framework. Their inclusion in a tradition of Mediterranean-style building, primarily aims to confirm the ‘technical markers’, called ‘technical fingerprints’3 in our text, that have already been characterized, but also to identify new ones.

More than ten years have passed since the site was discovered. This allows us to retrace the long journey of 1 The only 16th century wrecks of ‘Mediterranean construction’ whose architectural remains have been studied to date are: the wreck of Villefranche-sur-Mer (GUEROUT, RIETH, GASSEND, 1989) and the wreck of Calvi I in Haute-Corse (VILLIE, 1989, 1990 and 1991). The Ottoman wreck of Yassi Ada also contributes to the Eastern side of the Mediterranean (PULAK, 2005 and LABBE, 2010). Two other wrecks with high architectural potential have been the subject of excavation programmes: Gnalic (Croatia) in 2004 and since 2013 and Delta II (Cadiz, Spain) in 2012, but, to our knowledge, their architectural study has not been yet published. Two other 16th century Venetian wrecks found in Croatia, Sveti Pavao (Mljet Island) excavated between 2007 and 2016 (BELTRAME, GELISHI, MIHOLJEK, 2015) and the Brsecine wreck also seem likely to reveal architectural information. 2 The sites of the Mortella owe their name to the Genoese tower of the same name located at the western end of the Bay of Saint-Florent, which is the closest geographical reference point. The 16th century wrecks discovered in 2005 and 2006 were named Mortella II and III. The site of Mortella I is a mooring place located at the foot of the tower where ceramic shards attest to its activity over the past centuries.

Two other issues addressed in this book concern the ‘architectural project’, in other words, they are related to the design of the hull. Their main objective is to contribute to the knowledge of hull design in the 16th century, which is essential to understanding the transition that took place 3 We borrowed the word ‘fingerprint’ from O. Crumlin-Pedersen (CRUMLIN-PEDERSEN, 1991).

1

The Mortella III Wreck the tail frames was widely used. Archaeologically, it has been documented in the Mediterranean with the wreck of Cala Culip VI, which gave the oldest evidence of it (RIETH, 1998), or with that of Aveiro A, for example, for the Atlantic area (ALVES et al., 2001). The studies carried out in this field and over our period suggest moulded frames were used in the construction of the structure of the Mortella III wreck. Some observations made during the excavation have already led to remarks to this effect, particularly with regard to the nailing of the frame’s timbers. But a complete dismantling of an important part of the structure would have been necessary to find surviving traces of this moulding method, as far as it could be detectable. Unfortunately, this task was not possible in the programme of our excavation.

during this period between architectural design inherited from rules of the Middle Ages, and the definition of new criteria characterizing the ship of the modern era. The work presented here addresses three major issues related to ship design. They are studied in relation to spatial dynamics and, in this context, we have sought to highlight ‘architectural markers’—called ‘architectural traits’— that could constitute Mediterranean specificities.4 The first issue concerns the proportions of the ship. The written sources depict the 16th century as a significant period of transition in the field of naval architecture, with the evolution of the old ratio rule ‘As-dos-tres5.’ During the Renaissance this became generalised and produced very ‘round’ seafaring ships, built using more efficient proportions intended for Iberian navigation dated from 1580-1590 onwards.6 This question of ships’ proportions is an essential field of study in naval architecture because it determined, to a large extent, their shape and nautical specificities. However, archaeology has not yet given much attention to this theme. In order to situate the wreck of the Mortella III in the context of contemporary technological developments, it seems important to us, at the end of this study, to set out as precisely as possible the proportions of the ship and to draw conclusions about its nautical characteristics.

We address the issues that have been mentioned, relying on technical data and considerations, but also on documentary and historical references: This work seeks above all to contribute to the definition of the contours of technical culture and know-how highlighted by the excavation. The results of this study should therefore shed light on the constructive and architectural characteristics of a large 16th-century ship in the Mediterranean during a particularly important period of transition. The architecture of the Mortella III ship carries both the rules and architectural characteristics inherited from the Middle Ages as they are currently highlighted by archaeological research (see, among others, RIETH, 2016), and potentially the technical innovations that prefigured the modern era.

The second issue is related to her shape, and in particular that defined by her master-frame, which French manufacturers referred to as a ‘figure’. This is the founding element of the ship’s geometry, as is the case with all those—like the Mortella III—designed according to a transversal principle specific to the ‘frame-first’ construction. The study of the shape of the master-frame is carried out with particular attention to its design method and to the shape of its floortimbers.

Over time and during excavation campaigns, a delineation of this very general objective has gradually taken place and it is ultimately the profile of a Genoese commercial ship from the first third of the 16th century that has emerged, even if some of the conclusions that lead to this portrait still need to be refined. Finally, this study aims to place the ship in an Italian typology during a period that is now well circumscribed, and attempts to recover its essential nautical characteristics.

Finally, included in the study of the ship’s design, the excavation of the remains of the hull of the wreck of the Mortella III has been carefully observed in order to identify traces of the design method used to build the structure. During the Renaissance period, the Mediterranean method of predesigning the geometry of the frames located between

The limits of the archaeological study Once the outlines of the ambition of this work have been defined, it is immediately necessary to specify its limits. It must be said there were unfortunately many, which often acted as a brake, and sometimes a source of frustration, in the desire to go deeper into the study of the remains of the Mortella III wreck, and to draw further conclusions using archaeology. Here are the two main ones we faced:

4 We use the word ‘architecture’ in its original meaning, which links it to a notion of design to clearly distinguish it from construction techniques: ‘Ship architecture is the art of designing ships…’, Michel Benicourt, definition given in the article in the Encyclopedia Universalis. 5 The so-called ‘As, dos, tres’ is a rule of proportion whose origin dates back to the Mediterranean Middle Ages. It establishes a principle of proportionality between the beam (As), the keel length (dos) and the total length of the boat (tres). It was summarized in 1611 by the Spanish shipbuilder Thomé Cano (CANO, 1611) as follows: ‘…all the Spanish, Italian and other masters of shipbuilding have the practice of giving to one ‘codo’ at the beam, two to the keel length; and to another ‘codo’ at the beam, three to the overall length, and to three ‘codos’ at the beam, one to the flat wide; and for the depth of hold, three quarters of the beam.’ The rule of ‘As, dos, Tres’ is detailed in Chapter I of this book and the question of the proportions of the Mortella III wreck is addressed in Chapter V. 6 This movement, which was born in Spain under the impetus of manufacturers such as Juan de Veas or Rodrigo Ramirez, is known as ‘Nueva Fabrica’. Of Atlantic constructive origin, it leads to proportions that will be institutionalized in the Spanish Ordinances of 1607, 1613 and 1618 and which generate ships with more elongated shapes with a width/ length ratio between 3 and 4.

• First, the technical (and related financial) difficulties that have arisen in order to methodically excavate a wreck the size of the Mortella III in nearly 40 meters of water are enormous. The constraints posed by decompression and safety issues, the very short working times imposed by the hyperbaric environment (maximum 50 minutes per person per day) constituted a major limitation to the archaeological excavation. This prohibited, for 2

Introduction briefly reviewed. These are essential to pinpoint the chronology of the ship and without which any attempt at a constructive and architectural study would be in vain. The section on chronology also presents the 2013 dendrochronological study which achieved a precise dating of the ship.

example, the dismantling of the structure—except in very specific cases—and observations as exhaustive as those on sites located in shallow waters. Thus, the architectural study is based on partial observations that cannot be compared with sites where all the structural parts have been brought to the surface and recorded ashore. • A second major limitation to archaeological analysis is the product of the fire that destroyed the ship before it sank and the combustion of the deadworks. In underwater archaeology, wrecks often lie on the seabed leaning on one of their sides. If the conservation conditions are good, it enables the study of a high percentage of the hull. This was the case for significant 16th century wrecks such as that of Villefranche-sur-Mer (France, 1516) for the Mediterranean or that of Red Bay (Canada, 1565) or also that of Mary-Rose (United Kingdom, 1545) for the Atlantic. In all three cases, the study of the remains made it possible to reach at least to the third futtock of the frames. This is not the case for the Mortella III wreck. Although the anaerobic conditions of the sediment allowed it to be well preserved, the ‘flat’ position of the wreck on the seabed and, above all, the combustion of the dead-works prevented our access to the remains of the frames beyond the first third of the second futtock. This limits our confidence in hypothetical reconstruction of the ship’s hull form, which instead must be estimated according to several hypotheses above the waterline.

Finally, the general description of the layout of the Mortella III site allows us to reflect on the elements that contributed to the shipwreck, and ends with a presentation of the research programme with its various lines of research and the methodology of the excavation work. The study of the hull structure—which remains at the core of this book—is approached through consideration of construction methods and techniques, on the one hand, and the architecture of shapes and proportions, on the other. These two approaches are discussed in separate chapters. From a methodological point of view, it seemed important to separate everything related to construction and what F. Braudel defines as ‘this long series of human gestures that are at the root of what is called technical culture’ (BRAUDEL, 1979), from what is part of the ship’s architecture itself. In other words to make a clear distinction between what falls within the scope of the building technics (scarfs, fastening, caulking and all the technical solutions to the problems posed by the construction of a wooden structure intended for navigation in general) from its architectural and design concepts (geometry of its shapes and proportions) that set out the nautical characteristics of the ship. The notions of architecture and construction techniques are in constant interaction and, in fact, their boundaries are sometimes tenuous. But if we take them in their original sense as mentioned in Note 4, they touch on two conceptually different themes: design and manufacture.

Once these caveats have been made, the work in the next few pages are organized as follows: Before getting to the heart of the archaeological study of the Mortella III wreck, it seemed desirable that the first two chapters of this book be devoted to:

Construction techniques

1. A more detailed and in-depth presentation of the problems generated by the remains of the hull that were briefly mentioned above. The aim is to provide as complete an overview as possible of the various questions currently raised by the excavation of a ship from this period and to consider the answers— or elements of answers—that the wreck of Mortella III is likely to provide. These questions serve as an opportunity to survey the field of Mediterranean shipbuilding and architecture in the 16th century. 2. The general presentation of the site to help the reader to visualise it and fully understand its organization, chronology and dynamics. This general overview references the site of Mortella II, which we now know with certainty is linked to the same historical event as that of Mortella III.7 Although not very abundant and somewhat tangential to the present discussion, the artefacts of the Mortella III site are also

Construction techniques are covered in Chapters III and IV. Chapter III deals exclusively with hull structures. A study of transverse and longitudinal structures is carried out. In particular, the following are examined: • The types of connections used to join the timbers, the types of scarfs used in the extension of longitudinal and transversal structures; the way in which the pieces are joined, etc. • The systems for fastening the timbers, and in particular of the planking to the frames, alongside the nailing and pegging methods. • The caulking and sealing products used. We have also opted to address the ‘attributes of the hull’, in a separate chapter, Chapter IV. These are three architectural ensembles which are not formally part of the hull stricto sensu, but which are physically associated with it and which are closely linked to it considering their function:

Apart from their geographical proximity and similarity of artefacts, proof of the link between the two wrecks was provided by the lithic study of their ballast, which revealed that it shared an identical geological make up (see Annex VI)

7

3

The Mortella III Wreck study carried out by Brad Loewen, is very useful to characterize the construction methods and Mediterraneanstyle architectural design of the Mortella III wreck.

1. The mast-step which is, as we will see, a remarkable structure in many respects.8 This is a work of art related to the movement of the hull. 2. The water removal system. It is represented by the remains of the ship’s pump. We are dealing here with a device related to the hull’s flotation. 3. The rudder. This steering system is essential; its function is to control the direction of the hull.

Other wrecks such as the Calvi I, dated from the end of the 16th century, and the Cala Culip VI (RIETH, 1998, PALOU et al., 1998, PUJOL I HAMELINK et al., 1994) of an older period (early 14th century) are used as references for ‘Mediterranean construction’. For ‘Atlantic shipbuilding’, there are more wrecks and the Mary Rose (1545) (MARSDEN, 2009), is of course used as a reference. In addition to the Red-Bay wreck mentioned above, the milestones offered by the Ibero-Atlantic wrecks selected by Oertling are consulted , in particular those whose chronology is close to Mortella III, such as the Cattewater wreck, dated early in the 16th century (REDKNAP, 1984). Nevertheless, wrecks of older or more recent chronology, such as those of Nossa Senhora dos Martires dated 1606 (CASTRO, 2005) or Aveiro A (ALVES et al., 2001), dated from the 15th century, which have been meticulously studied, also provide very useful reference points and comparisons for the archaeological analysis of the remains of the wreck of Mortella III.

The architecture of the ship Next, the architectural question is addressed. Its aim is to return to the architectural project, using the study of the shapes and proportions of the ship to highlight what could constitute its Mediterranean characteristics, thereby moving beyond construction techniques. At this level, attention is focused on: • Questions of proportions: trying to approach the point of view of the builders of the time, particularly studying the relationship between the maximum breadth, the length of the keel and the total length of the ship. • The rising of the floors and the form of the hull. This will focus on the shape given to the master-frame and the way in which it is designed. This one is decisive in the general shaping of the hull and is at the heart of current issues in naval architecture.

Documentary sources Beyond the shipbuilding study, research into the literature has been based on written sources, construction treaties of the period and, when it has been possible, on iconography.

Archaeological sources To achieve the mentioned research objectives, the archaeological data available for the Mediterranean— beyond those provided by the Mortella III site—have been considered, in particular that of the wreck of the presumed Lomellina (1516). This was excavated in the bay of Villefranche-sur-Mer by the GRAN (‘Groupe de Recherche en Archéologie Navale’) between 1979 and 1988 under the direction of Max Guérout (GUEROUT, RIETH and GASSEND, 1989). It is one of the few Mediterranean wrecks of our period that has been the subject of an architectural study. Alongside a similar chronology, there are many commonalities between this wreck and the Mortella III ship. These make the Villefranche wreck an essential reference point to base a comparative account that allows many parallels to be drawn and considerably enriches archaeological reflection.

As archaeological and historical studies suggest the wreck of the Mortella III to be of Italian origin, it seems appropriate to give priority to its architectural analysis in the light of Italian written sources. Unfortunately, these are not very common for the 16th century. An important resource for this century, however, is the ‘Instructione sul modo di fabricare galere’ by the Venetian Pre Theodoro de Nicolò published in 1550 (PRE THEODORO, 1550). It is also worth mentioning the discovery, about ten years ago, of the manuscript of the treatise of the Ragusan (Dubrovnik) Nicoló Sagri, Il Carteggiatore (SAGRI, 2010), dated from the second half of the 16th century, found in an American library, which sheds new light on shipbuilding in the Western Mediterranean but also on navigation and life on board.9 Apart from the Sagri Treatise, Mediterranean references must be drawn essentially from Venetian sources of the 15th century. There are four main texts, which are primarily

In the same comparative perspective, this study is also based on documentation from the excavation of the wreck of the presumed Basque whaler San Juan (1565) carried out in Red Bay (Labrador, Canada) by the Parks Canada team between 1978 and 1985 (BERNIER, GRENIER et al., 2007). The use of a large volume of archaeological data recorded on this ‘Ibero-Atlantic’ technical tradition wreck, which has been given a prominent place in the architectural

Nicoló Sagri (1538-1571) was an officer of the Ragusan Navy (now Dubrovnik) in charge of training the personnel on board, a role that motivated the drafting of his treaty. This is a 105-folio document recently found in the United States in the James Ford Bell Library at the University of Minneapolis. Although it was lost, its existence was known because of its evocation by Bartolomeo Crescentio in his treatise Nautica Mediterranea (CRESCENTIO, 1607). In fact, this text, which is of great value for the knowledge of Mediterranean shipping and shipbuilding in the 16th century, had been sought in vain in Italy for years by Augustin Jal (JAL, 1840, Vol. I, 25). The Treaty of Sagri—written in Italian— provides an overview of Italian-influenced shipbuilding, with Ragusan shipbuilding being considered as a regional variant in the same way as Venetian, Genoese or Neapolitan shipbuilding.

9

8 The mast-step of the main mast is a good example of a device that could be analyzed from a technical and constructive point of view as a technical system participating in the ship architecture. This is the first meaning we have given priority to, considering that it does not directly influence its nautical characteristics.

4

Introduction concerned with ships of the galley family. The first two are dated from the first third of the 15th century. First, the ‘Fabrica di galere’ (ANONYMOUS, 1410) which is mostly a copy of Michele da Rodi’s ‘Libro’ (MICHELE DA RODI, early 15th century.) Then the third text, dated from the middle of the 15th century, is the ‘Libro’ by Zorzi Trombetta da Modon (ZORZI TROMBETTA, 1445). Finally, the ‘Ragioni antique spettanti all’arte del mare et fabriche de vasselli’ (ANONYMOUS, 15th century) written towards the end of the 15th century. These texts have sometimes been likened to inventories of ‘technical recipes’ for shipbuilding. But in fact, there is very little mention of construction techniques in these texts. While it is true that they follow the same model, setting out long lists of austere measures, their merit is to draw a portrait of ships of their time through relating all these measures. This information, which is sometimes very detailed, provides useful information on the dimensions, proportions and shapes of the ships and, in short, provides a profound insight into Italian, or more precisely Venetian, naval architecture.

of 1607, 1613 and 1618 are of great importance in understanding 16th century shipbuilding. We may be criticized for the consistency of our references to these texts, which chronologically date back to the beginning of the 17th century. It should be noted, however, that the latter codified the Laws of Spanish shipbuilding with an unprecedented level of precision and constructive detail. Their importance in the study of 16th century shipbuilding is immense because they constitute a remarkable synthesis of its evolution and the choices made by the promoters of the ‘nueva fabrica’ movement, which originated in the second half of the 16th century. Their richness lies in the fact that they enable the identification of the evolution of construction, but also the 16th century design principles caught between the Middle Ages and carrying the seeds of modernity of which ‘la nueva fabrica’ is an indisputable vector. On the Portuguese side we will mention three scholarly texts, now well known: Father Fernando Oliveira’s ‘Livro da fabricas das naos’ (OLIVEIRA, 1570), João Baptista Lavanha’s ‘Livro primeiro da arquitectura naval’ (LAVANHA, 1610) and finally, later, Manoel Fernandes’ ‘Livro de traças de carpintería’ (FERNANDES, 1616).

Finally, in terms of Italian sources, mention should be made of two later texts that clearly set out the method of predesigning the frames. These are: The ‘Visione’ by Baldissera Quinto Drachio (DRACHIO, 1594) written in 1594 which describes the method of designing a 43-metre galley and Bartolome Crescentio’s later treatise of 1607, (CRESCENTIO, 1607), which describes the features of a galley. In fact, all the texts we have quoted are less concerned with describing the characteristics of sailing ships than with those of rowing ships, such as galleys, a likely reflection of their primacy in the 16th century in the Mediterranean, at least in Venice.

All these Portuguese and Spanish texts form an Iberian corpus of early modern shipbuilding that contributes to the definition of an ‘Ibero-Atlantic’ type of ship. As for the Venetian texts, albeit through a somewhat simplistic regional prism, they nevertheless make it possible to sketch out a ‘Mediterranean’ vision. The architectural study of the remains of the hull of Mortella III wreck cannot be carried out without constant reference to these documents. The sources of the historical perspective

But our understanding of 16th century naval architecture would be fragmentary and incomplete solely using Venetian texts. When studying the wreck of the Mortella III, even if we focus upon the Mediterranean, the Iberian texts of the 16th and early 17th centuries, which refer to an ‘Atlantic design’ of shipbuilding, remain essential to understanding and interpreting its architectural remains.10 There are about ten of them, among which we would like to mention three Spanish treaties, three Spanish Ordinances and three Portuguese treaties:

Finally, this work ends with an historical contextualisation that remains at the core of archaeological study. This is the purpose of Chapter VI, which deals with historical issues, first with the aim of attempting to identify the ship of Mortella III, but also with a view to contextualising this ship within its historical period. The basis and nature of this work remains an archaeological work and, from this point of view, a resolutely technical study. It is nevertheless true that the intimacy of the links between archaeology and history would not allow a complete, intelligible and satisfactory analysis to be carried out without incursions into the field of historical sciences. The historical analysis is limited to an approach that pursues a double objective.

For Spain, Thomé Cano’s treatise, ‘Arte para fabricar…. las naos’ (CANO, 1611) and Juan Escalante de Mendoza’s, ‘Itinerario…’ (MENDOZA, 1575) are significant. These shipping and shipbuilding treaties are part of the ‘la nueva fabrica’ movement mentioned above. It is also worth mentioning Diego Garcia Palacio whose treatise, although published after Escalante’s, reflects an older and less innovative conception of Spanish shipbuilding (GARCIA PALACIOS, 1587). Moreover, the Ordinances

The first concerns the study of the various historical episodes likely to identify the wreck and its origin. This work is based on the results of literature research carried out in Italian, French and Spanish archives and libraries since 2005. As pointed out at the end of this study, there is a body of evidence suggesting that the wrecks of the Mortella are of Genoese origin and the documents located today enable us to link them to an episode from 1527 which

A particularly comprehensive technical study of these texts and Iberian shipbuilding was recently published by Cayetano Hormaechea (HORMAECHEA, 2018). On a more historical side, a collective publication led by David González Cruz (GONZÁLEZ CRUZ et al., 2018) also provides an in-depth insight into the topic.

10

5

The Mortella III Wreck

Figure 1. Geographical location of the Mortella wrecks (Illustration: Arnaud Cazenave de la Roche).

remains the main historical hypothesis favoured to date. These conclusions are essential to place the architectural remains in their geographical and chronological context, a crucial step for the construction methods and architecture of the ship to take on their full meaning. Once this chrono-geographical framework has been defined, the second objective of the historical approach is to place the shipwrecked vessel in the geopolitical and economic context of its time. As we will see, the latter can be helpful in understanding certain options chosen by the builder, both in terms of the constructive characteristics of the ship and its shapes and proportions.

6

1 Shipbuilding in the Mediterranean in the 16th century: State of the art and issues given to the concept of ‘maritime space’ the meaning of a ‘technical space’. But as Westerdhal’s definition emphasizes, this space is also conceived as a cultural space. This is where the notion of ‘technical culture’ becomes relevant. This makes it possible to create a link between various peoples who, although they may be culturally distant in political and social terms, share the same technical knowledge that brings them together and creates this ‘cultural’ link between them.

A key goal of the excavation of the Mortella III wreck is to contribute to the archaeological documentation of the shipbuilding of the early modern period, and to address related issues. Therefore, a review of the state of the art and the questions facing current research in the field of Mediterranean shipbuilding techniques and architecture during this period seems necessary in order to further the brief presentation made in the introduction. 1.1. The concepts of ‘maritime space’ and ‘technical culture’ at the heart of current issues

In the light of the above, the framing of the Mortella III wreck was studied with a view to placing it in a tradition linked to one of the ‘maritime spaces’ as previously identified, in this case the ‘Mediterranean maritime space’. It should be added that at the current stage of archaeological research, while a concept of a ‘Mediterranean maritime space’ has gradually emerged in recent years, it is still far from being representative of the diversity and complexity of shipbuilding in the Mediterranean in modern times. The work undertaken so far mainly concerns Italian shipbuilding, which dominates the western Mediterranean. But much still remains to be done to highlight the diversity of the Mediterranean space and in particular the study of shipbuilding on its eastern slope, especially that of the Muslim world of which the Ottoman wreck of Yassi Ada gives a brief overview.

As mentioned in the introduction, spatial dynamics are currently a major line of questioning in the nautical archaeology of the Renaissance. This is essential for our work and underlies all the themes related to the building and architecture of the Mortella III ship discussed in the following pages. The concept of ‘maritime space’ is relatively complex because it covers different meanings depending on whether one refers to its popular meaning, its legal meaning (which is used in maritime law), or its anthropological meaning, whether archaeological or ethnological. Ethnologists such as Olof Hasslöf have been working on the concept of ‘maritime space’ since the 1950s (HASSLÖF, 1958). This has been linked to the concept of ‘culture’. Indeed, C. Westerdhal points out that between 1975-80 in Sweden, the Norrland National Archaeological Survey Programme was carried out using the terminology of ‘Maritime Cultural Landscape’ (WESTERDHAL, 1992) whose definition is as follows: ‘human use (economy) of space through navigation: settlements, fishing, hunting, trade and the underlying cultures that result from it’. Therefore, under the prism of ethnology, the notion of ‘maritime space’ covers the meaning of a territory where economic and social activities are shared.

Therefore, the technical context of shipbuilding will be addressed in connection with a spatial dimension. This will be studied during its two main phases: • The first is conceptual and related to the architectural project. During its definition, two fundamental characteristics of the ship will be set out. First, its shape determined from the master-frame which remains the founding element of the hull design. The second aspect, in relation to the previous one, will set out the definition of the dimensions and proportions of the ship. In modern times, in line with a ‘transversal design’ specific to ‘frame-first’ style, the breadth of the ship is often the standard on which one relies to determine its size. The other dimensions, keel length, overall length and depth of hold will follow according to precise rules of proportion. • The second phase is related to the material execution of the architectural project. Here, two different aspects can be distinguished. The first concerns the method used to implement this project. The construction principle of the Mortella III ship still induces a design system that – to a large extent – involves the moulding of the frames located between the tail frames. Although this stage belongs to the construction, it is closely linked

In the field of naval architecture, the concept of ‘maritime space’ took on a specific meaning from the late 1980s when Th. Oertling paved the way for an ‘Atlantic’—and more precisely ‘Ibero-Atlantic’—shipbuilding concept in a founding article (OERTLING, 1989) which produced a definition of ‘traits’1 common to these ships. A total of twelve traits have been identified using seven sixteenth century shipwrecks whose origin is probably Iberian. In other words, it is in light of these common construction characteristics that an ‘Atlantic maritime space’ is envisaged in opposition to the ‘Mediterranean maritime space’. As such, we can see that naval architecture has 1

This the word used by the author.

7

The Mortella III Wreck a role in the ship’s build and it is ultimately in examining all these criteria that it is possible to identify the technical origin of the construction.

to the planning phase insofar as it is the mean used to achieve the geometric characteristics and desired shape of the hull. Finally, the second aspect concerns the construction processes themselves. In short, all the carpentry techniques used to build the structure. Unlike the design method, these have little or no impact on the shape and dimensions of the ship.

Finally, we would like to stress that a constructive tradition is not limited to the specificity of its construction processes. The shape of the ship and its proportions also contribute to defining a ship profile to a specific ‘nautical space’. To give an example, the geometry of the shapes of a Genoese trade nave and the relationship between the measures from which its proportions are derived will be different from those of a Basque trade ship of the same period. We have sought to identify these characteristics, as ‘architectural traits.’ In this case they are derived from the design of the Mortella III ship in order to differentiate them from ‘technical fingerprints’ that refer to the construction process, which we have sought to link – if possible – to the ‘maritime space’ to which they belong.

Our analysis was organised with regard to these phases in the construction of the ship, accompanied by a constant concern to highlight the specificities that could link them to a ‘technical culture’ within a ‘maritime space’, or in other words to a Mediterranean ‘technical space’ and allow to contribute to the definition of a model. This reflects the comparative approach yet does not detract from the specific and unique characteristics of the Mortella III ship, which is the result of its shipyard and builder. 1.2. Construction techniques: in search of ‘technical fingerprints’

1.3. The components of the architectural project: in search of an architectural model

The identification, study and inventory of the processes used for the construction of the hull and their connection to a technical tradition are therefore a first major issue. The corpus of construction processes that are linked to a Mediterranean tradition has been built up since the 1990s in relation to and in opposition to the ‘traits’ of ‘Atlantic construction’ identified by T. J. Oertling in 1988. E. Rieth compiled an inventory of these in the late 1990s (RIETH, 1998, 177-187). Essentially, four ‘technical fingerprints’ have been identified as indicators of a Mediterranean tradition:

1.3.1. Proportions: the relationship between dimensions Measurements and proportions—in other words, the relationship between these measurements—of a ship is a fundamental question of naval architecture, determining its shape and nautical proprieties. This question, which is at the root of the shipwright’s construction project and appears essential to the understanding of the shape of the ship and its characteristics, has nonetheless received little academic attention. A theoretical approach is necessary to understand the parameters the builder relied upon to give appropriate proportions to a ship. First of all, it involves exploring the proportion rules of the 16th century in the Mediterranean. Without such an approach, we remain disconnected from the reality of the technical knowledge and thought of their time, and express orders of magnitude linked to our anachronistic contemporary criteria.

1. Floor-timbers to first futtock scarf using ‘hook scarf’ between the tail frames as opposed to Oertling’s ‘trait’ nº1 which indicates the use of ‘dovetail scarfs’ in the ‘Atlantic construction’. 2. Fastening of the scarfs with metal nails as opposed to the mixed use of tree nails associated with metal nails (indicated in ‘line’ No. 1). 3. Fastening of the planking to the frames with metal nails as opposed to the mixed use of tree nails associated with metal nails (‘trait’ nº2). 4. Mainmast step arranged by two longitudinal timbers (or sister-keelsons) laterally reinforced by buttresses, as opposed to the shaping of mast step with a mortise carved an extended central area of the keelson (‘trait’ nº7).

Once this has been done, it is necessary to rely pragmatically on observations and measurements taken in the field to try to restore the proportions of the ship in accordance with the criteria of its time. This approach is undertaken in Chapter V. The measures that a shipbuilder gives to his ship in the 16th century traditionally involve five main dimensions linked together by a relationship of proportionality. These are the following:

Whilst this list of ‘technical fingerprints’ is not exhaustive, it is a base on which our observations relied to recognize the origin of the construction of the Mortella III wreck, while seeking to identify new ones. It should be added that these fingerprints remain indicators and that under no circumstances are they systematic.2 The builder’s free will, his technical influence and personal choices obviously play

• • • • •

To give an example, although the wreck of Calvi I (16th century, Corsica) was built in the Mediterranean tradition, it has dovetail scarfs to assemble the floor-timbers to the first futtocks whereas , the wreck of Arade 1 (16th century, Portugal) which, although it was built in the ‘Ibero-Atlantic’ tradition, has none.

the beam or maximum breadth the depth of hold the floor the keel length the overall length

2

The difficulty in accurately assessing the proportions set out in shipbuilding treatises and archival documents is that the way these dimensions are measured varies from 8

Shipbuilding in the Mediterranean in the 16th century one nation to another and from one era to another. For example, some builders measure the depth of hold at the beam, whereas others do so at the main deck, second deck, or do not specify it. There is the same uncertainty with regard to the keel length: does it include part of the stem and heel of the keel or not? Here again, some builders specify this, others do not. The concept of ship length is also open to different interpretations. Some builders—as is the case with the Spanish—consider it between the inner faces of the stem and stern posts, just as others will use a measurement on their outer faces. For these reasons, caution is required when calculating the ratio to express the proportions of a ship, whilst it should be kept in mind that these calculations only provide orders of magnitude.

involve the length of the keel, others the depth of hold, some neither, but as a general rule, they establish a proportional relationship between the beam and the overall length on which it depends. In this sense, the main breadth, or beam, appears to be the first and fundamental dimension of the ship, reflecting the transversal principle of the construction ‘frame-first’. We have listed eleven texts written between the 16th and 17th centuries that more or less clearly set out the ‘As-Dos-Tres’ rule.4 Such is the case with D. García de Palacio’s piece (GARCIA DE PALACIO, 1587) which implicitly states it through an example of shipbuilding. Most of these texts are of Iberian origin, but two authors of Italian influence, the Ragusan Nicolò Sagri and the Roman Bartolomeo Crescentio also formulate the rule. Whilst Sagri states the proportion (SAGRI, 1570, fº13R), the second refers to as an example of ship measurements (CRESCENTIO, 1607, 68).

The writings of the builders of the 16th and 17th centuries contain the formulation of a rule of proportion by the Spanish, named ‘As-Dos-Tres’ or ‘Tres-Dos-As’, which is widespread in the construction of merchant ships. It established a relationship between three of the mentioned five major dimensions of the ship with a proportionality ratio ranging from 1 to 3 between maximum breadth and overall length. It therefore generated quite ‘round’ vessels with a high carrying capacity and—according to all probability—relatively poor nautical qualities.

We also wished to include two 15th century Venetian texts in this study: ‘La Fabrica di galere’ (ANONYMOUS, c.1410) and Zorzi Trombetta da Modon (ZORZI TROMBETTA, 1445). These express different proportions from those of the ‘As-Dos-Tres’ rule using examples of navi measurements, more or less a century before the other texts, with the ratio located halfway between this one and that of the Spanish ‘nueva fábrica’ of the late 16th century.

We do not know exactly when this rule was established, at most we know that it was apparent in the 16th century, but perhaps well before that. In Spain, it was abandoned at the end of the 16th century to give way to the more stretched proportions of the ‘nueva fábrica’ institutionalized by the Ordinances of 1613. Nevertheless, at the end of the 16th century, the rule still seems to be widespread, according to the Spanish builder Thomé Cano (CANO, 1611):

For greater clarity, a summary of the proportions expressed by the texts that refer to the ’As-Dos-Tres’ rule is presented in Table 1. The proportions of Venetian texts are shown in Tables 2 to 4.The original texts are also presented in Annex VII. If most of the texts expressing the ‘As-Dos-Tres’ rule (six of them) give priority to the relationship between beam/ length, three of them (indicated in the table by a star) give priority to the rule by establishing the value of the maximum breadth in relation to the depth of hold.5 But regardless of the reference to the depth of hold to establish the maximum breadth, as we have pointed out, its value remains the main unit used as the basis for establishing the length of the ship and not the depth of hold.6

‘…todos los maestros españoles, italianos y de otras naciones que manejan estas fábricas de naos an tenido uso de les dar a un codo de manga dos de quilla; a otro de manga, tres de esloría, y a tres codos de manga, uno de plan; y el puntal tres cuarto de manga.’’ (See translation in footnote 5 of the Introduction). The exact origins of the ‘As, Dos, Tres’ rule are not known. It is possible that it was inherited from religious symbolism, where 1, 2, and 3 constituted an expression of the Trinity. We can also see how the discovery of the ‘Divine proportion’ in 13th Italian century by the mathematician Fibonacci, was transposed into naval technical culture3 whose sequence begins with 1, 2 and 3 (RUBIO SERRANO, 1991, vol. 1, 235). The authors who state it do not say it, they simply praise its perfection.

A particular expression of the rule set out by Escalante de Mendoza is also notable. He set out the value of the Thanks to Cayetano Hormaechea for sharing his work on the ‘AsDos-Tres’ rule, which mentions several of these texts and carries out a thorough and relevant analysis. 5 In this case, C. Hormaechea points out that the architectural representation of this ratio is frustrated by the vagueness of how the depth of hold is measured: some authors do not elucidate whether its measurement is taken at the beam or at the deck level. In this case, we don’t either know exactly how far the deck is located above the breadth line. 6 Nicoló Sagri’s text clearly establishes the maximum breadth as the basic measure from which all the others flow. This is also the case of the ‘Decreto del Consejo de guerra …’ which set out the length relative to the breadth. On the other hand, this is not the case of the anonymous text ‘el arqueo de Cristóbal de Barros’ which states that the ‘As’ is the measure of the depth of hold, but which, contradictorily, exposes, then, the example of a ship whose length (48 codos) is established in relation to the breadth and not to the depth of hold. In this case, the ratio given is: total length = 3 times the maximum breadth (16 codos). 4

We find various formulations of the ‘As-Dos-Tres’ rule in the texts of the 16th and 17th centuries. Some of these 3 Leonardo Fibonacci or his usual name, Leonardo Pisano, was an Italian mathematician of the 12th and 13th centuries, author of the algebraic suite that bears his name. This sequence, conceived as a series of numbers, each of which is the sum of the two previous ones (1, 2, 3, 5, 8, 13, 21, etc.), is called ‘the Divine proportion’ during the Renaissance period.

9

The Mortella III Wreck Table 1. ‘As-Dos-Tres’ proportions rule expressed by different authors between the 16th and 17th centuries Authors

Book / Text

Origin

Thomé Cano

Arte para fabricar, fortificar y aparejar Spain naos..., 1611, Dialogo segundo, p.67

J. A. Echeverri

Beam

Keel length

Overall length

Depth of hold

Flat

1611

1

2

3

3/4

1/3

MNM, Colección Vargas Ponce, T 3A Spain Doc. 108 fol. 391-395.

Mid-16th

1

2

3

2/3

1/3

Rodrigo Vargas

AGI, Real Patronato, leg. 260, 2º, rº 35: Text published by J.L. Casado (CASADO SOTO, 1988)

Spain

Around 1570

1

2

3

1/2 at deck

 

Diego García de Palacio

Instrucion náutica,... 1587

Spain

1587

1

2

3

 

 

Escalante de Mendoza

Ytinerario de navegación..., 1575, p.39

Spain

1575

1

2,27

3,18

 

 

Domingo de Busturia

AGS, Guerra Antigua, Leg. 347, nº 23. 1568. Published by J.L Casado (CASADO, 1998, T.2)

Spain

1568

1

 

3

1/2 at beam

 

Decreto del Consejo de guerra*

On papers brought by Diego Spain Brochero. AGS - Guerra y Marina, leg, 776 publish by Rodriguez M,, B. M. (RODRIGUEZ M,, 2008)

1612

1

 

3

1/2 at deck

 

ANON.*

Cristobal de Barros, MNM Vargas Spain Ponce, T.XXV B, doc.19 f. 42-43 (CRISTOBAL DE BARROS, 17th c.)

Before 1613

1

 

3

1/2

 

Fernando Oliveira O livro da fábrica das naos... , 1570, Chapter V

Portugal around 1570

1

 

3

2/3 at beam, 3/4 at 2d deck

 

Nicolò Sagri*

‘Il carteggiatore’, 1570, fº13v

Ragusa

around 1570

1

 

3

1/2 at 2d deck

1/3

Bartolomeu Crescentio

Nautica Mediterranea, 1607

Roma

1607

1

 

3

 

1/3

maximum breadth (manga) and that of the overall length (esloria) in relation to the length of the keel (quilla derecha7). Then he expressed this proportion with a ratio of 5 codos of keel for 2 codos and 1/5 at beam, which is exactly 2.2 codos. Then he indicates the value of the overall length equivalent to 7 codos for 5 codos of keel. From this point of view, Escalante Mendoza’s rule seems singular because he takes the value of the keel as its referent (As). The resulting ratio nevertheless has values close to the ’AsDos-Tres’ rule, although more stretched: 1 : 2.2 : 3.18 .8

Period

officially adopted in the Spanish Ordinances of 1613.9 Like all technical developments, the shape of ships in Spanish construction would change gradually and the two ‘schools’ (that of the ‘Nueva Fabrica’ and the old tradition) will coexist for a time. For example, Diego Garcia de Palacio who in 1587 (17 years after Escalante de Mendoza) gave his 400 toneladas ship the old proportions 1:2:3, alongside Thomé Cano who also recommended the ‘As-Dos-Tres’ rule in 1611. The text of the Ragusan builder Nicolò Sagri also reflects the heterogeneity of the rule’s application towards the last third of the 16th century when, after having stated the rule ‘As-Dos-tres’, he wrote in 1570:

In this way, could Escalante de Mendoza be the precursor of an evolution of the old ‘As-Dos-Tres’ rule towards a progressively more stretched shape? Proportions of the ‘nueva fábrica’ were inspired by the builders Juan de Veas and Rodrigo Ramírez, whose ratio for merchant ships in the last third of the 16th century was 1:3:3.75, which would be

‘[fº13V]…e questo è la più giusta migliore et più proporcionatta missura che si possi inmaginare sebene poche navi ogidi nel paesse nostro cossi si fabrichano ma Ile antiche chossì si fabrichavano (14R) et hoggi dei genovessi anchora cossi mantengono e li bischaini et portogalessi al medessimo e perciò quelle loro navi sono miglior veliere e spécialmente borinevolle e di miglior governo del timone…’

7 The ‘quilla derecha’ can be compared to the French concept of ‘quille portant sur terre’ (‘keel on land’), it means the ‘straight keel’, in other words, it includes the straight part of heel of the keel, and on the other hand, the lower end of the stem post tangent to the fore end of the keel piece. 8 According to J. L. Rubio Serrano, the rule enunciated by Escalante de Mendoza does not escape the symbolism of the Renaissance period, the ratio 22/7 being a good approximation of the figure Pi (RUBIO SERRANO, 1991, 148). In other words, a right-angled triangle with sides with a value of 5 will have a hypotenuse with a value of 7, which then allows the drawing of a circle with perimeter with a value 22 to pass through the end of the three corners of the triangle.

9 The captains Juan de Veas, Diego Rodriguez and Admiral Diego Brochero were, between the end of the 16th century and the beginning of the 17th century, the three builders who would bring fundamental innovations to Spanish shipbuilding and would be at the origin of what was called at the time the ‘nueva fábrica’.

10

Shipbuilding in the Mediterranean in the 16th century ‘[fº13V]…this is the fairest and best proportion that can be imagined although today few ships in our country are built in this way while the old ones were built in this way. [fº14R] Today the Genoese still build them in this way as well as the Biscayans and the Portuguese and that is why their ships are better under sail, especially to sail upwind, and they respond better to the rudder.’

adopted by Spanish and Italian builders as well as by those of ‘other nations’. The builder Domingo de Busturia in his 1568 Memorial highlighted it in Biscayan shipbuilding, Father Oliveira in the Portuguese and Bartolomeo Crescentio in the Italian in general. On the Italian side, most of the information on shipbuilding is provided by the Venetian treatises of the 15th and 16th centuries. Examination of the proportions they advocate reveals rules that govern shapes generally more stretched than those of the ‘As-Dos-Tres’ rule (Tables 2 and 3). The ‘Fábrica di galere’, for example,

As we have seen, we find the ‘As-Dos-Tres’ rule stated by authors of both Mediterranean and Atlantic technical traditions. Some of these authors also give it a universal character. Thomé Cano, for example, claimed that it was

Table 2. ‘Fabrica di galere’, beginning of the 15th c.Biblioteca Nazionale Centrale di Firenze, codex Magliabecchiano, XIX.7 Type

Nave latina

Breadth at first deck ‘bocha’ 

Keel length

Length at second deck ‘choverta’

Depth of hold at first deck

Flat

Ratio

1

2,50

3,33

0,40

0,38

Venetian feet

24

60

80

9,5

9

8,35

20,88

27,84

3,31

3,13

1

2,45

3,58

0,49

0,37

26,5 (2)

65

95 (1)

13

9,75

9,22

22,62

33,06

4,524

3,393

Meters Ratio Nave cuadra or Cocha Venetian feet Meters

Table 3. Zorzi Trombetta ‘da Modon’ c. 1444, ‘Libro’, British Library, Cotton MS Titus A XXVI , fº12a to 16a and 37a to 60b Type Nave 1000 botte (600 tons) Nave 700 botte A (420 tons) Nave 700 botte B (420 tons) Nave 500 botte (300 tons) Nave x botte Nave 300 botte (180 tons) Nave 250 botte (150 tons) Nave 200 botte (120 tons)

Breadth at first deck  ‘bocha’ 

Keel length

Length at second deck  ‘choverta’

Depth of hold at first deck

Flat

Ratio

1

2,50

 

0,35

0,32

Venetian feet

34

85

_

12

11

11,83

29,58

 

4,18

3,83

Meters Ratio

1

2,59

3,80

0,39

0,32

Venetian feet

28

72,5

106,50

11

9

9,74

25,23

37,06

3,83

3,13

Ratio

1

2,50

3,80

 

0,32

Venetian feet

28

70

106,50

_

9

9,74

24,36

37,06

 

3,13

1

2,90

 

 

0,36

Meters

Meters Ratio Venetian feet Meters Ratio Venetian feet

25

72,5

_

_

9

8,70

25,23

 

 

3,132

1

2,52

 

0,40

0,40

25

63

_

10

10

8,70

21,924

 

3,48

3,48

1

2,78

 

0,33

0,31

Venetian feet

22,5

62,5

_

7,5

7

Meters

7,83

21,75

 

2,61

2,44

1

2,93

 

0,41

0,41

Venetian feet

20,5

60

_

8,5

8,5

Meters

7,13

20,88

 

2,96

2,96

1

3,24

 

 

 

Venetian feet

18,5

60

_

_

_

Meters

6,44

20,88

 

 

 

Meters Ratio

Ratio

Ratio

11

The Mortella III Wreck Table 4. Pre Theodoro de Nicoló, c.1550, ‘Instructione sul modo di fabricare galere’, Biblioteca Nazionale Marciana di Venezia, manoscritti italiani, cl. IV cod. XXVI (5131) Type Great galleon

Breadth at first deck ‘bocha’ 

Keel lenght

Length at second deck  ‘choverta’ 

Depth of hold at first deck

Floor

Breadth at first deck ‘bocha’ 

Ratio

 

1

2,67

3,61

0,32

0,29

Venetian feet

33

37,5

100

135,50

12

11

11,48

13,05

34,8

47,15

4,18

3,83

Meters Commercial Nave of 10 pas

Ratio

 

1

2,17

3,11

0,28

0,30

Venetian feet

20

23

50

71,50

6,5

7

6,96

8,00

17,4

24,88

2,26

2,44

Meters

shows the measurements of two round ships, the nave latina (fº37) and the nave quadra (fº88), whose ratios (1: 2.50: 3.33 and 1: 2.45 : 3.58) reflect this well. In the 15th century, Zorzi Trombetta’s 700-botte (420-tons) nave appeared with even more elongated shapes: 1 : 2.59 : 3.80 (Table 3).

To return to the situation in Italy, it was perhaps Pre Theodoro de Nicolò who, in the middle of the 16th century, In his ‘Instructione’, after having studied the question of galleys at length, he presented two sailboats (Table 4): a warship of 33 feet of bocha (beam) and 20 pas (20 steps equivalent to 100 feet)11 of keel length which he calls ‘galion grande’ (fª26) to which he gives a ratio of 1: 2.67: 3.61 and a merchant ship, a ‘nave’ of 20 feet of bocha (fº27) whose ratio is 1 : 2.17 : 3.11.12

But here the question arises as to which type of ship these proportions correspond to? Indeed, the authors of the Venetian treatises are essentially the work of shipbuilders employed by the Arsenal. This was a stateowned building whose characteristics often differed from those of private shipyards specialized in the production of merchant ships. There is therefore still some doubt here. Generally, in the terminology of the time, the term ‘nave’ was used to characterize a merchant ship (GATTI, 1999, 145). Frederic C. Lane confirms this when he writes:

The values of Theodoro de Nicolò’s merchant ship are very close to the ‘As-Dos-Tres’ rule. As a result, it seems that this rule took precedence in the construction of merchant ships in the 16th century, at least for the first thirty years. As for those of the ‘galion grande’, they also seem to confirm Domingo de Busturia’s rule on the proportions of warships. However, it should be added that it was not systematic. One example is a valuable document reported by C. Hormaechea (HORMAECHEA, 2012, VOL.II, 217-222) published by M. Fernandez Navarrete (ANON., c.1570). This is a contract for the construction of twelve war galleons for the Spanish fleet of Illyria drafted around 1570. It gives a clear overview of the ‘As-DosTres’ proportions of these ships built in the arsenals of the Naples region (Castelammare, Salerno) and Ragusa by Venetian, Neapolitan and Ragusan masters.13

‘The large round-ships used in the commerce were commonly called ‘nave’. They were not built in the arsenal but in private shipyards of Venice.’ (LANE, 1934, 46). In these circumstances, do the ‘navi’ of the ‘Fabrica di galere’ and those of Zorzi Trombetta really meet the definition of merchant ships? Because, as we see on many occasions, the proportions that builders gave to warships differed from those of merchant ships.

Similarly, although the ‘As-Dos-Tres’ rule for merchant ships has been shown to be largely widespread, it nevertheless seems that it was neither universal nor systematic. Archaeology proves this. While most of the wrecks of the Renaissance Atlantic constructive tradition seem to adjust to the rule (such as the ship of Red-Bay, for

In Spain, for example, in the 16th century a length/width ratio of up to 4 was recommended for warships, as Domingo de Busturia indicated in 1565:10 ‘Las naos que en esta costa de Bizcaya se fabrican para mercancía son en la mayor parte de tres y a una (…) las naos o navíos que son hechos para la guerra han de ser de otra proporción, de a tres y medio y aun algo menos que cuatro y a una.’

11 One Venetian foot was equivalent to 34.8 cm, or 1.1 English feet. One step was equivalent to 5 feet, or 1.74 m. 12 The ratios presented in Table 4 are calculated on the assumption that for Theodoro de Nicolò, the definition of the ‘bocha’ is the measurement of the breadth at a height of 9 feet, as explained by F. Lane in his article ‘Venetian naval architecture about 1550’ (LANE, 1934, 27). In the text of Theodoro de Nicolò, it is indeed clear that the beam is located well above the ‘bocha’ and it is called ‘regia’. It is important to note that the definition of the ‘bocha’, here, differs from that given by the Venetian writings in 15th century. For both the author of the ‘Fabrica’ and Zorzi Trombetta, the ‘bocha’ takes the width value at the deck (segunda choverta = first deck). It is therefore the value of the beam – or close to the beam – that we took into account when calculating the ratios. 13 An analysis of this document was carried out by José Luis Casabán (CASABAN, 2017, 238-260).

Translation: ‘Most naos made in Biscay for commercial use have a ratio of tres to one (…) naos or ships for wartime use must have other proportions, from three and a half and a little less than four to one.’ 10

Thanks to C. Hormaechea for pointing out this text to me.

12

Shipbuilding in the Mediterranean in the 16th century example), the situation is less clear in the Mediterranean, as we will see in Chapter V: while the wreck of Calvi I meets the rule, those of Mortella III and Villefranche-surMer do not fit in.

‘… allí se asentará el primer madero de cuenta, el cual es formado por un palo que llaman estamenara y dos barraganetes, a manera de un medio círculo… ‘. ‘… placed here will be the first ‘madero de cuenta’[the master-frame], which is made up of a piece of wood called estamenara [first futtock] and two barraganetes [second futtocks] in the shape of a semicircle…’

Note (1): The length of the nave cuadra given in fº37 is 19 steps (passa), or 95 feet: ‘ E vole esser la dita chocha longa in choverta tante passa quanti pedi sonno in li due terzi de cio che laure in bocha che sonno passa 18. sera longa da roda passa 19 percio che la testa fara crescere passo 1.’ Note, however, that the sum of the keel length and the aft and fore rakes given in the text (22 and 2/3 feet and 5 feet, respectively) result in 92 and 2/3 feet. With this length value, the Width / Length ratio would be 3.5.

This approach is confirmed by the Spanish Ordinances of 1613 (fig. 2) and 1618. Article 15 of the 1618 Ordinances sets out how the master-frame’s timbers are to be built; and concludes that according to these prescriptions: ‘…saldrán los navíos redondos, con mucha bodega,…’.

Note (2): In the ‘Libro’ of Michele da Rodi, which is supposed to be the source of the ‘Fabrica di galere’, the value of the ‘bocha’ of the nave cuadra is 27 feet.14

‘…the ships will be round, as a result, with a large hold, …’. Far from being exclusively Spanish, we also find a circular design of the ‘figure’ of the master-frame in France, up to the level of the beam. It is described by Father Fournier as ‘the ancient method’ in his Hydrography, one of the first French treatises on shipbuilding (FOURNIER, 1643, 23). As Father Fournier’s work was written in the first third of the 17th century, it can be assumed that the ‘ancient method’ was in use at the beginning of the 17th century and perhaps even much earlier.

A final point to be taken into account is that builders will necessarily vary the proportions used according to the desired ship size. While the ‘As-Dos-Tres’ rule can be applied to relatively large ships, the more their tonnage decreases, the more—in principle—their shapes will stretch. This relationship can be verified by observing the ratios of the proportions that Zorzi Trombetta gave to his navi—except for the 500 botte nave—(Table 3). In Spain, these variations are institutionalized in the Ordinances of the early 17th century.

On the Portuguese side, the circular shape of the masterframe also seems to be required, at least for traditional size ships.

1.3.2. The form: the shape of the master-frame In French, the term ‘figure’ refers to the shape of the masterframe. This frame forms the basis of the architectural project insofar as it is on its model and through its ongoing modification that the shape of the frames located between the tailframes was set out.

In his Livro da Fábrica das Naus (OLIVEIRA, 1580) Oliveira draws the ‘figure’ of the master-frame of a ship 48 palmos breadth with a single arc-circle (fig. 3). Similarly, in 1616, Manoel Fernandes set out the ‘figure’ of the master-frame of his 500-ton galleon with a semicircular line (Fig. 4, MANOEL FERNANDES, 1616, Fº88). However, in this work Manoel Fernandes, as well as his compatriot João Baptista Lavanha, recommends a more complex master-frame shape for very large vessels. Fernandes proposes a shape with three circular arcs when building a 56 palmos ship’s breadth. Lavanha suggests a similar design for a large ship, the nau da India, which was 54 palmos breadth. C. Hormaechea points out, however, that these ships were of exceptional dimensions that exceeded those of the largest ships considered in the Spanish Ordinances of 1618. For smaller ships—which were the vast majority—the circular shape of the masterframe was the rule (HORMAECHEA, 2012, 168).

Examining the texts on the ‘Ibero-Atlantic’ side tends to show that in the 16th century, the galibo (this is the Spanish term for the master-frame shape) was set out along a single circle (HORMAECHEA, vol. 1, 168, 169). First of all, the Spanish authors advocate a single circle shape. Escalante de Mendoza is one of the first to recommend it in 1575 (ESCALANTE, 1575, 40): ‘… y con las dichas medidas, saliendo el costado redondo por su cuenta y razón…’ ‘… and with the said measurements, the shape of the side [of the master-frame] will be round…’ In the same way, Diego García de Palacio, writes in his ‘Instrucción Náutica’ (GARCIA PALACIO, 1587, fº92v):

The use of several tangent circular arcs in the master-frame shape design may have been influenced by the English model proposed by Mathew Baker in his Fragments of Ancient English Shipwrightry (BAKER, 1570) (fig. 5a.). Could this model have inspired the Ibero-Atlantic building tradition? In any case, this is what the archaeologists of the Red Bay wreck (Labrador, Canada, 1565) think. Indeed,

A detailed study of Michele da Rodi’s manuscript was carried out through a research project called ‘The Michael of Rhodes project’. An overview can be found on the web at: https://brunelleschi.imss.fi.it/ michaelofrhodes/index.html It has also resulted in a very comprehensive publication (LONG et al., 2009).

14

13

The Mortella III Wreck

Figure 2. Master-frame shape of the 16 codos galleon as prescribed by the Spanish Ordinances of 1613 (Courtesy of: Cayetano Hormaechea).

it (see, in particular, BARKER, 1986). This influence seems logical in light of Mathew Baker´s journey in Venice and the presence of several Venetian shipwrights in England from 1543 onwards (JOHNSTON, 1994, 128), including Augustino Levello who, hired to build Henry VIII’s galley, remained in the service of English shipbuilding for 40 years.15 In Baker’s work, the close similarities between the shape of the English and Venetian master-frames, its likeness to Venetian design methods and the use of Venetian terms to designate them make this influence likely. However, this influence should not make us forget the innovations that were certainly brought by the English builders, in particular the use of mathematics in the definition of the design. Finally, it must be taken into account that carvel building and its ‘frame first’ process are, at the time Baker wrote, a recent development in England while it was ‘ancestral’ in the Mediterranean. The likely Italian—and more specifically Genoese— origin of the Mortella III wreck led us to observe the shape of its master-frame, paying particular attention to those used in the Italian shipbuilding, as the texts on which we relied upon are Venetian. In his search for the perfect master-frame shape, among the 37 profiles he exposes, Mathew Baker describes the one he attributes to the Venetians by using four circle arcs (fig. 6). Nevertheless, an examination of the texts set out above seems to show

Figure 3. Shape of the master-frame of a 48 palmos ship, according to F. Oliveira, 1570, fº112 (Courtesy of: Biblioteca Nacional de Portugal).

they recognized the English design in the profile of its master-frame (fig. 5b.), by four circular arcs following the method described by M. Baker. Thus, they concluded that the English method had been adopted by Basque shipbuilding (LOEWEN, 2007, 97).

15 The archives of the Bodleian Library show that Augustino Levello was for 40 years a ‘Royal Master Shipwrights’ a title which only six manufacturers held (including Mathew Baker). The accounts of the Royal Administration show that no fewer than five Italians were working in English shipbuilding from 1543 onwards (cited by T. GLASGOW, 1970, 10).

The origin of the English method remains uncertain. There is a debate on the extent of influence that the Italian shipbuilding—in this case Venetian—may have had on 14

Shipbuilding in the Mediterranean in the 16th century

Figure 4. Shape of the master-frame of a 500-ton ship, according to Manoel Fernandes, 1616 (Courtesy of Academia de Marinha de Portugal).

Figure 5. Left (a): Shape of the master-frame recommended by Matthew Baker (Courtesy of: the Pepys Library, Magdalene College); Right (b): Representation of the master-frame of the Red-Bay wreck (Courtesy of: Carol Pillar, Parks Canada).

that circular arcs system design was not known or, in any case, not used by the Italian builders of the Renaissance.16 The shape of the master-frame they advocate – both for galleys and round ships – was obtained using a scale

of values, or offset, set out by the relationship between heights taken on a line running above the keel (ordinate) and the width of the frame (abscissa). In other words, the master-frame design was based on an algebraic method, as opposed to the later method of arc design method determined by geometry. For example, the ‘Fabrica di galere’ based the design of the master-frame of a round vessel on two reference values:

A design of the master-frame with three circular arcs can be seen in the Genoese influenced text by the German Joseph Furttenbach in 1629 (FURTTENBACH, 1629, 106).

16

15

The Mortella III Wreck

Figure 6. Shape of a Venetian ship master-frame obtained by four tangent arcs according to Matthew Baker (Illustration: Arnaud Cazenave de la Roche).

Figure 7. Zorzi Trombetta da Modon’s method for tracing the shape of the master-frame of a nave (Courtesy of: British Library).

1. That of the ‘trepie’ width, i.e. the distance between the center of the master-floor and the outer part of the frame at a height of three feet above the top of the keel. 2. That of the ‘bocha’, i. e. the width at the first deck17, being taken as a reference for the maximum breadth or beam.

bulge of the Venetian profile compared to the English one.19 The values provided by the manuscript of the ‘Fabrica di galere’ allow a reproduction of the master-frame shape of both the Latina and Cuadra navi. We present here that of the Nave Cuadra (fig. 11) because it almost perfectly coincides with the English shape of the master-frame recommended by Mathew Baker (fig. 12).

The shape of the frame is drawn using a curved line connecting the center of the keel to these two points, successively. Zorzi Trombetta da Modon introduced a fourth point in his ‘Libro’, which is the ‘Siepie’, i.e. the value of the width at six feet in height (fig. 7 and fig. 8).

Considering the above, a reflection can be formulated: 1. The shape of Mathew Baker’s Venetian model is identical to that of Zorzi Trombetta 2. The shape it recommends for English construction is almost identical to that of the ‘Fabrica di galere’ 3. These shapes are obtained by the Venetian treatises in the 15th century by means of offset scales, while they are obtained by the English builder in 1570 with four arcs-of-circles

As Trombetta da Modon’s drawing in fig.7 is only a clumsy sketch, the shape of the master frame is poorly represented. However, the values provided by the manuscript18 enable to reproduce a profile to scale with precision (fig. 8). Here is its result for the 700-botte ship: Surprisingly, the shape of the master-frame of the 700-botte nave, resulting from Zorzi Trombetta’s values (fº44v), corresponds—almost identically—to the Venetian masterframe depicted by Mathew Baker. The superposition of the two profiles (fig. 9) shows this eloquently.

In view of the above, it is therefore legitimate to wonder whether the Baker method would not be, in a way, a ‘geometrization’ of the Venetian method. It is in the light of the questions and issues examined during this overview that the design methods of the Mortella III

If we now compare Zorzi Trombetta’s model with the English one given by Mathew Baker (fig. 10), although we can see a proximity of the shapes, we notice an increased

19 If we stick to the sketch of Z. Trombetta (fig. 7), it seems that while the ‘Trepie` and ‘Siepie’ are measured on the inner face of the frame, the ‘bocha’ is measured on its outer face. However, as this is a guess, we have opted to build our profile by taking all measurements on the inside of the frame. If we had opted for an external measurement of the ‘bocha’, this bulge would certainly be less (about one Venetian foot) and there would be a greater proximity between M. Baker’s English profile and that of Z. Trombetta.

17 This is the seconda choverta in the Venetian texts, the first one being the orlop deck. 18 The ‘Libro’ of Zorzi Trombetta da Modon is very rich, it gives the proportion relationships of the measures of eight nave whose tonnages range from 1000 to 200 botte.

16

Shipbuilding in the Mediterranean in the 16th century

Figure 8. Restoration of the shape of the master-frame of the nave of 700 botte by Zorzi Trombetta da Modon (Illustration: Arnaud Cazenave de la Roche).

Figure 9. Superposition of fig. 6 and 8: similarity of Zorzi Trombetta’s shapes and Mathew Baker’s Venetian model (Illustration: Arnaud Cazenave de la Roche).

17

The Mortella III Wreck

Figure 10. Superposition of fig. 5a. and 8: forms of Zorzi Trombetta and M. Baker’s English model (Illustration: Arnaud Cazenave de la Roche).

wreck master-frame will be analyzed. But first, it was necessary to identify the Mortella’s master-frame and then reproduce its shape as accurately as possible. The stakes were high because, as we will see now, the overall shape of the hull is based on this frame.

according to the ‘frame first’ construction principle, i.e. in a transversal architectural logic.20 The ship of the Mortella III is in a way both an heir and a witness of this construction concept which characterized Mediterranean naval architecture throughout the Middle Ages and which continued into modern times. It is the result of a period when shipbuilders worked mainly by empirical methods, in other words, without using drawings and plans. As a result, the ship transverse design was based, to a large extent, on the use of moulded frames between the tailframes. The master-frame shape was used as the basis of their shapes to which several modifications gradually changed their form.

1.3.3. The design method: the moulding process A complete and understandable approach to the issues affecting the construction of a ship in the 16th century cannot be made without mentioning a central question on which nautical archaeology is currently focusing and which has mobilized—and still mobilizes—many researchers over the past few decades, namely the transition from carvel construction on ‘Shell first’ to ‘frame first’. In light of the latest results of the excavations carried out in the Dor lagoon (Tantura) in Israel by the University of Haifa, under the supervision of Professor Kahanov, this evolution probably took place in the Mediterranean between the end of the 5th and 7th centuries. It is beyond the scope of our work to describe the details of this research, we will only specify that the Mortella III wreck belongs to this family of ships built

This design method, already identified in Italian shipbuilding almost a century ago by C. Anderson (ANDERSON, 1925), then described and partially explained by F. C. 20 It seems to us that this notion of ‘transverse’ architecture that we owe to Richard Steffy (STEFFY, 1989, 419) finds a good illustration in the Spanish Ordinances of the first third of the 17th century, where the size of ships was classified according to their width, and not their length, as today it would seem natural.

18

Shipbuilding in the Mediterranean in the 16th century

Figure 11. Shape of the master-frame of the nave cuadra of ‘Fabrica di galere’ obtained from the measurements provided in folio 37 (Illustration: Arnaud Cazenave de la Roche).

Figure 12. Superposition of fig. 5a and 11: coincidence of shapes of the ‘Fabrica di galere’ and Matthew Baker’s English models (Illustration: Arnaud Cazenave de la Roche).

19

The Mortella III Wreck Lane (LANE, 1934), appeared to be widespread during the Renaissance. Since the late 1980s, several researchers have been interested in it, including Sergio Bellabarba (BELLABARBA, 1988), Richard Barker (BARKER, 1991 and 2001), Eric Rieth (RIETH, 1996, 2003), Filipe Castro for Portuguese construction (CASTRO, 2007) and Cayetano Hormaechea for Spanish construction (HORMAECHEA, 2012 and 2018), to mention a few. It should be added that the survival of the whole-moulding method continues to be identified nowadays in the Mediterranean shipbuilding in Provence and Greece (DAMIANIDIS, 1998), for example, but also outside the Mediterranean area, for example in Brazil (CASTRO, 2014). The size and shape of this group of frames located, it should be remembered, between the tailframes, met specific rules. The shape of the master-frame followed an algorithmic evolution, calculated on the site itself, using wooden measuring tools. There were four successive modifications made to the frames that succeeded the master-frame: • The narrowing, i.e. a progressive reduction of the floorframes length. • The rising, in other words the progressive raising of the floor-frames height by increasing the thickness of their heels. • And finally, the ‘trébuchement’ (French word meaning ‘tilting’), and the ‘hauling down/up the futtock’. These operations consisted of tilting the first-futtocks outwards (‘trébuchement’) together with an adjustment of their joint with the floor-frames (‘hauling down/up the futtock’). These last two processes allowed the builder to increase the volume of the hull and thus increase its buoyancy and capacity. This method of hull design, using moulded frames, is described in the Iberian shipbuilding treatises of the 16th and 17th centuries as ‘maderas de cuenta’ in Spanish or ‘madeiras da conta’ in Portuguese, whose terms ‘cuenta’ or ‘conta’ refer to the count or calculation carried out to modify the size and shape of the frames. The Portuguese authors, in particular Father Fernando Oliveira (OLIVEIRA, 1570) and João Baptista Lavanha (LAVANHA, 1610) describe the method in detail in their treatises. The narrowing and rising of the floor are obtained using diagrams, the graminhos, which enable the creation of simple geometric algorithms. On the Spanish side, Thomé Cano was the first to describe the method at the beginning of the 17th century (CANO, 1611). Subsequently, the Spanish Ordinances of 1613 officially established the method in Spain. It should be pointed out, however, that it is only in the Spanish text of 1613 that the notion of ‘trébuchement’ appeared for the first time under the name of joba. This operation was hitherto ignored in the Iberian texts. Cruz Apestegui attributes its introduction in Spanish construction to the shipbuilder Juan de Veas, captain and maestro mayor of the king’s shipbuilding.21

Figure 13. The ‘graminho’, an instrument for the narrowing (OLIVEIRA, 1570, Fº93) (Courtesy of: Biblioteca Nacional de Portugal).

It should be added that the moulding design method, as described above, has had variants. For instance, the methods described in the 18th century by the French P. Bouguer (BOUGUER, 1746) and the Spaniard Jorge Juan (JUAN, 1757). They are characterized by the use of ribands, but the basis of these methods is still based on the shape of the master-frame. A description of this can be found in Chapter 6 of the Maître-gabarit… (RIETH, 1996, 97-105). The first known references to the moulding design method date back to the end of the thirteenth century. First in a quotation for the construction of a nave for equestrian transport dated 1273 (FOURQUIN, 2001) and then in a letter confirming the construction of a galley in Brindisi dated 1275, the ‘galea rubra de Provincia’ intended for Charles I of Anjou’s court.22 In these texts appear the words ‘sextis’ and ‘sexto’, from which probably derives the Venetian term of ‘sesto’. Actually, it is on the Italian side -Venice, in this case – that we find the earliest descriptions of the moulding method. In the course of the 15th century, it was indeed characterized in several Venetian texts dealing with

Cruz Apestegui does not indicate the source that led him to attribute the introduction of joba in Spanish shipbuilding to Juan de Veas, perhaps it is simply based on the influence that this builder had in the drafting of the 1613 Orders?

21

22 Text quoted by S. Bellabarba (BELLABARBA, 1996), published by G. del Giudice in 1871 (GIUDICE, 1871, 25).

20

Shipbuilding in the Mediterranean in the 16th century

Figure 14. Fernando Oliveira illustrates the rising of the ‘madeiros da conta’: (OLIVEIRA, 1570, Fº103) (Courtesy of: Biblioteca Nacional de Portugal).

Two texts still evoke it in the 16th century: • The ‘Instructione sul modo di fabricare galere’ by Pre Theodoro de Nicolò (PRE THEODORO, 1570); • At the end of the 16th century, Baldissera Quinto Drachio’s ‘Visione’ undertakes a more detailed description of the method (DRACHIO, 1594). In this last text, the four modifications made to the shape of the master-frame are explained: the narrowing of the floor which appears under the terms of ‘partisone del fondo’, the rising of the floor-frames, the ‘stella’, the ‘trébuchement’, the ‘partisone del ramo’ and finally, the hauling down/up the futtocks, the ‘scorrer del sesto’.23 Although the method of designing ships using the moulding method nowadays appears to be linked to the carvel construction, that is to say to the ‘frame first’ process, it is not easily detectable archaeologically. Nevertheless, a remarkable attestation of this method has been found on the early 14th century wreck of Cala Culip VI (Catalonia, Spain) where a series of surmarks and numbers have been observed on the floor-timbers—or madiers, in Mediterranean language—to highlight the practice of the method (NIETO and RAURICH, 1998).

Figure 15. The ‘partisone’ system according to Zorzi Trombetta de Modon (TROMBETTA, 1445, Fº45) The ‘Mezzaluna’ (Courtesy of: British Library).

shipbuilding under the name of ‘Partisone’. It is more or less explicitly exposed in five texts. For the 15th century we have: • the ‘Fabrica di galere’, an anonymous text from the beginning of the 14th century (ANONYME, c.1410); • the ‘Libro’ by Zorzi Trombetta da Modon in 1445 (TROMBETTA, c.1445); • the ‘Ragioni antique spettanti all’arte del mare et fabriche de vassalli’, also published in the 15th century (ANONYME, 15th c.).

Frederic C. Lane undertook an explanation of the method set out in these texts in his article on ‘Venetian Naval Architecture’ (LANE, 1934, 24-49). E. Rieth takes up the analysis in Chapter 9 of the Maître gabarit… (RIETH, 1996, 133-148).

23

21

The Mortella III Wreck Even if we cannot assume the exclusivity of the moulding method on ‘frame first’ construction, it must be outlined that it seems generalized in modern period and going beyond the boundaries of the maritime spaces. Today, we know that the carvel construction was adopted by the Atlantic technical culture between the second half of the 15th century and the beginning of the 16th century.24 It can therefore be assumed that the moulding design method that appears to be closely linked to this approach, was adopted in the Atlantic maritime space at the same time.25 The builder of the Mortella III ship built the framework using a moulding process, although so far we have not found any surmark engraved on the frames that could demonstrate it formally. Nevertheless, some constructive characteristics, in particular the orientation of the nailing, confirm the use of the predetermined frames, we will come back to these aspects in Chapter III.

24 In his Documents inédits pour l’histoire de la Marine, Auguste Jal published Antoine Conflans’ text, which dates back to 1512 and which refers to several types of ships built ‘on carvel’ that can be found in different ports on the ‘Ponant’ coast (JAL, 1842, 27-59). A presentation of this text was made by Michel Mollat du Jourdin and Florence ChillaudToutée at the 107th Congress of Sociétés Savantes in 1982 (MOLLAT, 1984, 9-44). Finally, in his note 50, Jal reports that, according to the Dutch author Theodore Vélius, the first ‘carvel’ ship to appear in Holland was built in Horn in 1460. 25 The moulding method nevertheless gave rise to a debate born in the early 2000s on the question of its origins in the Atlantic area. Archaeologists of L’archéologie subaquatique de Red-Bay say they have detected the use of a ‘hauling down’ of the futtocks of the wreck of the San Juan consistent with the system evoked by the British builder Mathew Baker. Brad Loewen draws on this fact and on the absence of any ‘trébuchement’ to hypothesis that the Basque system would proceed from an Atlantic tradition independent of any Mediterranean influence (LOEWEN, 2007, 70 and 71). But can we consider an Atlantic tradition independent of the Mediterranean one in the sixteenth century if we admit the weight of Venetian influence on English shipbuilding?

22

2 Portraits of the Mortella wrecks: The discovery of the sites, their characteristics, their chronology & excavation methods 2.1. The Mortella II and III: discovery of two sites of the same origin. Characteristics and background

Bay of Saint-Florent (Upper Corsica). The second, called Mortella III—which is the subject of this work—was discovered in November 2006, 38 meters below sea level and 650 meters to the south-east of the first site.

2.1.1. The discovery of the sites

2.1.2. General description of the Mortella sites at the time of their discovery

The Mortella II and III sites were uncovered during a multi-year archaeological survey programme in collaboration with the DRASSM (French Department of Underwater Research) as part of their wider project to draw up an archaeological map of Corsica’s underwater heritage. This programme was carried out on the coasts of the Nebbio region (Upper Corsica) by the Centre d’Etudes en Archéologie Nautique -CEAN (Center for Nautical Archaeology Research) between 2005 and 2008, surveying a total surface area of around 15 km² off the coast and 50 meters deep. The two sites were discovered with a side-scan SONAR during this geophysical survey. They held the remains of large shipwrecks dated back to the 16th century. The first, called Mortella II was located in October 2005, 48 meters deep in the middle of the

2.1.2.1. Site characteristics At the time of their discovery, the Mortella sites were characterized by: • The presence of several tumuli: one on the Mortella II site and two on the Mortella III site. They were covered by gravel and ballast stones. We will come back to their description in Chapter Four. • An outcrop of wooden structures that appeared to belong to the hull of a ship, both on the Mortella II and Mortella III sites.

Figure 16. SONAR image of the site of Mortella III that led to its discovery (Photo: Arnaud Cazenave de la Roche).

23

The Mortella III Wreck

24 Figure 17. General organization of the site at the time of its discovery (Illustration: Arnaud Cazenave de la Roche).

Portraits of the Mortella wrecks

Figure 18. Site of Mortella III. Tumulus B at the time of its discovery (Photo: Christoph Gerigk).

• Three large anchors, two 4.5 m long on the Mortella III site, and one 4 m long on the Mortella II’s.1 • Wrought iron artillery found with stone shot. • A variety of encrusted artifacts scattered on the bottom, including shards of highly eroded ceramics. Notably, many well preserved ropes were found in the fore part of the Mortella III wreck.

contrasts with the poor amount of artefacts, with the exception of the well-preserved artillery. 2.1.2.3. The link between the two sites The geographical proximity of the Mortella II and III wrecks combined with the similarity of their artefacts, in particular their artillery and anchors, led to the conclusion that they were likely involved in the same historical event. The petrographic study of the gravel and ballast stones of the two sites carried out by François Gendron (Muséum National d’Histoire Naturelle -MNHN-) confirmed this link by highlighting the resemblance of the ballast stones (GENDRON, GENDRON-BADOU, 2008). In the conclusion of this study, it states the following:

2.1.2.2. Nature of the sites In 2007, a sub-bottom profiler survey revealed a particularly deep layer of archaeological remains (several meters) of small extent at the Mortella II site, while conversely it appeared shallow (about 1 meter) and extensive for Mortella III. This latter site is located on a relatively flat bottom, with a slight slope to the northwest. The minimum site depth recorded was 36.4 m to the southeast and the maximum depth was 37.6 m to the northwest. As the distance between these points was 40 m, it can be deduced there was a general slope of 2.25%.

‘These observations suggest that the stones from Mortella II are similar to those from the tumuli A and B of Mortella III. We concluded they were made of sandstone limestone more or less rich, depending on the sample, in sandstone or limestone while the white veins were calcite… It therefore appears that the Mortella II and III wrecks were certainly weighted with the same ballast stones, which could mean a common origin.’

Both sites are located amongst compact and deep mud. Its anaerobic nature is conducive to the conservation of organic artefacts. The 2007 DRASSM survey of the Mortella III site confirmed this by revealing a large and generally well-preserved hull. This architectural richness

2.1.3. Background and history of interventions The scientific and archaeological significance of the Mortella sites has prompted several interventions and studies since 2005:

A study of the Mortella III’s western anchor was carried out by Fabrizio Ciacchella, NavLab – Università di Genova (see Annex VIII). 1

25

The Mortella III Wreck identification of the wrecks of Mortella, a topic addressed in Chapter VI.

During the 2005 and 2006 survey campaigns, CEAN carried out first observations on the Mortella II and III sites, which was recorded in two documents submitted to the ‘Department of Underwater Research’—DRASSM (CAZENAVE DE LA ROCHE 2005 and 2006). This preliminary study was complemented in 2007 by a review organized by the DRASSM in which CEAN participated. This intervention resulted in the writing of a report (BERNARD, 2008) and an archaeological study (CAZENAVE DE LA ROCHE, 2008).

2.2. Archaeological material and wood of the hull: the artefacts and the material helping the chronology As always in archaeology, the ability to insert the object of study into a chronological framework is essential. Without it, observations and analyses on shipbuilding and ship architecture could not be historically contextualized and would therefore lose much of their meaning. For this reason, from the very beginning of our work on the site, we have endeavored to gather all the evidence likely to provide information that may aid dating.

The archaeological potential of the wrecks and their scientific interest, in particular for naval architecture, was demonstrated by this preliminary work. Hence, in 2010 the first archaeological excavation of the Mortella III site was organized. This site was selected because it was located in shallower waters than the Mortella II. Then, four excavation campaigns were carried out between 2012 and 2015. At the time of writing this book, we believe that the most important part of the archaeological information has been collected. However, the complete study of the site has not yet been fully achieved and a final excavation campaign is scheduled at the end of the summer 2019.

First of all, the artefacts provided us with a guideline to the wreck’s chronology. Although seemingly foreign to our work, their ability to reflect a particular historical period necessitates a brief description of their main characteristics and to deduce from their typology a relative dating. Secondly, thanks to the collection of several samples of wood, it has been possible to carry out a dendrochronological study which has led to an absolute dating. This has undoubtedly been an important in comprehending the chronology of the site yet triggered a new issue concerning the identification of the wreck, to which we will return.

Finally, it should be outlined, that in parallel with the field work, CEAN has organized a programme of historical research—still ongoing—in the French, Italian and Spanish archives. This has led to a probable

Figure 19. Wrought iron bombard (Cn3) (Photo: Christoph Gerigk).

26

Portraits of the Mortella wrecks 2.2.1. The contribution of the artefacts to the understanding of the wreck and to its dating

with a mass of 16 to 17 kg. Three other calibers were found, in smaller numbers: 158 mm, 125 mm and 96 mm in diameter. It is worth mentioning that several of the shot were found half built during shaping. Petrographic analysis (GENDRON, 2008) shows that they have been carved from serpentine, a metamorphic rock of the phyllo silicate family that is particularly hard. It is widespread in the Piedmont area and further south, on the edge of the Gulf of Genoa. The various petrographic analyses carried out during each excavation campaign (GENDRON, F., 2010, 2012 and 2013) show that, apart from the serpentine which is native to northern Italy, the rock used to make the shot is frequently associated with a volcanic origin of southern Italy.

2.2.1.1. Artillery The first chronological milestones were set at the time of the site’s discovery by observing the artillery. Further study was carried out by Max Guérout (Groupe de Recherche en Archéologie Navale – GRAN) and attached to the 2010 excavation report (CAZENAVE DE LA ROCHE et al., 2010, 59-70). The artillery is composed by wrought iron stave-type cannons with a removable breech. The barrels are about 2 meters long and 35 to 40 cm in diameter. They are made of staves encircled by barrel hoops consolidated by fitting together cylindrical tubes (sleeves), themselves reinforced at their seams by rings (ribs). Max Guérout’s study specifies their typology: these cannons were named bombardas whose use in navies date from between the end of the 15th century to the 16th century. From the second half of the 16th century, they were gradually replaced by cast iron artillery, which could be produced by an improvement in the performance of the furnaces:

2.2.1.2. Ceramics Dating the wreck was made possible by studying the ceramic artefacts found on the site. These were a set of shards, mainly located on the aft part of the wreck and an initial study was carried out by Emilie Thomas following the excavation in 2010. On this occasion, the reference production site was located in northern Italy (CAZENAVE DE LA ROCHE et al., 2010, 55). More shards were collected during the 2012 season which enabled a more in-depth study by Franck Allegrini Simonetti (Collectivité de Corse – CdC) and allowed both the chronology to be refined, but also to specify their origins (CAZENAVE DE LA ROCHE et al., 2012, 49). The shards are divided into four groups, all of which are related to food preparation:

‘Let us recall that the mention of the first cast-iron cannon appears in the inventories of the Tower of London in 1559. The last inventory which still mentions 15 pieces in wrought iron is that of 1595. Although covering most of the 16th century, the period in which wrought iron bombards were predominantly used was in its first half. This particularly true for warships, which were the first to be equipped with such innovations, while the use of bombards continued on board of commercial ships, but sometimes also on certain armed ships in war, even though they had become obsolete.’

• • • •

Cups Jugs Pots Jars

These cannons were found with stone shot. The majority of them (about twenty) were of about 220 mm in diameter

The morphology of the vessels may be compared with the crockery products called ‘Pisano Ligurian monochrome’. As for the incised decoration, it brings to mind the output of ‘Ligurian monochrome graffiti’ products in the Savona region.

Figure 20. 225 mm caliber shot. Note the inscription carved in stone (Photo: Christoph Gerigk).

Figure 21. Medium cup. Edge with incised decoration, the colour of the paste is buff to pink (Illustration: Franck Allegrini-Simonetti).

27

The Mortella III Wreck north to Norway, to 40º in the south and can even be found growing below in its easternmost area of diffusion (Fig. 22). The sessile oak is the most common species in France, despite being absent from its Mediterranean coast, unlike Italy where it can be found on the Tyrrhenian Sea in the north and also on its Adriatic side.

The jugs: a single spout fragment with green decoration seems attributable to Northern Italian manufacture, perhaps also from Liguria. Although showing a glaze with lighter pink hues, this part is comparable with examples of Italian jugs in the early sixteenth century wreck, Lardier 2, found between Hyères and St. Raphaël.’ On this point, the ceramic study by Franck Allegrini concludes: ‘Despite the sparseness of the remains, the chronology of the ceramic entity recalls the first half of the sixteenth century.’

Nevertheless, Fabien Langenegger’s dendrochronological study (see Annex III) was able to determine that the sessile oak used in the construction of the ship could originate from an area north of the Alps. He specified that the average curve of the samples matches that of the Burgundy region (South of Paris, France).

2.2.2. The wood: its characteristics and contribution to dating using dendrochronology 2.2.2.1. The wood and its origin

The characteristics of the oak used for the construction call for some remarks:

The species used in the construction of the ship: Samples taken from the planks, frames and the longitudinal timbers of the maststep arrangement reveal a sessile oak structure (Quercus petrae). However, it should be noted that a larger sampling has yet to be carried out and therefore the potential presence of secondary species of wood in the construction of the hull cannot be discounted. Three other species were also identified: beech, Fagus sylvatica (ceiling and pump tube), chestnut, Catanea sativa (pump well). Finally, the buttresses of the main-mast step were made of juniper, Juniperus communis.

Most of the timbers contain heartwood. Part of the timbers used in the framing was made of logs that were not split but simply squared. This suggests the oaks employed were generally of a small diameter. It probably also explains the small widths of the planks, with an average of only 18 centimeters, compared for example to the average of 34 centimeters on the Red Bay wreck (LOEWEN, 2007, 111). Relatedly, the growth rings are particularly thin and plentiful; the distance between them is about one millimeter. In fact, the timbers used for the planking or the framing are often more than 100 years old. The small diameter of the oak logs used is therefore not due to the youth of the trees, but rather to their slow growth.

The origin of the wood: Sessile oak is a species of the fagaceae family widely distributed in the temperate regions of the northern hemisphere. It is widespread throughout Western Europe, ranging from around 60º longitudein the

Figure 22. European distribution of sessile oak (Quercus petrae) (Photo licensing: Wikipedia).

28

Portraits of the Mortella wrecks

Figure 23. 300-year-old sessile oak from Burgundy (Photo licensing: Arnaud 25).

of oak as the predominant species is an indication of ‘Atlantic construction’ (LOEWEN, 2001). However, careful observation of the types of oak used in the construction of ships in the ‘Atlantic’ and ‘Mediterranean’ areas calls for caution. First, it should be noted that the Portuguese authors who have defined an ‘Ibero-Atlantic’ shipbuilding model, advocated the use of pine for the planking (OLIVEIRA, 1570, 63, 64 and 140, 141) and, in accordance with this recommendation, Portuguese wrecks have been found with pine planking. Such was the case with Nossa Senhora dos Martires, which was shipwrecked in Portugal in 1606 and built using umbrella pine (Pinus pinea) (CASTRO, 2001, 400).

This can be explained by the very nature of the species, whose growth is naturally slow and is associated with an unfavorable natural context, that is to say climatic conditions and/or a soil type unfavorable for growth. 2.2.2.2. Its use in shipbuilding The outstanding qualities of oak, both in terms of its physical and mechanical properties alongside its durability, are undoubtedly what have earned it its place as one of the most sought-after species throughout the history of shipbuilding. It is a hard and dense wood whose density (700 to 800 kg/m3) ensures a great strength but also due to its high tannin content it provides good resistance to an attack of micro-organisms or xylophages organisms. There is a wide variety of oaks, about 300 of them of the genus quercus. But there are about ten European varieties that have mainly been used in navies, coming from two main categories, the one with deciduous foliage (Red oak, Turkey oak, Downy oak, Pyrenean oak, English oak, Sessile oak, etc.) and the one with persistent foliage, which is found more particularly in the Mediterranean region (Holm oak, kermes oak, cork oak, etc.).

In the Mediterranean space, it is clear that oak remained the species unanimously used in the construction of frameworks2 (Table 5). For the planking, the situation is more varied, but the predominance of oak seems to persist. However, the presence of pine has sometimes been observed. Such was the case with the planking of The Mediterranean preference for oak, at least for the construction of hull, is stated in Provence in a text from the Middle Ages dated 1318. This is an offer of charters from the city of Marseille to the Count of Clermont for a crusade project entitled ‘Informationes Civitatis Massiliae pro passagio trasmarino’ in which we can read: ‘Naves debant esse de quercus, sive de robore faxate usque ad copertam de media’: ‘the ships must be made of heated oak to the middle deck. ‘(FOURQUIN, 1990, 235).

2

In terms of archaeological research, can the use of a particular species be considered as an indicator of the construction origin? Brad Loewen believes that the use 29

The Mortella III Wreck Table 5. Species used in some wrecks of Mediterranean building tradition Species used in some wrecks of Mediterranean building tradition

Framework

Planking

Cala Culip (Spain)

XIVème s.

Holm oak (Quercus ilex)

 

Mortella III (France)

XVIème s.

Sessile oak (Quercus petrae)

Sessile oak (Quercus petrae)

Villefranche-sur-Mer (France)

Deciduous oak

Umbrela pine, Aleppo pine and Deciduous oak

Calvi I (France)

Deciduous oak

Deciduous oak

Delta II (Spain)

Oak

Oak

Yassi Ada XVIème (Turkey)

 

Turkey oak (Quercus cerri)

Sveti Pavao, Mljet (Croatia)

English oak (Quercus robur)

Outer planking Umbrela pine (Pinus pinea).

the Villefranche-sur-Mer wreck3, or the outer planking of the Sveti Pavao wreck (Mljet, Croatia) which nevertheless remains a special case because of its double planking.4

In the 16th century, builders preferred local wood, which seemed natural. Authors such as Brad Loewen (LOEWEN, 2007, 287-292) and Michael Barkham (BARKHAM, 1985, 211-276) have shown how shipbuilding in the Basque Country during the 16th century was organized in the immediate vicinity of shipyards, with the help of state subsidies (LOEWEN, 2007, 287). The management of local Basque forests was entirely oriented towards shipbuilding, in particular growing trees adapted for the production of suitable timbers. B. Loewen writes on this subject: ‘The notion of training trees to obtain the desired shapes is certainly one of the most original hypotheses that emerged from the Red Bay study.’ (LOEWEN 2007, 191)

The type of oak used in some cases may indicate the region of construction. The use of holm oak, also called Green oak (Quercus ilex), in the construction of the framework of the wreck of Culip VI (14th century) is such an example, with this species being typically Mediterranean. Archaeology has offered many other examples, such as the Ottoman wreck of Yassi Ada (16th century) built with Turkish oak (Quercus cerri). Similarly Cork oak was (Quercus suber) recommended by the Portuguese authors for the construction of frames. In this sense, the Nossa Senhora dos Martires wreck with its Cork oak frame and pine planking perfectly meet the theoretical model adjusting to regional species.

What about the wood supply to the Genoese builders, to which the Mortella III wreck now seems to be linked? In modern times Italy possessed important forest resources for its shipbuilding: the regions bordering the Alps and Liguria provided oak. Further south, large forest estates existed, mainly in the Gargano and Calabria regions. According to R. Barker, there were forest resources used for shipbuilding in Italy which utilized a management model similar to that of the Basque Country: ‘But the practice of pruning and training forest trees is not unknown elsewhere. It clearly occurred in Italy from the sixteenth century at latest’ (BARKER, 1998). Furio Ciciliot, on the other hand describes this model of forest management in Genoa during modern times using a variety of Genoese manuscripts written between the 13th and 17th centuries (CICILIOT, 1999).

Still in Portugal, the wreck of Arade 1 is notable. Although not following the recommendations of contemporary authors, this ship was built using a regional species that helps to reveal its origin. Vanessa Loureiro concurs: ‘The integration of the Arade 1 wreck into the Ibero-Atlantic family is based, on the one hand, on the concordance between its design process and the standards set out in the 16th and 17th century treatises and, on the other hand, on the geographical origin of the species used in its construction. Although the wreck is the opposite of the recommendations in the Portuguese shipbuilding treatises—which indicate that the structure of the ships should be made of Cork oak and planking with pine— the Arade 1 was built mainly of Portuguese oak (Quercus faginea), a species characteristic of the Iberian Peninsula.’ (LOUREIRO, 2009).

But it appears that in the 16th century, Italian forest resources were in crisis. The importance of shipbuilding and the production of new ships, in the Italian States had continued to increase for centuries and led to the overexploitation of forests and the depletion of raw materials. Fernand Braudel describes this phenomenon in his book La Méditerranée : ‘Didn’t the ship, that was one of the main drivers of deforestation, finally fall victim to this process? One day, the Calabrian forests, or those of Mount Gargano, ceased to be exploitable for the major shipyards in Dubrovnik or the coast near Naples.’ (BRAUDEL, 1977). Local supply therefore transitioned towards an import supply. The dendrochronological study

In Villefranche, although the presence of deciduous oak was documented, the planking is mainly composed of coniferous, pinion and Aleppo pine (GUEROUT et al., 1989, 63). 4 The planking of the Sveti Pavao wreck is curios as it consists of two woody layers that form a double hull (BELTRAME, 2014, 48). This double planking is currently—to our knowledge—the only one documented for the period in the Mediterranean. Carlo Beltrame notes that this technique was used in modern Dutch shipbuilding in modern times. However, it can be assumed that its origin is Mediterranean since it is found in ancient construction, the case of the Roman wreck of Giens (Var, France) being a famous example (GIANFROTTA, POMEY, 268270). 3

30

Portraits of the Mortella wrecks it did not reach a dating. A second dendrochronological study done in 2013 by Fabien Langenegger (archaeologist and dendrochronologist at the Office of Heritage and Archaeology of the Canton of Neuchâtel, OPAN, Switzerland) succeeded. The growth rings of twelve Sessile oak samples taken from the wreck frames were central to the dating. The particularly slow growth of this species (1 to 2 mm per year) made it possible to date 3074 rings, creating a 206-year date range between 1310 and 1517 (see Annex III). Dating when the trees used for the construction were felled was reached using the second futtock M20 frame (twentieth visible frame from the aft) which had four sapwood rings. The growth width of the last rings of this sample was particularly high (between 4 to 5 mm). The preserved width of the sapwood was 2.14 cm. Considering that a normal width reaches 3 to 3.5 cm, it has be estimated that 3 sapwood rings would be missing up to the cambium. Three to six missing rings were eventually noted in this sample. On the basis of this calculation, the dendrochronological study shows that the tree used to build this frame was felled in the first third of the 16th century. The study concluded: ‘Thus, the beginning of construction could begin at the earliest between 1517 and 1520.’

of the Mortella III shows that it was built with wood coming from the north of the Alps, probably from Burgundy. It is perhaps, therefore, an example of this phenomenon. Braudel goes on to underline the political effects of this economic dependence: ‘the Christian Mediterranean in the ‘Ponant’ (Western), … would lose the domination of the inland sea where the British and Dutch would start to dictate their conditions.’ In fact, from the 17th century onwards, Genoese construction would fade in favor of the Dutch construction, which would have a total hold on the Italian city. 2.2.2.3. Woodworking and tool marks Once sawn and processed, the pieces of wood were actively shaped to obtain their final form. This finishing work was carried out with a sharp tool like an adze. The marks left on the wood by this tool are omnipresent. The woodwork with the adze has been executed with different degrees of care, depending on the needs: rough on the faces of the timbers, the adze marks are clearly visible, each shot leaving a characteristic mark. On the other hand, the shaping is done much more finely on the scarfs areas, which are carefully smoothed. In this case the marks of the tool are sometime visible, but much more difficult to detect.

A first dendrochronological study was carried out in 2011 by Marta Dominguez, a Spanish dendrochronologist from the Nederlands Centrum Vor Dendrochronologie. However,

Finally, it should be stressed that the date the trees used for the construction of the ship were felled must be clearly distinguished from the date of her sinking, which may have taken place years later. But it allows us to get close enough to estimate that it probably took place in the first half of the 16th century. As we have seen, this chronology corresponds to the results of the artefact study. However, it should be underlined that the period during which the vessel was constructed is more relevant for the architectural study than the date of her sinking. The sinking date does, however, remain to be key information for the identification of the vessel and the historical event which caused its loss.

Figure 24. The adze, an essential cutting tool in the shaping of the timbers (Photo: Arnaud Cazenave de la Roche).

Figure 25. Cross-section of the M20 futtock that gave an absolute dating (Photo: Fabien Langenegger).

There are also marks of sawing carried out with a pit-saw. They are mainly visible on the inner part of the planks, demonstrated by vertical streaks characteristic of sawing. On the other hand, their external faces were shaped using an adze. Finally, we can see frequent scars left by a caulking iron on the edge of the planking. 2.2.2.4. Dendrochronological study

31

The Mortella III Wreck 2.3. General layout of the Mortella III site and context of the shipwreck

• In 2010, determining whether or not a second wreck was located under the ballast of tumulus B was included in the excavation campaign’s list of objectives. An SB/10 survey zone was located at the foot of the tumulus B, on its northern slope. Covering an area of 2.5 m², its objective was to verify the nature of the remains under the gravel, and to verify if whether or not there was a second wreck. Several structures were uncovered, but it was not clear whether they were the remains linked to those of tumulus A or not. (CAZENAVE DE LA ROCHE, 2010, 41). • In 2012, the decision to excavate on the top of tumulus B (AF12/B) made it possible to then identify structures that could be linked to tumulus A. It was therefore confirmed that we were not dealing with a second wreck, but with the remains of the wreck’s port side portion which was located below tumulus A. The arrangement of the frame parts in AF12/B showed that these timbers were connected to a part of the hull located in tumulus A before midship. These remains were confirmed to have belonged to the forward port side of the ship upon observing the scarfs. The first futtocks were fixed on the fore faces of both the floor timbers and second futtocks, indicating that this ensemble was located before the master-frame.

2.3.1. General layout of the remains of the hull (fig.26) The Mortella III site comprises two archaeological areas formed around two separate tumuli thirty meters apart at the southwestern part and meeting at its northeastern part. The first, called tumulus A, extends over an area of about 20 x 10 m along a north-east/south-west axis. The second, called tumulus B, extends over an area of about 12 x 7 m along a generally North-West / South-East axis. Tumuli A and B are formed by archaeological remains, mainly wood, covered by a layer of ballast gravel more than one meter thick at the top of tumulus A and 80 cm at the top of tumulus B. The calculations made by Guillaume Martins show that we are dealing with a large volume of ballast, estimated at a little less than 100 m3 and weighing a little less than 150 tons. As soon as the site was discovered in October 2006, the dualistic organization of the Mortella III site raised a question, only answerable once the excavation in 2012 had uncovered a portion of tumulus B. This then suggested the presence of a single wreck fractured longitudinally and whose two sides separated to form two distinct areas of archaeological remains. We will provide further detail on this aspect later, as it gives the site its particularity.

To complete the description of the remains, it should be noted that the longitudinal rupture along the hull occurred at two levels:

Under tumulus A, we found the remains of a carvel built, shipwrecked hull, oriented along axis 225°/45°. The heel of the keel is located in the northern part, the fore end of the keel in the south. The aft area of the wreck that emerges from the sediment was first observed during the survey conducted by DRASSM in 2007. In the north, beyond the keel heel, the degraded remains of the rudder were uncovered and were subsequently studied during the 2015 excavation campaign.

• The first was along the assembly of the floor-timbers to the first-futtocks of the port side, which resulted in the separation of the hull structures and the formation of tumuli A and B. • A second fracture line symmetrical to the first was detected in 2007 with the discovery of the first structures of tumulus A. It runs along the heads of the floor timbers on the starboard side of the wreck. They could be observed during the further excavations on all the frames from aft to the fore. Unlike the rupture observed on the port side, on starboard the frames broke but did not separate.

The longitudinal axis formed by the keel and the keelson has been entirely preserved. The starboard side of the wreck is preserved up to a part of the second futtock, where charred ends clearly indicate that a severe fire consumed the ship before it sank. On the port side only the ends of the floor timbers, sometimes broken, have been preserved. The central part of the wreck is covered by a layer of gravel and ballast stones that sometimes is thicker than one meter.

To conclude this general presentation of the architectural remains of the Mortella III site, we should add that the 2013 excavation of the fore part of the wreck located under tumulus A allowed us to clear the fore end of the keel and uncover the remains of the port planking strakes extremities. However, no trace of the bow post could be located.

Under tumulus B, as we have pointed out, the excavation of the remains under the ballast enabled us to understand the dual organization of the site. The steps were as follows:

2.3.2. The context surrounding the sinking At this stage of our research, the way in which the remains of the Mortella III are organized only allow to propose hypothesis on the scenario surrounding the ship’s sinking. Envisioning the sequence of events that led the Mortella III ship to rest fractured in half 38 meters below sea level, may be interesting, it however would have no direct repercussions on any factual analysis of the way in which the ship was built.

• In 2007, during the assessment conducted by DRASSM, a survey called SF, conducted between the two tumuli A and B, was negative. It showed the sterile nature of the sediment and confirmed the presence of two very distinct archaeological areas (DRASSM, 2008). 32

33 Portraits of the Mortella wrecks

Figure 26. General planimetry of the Mortella III site (Illustration: Arnaud Cazenave de la Roche).

The Mortella III Wreck

34 Figure 27. Remains uncovered under tumulus A. Extract from the photomosaic (Photo: Christoph Gerigk).

Portraits of the Mortella wrecks

Figure 28. Top: tumulus B (extract from the mosaic photo). Bottom: front part of tumulus A (crutches and fore end of the keel) (Photos: Christoph Gerigk).

With these clarifications in mind, it will therefore be sufficient to set out a number of elements that describe the context of the sinking.

two sections. Only an extremely powerful shock wave could have caused the serial rupture of the floor timbers heads on both port and starboard. On starboard, it was not the floor-timbers/first futtock scarfs that gave way, but the floor-timbers themselves, made up of oak pieces averaging 17 × 17 cm in section, fractured a

1. The most striking feature of the Mortella III site is undoubtedly the separation of the hull structures into 35

The Mortella III Wreck little before their assembly. To port, while it is true that the scarfs were broken in central part of the wreck, we were surprised to discover the heads of floor-timbers still jointed with their first-futtock under the tumulus B. This undoubtedly reveals the frames robust build. Their timbers were fixed together by two crossing iron nails and then held in place by two clamps. As a result, the ship’s hull literally ‘ripped’ lengthwise to port and fractured to starboard. In both cases, this generally took place just before the floor-timbers/first-futtock connection. 2. The traces of calcination on the second-futtocks are generally at the same height both on the port and starboard sides. This suggests that the ship remained balanced in the water as it burned and that it probably did not list before sinking. If we accept the historical hypothesis as set out in Chapter VI, we therefore must consider texts that suggest a shipwreck in the total absence of wind. 3. The ship burned to a relatively low part of its hull (about the middle of the second futtock) reaching a level below the overlop. The reconstruction of the masterframe shape indicates that, at the time of the sinking, the hull’s consumption line was only 2 meters above the upper face of the keel. 4. When it sunk, the entire hull was covered by the ballast gravel in the ship’s hold. An estimate of its volume suggests that it weighed about 150 tons. This mass of ballast was not dispersed by the sinking, and it divided into two groups, the first—the most important— covering the starboard part of the wreck and a small part of its port side to the end of the floor-timbers (tumulus A). The second one covered the remains of the ship’s port hull, between the first and half of the second futtock (tumulus B).

since the first intervention in 2007 until the last excavation campaign in 2015 (AF07 to AF15). In total, six campaigns carried out in 2007, 2010, 2012, 2013, 2014 and 2015. In 2007 focus was placed on the aft part of the wreck, along the first ten meters of the keel, and as result it was decided to excavate this area first, with the aim of uncovering and studying the central part of the tumulus A remains. The objectives determining this choice were: • to uncover the central frames and study their construction, looking at the scarfs and fastening methods, etc., • to reach and study the main maststep, • to identify, if possible, the master-frame and restore its design and shape. The 2010 campaign reached the pump well and discovered the remains of the bilge pump, but the structures of the main maststep were not uncovered until 2012. As mentioned, another priority was to identify the remains under the tumulus B. This led to an excavation of its middle area in 2012 which reached the conclusion that we were dealing with a single wreck. In 2013, the excavation strategy was working on two fronts: • The continuation of the study of the midship area. This allowed us to identify the master-frame and to carry out a complete study of the maststep. • The excavation of the fore part of the wreck, aiming to uncover the fore end of the keel, to measure its total length, and to search for the stern post. To recover the precise shape of the master-frame in 2014 we dismantled it and studied it on land. A second objective that year was to observe the keel through a gap in the planking.

2.4. Defining the research programme: research axis, methodology and excavation strategy The main issues addressed in this book are part of a research line on shipbuilding and naval architecture, the details of which were first specified in 2009. However, when defining the objectives of the excavation project, two other lines of inquiry on the fringe of the present subject were also set out: the first concerns artillery, the presence and subsequent importance of which has required further investigation in its own right. This one was done by Max Guérout (GRAN). Another research area focused upon the artefacts and material culture of the period, with particular focus on ceramics, which was studied by Franck Allegrini Simonetti (Archaeology Department, Government of Corsica, CdC).

In 2015, attention was focused on the fore end of the keel and the area between tumuli A and B. During this campaign the remains of the rudder were discovered and studied. As mentioned in the introduction, the excavation of the Mortella III wreck has not been fully completed. The fore tail-frames of the ship still require excavation. Locating the master frame’s port side of the under tumulus B would also be useful. Finally, it remains necessary to check if any remains of the stern post at the southern end of the tumulus B can be found. The next excavation campaign planned to continue the work is scheduled for 2019.

2.4.1. Methodology and excavation strategy

2.4.1.2. Means and excavation methodology

2.4.1.1. Objectives and excavation strategy

Diving. The study of the remains was carried out by scuba diving with air, with decompression stops using pure oxygen. The time available for underwater archaeological activity—excluding decompression – was 30 minutes for the first dive and 20 minutes for the second, for a total time

The research focus on shipbuilding has dictated the excavation strategy and how the areas of excavation were programmed. Fig. 29 shows those that have been set up 36

Portraits of the Mortella wrecks

Figure 29. Excavation areas of the Mortella III site from 2007 to 2015. (Illustration: Arnaud Cazenave de la Roche).

underwater of 50 minutes per diver per day. These low daily working times required large teams, high turnover and complex logistics and security organization. These conditions have made it difficult to study the remains and in particular, they prohibited the dismantling of structures to study on land, with few exceptions.

with the baseline and was then used for each following campaign. It was made of two 350 × 350 cm squares levelled with bubble levels and was used for multiple tasks. The movement of a sliding ruler equipped with a trolley allowed for: • Accurately positioning the features located inside the frames at a right angle • The addition of a vertical ruler on the trolley was also used to produce the longitudinal and transverse recording (fig. 31) • A plate fixed on the trolley was used to create photo mosaics inside the grid.

The removal of sediment and ballast gravel. Airlift and water-dredge systems were used to uncover the framework. The large amount of ballast gravel (about one meter deep at the top of the A and B tumuli), and the high density of the material necessitated the use of a powerful airlift device in the sterile gravel layers, which was fed from the surface by a low-pressure compressor (15 bars). In the muddy areas and on the remains, two water-dredges powered by high-pressure pumps were used.

All the measurements identified resulted in the production of a site map, assisted by the use of Adobe Illustrator software and updated at the end of each campaign.

The topographic survey. The structures were recorded using a geodetic network of survey points. A baseline was drawn along the keel along a global axis 45º – 225. The survey points and first features were positioned by trilateration techniques. A grid composed of aluminum templates fixed on sliding feet was set up in correspondence

The photomosaic. Roughly half of the time, the water was transparent with good visibility. This allowed us to set up a photomosaic of the site which partially compensated for the low working times at the bottom, and certainly helped to understand how the remains were organized. 37

The Mortella III Wreck Photomosaic is a technical photographic document produced by a specialized professional photographer, Christoph Gerigk, using a large number of high-definition photographs (several hundred) taken vertically from the site and assembled afterwards. To convert this image into a working document to scale, about thirty graduate scales were placed on the bottom before the photos were taken, allowing the images to be adjusted and to account for the curvature of the lens during the post-processing phase. Campaign after campaign, secondary photomosaics, those of the excavated surfaces, were embedded in the primary image. In this way, it was possible to obtain a complete image of the wreck, in such a way that would have been impossible to produce in the field, given each excavation area was systematically reburied after intervention. The high definition of this photographic document enabled very interesting background work, observing structural details and objects that went unnoticed while diving. A reproduction of the photomosaic in A3 format is visible in Annex III. Figs. 27 and 28 are excerpts from this. It should be emphasized that photomosaic is a 2D document. During the excavation of the Mortella III wreck, 3D image processing software, such as Agisoft Photoscan, has emerged and has made a considerable contribution to archaeological methods and techniques. Indeed, these computer programs now allow for the creation of photogrammetry by processing a simple image mosaic. One of the advantages of this technique is to provide 3D images to scale. This progress in scientific imaging

Figure 30. Sediment clearance on the remains of the hull (Photo: Christoph Gerigk).

Figure 31. Transverse cross-section survey at mid-ship (Photo: Christoph Gerigk).

38

Portraits of the Mortella wrecks

Figure 32. The grid square in place on the remains of tumulus A (Photo: Christoph Gerigk).

Figure 33. Scales placed on the bottom to allow calibration of the photomosaic (Photo: Christoph Gerigk).

39

The Mortella III Wreck

Figure 34. 3D survey of the heel of the keel (Photo: Bérenger Debrand).

Figure 35. Photography, drawing and information processing in the excavation laboratory (Photo: Arnaud Cazenave de la Roche).

marks a milestone in the methodology of archaeological surveying in that it eliminates the need for a whole series of manual measurements. It therefore saves precious time, especially in the case of a deep site like ours. We have used this 3D imaging since 2014 to visually reconstruct the main part of the framework and other structures as the rudder, for example.

1. The sketches and measurements were recorded underwater on polyethylene millimetric velum sheets. The drawings were then cleaned up on Microsoft Illustrator. The artefact inventories were managed on a Microsoft Access database. • Photography was used underwater as a recording tool and was also used in the excavation laboratory to help create the artefact inventory sheets. • Video recordings were made throughout the campaigns. In particular, they were used to keep a record of the different phases of the excavation. The recordings

Recording information. Classically, we have used drawing, photography and video as means of recording information during the excavation program. 40

Portraits of the Mortella wrecks

Figure 36. Video recording by a R.O.V. and manual recording on the bottom (Photo: Christoph Gerigk).

were made using GoPro cameras fixed both to tripods placed on the site and by means of an R.O.V. (Remote Operated Vehicle) operated by Charles Pinelli (Les Amis des Agriate).

likely be permanently lost, have been brought to the surface. Schematically, the operating chain was as follows:

2.4.2. Conservation and protection of the site

(a) The selected artefacts were brought to the surface and taken on-board by a preventive conservation officer who assigned them an inventory number and recorded their first observations on an inventory form. Then these artefacts were immersed in fresh water tanks and sorted according to their nature (ceramic, lithic, etc.)

2.4.2.1. Preventive conservation of the artefacts The archaeological material is scarce on the Mortella III site and its extraction has been targeted. Only artefacts likely to provide important chronological information, or remain of archaeological or historical value and would

Figure 37. Covering the excavation areas with a geotextile fabric (a.) in turn covered by a layer of sediment (b.) (Photo: Christoph Gerigk).

41

The Mortella III Wreck (b) Once returned to the port, the tanks and inventory forms were transferred to the preventive conservation laboratory and taken over by those responsible. From then on, the artefacts were observed, drawn and photographed. They were then stored away from light, heat, temperature variations and submitted to a controlled dechlorination process. The inventory forms were then computerized. This preventive conservation protocol was developed in collaboration and under the supervision of the specialized laboratory A-Corros (Arles, France), directed by JeanBernard Memet and Philippe de Viviès. 2.4.2.2. Conservation of the remains on the site The search program was designed to be as non-intrusive as possible. The areas excavated during each campaign (between 20 and 35 m2, per campaign) were covered with a sheet of geotextile material on which a layer of sediment was then spread out in order to recreate an anaerobic environment and allowing for the conservation of the hull remains uncovered during the excavation.

42

3 The hull of the Mortella III wreck and its construction method 3.1. The transversal framework: organization, dimensions, morphology and scarfs methods

adjusted the irregularities of the futtocks by using wooden wedges, shims and blocks, and by using heel wooden pieces to properly sit the floor-timbers on the keel. This building method used for the transverse structure gives it a heterogeneous aspect which leaves the feeling of an apparent carelessness during its construction.

3.1.1. General layout of the transverse framing The transverse framework of the ship was made up of a series of frames composed of three timbers on each side: a frame-timber resting on the keel which at one end was attached to a first-futtock and whose other end was joined to a part of a second-futtock. The second-futtocks are only visible in the central part of the wreck. Their outer ends were charred on both port and starboard sides, bearing witness to the violent fire which preceded the sinking.

This constructive characteristic seems to have been observed in a rather similar way on the wreck of Villefranche-sur-Mer as the authors of the ‘Le navire génois de Villefranche’ wrote: ‘The irregular dimensions of the framework seem to meet a major requirement: to make maximum use of wood supplies without meeting very strict dimensional standards. When the sizes of a piece seem too small, shipbuilders put in wedges with apparently little concern to get a composite piece with precise scarfs. It should be noted that the sometimes rudimentary nature of the structure of the internal framework is not specific to first and secondfuttocks, but is also found in the floor-timbers and thirdfuttocks.’ (GUÉROUT, RIETH, GASSEND, 1989, 47).

Most of the excavation programme between 2007 and 2015 was focused under the tumulus A, where most part of the framework was observed. The view we have of it is therefore mainly that of its starboard side. Under tumulus B only a portion of the framework was observed during the 2012 excavation. This section was probably located in the fore tailframes area. We will come back to this set. The species used for the entire transverse structure was sessile oak. Its main dimensions have been reported in two summary tables No. 9 and 10. These tables contain the main measurements of the 42 frames found in tumulus A. In order to represent them according to their location, we have chosen to average the measurements of the timbers after splitting them in three large groups delimited by the aft tailframes, fore tailframes and between tailframes areas.

This remark can literally be applied to the wreck of Mortella III and is one of the many common points shared by these two wrecks. 3.1.1.1. Definition of the frames groups As with the wreck of Red Bay or Villefranche-sur-Mer, it is not possible to know the precise rule used by the builder to determine the location of the tailframes (LOEWEN, 2007, 58). The Iberian treatises, for example, that of Father Oliveira, gave indications as to where on the keel they should be placed. In Spain, it was only in 1618 that a precise rule was written in this way for the first time. Article 20 of the 1618 Spanish Ordinances stated that the fore tailframes should be one ‘codo’ on the fore part of the point, located at a quarter of the total length of the keel when starting from the fore. Then, the aft tailframe would be placed two ‘codos’ on the fore part of the point located at a quarter of the total length of the keel starting from the aft.2

Each frame was fastened to the keel and the keelson by a circular iron bolt of 30 mm diameter passing through these pieces and the center of the floor-timber.1 Before describing the timbers that made up the transverse structure of the ship, a general remark on its construction is useful. A striking characteristic was the irregularity in their dimensions and shapes. This should be said at the outset because the descriptions accompanying the following pages are often based on average dimensions and do not necessarily adequately reflect their disparities. For example, whereas the master floor-timber was made up of a timber with rather modest sizes (15 cm moulded, and 18 cm sided), three frames further forward, there was a floortimber (V30) whose dimensions were 27 cm moulded and 19 cm sided, at the level of the sister keelsons (fig.40).

If this rule is applied to Mortella III wreck, the position of the tailframes would be as follows: Position of the fore tailframe:

These differences were often significant and resulted in an inhomogeneous set where the shipwright constantly

(25 ÷ 4) × 3] + 0.5747 = 19.32 m from the aft end of the keel. In 1590, the adoption of the ‘codo de ribera’ brought it to 1/3 of one ‘vara’ and thirty-two hundredths, or 57.47 cm.

1 All the floor-timbers whose connection to the keel could be observed (ie 34 out of 42) were bolted.

2

43

The Mortella III Wreck

Figure 38. a. The frame of the starboard side at the level of the beam. Two types of wedges: Left (b.), ‘duck head’ shaped fitted between first-futtock G28 and the foot-wale S1. Right (c.) ‘whistle shaped’ fitted between G27b (already in place on the picture). (Photo: Christoph Gerigk).

Figure 39. Framing at mid-ship (Photo: Christoph Gerigk).

44

The hull of the Mortella III wreck and its construction method

Figure 40. The floor-timber V30 (Photo: Christoph Gerigk).

Position of the aft tailframe:

The master-frame had two first-futtocks attached to its aft and fore faces that marked an inversion of the assembly sequence. This floor-timber / first and second-futtocks sequence was organized as follows: the first-futtocks were attached to the aft faces of the floor-timbers and second-futtocks in the area located after the master, and conversely, they were arranged on their fore faces forward of the master.

(25 ÷ 4) + (2 × 0.5747) = 7.40 m from the aft end of the keel. If we assigned the number 0 to the master-frame, as shipbuilders did in the 16th century, the aft tailframe would be the 19th and the fore one the 17th from the master. The distance between them would be 12 meters, so it would cover a little less than half of the keel. In this case, the aft tailframe would be the M9 frame and the fore tailframe would be the M44 frame. The fore tailframe has been established by deduction, because the area where it is located remains to be excavated.

In other words, the Mortella III ship was built according to a traditional constructive layout, with the floor-timbers facing the master-frame (fig. 43): this type of sequencing is found both on wrecks of Atlantic construction tradition such as Cattewater (early 16th century) or Red Bay (1565) and on wrecks of Mediterranean constructive tradition such as Calvi 1 (last third 16th century). It seems that the only exception to this rule highlighted by the archaeology of the period is the case of the Villefranche-sur-Mer wreck whose first-futtocks were facing the master-frame.3

Of course, this example is relative, since this rule was applied only to Spanish ships at the end of the 16th and beginning of the 17th century. It is therefore not possible to deduce from it a precise location of the tailframes of the Mortella III ship. However, it is a useful indicator that has been used in this work to mark a division between the frames of the central part of the wreck and those of the aft and fore areas.

The intervals between the frames (fig.42) The intervals between the frames was on average 19 cm for the central group, which in view of their moulded sizes gives an average of one room for a space. In this area,

3.1.1.2. Sequencing of the frame components, room and space

The authors of ‘l’épave de Villefranche-sur-Mer’ wrote on this subject: ‘However, the wreck of Villefranche has a particular characteristic. Indeed, … [the first-futtocks] are located facing the master-frame… However, the usual practice is to arrange the first-futtocks forward the floor-timbers making them facing the master-frame. We found no mention of a layout of the frame similar to that observed at the Villefranche site.’ (GUÉROUT, RIETH, GASSEND, 1989, 39).

3

The master-frame was identified within the central group. This is the M27 frame – the twenty-seventh from the aft, which passed under the main-mast step. It has considerably helped in understanding the shape of the ship and the study of its design is presented in Chapter IV. 45

The Mortella III Wreck

46 Figure 41. Planimetry of the central and fore areas (Illustration: Arnaud Cazenave de la Roche).

The hull of the Mortella III wreck and its construction method the overlap between the floor-timbers and the first-futtocks formed an almost continuous wooden surface where the spaces between the pieces were scarce (intervals of 6 cm on average between 7 frames out of the 19 of the group). A decrease of the intervals in the groups of fore and aft frames was also observed: The space between each frame was 15 cm on average. Similarly, there was almost no gap between the floor-timbers/first-futtock connections, in the fore area, while at the aft the absence of a first-futtock prevented observation.

III wreck, its value increased fore and aft: 42 cm on the fore frame group and even more so aft, at 45 cm. This increase in the distance between each frame on the Red Bay wreck was explained because of the increasing space required for the canting of the first-futtocks (BERNIER et al, 2007, vol. III, 62). In the case of the Villefranche-sur-Mer wreck, the distance between the first-futtocks varied between 17 and 35 cm. Whereas it was between 5 and 22 cm at the crossing between the first and second-futtocks. At this level, however, it was sometimes non-existent.

The room and space:

3.1.1.3. The frames of the aft part of the hull

We now turn to the distance between the center of each floor-timber. The advantage of this alternative method of evaluating these intervals is to grasp the variation distance between the frames whilst taking in account their different sizes. As we have seen, this space was decreasing forward and backward, but the measurement of the room and space shows that the distance between the center of each frame did not vary much: It was on average 34 cm in the central and aft area of the ship and 35 cm in the group of fore frames. Therefore, it can be deduced that the reduction in the intervals, that is to say the distance between the frames observed towards the fore and aft, was due to an increase in their moulded sizes.

The first traces of frames appear on the keel 4.80 meters from its heel where the remains of the floor V1 were located. The crutches that were located between the heel of the keel and V1 are no longer visible. There are only bolt traces left, of which we have counted 9. The first first-futtock that could be observed appeared from the starboard M7 frame, at a distance of 6.60 meters from the heel of the keel. Before it, the remains of the frame were composed only of floors whose upper face and ends had been degraded by woodborer organisms. The lacunar ends of these floor-timbers followed the lines established by the remains of the port and starboard planks which were separated by only 30 centimeters at the M1 frame to 80 centimeters at the M7 frame (Fig. 41).

Fairly similar values were recorded at Red Bay, where the room and space was 36.8 cm in the central part of the wreck. However, unlike what was observed on the Mortella

Figure 42. Room and space and intervals at mid-ship (Photo: Christoph Gerigk).

47

The Mortella III Wreck Table 6. Mortella III – Room and space of the framing (tumulus A -Starboard) Frames

Sequential distance

Room & Space (1)

Sequential distance (3)

Intervals (2)

Fore area

measures in meters M 7P – M 8P

23,81

0,35

23,50

23,66

0,16

M 6P – M 7P

23,46

0,39

23,12

23,33

0,21

M 5P – M 6P

23,07

0,39

22,74

22,91

0,17

M 4P – M 5P

22,68

0,26

22,51

22,53

0,02

M 3P – M4P

22,42

0,37

22,10

22,31

0,18

M 2P – M3P

22,05

0,33

21,78

21,93

0,15

M 1P – M2P

21,72

0,34

21,42

21,59

0,17

M34 – M 1P

21,38

5,28

15,99

21,27

5,28

Aft area

Mid-ship area

Average

 

0,35

 

 

0,15

M33 – M34

16,10

0,36

15,78

15,99

0,21

M32 – M33

15,74

0,37

15,42

15,62

0,20

M31 – M32

15,37

0,40

15,05

15,28

0,23

M30 – M31

14,97

0,40

14,68

14,88

0,20

M29 – M30

14,57

0,34

14,29

14,43

0,14

M28 – M29

14,24

0,32

13,97

14,12

0,15

M27 – M28

13,92

0,35

13,62

13,83

0,21

M26 – M27

13,57

0,31

13,33

13,48

0,15

M25 – M26

13,26

0,28

13,04

13,17

0,14

M24 – M25

12,98

0,26

12,79

12,89

0,10

M23 – M24

12,72

0,27

12,48

12,63

0,16

M22 – M23

12,45

0,39

12,10

12,33

0,23

M21 – M22

12,07

0,35

11,79

12,33

0,54

M20 – M21

11,72

0,26

11,50

11,64

0,14

M19 – M20

11,46

0,29

11,21

11,34

0,13

M18 -M19

11,17

0,34

10,87

11,04

0,17

M17 – M18

10,83

0,34

10,55

10,76

0,21

M16 – M17

10,49

0,37

10,24

10,40

0,16

M15 – M16

10,12

0,42

9,77

9,97

0,20

M14 – M15

9,70

0,36

9,40

9,61

0,21

M13 – M14

9,34

0,37

9,04

9,24

0,20

M12 – M13

8,97

0,35

8,71

8,86

0,15

M11 – M12

8,62

0,42

8,31

8,49

0,18

M10 – M11

8,20

0,34

7,95

8,06

0,11

M9 – M10

7,86

0,37

7,58

7,77

0,19

Average

 

0,34

 

 

0,19

M8 – M9

7,50

0,37

7,20

7,38

0,19

M7 – M8

7,13

0,36

6,86

7,05

0,19

M6 – M7

6,77

0,32

6,54

6,69

0,15

M5 – M6

6,45

0,33

6,23

6,33

0,11

M4 – M5

6,12

0,36

5,85

6,00

0,15

M3 – M4

5,75

0,36

5,47

5,62

0,15

M2 – M3

5,40

0,28

5,18

5,31

0,13

TQ(1) – M2

5,12

 

0,00

5,05

5,05

 

0,34

 

 

0,15

Average

Note (1): Room and space: distance measured between the center of one floor-timber to another. Note (2): Intervals: distance measured between each frame (at the level of the floor-timbers). Note (3): Sequential measurements are defined by the addition of distances between each point starting from the heel of the keel (TQ).

48

The hull of the Mortella III wreck and its construction method

Figure 43. Reverse assembly sequence from master-frame M.27 (Photo: Christoph Gerigk).

3.1.1.4. Frames located between the tailframes

All the starboard frames between the tailframes were broken at the level of the floor-timbers heads. As a result, the entire framework was bent downwards. The issue raised by this situation was the difficulty of restoring the original transverse shape of these collapsed structures. After an unsuccessful attempt at a theoretical approach, it was finally the study of the M27 master-frame on land in 2014 that allowed us to understand the initial shape of the frames in the central part of the wreck for the first time.

The transverse structure between the tailframes was made up by a succession of about 30 frames (34 if we refer to the Spanish rule of 1618). They were complete up to a part of the second-futtock on the starboard side (preserved for about one m after their scarfs to the first-futtock), and interrupted at the end of the floor-timbers on the port side. The frames were arranged on either side of the master-frame, identified as M27. Its V27 floor-timber passed exactly under the second key of the main mast-step. Classically, the master-frame was framed by two first-futtocks named G27A and B. The total length of this piece, which lost its initial curvature with the rupture of the head of floor V27, was measured at 5.20 meters in planimetric view from the center of the floor-timber to the end of the second-futtock.

The fracture observed in 2007 on the floor-timbers M10 and M11 frames, about one meter from the keel axis, was then later observed on the frames M16 to M19, at a distance of 1.7 and 1.9 m from the keel axis. These breaches were located closest to their assembly with their first-futtocks. In most cases, it was not the first-futtock/ floor-timber union that gave way, but the floor-timbers themselves that broke. At the M20 and M21 frames, the fracture was approaching the keel axis, from which it was only 50 and 80 cm respectively.

The overlap of the pieces of the frame occurred over large portions of their length: the ends of the floor-timbers were scarved on about 1 meter with the first-futtocks. The overlap between first and second-futtocks was on an average of 80 centimeters. In total, 57% of the length of the first-futtocks were overlapped by the floors and second-futtocks ends.

In the following pages eight frames shapes are presented. They were obtained by transverse and longitudinal topographical surveys made at regular intervals between the tailframes. These surveys were also used to evaluate the slope of the wreck:

Although the wreck of Red Bay is much smaller than the Mortella III, there is a similar percentage of the frame timbers overlapping.

– The cross-sections highlight a low degree of inclination of the wreck on its starboard side. It was estimated to 49

The Mortella III Wreck

Figure 44. The master-frame on the starboard side (Photo: Christoph Gerigk).

Figure 45. The fracture of the floor-timbers V16 and V17 (Photo: Arnaud Cazenave de la Roche).

50

The hull of the Mortella III wreck and its construction method

Figure 46. The fracture of the floors-timbers V20 and V21 (Photo: Arnaud Cazenave de la Roche).

Figure 47. Top: Broken head of floor-timber V27. Bottom: Breakpoint indication at the level of the transverse profile of the M27 frame measured on the bottom (Photo: Arnaud cazenave de la Roche).

51

The Mortella III Wreck

Figure 48. Cross-section of frames M20 and M22 (Illustration: Arnaud Cazenave de la Roche).

have settled at 8% relative to the level of the sisterkeelsons. – The comparison of the underwater and land surveys of the M27 frame shows that the degree of collapse of the starboard framing was 40º relative to its original position.

cross-section. They were quickly identified as frames, part of the port side hull framework missing from the tumulus A. Under tumulus B the heads of the floors-timbers were found, preserved up to 20 to 40 cm. They were joined to the degraded ends of the first-futtocks, themselves linked to remains of the second-futtocks. These broken floor heads are the physical evidence of the ship’s hull fracture line that caused the separation of the ship’s hull structures. This lengthwise split of the framework took place before the floor-timbers/first-futtocks connection on the vessel’s port side.

The longitudinal surveys revealed a forward slope of the ship structure of 2.6% which is consistent with the general slope of the site estimated to be at 2.25% (CAZENAVE DE LA ROCHE, 2009, 10). 3.1.1.5. Frames of the fore part of the hull – Around the fore tailframes area (tumulus B): The first overview of the transverse structures was provided by the framework uncovered in 2012 under the tumulus B. Clearance work was undertaken on the AF12/B excavation area located on the highest part of the tumulus B, which culminated at a depth of 36 meters, under a layer of 70 cm of gravel and ballast stones.

In the same way as on starboard, on port side the ends of the second-futtocks were heavily charred. Nevertheless, they were preserved slightly higher than those on the starboard side. Actually, they still were provided with their second clamp (S6), while on the starboard side, it had disappeared. In practice, we do not know exactly where the frames found in the AF12 area were initially located on the keel. By deduction, however, we can make some conclusions:

Finally, the architectural structure was symmetrical to the starboard side of the tumulus A that was uncovered under the gravel of the tumulus B alongside the discovery of nine half-frames. These half-frames were oriented in an EastWest direction along an axis 346°-166º.

First, as we know that the floor-timbers were arranged facing the master-frame, and as the uncovered first-futtocks were fixed to the floor-timbers and second-futtocks on their fore side, therefore, the whole B12 structure was necessarily located forward the master-frame.

In the western part of the grid-frame, the first structures that appeared were the degraded ends of timbers 15 to 18 cm in 52

The hull of the Mortella III wreck and its construction method

53 Figure 49. Cross section of the frame M24 (Illustration: Arnaud Cazenave de la Roche).

The Mortella III Wreck

Figure 50. Finding of the port side frame under the tumulus B in 2012 (Photo: Christoph Gerigk).

the M34 frame of the tumulus A, last frame to have been uncovered on the fore part of the master-frame. The group of nine frames studied under the tumulus B was therefore likely to have been related to nine of the fifteen frames still to be studied in the fore tailframe area (see fig. 29).

The lower extremities of the first-futtocks uncovered in this area were mostly still attached to the broken end of a floor-timbers (fig. 51). However, in tumulus A, as the floors uncovered foreward the master floor were all found with their port heads, it can be deduced that the entire tumulus B framework was necessarily located before

Figure 51. Planimetry of the excavation area AF12/B (tumulus B) (Illustration: Arnaud Cazenave de la Roche).

54

The hull of the Mortella III wreck and its construction method

55 Figure 52. Cross-sections of frames MB2, MB5 and MB (tumulus B) (Illustration: Arnaud Cazenave de la Roche).

The Mortella III Wreck – The fore end of the ship (tumulus A, excavation area AF13/1). The transverse framework is represented by the remains of a series of eight frames, the last trace of which was visible towards the bow under artillery gun Cn9, about one meter fore the end of the keel.

Finally, the eight frames found in the AF13/1 area were all bearing traces of keel attachment with circular iron bolts of approximately 30 mm in diameter.

The first three frames, moving from the stern to the bow, were composed of risen floor-timbers (V1P to V3P) provided of cant first-futtocks on starboard, whose outer end was lost in a concretion zone (G1P to G3P). On the port side, the edge of the remains consisted of the levelled ends of the crutches and no futtock was visible. The remaining length of the V1P to V3P floor-timbers was approximately 1.40 m. Given their starboard length was completely preserved between 90 to 98 cm, it can be estimated that their initial overall lengths were 1.8 to 2 meters. Their sided sizes were 29 cm for V1P and V2P and 26 cm for V3P. The V4P to V8P floors evolved as crutches whose forks were levelled. Three of them were still extended by their first-futtock (G4P to G6P). They were composed of large single timbers probably coming from rising double branches. Only the base of the fork of these Y shaped crutches remained and had a maximum height of 40 cm. Their moulded dimension varied from 20 to 25 cm. We did not observe any ruptures of the floors/crutches in their connection to the futtocks, as is the case throughout the framework studied in the excavation areas further aft.

Figure 53. Location plan. Excavation area AF13/12 (Illustration: Arnaud Cazenave de la Roche).

Figure 54. Planimetry of the fore end of the wreck (Excavation area AF 13/1) (Illustration: Arnaud Cazenave de la Roche).

56

The hull of the Mortella III wreck and its construction method

Figure 55. Cross section of the M2P, M4P and M5P frames (Excavation area AF13/1) (Illustration: Arnaud Cazenave de la Roche).

The interval measured between the frames was variable due to the irregular shape of the pieces of wood. It varied from 9 to 25 cm for an average of 15 cm. The interval measured between the Centre of one floor-timber to the next (room and space) varied between 30 and 45 cm, with an average of 34 cm.

The floor-timbers were about four meters long at midship (fig. 58). Then, they gradually decreased in size after and forward the master-floor to about two meters at the level of the tailframe areas (measurement taken on floor V9). The gradual decrease in the length of the floor-timbers is testimony to the narrowing process of the dead flat.

3.1.2. Components of the frames (see Table 10 for a summary of dimensions)

Morphology It should be stressed that, in our case, the dead flat area set out by the floors-timbers and located between the tailframes remains a theoretical concept insofar as these were morphologically characterized by a curved and regular arcuate shape, even at mid-ship. On this curved profile the bilge points were not visible, hence the turn of the bilge was not marked by a line break. These morphological features reflect a rounded hull shape. This profile can be compared to the wreck of Villefranche-surMer and can be juxtaposed to that of the Red Bay wreck for example, whose moulded floor-timbers all had a flat surface and the master floor a lack of rising. The roundness of the hull of the Mortella III’s ship is a major trait of the ship’s design and will be discussed in detail in chapter 5.

3.1.2.1. The floor-timbers Dimensions The floor-timbers had generally square sizes at midship where their sided and moulded average dimensions were respectively 20 and 18 cm (measurements taken next to the starboard side of the keelson). Notably, sections of the aft and fore groups were larger, with the sided and moulded average sizes measuring 21 and 20 cm respectively. The sizes of the Mortella III floor-timbers are comparable to that of the wreck of Villefranche-sur-Mer or that of Red Bay (Table 7). This similarity has been observed for ships of very different dimensions (800 tons for Villefranche and 250 tons for Red Bay) leads to the conclusion that the dimensions of these timbers had no major relationship with the size of the ships.

The master floor-timber V27 The discovery of the master-frame in 2013 provided an opportunity to carry out a detailed study of its structure on land. Here are the main characteristics of its construction:

Table 7. Comparison of the moulding sizes of the frame-timbers between Mortella III and other wrecks Sizes in meters

Mortella III

Cattewater

Villefranche

Mary-Rose

Red Bay

Moulded

0,18

0,20

0,20

0,25–0,49

0,20

0,16

0,16

Sided

0,20

0,20

0,20

0,32

0,20

0,17

0,17

57

Molasses Reef Highborn cay

The Mortella III Wreck

Figure 56. Planimetry of the excavation area AF12/B (tumulus B) (Illustration: Arnaud Cazenave de la Roche).

58

The hull of the Mortella III wreck and its construction method

Figure 57. Photograph of the master floor-timber V27 (Photo: Arnaud Cazenave de la Roche).

Located under the forward key of the maststep, it was broken at both ends of its floor-timber: its head was broken at 65 cm from its end on the port side, and at 70 cm on the starboard side where it collapsed by about 30° down.

a satisfactory degree of accuracy, and to go beyond a working hypothesis. Dimensions The total length was 4.05 meters (the largest of the 42 floor-timbers observed to date). The length of the half floor-timber on the starboard side was 2.01 meters and 2.04 meters on the port side.

The land-based study of the floor-timber V27 allowed us to replace the wood grains into their original position. The positioning and fastening of the two first-futtocks that framed the sides of the floor-timber validated the accuracy of the restoration of the original position of this broken floor-timber head. The possibility of restoring the original shape of the V27 floor-timber, at least on the starboard side, occupies an important place in our study of the M27 frame. It was an essential condition for being able to restore the original shape of the transverse ship with

The sided dimension was fairly regular over the entire length of the piece, 15 cm on average. The molded dimension at mid floor was 18 cm, the starboard branch was larger than that on the port side; its sided size remained 16 cm up to one meter from the center compared to an average of 14 cm on the port side.

Figure 58. Master floor-timber record V27 (Illustration: Arnaud Cazenave de la Roche).

59

The Mortella III Wreck Table 8. Overlaps of the timber framing components

Morphology

Floors-timbers / first-futtock

One of the main characteristics of the V27 floor-timber was its curved and regular shape which resembled a circle of 5.65 meters radius. This shape, conferred by a marked rising of the heads, will be analyzed in Chapter V. To complete the morphological description of floor-timber V27, it is necessary to mention – A part of wood missing on its base lying on the keel, which has been compensated by the addition of a wedge. This practice of compensating for the irregularity of the floor-timbers bottom seems common since we have also observed the presence of a wedge under the floor-timber V25. This was also documented on the wreck of Villefranche, where wedges were found on five floor-timbers (GUÉROUT, RIETH and GASSEND, 1989, 42). – Two roughly rounded limber-holes located on both sides of the keel next to the garboards.4 The one on the starboard side is 6.5 cm wide and 3.5 cm high. The one on the port side 5 cm x 3 cm. Finally, it should be noted that there is an orifice about 3 cm in diameter located on the central axis of the floor-timber, 8 cm above the upper face of the keel. We wondered about the functional of this through hole. It seems to us that it could have participated in the water circulation system between the floors and constitute a 3rd circulation lane. In this case, it should be compared to the typology of the system observed at Villefranche-sur-Mer.

cm

First/Secondfuttocks  

cm

V18/G18

87

 

V19/G19

112

V20/G20

124

G20/A20

88

V21/G21

109

G21/A21

76

V22/G22

109

G22/A22

130

V23/G23

112

G23/A23

79

V24/G24

108

G24/A24

88

V25/G25

95

G25/A25

100

V26/G26

93

G26/A26

101

V27/G27

85

G27/A27

83

V28/G28

92

G28/A28

73

V29/G29

95

G29/A29

101

V30/G30

87

G30/A30

74

V31/G31

94

G31/A31

95

V32/G32

88

G32/A32

45

V33/G33

78

G33/A33

 

Average

98

 

87

 

assembly appears in other French texts as ‘écart à dent ‘or ‘écart à croc’, a method of connecting timbers typical of the Mediterranean constructive tradition. B – The fastening methods: in general, as shown in fig.59, fastening was ensured by two iron nails with a circular section of approximately 12 mm in diameter. They crossed the first timber and finished their trajectory in the wood of the second timber. The nailing was alternated: the first nail was driven from the floor-timber to the first-futtock, the second from the first-futtock to the floor-timber. Here again, we can underline the similarity of this method with that observed in Villefranche (GUÉROUT, RIETH and GASSEND, 1989, 43).

3.1.2.2. Scarfs of the floor-timbers to the first-futtocks Between the tailframes: A – Scarfs methods: The assembly of the floor-timbers to the first-futtock was characterized by scarfs which had a single hook with a notch of about 15 mm deep. This provided a careful levelling of the contact surfaces between the overlapped areas (fig. 59) over a length of about one meter, on average (Table 8).

A particular feature of the nailing of the floor-timbers to the first-futtocks, mainly on the port side where many floor heads were bare, is that the first nail had been driven horizontally, which would logically suggest a preassembly. But, on the contrary, the second nail, the one closest to the end of the piece, had been pushed obliquely from the upper face of the floor to the lower part of the futtock side.6

This type of scarf is known as ‘hook scarf’, ‘écart à cadeau’ in French, a term that appears in an anonymous text from the end of the 17th century dealing with the construction of galleys.5 Its description is given in fº25:‘…They are joined [‘madiers’ and ‘estamenaires’] by a two-and-a-half foot scarf, in the middle of which there is a tooth called ‘à cadeau’, by means of which these two pieces perfectly fit together[…]’ The type of scarf that characterizes this

It is difficult to interpret the constructive function of this nailing method; at most it can be said that it induces the possibility that the first-futtocks were pre-assembled to the floors by a single nail, the second having been placed

The wreck of Villefranche-sur-Mer also had two limber holes on the port and starboard sides of its floor-timbers. However, the two holes were located higher on the timber than in the case of the wreck of Mortella III. In the Red Bay or Cattewater wreck, there is only one shaped limber hole above the upper face of the keel. 5 ‘Traitté de la construction des galères’, 1691 (Service Historique de la Marine, Vincennes, ms SH 134) ; Edition commented by Fennis, J., 1983, Un manuel de construction des galères, 1691, Amsterdam (ANON., 1691). This text is quoted in ‘Le navire Génois de Villefranche…’ (GUÉROUT, RIETH et al., 1989, 43). 4

6 A notable exception to this nailing method is the master-frame’s floortimber V27 to its first-futtock: Both of them have been horizontally driven in. It confirms, thus, this frame was intended to be the first frame to be placed on the keel with the tailframes.

60

The hull of the Mortella III wreck and its construction method

Figure 59. Scarf of the floor-timber V18 with its first-futtock G18 (Illustration: Arnaud Cazenave de la Roche).

Hence we had a mixed type scarf provided with a mortise on one side, and a single hook of about 15 mm, on the other. In fig.63, the timbers of fig.62 have been superimposed in transparency to help to visualize this diagram.

afterward. This technique is not documented – to our knowledge – on any other wreck. Finally, it should be added that the scarfs were reinforced by two clamps, a foot-wale and a bilge-clamp (S2 and S3) of about 15 cm moulded/sided, notched on the frames. We will come back to their description when addressing the longitudinal framework.

Strictly speaking, we are not dealing with a mortise and tenon scarf ‘dovetail’ type, as there was no tenon. From a mechanical point of view, a tenon and mortise joint offers resistance to the forces exerted along the longitudinal axis of the piece, both in compression (force exerted inwards) and in tearing (force exerted outwards). In this case, we are dealing with a type of assembly that is more like the family of hook scarfs as the only resistance offered is to tearing away. In this case, it is the simple hook located on the floor-timber V27 in connection with the edge of the mortise located on G27 A that ensures this resistance to tearing. Mortise had no other function than to accommodate the head of the floor-timber and to ensure good cohesion of the surfaces. For this reason, in our opinion, this scarf could be assimilated to a variant of hook scarf type.

C. Variations: The detailed study of the M27 master-frame in 2014 revealed two variations of scarfs typology (a complete drawing of the master-frame is shown in fig. 72): a) The first at the assembly level of the master-floor V27 with the first-futtock G27A, as well as the first-futtock G27 with the second-futtock A27 b) The second one at the level of the master-floor V27 with her second first-futtock G27B. (a.) Prior to disassembly, the V27/G27A scarf had the characteristic traits of a tenon and mortise scarf (dovetail typology). Visually, it seemed that the floor-timber was fitted with a tenon embedded in a mortise on the first-futtock (fig. 61). However, as discovered at its dismantling, the reality was different. A large mortise, extended by the size, but not very deep, had been carved over a length of 60 cm. It was located at distance of 50 cm from the fore end face of the first-futtock G27 A (fig. 62). The shape given to this mortise enabled it to fit the aft end face of the V27 floor-timber.

(b.) The other type of connection highlighted by the assembly of the G27 B first-futtock with the V27 floortimber was its juxtaposition with the timbers that were roughly flattened but without mortises or hook, then nailed. However, curiously, we there was a small hook of about one centimeter thick cut on the fore face, thus on the opposite side to the assembly (fig.65 right). There were also two nails on either side of the notch, separated by 60 61

The Mortella III Wreck

Figure 60. Oblique nailing of the second nail joining the floor-timbers to the first futtocks (Photo: Christoph Gerigk).

Figure 61. Assembly of floor-timber V27 with first firstfuttocks G27A and G27B (Illustration: Arnaud Cazenave de la Roche).

cm with no apparent function. G27 B therefore had the characteristics of a timber that could be reused.

Figure 62. Photographic representation of the assembly of the floor-timber V27 with the first-futtock G27A (Photo: Arnaud Cazenave de la Roche).

At the fore end of the wreck As it was impossible to study the scarfs of the framing pieces in the aft part of the wreck where the first-futtocks are absent, we did so in the forward area on the last eight frames that were resting on the keel. At this extremity of the ship, only the assemblies between floor-timbers/crutches and first-futtocks were visible. They were no longer made with a hook but by a simple juxtaposition of the pieces.

As can be seen in the picture in fig. 66, the first-futtock ends had been conscientiously shaped to fit the form of the heads of the floor-timbers. The first-futtocks G2P and G4P and GP5, in particular, had been cut in the shape of spikes to be slid between the floors. They were ‘floating,’ as no nails ensured their connection with the floors. 62

The hull of the Mortella III wreck and its construction method 3.1.2.3. The first-futtocks In the central part, the first-futtocks moulded and sided dimensions were, on average, 15.3 cm and 13.7 cm, respectively. The sided dimension tended to decrease moving forwards (12.7 cm). In the aft area the number of timbers preserved was too small to draw a conclusion. In general, these dimensions show thinner timbers than at Villefranche-sur-Mer (first-futtock section of around 20 x 20 cm) and at Red Bay (cross-section 19 to 20 cm in their lower part and 16 to 17 cm at their top).

Figure 63. Photomontage: Transparent overlay of the two timbers V27 and G27 (Photo: Arnaud Cazenave de la Roche).

The length of the first-futtocks varied between 3.20 m (G31) and 3.82 m (G19), with an average of 3.47 m. During the study of the master-frame, the G27 A and B first-futtocks were observed in detail. Here are their characteristics (figs. 67 to 69): Figure 64. Transparent representation of the V27 - G27A assembly and nailing (Illustration: Arnaud Cazenave de la Roche).

The lengths of G27 A and B were 3.34 m and 3.32 m respectively. The moulded dimensions of both pieces

Figure 65. G27B : face attached to V27 (Left) and fore face of the floating end (right) (Photo: Arnaud Cazenave de la Roche).

Figure 66. V1P to V8P frames of the aft of the ship (Photo: Christoph Gerigk).

63

The Mortella III Wreck

Figure 67. First futtocks G27 A and G27 B. Note the tapering of the ends (Photo: Arnaud Cazenave de la Roche).

Figure 68. First futtock recording G27A (Illustration: Arnaud Cazenave de la Roche).

(outside the scarf area) varied from 14 to 16.5 cm; their average being about 15 cm. The sided dimension was on average 13.2 cm for G27 A and 14.8 cm for G27 B.

hook scarf – was also found between the first and secondfuttocks.

3.1.2.4. Scarfs from first to second-futtocks

The same method of fastening the two together, with two circular iron nails of 12 mm in diameter, was similarly used. The first one was driven from the first into the second -futtock and then, conversely, the second nail was driven from the second into the first-futtock in the wood. The only notable difference was the systematic tilt of the nailing which went from the upper face of the first timber to the lower side of the second one, inducing, as in Villefranche-sur-Mer, the fastening of the second-futtock once the floor-timbers and first-futtocks were in place.

The same method of assembly as that was observed between the floor-timbers and the first-futtocks -with a

The same type was found in the scarf of the first-futtock G27 A as with the second-futtock A27 (fig.70 and 71):

Wood: The G27 A and B first-futtocks were roughly carved from sessile oak logs that grow particularly slowly (average 120 years). G27 A came from a quarter-slit log while G27 B was cut from a single roughened log. From a morphological point of view, the shape of the first-futtocks G27 A and B generally followed the line of the floor-timber V27 with a very similar circular arc.

64

The hull of the Mortella III wreck and its construction method

Figure 69. First futtock recording G27B (Illustration: Arnaud Cazenave de la Roche).

In this instance, the mortise was cut on the secondfuttock, the contact face of the first-futtock having only one hook of about 20 mm. Like the floor-timber V27, the mortise extends over a long length, about 60 cm. It had a triangular shape that followed the shape of the G27A first-futtock head exactly. The connecting surfaces of the first and second-futtock had also been leveled in order to

ensure the largest contact surface and a perfect adhesion (fig.70). The fastening of the G27A first-futtock to its A27 secondfuttock was also achieved using two nails (fig.72). The first one went from the first to the second-futtock. The direction of nailing was attested by the notch cut in the first-futtock

Figure 70. View of the lower face of the assembly of G27A to A27 (Photo: Arnaud Cazenave de la Roche).

65

The Mortella III Wreck

Figure 71. Photographic representation of the assembly G27A to A2 (Photo: Arnaud Cazenave de la Roche).

made to hold the head of the nail. The direction of nailing was similarly alternating with a second nail driven from the second towards the first-futtock.

3.1.2.6. Summary tables of the frames measurements

3.1.2.5. The second-futtocks

3.1.3.1. A technical choice

The remains of the second-futtocks are incomplete. As we have already pointed out, their ends were charred less than one meter after their overlap with the first-futtocks, located under the S4 and S5 clamps. Their curvature was relatively slight. Their moulded/sided dimensions were similar to that of the first-futtocks: 15 × 15 centimeters calculated on the average of the fifteen visible timbers in the central area of the wreck.

Observing the scarfs of the frames of the shipwrecks of Mediterranean construction highlights the predominance of hook-scarfs, so-called ‘écart à cadeau’ in French, which constitute a technical ‘fingerprint’ typically linked to the Mediterranean nautical space. In fact, this assembly system observed on the Mortella III is also observed on wrecks such as those of Cala Culip VI dated from between the end of the 13th to the beginning of the 14th century (RIETH, 1998). This is also the case of the Villefranche-sur-Mer wreck, that of Yassi Ada I (LABBE, 2010), both dated from the 16th century and, from a later period, the wreck of Les Sardinaux which has been dated from the end of the 17th century (JONCHERAY, 1988). It differs from traditional Atlantic or Ibero-Atlantic assemblies, such as the Red Bay wreck, which are traditionally made with a dovetail mortise scarf. This last typology is also found on the wrecks of Highborn Cay (first third of the 16th century), Molasses Reef (early 16th century), or Western Ledge Reef (last third of the 16th century), to mention just a few examples.

3.1.3. Scarfs typology of the frame components

The characteristics of the A27 second-futtock of the master-frame were as follows: Dimensions: The preserved length of A27: 1.80 meters. Sided: average of 15.5 centimeters. Moulded: average of 14.8 centimeters. Morphology:

Using the scarf typology as a ‘technical marker’ must, however, be taken with caution. Whilst up to this point, hook scarfs have not been found on wrecks in the Atlantic construction tradition, it has been possible to find dovetail scarfs on wrecks of Mediterranean construction. This is the case in Villefranche-sur-Mer where hook and dovetail scarfs coexist, and in Calvi I where the scarfs of the floor-

The A27 second-futtock was cut from a sessile oak log about 170 years old which had been split into quarters. The charred end of this timber testifies of the fire that consumed the ship. Its degree of curvature was close to V27 and G27B. 66

The hull of the Mortella III wreck and its construction method

67 Figure 72. Frame M27 and timbers assembly layout (Illustration: Arnaud Cazenave de la Roche).

The Mortella III Wreck

Figure 73. Record of the first futtock G27A (Photo: Arnaud Cazenave de la Roche).

Figure 74. Record of the second futtock A27 (Illustration: Arnaud Cazenave de la Roche).

68

The hull of the Mortella III wreck and its construction method timbers and the first-futtocks are all dovetail-shaped (RIETH, 1998, 186). In other words, when hook scarfs are observed on a wreck, there is a high probability that it is of Mediterranean origin, whereas if dovetail scarfs are found, the wreck is not necessarily of Atlantic origin.

simple technical choice. Archaeologists have wondered whether, more than a mechanical function, scarfs played a role in the design of ships by predetermining the frames between the tailframes (BARKHAM, 1985). In fact, this relationship was highlighted in the case of the wreck of Cala Culip VI (RIETH, 1998).

3.1.3.2. A possible architectural design function

In the case of the Red Bay wreck, for example, only in fourteen frames in the central area of the ship had their floor-timber attached to first-futtocks with dovetail scarfs

It is important to note that the scarfs of the floor-timbers to the first-futtocks may have a deeper meaning than a

Table 9. Mortella III – Main measurements of the frames (tumulus A - Starboard) Measures in meters

Fore area

Frames

Floor-timbers Length

First-futtocks

Second-futtocks

Moulded

Sided

Length

Moulded

Sided

Moulded

Sided

42

M 8P

23,68

 

x

x

0,37

x

x

 

x

x

41

M 7P

23,35

23,54

1,70

0,19

0,35

x

x

 

x

x

40

M 6P

22,93

23,15

1,36

0,22

0,34

2,08

x

 

x

x

39

M 5P

22,56

22,77

1,30

0,21

0,35

1,95

0,135

 

x

x

38

M 4P

22,34

22,58

1,00

0,24

0,31

x

0,135

 

x

x

37

M 3P

21,95

22,18

0,95

0,23

0,30

x

0,110

 

x

x

36

M 2P

21,62

21,82

0,74

0,20

0,23

x

0,120

 

x

x

35

M 1P

21,30

21,46

0,58

0,16

0,22

x

0,135

 

x

x

 

 

 

0,21

0,31

2,02

0,13

 

 

 

Average

Mid-ship area

Sequential distances

34

M34

15,99

16,18

3,18

0,190

 

 

 

 

 

 

33

M33

15,63

15,81

3,03

0,180

 

2,80

0,130

0,148

0,190

0,170

32

M32

15,28

15,45

2,70

0,170

0,190

3,36

0,173

0,141

0,200

0,180

31

M31

14,89

15,07

3,13

0,180

0,160

3,20

0,144

0,137

0,165

0,160

30

M30

14,45

14,71

3,80

0,260

0,170

3,27

0,100

0,141

0,155

0,160

29

M29

14,14

14,31

3,35

0,170

0,180

3,46

0,190

0,160

0,175

0,145

28

M28

13,83

13,97

3,81

0,140

0,180

3,27

0,170

0,144

0,165

0,150

27

M27

13,49

13,65

4,05

0,150

0,180

3,34 B 3,36 A

0,150

0,148—B 0,132 A

0,148

0,155

26

M26

13,19

13,34

3,66

0,150

0,170

3,47

0,150

0,120

0,110

0,140

25

M25

12,91

13,06

3,56

0,150

0,170

3,70

0,150

0,120

0,120

0,135

24

M24

12,64

12,80

3,38

0,160

0,180

3,10

0,150

0,140

0,110

0,130

23

M23

12,35

12,50

3,13

0,150

0,180

3,70

0,136

0,126

0,135

0,150

22

M22

12,00

12,12

3,21

0,120

0,210

3,82

0,140

0,130

0,100

0,140

21

M21

11,65

11,81

3,26

0,160

0,185

3,50

0,200

0,127

0,160

0,150

20

M20

11,36

11,52

3,22

0,160

0,220

3,60

0,136

0,134

0,120

0,135

19

M19

11,06

11,25

3,25

0,190

0,220

3,59

0,135

0,142

x

x

18

M18

10,75

10,87

3,10

0,120

 

x

0,194

0,135

x

x

17

M17

10,40

10,57

3,08

0,170

 

x

0,150

0,134

x

x

16

M16

9,98

10,25

3,04

0,280

 

x

0,153

 

x

x

15

M15

9,62

9,79

2,93

0,170

 

x

x

 

x

x

14

M14

9,25

9,41

2,8

0,160

 

x

x

 

x

x

13

M13

8,87

9,04

2,66

0,170

 

x

0,160

 

x

x

12

M12

8,50

8,70

2,44

0,200

 

x

0,100

 

x

x

11

M11

8,06

8,31

2,35

0,250

0,320

x

0,140

 

x

x

10

M10

7,77

7,95

2,1

0,180

0,350

x

0,140

 

x

x

9

M9

7,38

7,58

1,96

0,200

 

x

x

 

x

x

 

 

3,25

0,18

0,20

3,49

0,15

0,14

0,15

0,15

Average

69

Aft area

The Mortella III Wreck 8

M8

7,06

7,19

1,09

0,150

 

x

x

 

x

x

7

M7

6,70

6,85

0,78

0,150

 

x

x

 

x

x

6

M6

6,33

6,54

0,76

0,210

 

x

x

 

x

x

5

M5

6,00

6,23

0,56

0,230

 

x

x

 

x

x

4

M4

5,47

5,86

0,47

0,240

 

x

x

 

x

x

3

M3

5,31

5,48

0,4

0,170

 

x

x

 

x

x

2

M2

5,04

5,16

0,35

0,120

 

x

x

 

x

x

1

M1

 

 

X

X

 

 

 

 

 

 

 

 

 

0,19

 

 

 

 

 

 

Average

Note (1): Room and space: distance measured from the center from one floor-timber to another. The measurements given for each indicate the distance between this frame and the previous one in the table. E.g.: the measurement 0.350 given for M33 means that the distance between M33 and M34 is 35 cm, and so on. Note (2): Intervals: distance measured between each frame (at the level of the floor-timbers). The reading of the measurements must be done as follows indicated in Note (1). Note (3): Black figures are those observed. Red figures are not significant (pieces not complete). Blue figures are estimated. X indicates that the measure could not be identified (gaps, absences, concretions, etc.)

Table 10. Average of the main transverse framework measurements Location / timbers (measures in meters)

Framing Room & Interval Space

Floor-timbers

First-futtocks

Length

Moulded

Sided

Length

Moulded

Second-futtocks Sided Moulded Sided

FORE AREA: averages M1P to M8P

0,35

0,15

X

0,21

0,31

 2,02

0,13

X

X

X

MID-SHIP AREA: AVERAGES M9 to M34

0,34

0,19

3,47

0,18

0,20

3,46

0,15

0,14

0,15

X 

AFT AREA: averages M1 to M8

0,34

0,15

X

0,20

X

X

0,13

X

X

X

and a fastening system combining iron nails and tree nails. The rest of the overlaps between the floor-timbers and first-futtocks fore and aft were devoid not only of scarfs, but also of any fastening between them: ‘An important feature of the frames is that their components are not scarved between them. With a few exceptions, they were only held in place by their fasteners to the hull’s planks and internal structures (LOEWEN, 2007, 57). The Mortella III wreck seems to follow this principle. A scarf connection is used only for frames in the central area of the ship since the branches of the six crutches uncovered in the fore area of the wreck were simply juxtaposed to their first-futtocks, without scarf. However, the completion of the excavation forward the master-frame will allow us to specify this organization.7

Figure 75. Hook scarfs of the frames of the wreck of Yassiada I (Drawing: Jay Rosloff (PULAK, 2005, 141) (Illustration: Cemal Pulak).

This model for connecting floor-timbers and first-futtocks does not follow the same approach of three other wrecks built in the Mediterranean constructive tradition. The floor-timbers/first-futtock connections of the Villefranchesur-Mer and Calvi I wrecks used dovetail scarfs beyond

the tailframes, on the one hand, whereas on the other hand the crutches at the fore end of the Yassi-Ada I wreck are assembled to their first-futtocks with hook scarfs.

7 It will also be necessary to observe the way in which the first-futtocks are joined to the floor-timbers in the part of the wreck located between frames M34 and M1P, an area that has yet to be excavated, to determine from which frame exactly the scarfs are no longer being used.

The functional link between scarfs and preassembly of the frames deserves to be explored in greater depth. Nevertheless, the disparity of the cases we have 70

The hull of the Mortella III wreck and its construction method timbers were fastened to it with iron bolts with a circular section of 30 mm in diameter.8

mentioned and the small number of wrecks still likely to be studied does not allow any convincing conclusions to be formed here. Twenty years ago, E. Rieth wrote about the wrecks of Villefranche-sur-Mer and Calvi I: ‘In the current state of progress of our research, and considering also the very small number of sites, it seems difficult and hazardous to propose a coherent interpretation of these two ‘cases’ (but are they really ‘cases’?) of Villefranchesur-Mer and Calvi I. Could the presence of dovetailshaped mortices reflect an influence of Westward (Ponant) practices? Would the presence of these same mortices in the floor-timbers and crutches indicate a different design method than those known until then?’ (RIETH, 1998, 186 and 187). The question remains unanswered today.

3.2.1.1. Search for keel length The search for the forward end of the keel led to the opening of an excavation area in the southwestern part of the site in 2013 (AF13/1). The discovery of the mast-step in 2012 had enabled an estimation of its theoretical length which was thought to be around 25 meters. By calculating its length, we were able to determine a probable area where the forward end of the keel would have been located. This area, named AF13/1, was cleaned up and revealed several crutches which were described above. Upstream, towards the stern, an imposing longitudinal timber was uncovered which passed under the breech chamber Cl.6 and turned out to be the fore end of the keel we were looking for. Its longest length was exactly 24.98 m on its upper face. Its fore end was eroded and was generally in the form of a whistle (fig. 76); a cutout that could be the remains of a scarf with the missing stem post.

3.2. The longitudinal framework: layout, dimensions, morphology and scarfs methods 3.2.1. The keel, the heel of the keel and the keelson As is always the case in a ‘frame first’ construction, the keel, or primo in the language of the Genoese builders of the time (LO BASSO, 2012, 273), constitutes the major longitudinal axis of the ship and the keystone of the architectural edifice, which is well reflected by the Italian term. The frames rested on the keel and whose floor-

It should be noted here that the measurement was taken from the aft face of the heel of the keel, so on the one hand it includes the length of the heel of the keel, yet on the other hand it does not include the lower part of the stem bearing on land, with the stem post having disappeared.

Figure 76. Port side of the keel end (Photo: Christoph Gerigk).

8 All the floor-timbers whose joint to the keel could be observed (34 out of 42) were bolted.

71

The Mortella III Wreck

Figure 77. End of the keel seen on its upper face (Photo: Christoph Gerigk).

Figure 78. Keel scarf as revealed under floor-timber V26 (Photo: Antoine Couppey).

72

The hull of the Mortella III wreck and its construction method Keel scarf: A butt scarf was the typology chosen for the keel assembly. It consisted of butting the two keel pieces together without any hook, they were simply brought together and placed end to end.

In other words, the exact length of the keel piece itself – what the Spanish call the quilla rigorosa – is unknown. As well as this up this point we have omitted the length of the keel ‘laying on ground’, the quilla derecha (‘straight keel’) for the Spanish, whose measurement is necessary for calculating the proportions between the different dimensions of the ship. In order to restore the length of the keel ‘laying on ground’, we added one meter to our measurement, i.e. 25 m + 1 m = 26 m, an estimate intended to compensate for the flat lower part of the missing stem post.

The weakness of the scarf, it being a timber as mechanically important as the keel, may seem surprising, as it surprised the archaeologists of the wreck of Cais do Sodré (early 16th century, Lisbon) when they discovered a similar scarf on this wreck. They then wrote: ‘…this very rare butt joint system was observed on the coaster of Cala Culip VI (late 13th, early 14th century)…’ (RODRIGUES, ALVES et al., 1998, 357).

3.2.1.2. Morphology of the keel and types of scarfs During the 2014 excavation campaign the discovery of a scarf on the keel made it possible to highlight its dual morphology, which is both original and unusual, at least archaeologically. The scarf was located under the V25 floor-timber, or to be more accurate, in line with the bottom face of this floor. It was marking an assembly point of the keel pieces in its middle, 12.88 meters from the heel of the keel and 12.12 meters from its forward end, almost half a keel away.

However, studying shipbuilding texts shows that this assembly technique has not been an exception in the history of shipbuilding and that it has even been a rule, perhaps originating in the Mediterranean tradition widespread in the 16th century. A first text has been quoted by the archaeologists of Cais do Sodré themselves; a French manuscript from 1691 which describes this system of union for the construction of galleys.9 In a very comprehensive study carried out on this subject by Cayetano Hormaechea (HORMAECHEA, 2012, 284), we see that this system was recommended at the beginning of the 17th century by the Spanish Ordinances which generalized its use for all ships built in Spain. Article 20 states: ‘Puesta la quilla, que ha de llevar la union a tope.’ that is to say, ‘the keel placed which must have its assembly end to end (REAL

In order to observe the nature of this assembly, a window was cut out of the port side planking facing this gap to provide access to the keel side. What we have discovered can advance our knowledge of the longitudinal framework organization and the associated construction techniques:

Figure 79. Butt scarf discovered under floor-timber V25 (Photo: Antoine Couppey). 9

73

‘Traité de la construction des Galères’, 1691, op. cit.

The Mortella III Wreck Iberian world. However we have noticed that it was also recommended in France in the 18th century, as expressed by two French authors.

ORDENANZAS, 1618, Libro IX, Titulo XXVIII, art. 20). It can therefore be assumed that many ships were built in Spain with keel scarfs which adopted this principle. C. Hormaechea quotes three other 17th century Spanish authors who advocate this system: Juan de Amassa, in 163510, Diaz Pimienta in 1645 (DIAZ PIMIENTA, 1645, T.3, doc.102) and Francisco Garrote who in 1691 gives an instruction explaining why the butt scarf is preferred to others (GARROTE, 1691):

It was Duhamel du Monceau who initially mentioned the addition of a second keel timber to reinforce butt scarfs. He did not describe this system as a common practice, but proposed it as a simple suggestion: ‘The length of the scarfs is usually four times the thickness of the keel. Perhaps there would be no disadvantage in removing the scarfs and matching the keel pieces end to end, by reinforcing the scarfs with the adjunction of a second keel (‘contre-quille’) and keelson timbers.’ (DUHAMEL DU MONCEAU, 1758, Chap.II, art.9)

‘…la opinión que siguen los Españoles, que es el que dichas juntas deben ser de tope por haber hallado que aunque alquiebran mucho los bajeles, tienen la facilidad de atajar las aguas. (…) …que es fuerza que lleve dos juntas, la una con el pie de roa de popa y la otra el de proa. (…) no apartándome en el todo de la opinión que siguen los españoles que dichas juntas deben ser a tope.’

The second author was Vial de Clairbois, who suggested the use of a 4-inch thick ‘false-keel’ to protect the keel, but also to strengthen the scarfs. However, their type is not specified here. (VIAL DE CLAIRBOIS, 1787, 13).

Translation: ‘…the Spanish opinion is that these scarfs must be ‘end to end’ (‘a tope’) because although they favor the tendency of many ships to bend, they facilitate their sealing (…)

These two examples, in addition to showing the longevity of these technical practices of assembling keel timbers, also raise a conceptual question regarding the multiple pieces which made up of the keel.

…that it is necessary that[the keel] be provided with two unions, the first at the level of the heel piece, the second with the stem foot (…), maintaining the opinion which is that of the Spanish, that these scarfs must be end to end.’

– in the case of the Villefranche-sur-Mer wreck, the butt scarf is located on the keel, where the heel of the keel meets with the keel, and the upper timber acts as reinforcement. This arrangement would therefore coincide with the technical solution suggested by Duhamel du Monceau. – On the other hand, in the case of the wreck of the Mortella III, it is a lower timber placed under the keel that reinforces the upper scarf. We would therefore be closer here to the organization mentioned by Vial de Clairbois: the lower reinforcement piece could be defined as a ‘false keel’, even if its sided size is twice that recommended by the author (20 cm against 4 inches, or about 10 cm).

So here we have the main reason for this choice: better waterproofing. It should also be noted that the butt scarf of the wreck of Cais do Sodré was provided with a dowel to cut off the passage of water. In the case of the wreck of the Mortella III, we did not see any trace of it, but an unbundling of the keel timbers would be necessary to ensure this, which was not possible for us. Finally, it should be added that this type of scarf is still found on the wreck of Villefranche-sur-Mer at the level where the keel is added to its heel piece. But its particularity is to be associated with hook scarf located in the upper part of the timber, we will come back to this assembly in the part concerning the heel of the keel.

There are two other notable wrecks which had a double keel:

As the Spanish texts show, the principle of joining the keel timbers with a butt scarf was well established in the

– The Les Marinières (15th century), located in Villefranche-sur-Mer. Michel Daeffler, who led its excavation, explained it had constructive characteristics informed by both ‘Mediterranean’ and ‘Atlantic’ traditions. This wreck had a double keel of unusual morphology. Its rabbets were formed by the difference in width between the two keel timbers where the garboards were fitted. (DAEFFLER, 2007, 19). – Another wreck provided with a double keel is the presumed Nossa Senhora dos Mártires (1606) of Ibero-Atlantic tradition. However, this conclusion was deduced, as opposed to observed, due to the poor condition of the keel.

‘La quilla ha de ser de quatro pedazos y las juntas a tope en escuadra… ’, quoted by Martinez Ruiz, E. (MARITINEZ RUIZ, 2008)

Archaeology is currently unable to determine the origins of the use of a double keel. If the adoption of butt scarfs

The double keel morphology: The morphology of the keel we were able to partially observe through the window opened on the port side provides an answer to the problem of the scarf’s mechanical weakness that we have described above. Indeed, the keel was not composed of one timber, but of two large timbers of similar size. The upper piece was 24 cm moulded x 26 cm sided, reinforced by a lower piece whose dimensions were 24 cm moulded x 20 cm sided (fig.80 and 81).

10

74

The hull of the Mortella III wreck and its construction method

Figure 80. Two overlapped keel timbers (Photo: Antoine Couppey).

Figure 81. Schematic representation of the keel (Illustration: Arnaud Cazenave de la Roche).

to assemble keel timbers was of Mediterranean origin, it could logically be considered that its reinforcement by the addition of a false-keel would also belong to the same maritime space. However, we can only point out, once again, that the limited number of wrecks with a ‘Mediterranean’ constructive tradition documented prevents a clear conclusion to be reached on this subject.

general plan, Annex I). It led to its heel, cleared during the first survey of the site in 2007. Then its characteristics were clearly defined. The aft end of the keel was made of a strong piece of wood that ended in a heel and above which was a notch very similar to the one seen on the keel-heel of the Villefranche wreck (see fig. 84). A piece of wood trimmed in its upper part, a possible remainder of the stern post, was still in place at the time of its discovery, but it didn’t survive to the excavation. The height of the notch was about 20 cm from the heel of the keel, but it may have been much higher initially if – as in Villefranche – the heel had been

3.2.1.3. The heel of the keel, its scarf to the keel and the stern A portion of the keel, in the aft part of the wreck, was the only visible remain of the hull when it was discovered (see 75

The Mortella III Wreck

Figure 82. Diagram of the keel morphology (Illustration: Jesús Guevara (Aingurak)).

shaped from a piece of twisted wood whose upper fork would correspond to the start of the stern post (fig. 84). On the site of the Mortella III, however, the angle of the aft part of the keel prevented an observation of the fork.

This was also documented by the Portuguese author Joao Baptiste Lavanha, one of the very first authors to describe how this piece, called ‘couce de popa’ in Portuguese, should be manufactured (LAVANHA, 1608, fº63) (fig.86). It seems that over the following centuries, the trend moved towards using less twisted wood and more complex assembly techniques (LOEWEN, 2007, 45).

Archaeology has provided several keel-heels to observe in recent years. Generally speaking, during the period under discussion, whether in the Atlantic or the Mediterranean, large and single timbers are made from the trunk of a tree and one of the very first branches.

There are other examples of keel-heels such as that the wreck of the San Esteban, a Spanish ship of the New Spain fleet, sunk in 1554 off Padre Island (ROSLOFF, BARTO ARNOLD III, 1984, 286-296). Similarly, there are also keel-heels on much older ‘Ibero-Atlantic’ shipwrecks whose morphology and characteristics are similar to

Akin to the heel of the Villefranche wreck, this process was used to manufacture the heel of the Red-Bay wreck, for example. 76

The hull of the Mortella III wreck and its construction method

Figure 83. The heel of the keel (Photo: Arnaud Cazenave de la Roche).

Figure 84. Heels of the keel of the wrecks of Villefranche and Mortella III (Photo: GRAN/Arnaud Cazenave de la Roche).

those recommended by Lavanha. Examples include the Portuguese wreck of Aveiro A (ALVES, RIETH et al., 2001, 325) dating from the middle of the 15th century (fig. 87) and the even older (14th century) Corpo Santo wreck (fig. 88). Finally, we should mention the keel-heel of the Red-Bay 29 M wreck which, like the 24 M (San Juan), is dated to the 16th century (STEVENS and WADDELL, 2007, I-223). It is a Basque whaler whose of the keel-heel was unique in that it was adjoined to the stern piece by means of a curved lateral scarf (fig.89).

The similarity of their scarfs and stern posts is one of the common points: another is how their two heels were provided with two very similar notches intended to fit the stern post (fig.84). These notches were placed on the back side of the ascending branch of the heel, while in the case of Red-Bay and the model recommended by Lavanha, the flat vertical scarf was placed on its inner side (fig. 86). Finally, the wreck of Calvi 1 highlights a new variant of scarf adjoining the heel with the stern post (fig.90).11 As on the Mortella III and Villefranche, this scarf is located on

The assembly of the heel of the keel to the stern. In several respects, the Mortella III keel-heel morphology appears to be the closest to the wreck of Villefranche-sur-Mer.

11 The wreck of Calvi 1 is unique in being provided with a single keel timber. Its aft end forms a heel of the keel with a rising branch, without any assembly (VILLIÉ, 1990, 84).

77

The Mortella III Wreck

Figure 85. The heel of the keel of the Red-Bay wreck (Courtesy of: S. Laurie-Bourque, Parks-Canada).

Figure 87. The heel of the keel of the Aveiro A wreck (Photo: CNANS).

see a technique common to both wrecks in view of the fact that – as we have seen previously – we observed the same type of scarf on Mortella III keel timbers at midship. At the time of the discovery, the archaeologists of the Villefranche wreck were surprised by the morphology and the apparent mechanical weakness of the vertical and smooth butt end of the keel-heel. . This is the reason behind the hypothesis that the missing keel timber could have had an upper end that would have fitted with the diagonal scarf of the upper keel timber, as seen in Figure 92a. The nature of the scarf found on the middle of the keel of the Mortella III wreck opens the way for the hypothesis that it could simply have

Figure 86. The heel of the keel according to Lavanha, fº43 (Courtesy of: Academia de Marinha de Portugal).

the fore part of the ascending branch. Nevertheless, here, it is the stern post that forms the keel-heel by covering it and connecting it with a large mortise. The scarf of the keel-heel to the keel. On the wreck of the Mortella III, this scarf remains to be discovered, if it exists. On the wreck of Villefranche-sur-Mer, this assembly has been made up with a simple butt scarf. Here again, we can 78

The hull of the Mortella III wreck and its construction method

Figure 88. The heel of the keel of the Corpo Santo wreck (Photo: CNANS).

Figure 90. The heel of the keel of the Calvi 1 wreck (Illustration: Pierre Villié).

Figure 89. The heel of the keel of the 29 M wreck (Red-Bay) (Courtesy of: D. Kappler, Parks Canada).

been covered by a second independent timber, an upper keel piece, constructed according to the same principle (fig. 92 b). Figure 91. Assembly of the heel of the keel to the stern on the Red-Bay wreck (Courtesy of: R. Chan, Parks-Canada).

The keel / heel of the keel scarfs that have been observed on the Ibero-Atlantic wrecks mentioned above reflect a different arrangement. The principle was recommended by J. B. Lavanha (fig. 94). It used a vertical cut with a notch at mid-wood. This process was used on the Red-Bay wreck as well as on the Aveiro A wreck (fig. 87). This type of scarf can be found on later Ibero-Atlantic shipwrecks, such as Nossa Senhora dos Mártires, 1606? (CASTRO, 2005, 156) (fig.93).

The keelson consisted of an oak timber about 20 cm sided and 14 cm moulded. Its lower part was notched on the upper faces of the floor-timbers. It extended over 14.30 m between the floors V10, at the aft (fig.95), around the tailframe area, and VP3, at the forward side. Its degraded ends suggest that it initially extended over a longer length. On the keelson ran a beech joist. Its upper edges were chamfered and the ceiling, also made of beech, was leaning on it. Between the floors V19 and V34, over a length of 5 meters, the keelson was enclosed by two strong pieces of oak, two longitudinal timbers of the mast step arrangement. We will come back to the description of this device.

3.2.1.4. The keelson The longitudinal axis of the wreck consisted of the keel at the lower level and the keelson at the upper level, which covered the floor-timbers. This set was fastened at the level of each floor-timber by a 30 mm diameter iron bolt. 79

The Mortella III Wreck

Figure 92. Scarf keel / heel of the keel of the Villefranche-sur-mer wreck Hypothesis a. and b (Illustration: GRAN).

Figure 93. Scarf of timbers of the keel of Nossa Senhora dos Mártires (Illustration: Filipe Vieira de Castro).

Finally, it is worth mentioning the presence of two 7 x 7 cm mortises carved on the upper face of the keelson, intended to receive the tenons of two 20 cm side stanchions, whose remains have been found (fig.95). This type of stanchion attached by a tenon is similar to the one observed on the wreck of Calvi I (VILIÉ, 1989, 24). 3.2.2. Clamps, sill and ceiling 3.2.2.1. Clamps The first two clamps starting from the keel, the foot-wale and the bilge-clamp, were named S1T and S2T on the starboard side. These longitudinal timbers had the function

Figure 94. The scarf recommended by J. B. Lavanha, fº43 (Courtesy of: Academia de Marinha de Portugal).

80

The hull of the Mortella III wreck and its construction method

Figure 95. The keelson of the fore part of the wreck (Photo: Christoph Gerigk).

The two sets of clamps were separated by a stringer, smaller in size, (15 cm moulded, 11 cm sided) called S3. Its function was to strengthen the cohesion of the firstfuttocks and to uphold the framing.

of maintaining and reinforcing the frame where the floortimbers overlapped with the first-futtocks. At mid-ship, the first clamp S1 was also used to support the buttresses or ‘coignets’ in Mediterranean vocabulary, which were arched between the clamp and the sister keelsons. Its marks can still be seen on the inner side of the clamp (fig. 96).

The clamp was assembled using a large flat vertical scarf over 60 cm long (fig. 97). The clamps were fastened to the frame with iron nails with a circular cross-section of 12 to 13 mm in diameter.

The moulded dimension of S1 varied between 21 and 24 centimeters, with S2 varying from 16 to 19 cm. The average sided dimension of the two clamps was 10 cm. They were partially notched on the upper face of the floor-timbers and first-futtocks. Nevertheless, none of the pieces of the frames were of the same height, hence the notches were shaped accordingly. When the lack of material was too great, wedges were cut into appropriate shapes to fit into the gaps and forced in with no other fastening than the pressure of the pieces between them (fig.96).

On the fore part of the wreck (see plan of excavation zone AF13/1): – On the starboard fore side, the degraded end of clamp S1 could be seen, located a short distance, about 35 cm, from the keelson. The trajectory of the clamp therefore got closer to the keel axis. At this fore end, it was also thicker in cross-section than in the central area: it measured 16 cm moulded and 12 cm sided. This clamp covered the first two floor-timbers V1P and V2P and reinforced their assembly with their respective firstfuttock G1P and G2P. – 25 cm further to starboard, there was the degraded end of the second clamp S2. Alongside this, the degraded end of the sill plank was visible 1.50m towards the first-futtock G4P, but with a gap between G1P and G2P. Clamp S2 had similar dimensions to clamp S1.

S1 and S2 were separated by just under 30 cm when covering the M22 to M24 frames. The S2 clamp was surrounded on its outer side by a wide sill plank (fig. 98). About two meters from the first two clamps S1 and S2, were the clamps of the first and second futtocks scarfs named S4 and S5, which had approximately the same dimensions as S1 and S2. 81

The Mortella III Wreck

Figure 96. The foot wale S1 and the bilge clamp S2 (Photo: Christoph Gerigk).

Figure 97. The clamps S4 and S5 (starboard side) (Photo: Christoph Gerigk).

82

The hull of the Mortella III wreck and its construction method

83 Figure 98. Longitudinal frame layout (Illustration: Arnaud Cazenave de la Roche).

The Mortella III Wreck 3.2.2.2. The ceiling and the sill.

Towards the forward part of the wreck: The degraded sill plank has been found at the fore end of the wreck where its width seems to have decreased a little; at this level it was about 25 cm. Its notches on the first-futtocks G2P and G3P, and G3P and G4P were visible but the filling planks had disappeared.

The ceiling. The presence of a ceiling was covertly observed because it was so fragile that it didn´t resist to the excavation. It was made of beech plank (fagus sylvatica) about 30 mm thick and was located in the most central part of the hull between the port and starboard S2 clamps. Beyond that, the inner part of the hull did not appear to be covered with ceiling. As far as we could understand from its arrangement, the ceiling was placed transversely in the center of the wreck, between the keelson -where the end of the planks were accommodated in the chamfer of a joistand the clamp S1. Then, they were placed in a longitudinal direction between S1 and the so-called sill, in oak, as the third and final plank of the ceiling.

3.2.3. The planking and its fastening method to the frames Without dismantling, the planking could only be partially observed. The length of the strakes, for example, could not be measured. The information that could be gathered came mainly from information that could be accessed between the first-futtocks on the starboard side. 3.2.3.1. The wood

The sill. This laid longitudinally alongside the entire internal length of the hull. The sill plank consisted of an oak piece about 30 centimeters wide and 6 cm thick. Its inner part rested along the S2 bilge clamp. It was given notches on its outer part to allow 6 cm thick filling planks to be inserted between the frames. These were leaning on the planking in a slope of about 40° and their lower and upper edges were beveled to rest flat on the planking and flush with the upper side of the sill. A small rectangular piece of wood measuring about 15 x 15 cm placed above each first-futtock sealed the whole. The careful adjustment of these pieces ensured the keel/floor-timbers insulation (fig. 99).

Like all the structures of the framework, the planking was made of sessile oak. However, the only surviving remains were that of the lower part of the hull. In the case of the Red Bay wreck, for example, the planking consisted exclusively of oak up to strake 25, then other species were used (LOEWEN, 2007, 111). If the Mortella’s ship had been built similarly, other species could have been used in the upper works. In the case of the planking of the Villefranche nave, a mixture of several species has been observed (GUEROUT, RIETH and GASSEND, 1989, 63 and 65).

Figure 99. The sill: the inboard side next to the bilge clamp S2 (Photo: Christoph Gerigk).

84

The hull of the Mortella III wreck and its construction method

Figure 100. Carpenters’ mark (Photo: Christoph Gerigk).

the hull of the Villefranche wreck: at Villefranche, the first fourteen strakes were 12 cm thick, those following 10 cm thick until the thirtieth strake. The distance from the keel at which our samples were taken puts us at the level of this second group of strakes. Hence their thickness can be compared: 8-9 cm at Mortella III / 10 cm at Villefranche. The planking of the Red Bay wreck was half as thin: 5 to 6 cm at the level of the first 18 strakes, a ratio that can be explained by the much lower tonnage of this ship in comparison with that of the Mortella III or Villefranche.

After sawing, the outer and inner faces of the planks were regularly shaped, using an adze-like tool. There are also clear and deep marks of a caulking iron between the edges. Like most of the pieces of wood that could be studied during this excavation, many include the core of the log from which they originate, thus confirming the relatively small diameter of the trees from which the planking is made. 3.2.3.2. Morphology and dimensions

On the fore end of the wreck. At the level of the fore crutches on the port side the remains of three strakes were found. The width of the two planks following the garboard was 20 cm each. The upper strake consisted of two planks between 8 and 10 cm thick. Two carpenter’s marks were observed on the inner side of the first upper plank. The first one was located just before the extremity of the sternest plank. The second one was at the level of the V4P floortimber (see Figure 100). These two marks are similar; they have a VI shape which probably reflects a system used to assist the assembly of the planking. It is possible that this was used to number the strakes moving forward from the keel, with strake I being the garboard.

This may explain the narrow width of the planks, which only varied from 16 to 20 cm, whereas, using the example of the Red Bay wreck, the average width of the planks was 33 to 34 cm and of which only 8% had a width between 10 and 20 cm (LOEWEN, 2007, 111). The plank strakes were carvel jointed, with a slight beveling of the edges on their outer part, so that the joint of each plank was opened by about 10 mm on the outer side while the edges met on the inner side. These are classic features of the carvel construction system, which are well documented in Red Bay (LOEWEN, 2007, 112 and 113). Between the tailframes. The thickness of the planking, found in the central part of the hull, ranges from 8 to 9 cm at first-futtock level. This thickness can be compared to

A structure that would correspond to a port end planking panel (see fig. 101): Two meters south of the keel end a

Figure 101. Oblique cutting line from the fore end of the strakes of planks (Photo: Christoph Gerigk).

85

The Mortella III Wreck hole 2 to 3 cm in diameter was made on the outside of the planking to allow its head to be driven into the wood.

wooden structure composed of a series of planks assembled edge to edge appeared under a few centimeters of sediment. A series of timbers – four of which were cleared – of 15 to 17 cm sides were placed perpendicularly. Strongly eroded by wood borers on their upper side, they were arranged at intervals ranging of 15 to 35 cm. There were also many traces of carbonization of the wood.

The diameter of the nails was between 10 to 12 mm. An original feature of this fastening system is that the nails were passing through the planking and the frame, and their ends were folded down over 3 to 4 cm on the inner face of the frame. Many little concretions located on the inside face of the frames evidenced this through nailing and kept the imprint of the clenched end of the nails (fig. 103). The length of the nails used can be estimated to be about 28 cm long by combining the thickness of the planking and the frame with the length of the folded end (9 cm + 15 cm + 4 cm = 28 cm).

The size and arrangement of the wooden pieces of this panel suggested it was a hull structure composed of planking attached to remains of frames. A particular detail of this structure has made it possible to refine the interpretation. We noticed that the end of the planks, on the side of the axis of the building, had been sawn along a regular line that is not at right angles to the axis of these planks. Instead, they were following an oblique line inclined at about 125° with relative to this axis, or 35° with relative to the vertical axis. This observation suggested that this oblique cut corresponds to the end of the plank strakes of the ship that were inserted into the rabbet of the stern post.

The use of iron nails and the absence of tree nails set out a typical fastening system of the Mediterranean technical tradition of construction. Actually, we know that the dowelling of planks was not in use in the Mediterranean in the Modern era. This characteristic brings the Mortella III wreck closer to that of Villefranche and separates it from wrecks of Atlantic tradition, such as the MaryRose, for example, whose planks were pegged, or the Red-Bay wreck, whose fastening system used both iron nails and tree nails. We will come back to these aspects in

3.2.3.3. Fastening method of the planking to the frames The planks were attached to each frame by two iron nails located near their edges. For each nail, a circular pilot

Figure 102. Nailing traces on the planking (Photo: Arnaud Cazenave de la Roche).

Figure 103. Right, Concretion formed by the tip of a clenched plank nail Left, Impression of the nail left in the concretion (Photo: Arnaud Cazenave de la Roche).

86

The hull of the Mortella III wreck and its construction method ‘La clavazón se debe advertir que para la fortificación de las naos ha de revitar, y en caso que no pueda ser toda, se ha de entremeter clavazón que alcance al revite, y en las cabezas de las tablas, donde se ponen al tope, se han de clavar con clavos que reviten en el madero de popa y de proa, porque en el de la junta suele rajar, y la cuenta con que se debe clavar la tablazón del costado es, que si la tabla tiene de ancho un jeme, que viene a ser una tercia de codo, se ha de dar uno en el canto bajo de la tabla y otro en el alto; y si fuese el ancho de la tabla de más de tercio de codo, hasta llegar a medio, será bien clavarlo con tres clavos; y si es más ancha que de medio codo la tabla, será bien que se clave a hecho en todos los maderos con tres cla­vos en cada uno; y si es de dos tercias de codo, con cuatro clavos en cada madero,…’

Figure 104. Section of a hole left by nailing the planks (First futtock G 20) (Photo: Arnaud Cazenave de la Roche).

Translation : ‘It must be stressed that to fortify ships, the nails must be folded down, and in case the not all can be, care must be taken to ensure that its length enable it to be folded down and especially at the level of the joint of the ends of the planks, the nails must be folded down, because the joint has a tendency to split[the wood]; the calculation of how the plank is to be nailed is that if the plank is one ‘jeme’ wide, which represents one third of a codo, a nail must be placed near the lower corner of the plank and another near the upper corner. If the width exceeds half a codo, it should be nailed with three nails; and if it is more than two third of a codo, with four nails in each plank,…’

the description of the Mediterranean shipbuilding model illustrated by the wreck of the Mortella III. In addition, the option chosen by the shipbuilders of the Mortella III to fasten the planking with iron nails crossing the whole frame-timbers is a significant difference from the technique used in Villefranche-sur-Mer or Red Bay where iron nails were driven in without reaching the furthest side. We would like to mention two Spanish texts at this point that deal with this subject and that show that the technique of clenched through nails was common in the Spanish shipbuilding tradition of the Modern era.12 The first was written by Juan de Lasalde to the King of Spain in 1581 (LASALDE, 1581, Fº299-301). It deals with the technical characteristics of the construction of eight galleons. He tells us that this fastening technique was recommended in Spain in the 16th century:

In the language of the time, the term ‘revitar’ is synonymous with ‘remachar’, so it also means ‘to fold the tip’. As we can see, this practice is recommended in order to give more strength to the whole structure and limit the appearance of slits in the wood. The use of two iron nails placed near the corners was also recommended to fasten the ends of the planks, whose width is equal to or less than one ‘jeme’, a measure equivalent to one third of a codo, i. e. slightly less than 20 cm.13 Then, for a width over two third of a codo, four nails were recommended. .

‘Que cada uno de los dichos galeones lleve, desde la quilla hasta la primera cinta, pernos de fierro en lugar de gavillas de palo, barrenado con tres barrenos cada agujero; y que haya de pasar cada perno dos dedos más del grosor que tuviere todo costado del dicho galeón, de manera que por dentro se pueda remachar,….’

Later in the text, the author compares the fastening systems of the planking with nails of circular cross-section to those made with square cross-section:

Translation: ‘Let each of the said galleons have iron nails instead of tree nails from the keel to the first wale, each hole to be prepared by means of three pilot holes, and let each nail be two inches longer than the thickness of the hull, so that it is possible to clench the point inside…’

’En el reino de Levante hacen la cla­vazón redonda y de un mesmo grueso desde la cabeza hasta los dos tercios del clavo, y el otro tercio lo hacen esquinado, para que mejor se pueda clavar y revitar; y razón que dan para hacer la clavazón redonda, que no rasga tanto la madera y entra más ajustado en el barre­no, y es más estanco y no se carcome y gasta tan presto donde hay humedad, respecto de que el barreno por re­dondo está más lleno de fierro y ajusta más.

Here, the term ‘remachar’ indicates the action of folding the tip of the nail. The second text, which is dated from the middle of the 17th century, is an anonymous treatise on building hulls for ships and was published in Cesareo Fernandez Duro’s Disquiciones náuticas (FERNANDEZ DURO, 1996, vol. VI, 243). This is the part that deals with this question:

13 Until 1590, the Spanish codo normal is equivalent to two-thirds of a vara castellana, or 0.5573 m. From 1590, a Royal Ordinance replaces the codo normal or codo castellano by the codo de ribera which is equivalent to 2/3 of a vara plus 1/32, that is 11/16 of vara, that is to say 0.5747 m.

12 These two texts were kindly reported to us by Cayetano Hormaechea, author of Los galeones españoles del siglo XVII, op. cit.

87

The Mortella III Wreck De esta opinión era el capitán de la maestranza de la armada Real Vicente de Bartolosi, y así en todos los navíos que fa­ bricó en Vizcaya les hizo hacer la clavazón redonda, y en todo lo demás de altos y cubiertas esquinada; y en la carena que se dio al galeón San Juan Evangelista, que fabricó el dicho Bartolosi, se experimentó que en una tabla, queriéndole sacar los clavos que tenía con el pie de cabra, los que eran esquinados se sacaban con más facilidad que los redondos, aunque todos eran de un largo y de un grueso; y aunque alguno podrá decir que siempre en Vizcaya y en otras partes se ha fabricado con clavazón esquinada, con la experiencia se ha visto ya la bondad que tiene la redonda, por las razones que tengo dichas, y porque para la clavazón redonda es me­nester más delgado el barreno, con que llama más y ajusta mejor los maderos.’ (FERNANDEZ DURO, 1996, vol.VI, 246, 247)

Nailing of the planking to the M27 master frame. During the study of the master frame, the nailing of the planking was studied in detail. The planks were attached to the M27 frame by means of 2 circular nails that passed through them, the tips of which were folded down on its upper tower face. Each nail had a length of 28 cm and a diameter that fluctuated from 10 to 13 mm. The irregularity of the morphology of the nails is a characteristic which – at the time of the study of the M27 frame – could have cast doubt over their exact shape. The irregularity of their diameters gave the nails a shape whose roundness was sometimes doubtful. Some nails, for example, seemed to have an octagonal shape that may have led us to believe initially that they were square nails. These doubts were dispelled after a series of mouldings were made using polyurethane mastic (Figs. 105 and 106).

Translation : ‘In the reign of the ‘Levant’ (Eastward), they made the nails using a cylindrical shape of consistent thickness from the head up to two thirds of the nail, and the other third using a square shape so that it can be better nailed and folded down; and the reason they give for doing round nailing is that it scratches wood less and fits better in the pilot hole, and is more waterproof, less subject to the development of shipworms, and that moisture wears it out less as the circular hole is filled with more iron and fits better.

Number of nails counted on the M27 frame: – – – –

Floor-timbers: 18 First-futtock G27A: 31 First-futtock G27 B: 24 Second-futtock A27: 7 (not significant)

3.2.3.4. The forward end of the planking These were the remains of a fore part of the hull of about 6 m², composed of a series of planks and the remains of seven degraded frames, two of which had practically disappeared. The whole was very eroded, gnawed by woodworms, and the frames are largely charred, especially in their upper part (fig.108).

This opinion was that of the captain of the Royal Navy shipyard, Vicente de Bartolosi, and so he adopted round nails for all the ships he manufactured in Vizcaya, except for the upper parts and decks where a square section was used. Bartolosi experimented with extracting nails with a crowbar from a plank, and those of square section were removed more easily than those of round section despite identical their dimensions were; and although some may say that in Vizcaya and elsewhere, nails have always been made in a square shape, experience has shown the quality of round nailing for the reasons I have stated and because the pilot holes are thinner and the wood fits better.’

The particularity of this piece of hull is that the ends of the planks on the west side are beveled over about 2 meters. These remains were identified as the end of the plank strakes on the port side that were inserted into the stem post rabbet. Examination of the ends of these plank strakes reveals several important elements:

This text is notable because it specifies that circular nails were manufactured and used in the ‘reino de Levante’. This geographical term refers to the Mediterranean coast of the Iberian Peninsula and its Mediterranean territories (Sardinia, Sicily and Naples). The author then indicates that, in contrast to the builders in the provinces of Levante, the builders of Vizcaya – and elsewhere – used square section nails.

– The angle formed by cutting the boards on an incline is 140° (fig.107) – The edge of the end of the planks is also beveled – The end of each plank was secured with nails of circular cross-section – The 2 meters on which the ends of the planks were cut in an inclined plane and form a slightly curvilinear line (fig.108) – This line drawn by the end of the planks formed an angle of 40° with the axis of the frames

In addition, the argument put forward to justify the superiority of circular nails is that they do less damage to the wood and deteriorate slower than square nails. They were also recommended due to their close fit to the ‘barreno’, a term that refers to the pilot hole that was made throughout the thickness of the wood by means of a drill, used to facilitate the driving of the nails.

The assemblies were too degraded to be able to observe the fastening method. However, the nailing of the planking to the frame remained visible. As in the central part of the ship, it was done with circular nails of 8 to 10 mm in diameter.

88

The hull of the Mortella III wreck and its construction method

Figure 105. G27B: Moulding of a plank nail (polyurethane mastic moulding) (1) (Photo: Arnaud Cazenave de la Roche).

Figure 106. G27B: Moulding of a plank nail (polyurethane mastic moulding) (2) (Photo: Arnaud Cazenave de la Roche).

The obliquity of the cut-out of the planking end was measured at each end and varied between 140° and 145° in relation to the longitudinal axis of the planking. This means that we were dealing with the lower part of the hull since, logically, the further we go down the stem, the more the plank ends form an open angle (fig.110) Figure 107. AF15/2 – Diagram of the fore end planking slicing (Illustration: Arnaud Cazenave de la Roche).

Unfortunately, it seems difficult to go further with this analysis due to the small size of the piece of hull that was found, combined with the lack of knowledge surrounding the shape of the bow. Thus it was impossible to determinate a precise position of these planks on the stem post. One clarification can be made, however. The remains of the seven frames were most probably first-futtocks, given that the two timbers MP4 and MP5 were attached on the fore faces of the upper pieces in accordance with all the first-futtocks placed forward to the master frame. We can therefore deduce that these two remaining wooden pieces that appeared in the upper part of the panel were the beginning of second-futtock.

Analysis and interpretation: One of the most interesting features of this wooden panel was the end of the planking, the complete clearance of which has allowed us to observe the continuous beveling of the edges of the planks ends. This bevel cutting makes it possible to consider that edges were fitted into the stem post according to a scheme that could be similar to that shown in Figure 24.

89

The Mortella III Wreck

Figure 108. AF15/2 - Photomosaic of the port side fore planking panel (Photo: Christoph Gerigk).

3.2.3.5. Caulking and sealing products The joints between the planks contained remnants of caulking material of a fibrous nature (fig. 111). The Mortella III ship’s hull was not covered with protective lead sheeting as the wreck at Villefranche-sur-Mer was (GUEROUT, 2005). On the other hand, the inner and outer sides of the planks were coated with an ochre-colored material that formed a 2 or 3 mm thick crust. The analysis of this material was carried out in 2012 by Carole Mathe (University of Avignon, IMBE UMR 7263/ IRD237). This analysis can be found in Annex V. It concludes with a mixed composition formed by three materials:

Figure 109. Hypothesis of fitting of the fore end planks into the stem rabbet (Illustration: Arnaud Cazenave de la Roche).

– Tar (obtained from resin heated in an oxidizing atmosphere) 14 – Pine resin

– Fat, probably animal The Spanish caulking treatise cited above mentions the use in Spain of a sealing product made from a mixture of pitch15 and sulfur:

The pitch is a substance widely used since antiquity, especially as sealant. It was used both to seal the surfaces of ceramics such as amphorae and hulls of boats. Two pitch preparation techniques are known. The first consists of collecting the pine resin and heating it. This technique described in detail by Pliny the Elder is called ‘gemmage’ in French. The second is the distillation of softwood branches. 14

The pitch would correspond to the French word brai which is tar solidified under the action of air.

15

90

The hull of the Mortella III wreck and its construction method

Figure 110. Fore end of the planking of a 16th century ship (Illustration: Arnaud Cazenave de la Roche).

Figure 111. Edge of a plank with remains of caulking products (Photo: Arnaud Cazenave de la Roche).

‘Habiendo de ser el viaje de las naos a la Nueva Es­ paña, puertos de Honduras, Santo Domingo y otras par­tes donde hay broma, se deben emplomar las naos; y para mayor seguridad, será más conveniente que debajo del plomo lleve su lienzo alquitranado, porque suele ro­zarse el plomo con los cables y otras cosas, y queda el lienzo pegado a la tabla, con que resiste que no se pase de broma. También se usa para defensa de ella darles a las naos un betún que se hace de azufre molido y brea engrasada, con que se hace un género de costra que re­siste la broma.’(FERNANDEZ DURO, 1996, vol. VI, 262).

Figure 112. Outside face of a plank covered by sealing material (Photo: Arnaud Cazenave de la Roche).

91

The Mortella III Wreck Table 11. Summary of the dimensions of the longitudinal framework Measures in meters Keel 1 (upper)

Lengths

Moulded

Sided 0,26

25

0,24

Keelson

 

0,20 à 0,22

aver. 0,14

Clamps

 

average 0,16

aver. 0,14

 

0,06 à 0,29

 

average 0,14

Keel 2 (lower)

Planks

0,46

0,20

0,08 à 0,09

Note: the measurements taken on the planking and keel at mid-ship.

Translation: ‘Ships travelling to New Spain, the ports of Honduras, Santo Domingo and other places where shipworms are found must be sealed; and for greater safety, it will be desirable that their hulls be coated with a tarred substance under the lead because the lead is worn out in places by ropes and other things, and the tar remaining glued to the planking, it prevents attacks by shipworms. It is also used to protect the naos with a bitumen made of grinded sulfur and greased pitch, with which a kind of crust is made that resist to the shipworms’

iron nails and treenails has been highlighted as a constant technical practice both for fixing the floor-timbers to the first-futtocks and for fixing the planking to the frame (characteristics 1 and 2 in T.J. Oertling’s table). The system of fixing these same timbers observed on the Mortella III wreck using iron nails and the absence of treenails differs radically from the Atlantic scheme and thus constitutes a ‘technical fingerprint’ typical of Mediterranean shipbuilding. Indeed, to our knowledge, no known wreck with a Mediterranean constructive tradition is exempt from this rule.

3.3. Fastening system for the timbers 3.3.1. Iron nails and absence of treenails

The written documents do not seem to shed any light on the advantages and disadvantages of using nails and treenails to join the frame components. Therefore, we don’t know the reasons that may have led the Mediterranean shipbuilders to opt for the exclusive use of iron nails to fasten the floor-timbers to the first-futtocks and renounce the use of treenails. Conversely, we cannot establish the advantage that the Atlantic shipbuilders saw in the joint use of the two fastening systems. Nevertheless, the texts shed light on the reasons that led the Mediterranean builder to use iron nails rather than treenails to attach the planking to the frame. First of all, it is the anonymous Spanish treatise published by Fernández Duro at the beginning of the 17th century which provides an explanation (FERNANDEZ DURO, 1996, vol. VI, 243): nailing would make it less easy for shipworms to attack the hull. This is mainly due to the fact that shipworms dig their holes following the grain of the wood, hence the danger of treenails crossing the hull of the ship from side to side. This motif was also mentioned at the beginning of the 17th century by the Portuguese author João Baptista Lavanha in his ‘Livro primeiro d’arquitectura naval’ (LAVANHA, 1608). The author indicates that the use of treenails to fasten the planking to the framing is only used in cold water nations because those with warm water are conducive to shipworms. They would be quickly destroyed by them and would give way to waterways.16 This is an attractive explanation; the xylophages’animals of the terenidae family are more commonly found warm waters and, as a result, pose a much more serious threat to ship hulls in the Mediterranean than in European Atlantic waters.

Beyond the type of scarf, both the type of attachment for the framing timbers and that used to attach the planking to the framing seem to constitute a fairly reliable technical marker of differentiation between the two Mediterranean and Atlantic traditions. For the latter, the combined use of

Figure 113. Sketch of the attachment of the Red-Bay wreck planking to the frames (Courtesy of: Carol Pillar, Parks Canada).

Shipworms attacking the wood by following its grain, the use of tree nails to fasten the planking exposes it particularly to waterways.

16

92

The hull of the Mortella III wreck and its construction method

Figure 114. Impression of the nail tips on the inner side of the frames of the wreck of Nossa Senhora dos Martires (Photo: Filipe Vieira de Castro).

However, this argument raises questions, at least as far as Iberian shipping is concerned, if we take into account the large volume of shipbuilding in modern times destined for travel in Asia, by the Portuguese and the Americas, by the Spanish. H and P. Chaunu showed the importance of shipbuilding, particularly in Biscay, for the manufacture of ships for the warm waters of the American coasts in the 16th century (CHAUNU, 1955-1960). Great navigators and discoverers such as Christopher Columbus and Hernan Cortés wrote about the considerable damage inflicted on their ships by shipworms.17 Under these conditions, one can therefore question the reasoning behind the consistent use of treenails in Ibero-Atlantic construction if they facilitated shipworms to attack the hull.

to the ‘Ponante’ (Mediterranean) tradition which mainly used nails of quadrangular shape (FERNANDEZ DURO, 1996, VI, 247). We will not go back over the arguments that – according to this author – motivated this choice on the part of Mediterranean builders. We will just outline that the archaeological evidences seems to validate him. As we have seen, the nails used to fasten the pieces of the Mortella III wreck were circular. This was also the case for the wreck of Villefranche, as opposed to the nails of the Red Bay wreck, which were quadrangular. A detail still needs to be restated. The anonymous author of the treatise adds that in the ‘Levante’, it was customary to use circular nails, the first third of which is quadrangular, which facilitated the folding of the tip. In the case of the Mortella III wreck, the irregularity and complexity of the impressions left by the nails in the wood initially led to the suggestion that nails of this mixed nature could have been used. But the circular impressions left by the tip folded down on the frame (fig. 103) led to the rejection of this hypothesis.

3.3.2. Circular section of the nails In the part of this chapter dealing with the fastening of the structural pieces of the wreck, a passage from the Spanish treatise of caulking cited above states that the circular shape of the section of nails used to build the hull was a ‘Levante’ (Atlantic) characteristic, as opposed

To conclude this part on the nature of nails, it therefore seems that their circular morphology can be added to the list of ‘technical fingerprints’ that could suggest a Mediterranean constructive tradition. However, this proposal must be put forward as a hypothesis and with caution due to the small number of sites that still limits statistical studies for the Mediterranean. In the current state of our research, the nails used on wrecks with a Mediterranean constructive

In the description of his fourth voyage, C. Colomb writes: ‘In April, the ships were devoured by the shipworms and they could no longer support themselves on the water … I left, on behalf of the Holy Trinity on Easter night, with rotten ships, gnawed by shipworms and all pierced with holes … ‘. (COLOMB, 1979, 196-202). Hernan Cortes, for his part, wrote in 1519 that he persuaded ‘the pilots to report stating that the ships anchored in Veracruz had suffered severe damage from violent gales, and that the worms had gnawed at their hulls, that most of them were unable to support the sea, some even staying afloat.’(PRESCOTT, 1843).

17

93

The Mortella III Wreck tradition nevertheless seem to meet this characteristic. One exception has been the Ottoman wreck of Yassi Ada (16th century), whose planking was fixed by quadrangular nails, as well as its floor-timbers/first-futtock joints (LABBE, 2010, 63 and 72). 3.3.3. Clenched nails tips An original feature of the nailing observed on the Mortella III wreck is the folding of the nail tips used to attach the planking to the frame. This procedure undoubtedly strengthens the cohesion of the strakes on the frame. Does this nailing technique have a Mediterranean origin? The Spanish caulking treatise quoted earlier suggests that it was a common practice in the Mediterranean in the period since the reason given for the use of nails by the ‘Levante’ nail manufacturers with a quadrangular section, was precisely their suitability fold the tip. However, this technique, described by the Spanish terms ‘remachar’ or ‘revitar’, is not described as exclusively ‘Levante’. In fact, archaeology shows a relative disparity in cases and does not allow – for the moment – a clear conclusion to be reached: – The few wrecks of Mediterranean constructive tradition on which the detail of the nailing could be observed do not follow the technique used on the Mortella III wreck. On the wreck of Villefranche-sur-Mer, on that of Calvi I or on that of Yassiada, the nailing of the planking was driven into the frame. On this last wreck, however, the tips of the three nails that joined the floor-timbers to the first-futtocks were clenched, which was not the case on the Mortella III wreck. , This is excepting the timbers of the master frame whose holes when passing through these scarfs suggest that the tips were clenched, although the impressions are not clear. – There are also a heterogeneity of cases for wrecks with Atlantic traditions: the wreck of Red-Bay, for example, had a planking with the nails inserted into the frames wood, while another ship of Ibero-Atlantic tradition, Nossa Senhora dos Martires, although later (1606) and of similar dimensions to the Mortella III, revealed a fastening of its planking on a model comparable, although only a part of the nails were cast through the frames (CASTRO, 2005).

94

4 The attributes of the hull: The mainmast step, the pump, and the rudder 4.1. The main mast-step

of the wood. Their outer edges had been beveled to form a chamfer, possibly to fit a ceiling.

The base of the main mast-step was uncovered in 2012. It is an architectural ensemble located on the center of the keel whose function is to firmly attach the mast-foot. It was composed of two strong longitudinal timbers of sessile oak wood, also named ‘sister keelsons’ or escasses, in the French Mediterranean language, and twelve transverse buttresses in juniper (Juniperus communis), six on each side. As we will see, this technical mast-step system is of a typical Mediterranean tradition. It has the particularity of being independent of the ship’s structure and of being grafted onto it.

These longitudinal timbers were provided with two keys in their middle part, holding them together. They were laterally maintained by six buttresses on each side that were fitted into notches shaped on their outer faces. These notches were 3 cm deep, about 15 cm high and 11 to 17 cm wide. Finally, the sister-keelsons were fastened to the keelson by long nails crossing them laterally and finishing their trajectory with a lost point. 4.1.2. The buttresses

4.1.1. The longitudinal timbers of the mast-step arrangement (or sister-keelsons).

The first side reinforcement buttress of a series of six (only four of which were still preserved) appeared at the level of the floor-timber V23. These buttresses were roughly trimmed and their ends shaped to fit them into the notches of the outer sides of the sister-keelsons, on the one hand, and on the inner side of the foot-wales, on the other. On the port side the foot-wale was no longer present, but it was on the starboard side (S1) and its notches could be recorded. The buttresses were just over 80 cm long and had a cross-section of about 18 cm. The cross-section of their ends had been reduced to 9 cm so that they could be slid into the notches. It should be noted that the intervals

The port sister-keelson was 25 to 28 cm sided and 20 cm moulded. The sided size of the starboard sister-keelson was 22 to 24 cm and 18 cm moulded. Their length was 5.10 m. Their bottom surface was notched—in the same way as the keelson—in order to lean on the floor-timbers, enclosing the outer faces of the keelson. They were intensively shaped, and many tool marks were visible on the surface

Figure 115. Diagram of the main mast-step (Illustration: Jesús Guevara (Aingurak)).

95

The Mortella III Wreck

Figure 116. Buttresses T1 to T3 arrange to wedge the port sister keelson (Photo: Christoph Gerigk).

between the buttresses was not the same throughout the length of the sister-keelsons, it increased steadily from aft to fore: 40 cm between the first two, then 46 cm, 49 cm, 54 cm and finally 56 cm.

were dovetail shaped wood pieces, about 40 cm long, 20 cm wide and 15 cm high. Located 65 cm apart, each one fitted into two mortises trimmed on the upper face of the sister-keelsons. These mortises were about ten centimeters deep. There was therefore a six cm day between the upper face of the keelson and the lower face of the keys (fig.118).

The buttresses were used for a lateral reinforcement of the sister-keelsons on which the spindle of the mast-foot transmitted considerable forces. In fact, they were made of juniper, a heavy and dense wood, a shrub common in the Mediterranean region. This species is characterized by high durability (‘a wood that lasts more than a century without being spoiled’) and excellent mechanical properties, including high elasticity. The choice of this species should therefore not be a matter of chance.

The keys fitted perfectly into the space carved into the sister-keelsons to accommodate them, except for the port dovetail, to which a wedge had been added to fill a gap in the wood (fig. 119). 4.1.4. A major Mediterranean technical ‘fingerprint’  The main mast-step device uncovered during this excavation appeared to be a remarkable and highly technical structure. The typology of this system seems to constitute a reliable marker of distinction between both technical traditions, Mediterranean and Atlantic.

We were surprised to find no trace of any fasteners for these pieces, neither nails nor dowels: they were simply placed in the notches that served as slides. They were not supported by any timber, since there was a space of several centimeters between their lower face and the floor-timbers underneath (fig. 117).

The Mortella III’s mast-step morphology is very similar to that of the wreck of Villefranche s/mer. It is perfectly identical in terms of its design and only a few details differ in terms of dimensions and organization (fig. 121): in Villefranche, the mast-step was a little larger, the length of the sister-keelsons was estimated at 5.60 meters (as a reminder, 5.10 m for Mortella III). There were 18 buttresses, 9 on each side, compared to 12 on the wreck of the Mortella III. Five of them were also concentrated around the mast-foot, while in the case of the wreck of the Mortella III, they were distributed along the length

On the other hand, under each buttress, there was important concretions caused by the diffusion of metal, which may suggest the initial existence of a metal structure under each buttress. But no visible remain could be interpreted. 4.1.3. The keys The second device for holding the mast-foot and joining the sister-keelsons together was made up of two keys. They 96

The attributes of the hull

Figure 117. Buttress T2 fit into the groove of the port sister keelson (Photo: Christoph Gerigk).

of the sister-keelsons with increasing forward intervals. In Villefranche, we also had the same system of floating buttresses which were only held by the notches in which they were fitted. Finally, the keys to the sister-keelsons were spaced 1.50 m apart in the case of the Villefranche wreck, while in this case they were only 0.65 m apart.

used on galleys and some Mediterranean ships, such as the chebecs. A French treaty on the construction of galleys dated 1691—which we have already mentioned— gives a good account of the organization of this system and the French Mediterranean names of its components (ANONYME, 1691).

This morphology main mast-step device is part of a Mediterranean architectural tradition inspired by that

This conception of this mast-step is very old, it was already used in antiquity and, actually, it can be found on

Figure 118. Sketch of the keys joining the two sister keelsons and grooves (Illustration: Arnaud Cazenave de la Roche).

97

The Mortella III Wreck

Figure 119. The keys joining the two sister keelsons (Photo: Christoph Gerigk).

Figure 120. Drawing of the main mast step of the Mortella III wreck (Illustration: Jesús Guevara (Aingurak)).

98

The attributes of the hull

Figure 121. Diagram of the main mast step of the wreck of Villefranche-sur-Mer (Illustration: Jean-Marie Gassend, GRAN).

Mediterranean wrecks of the classical period. It is also the traditional means of subjection of the main-mast used on galleys (GUEROUT, RIETH, GASSEND, 1989, 78). It differs from the system of the ships in the Atlantic area, which is traditionally characterized by a widening of the keelson in which a mortise is shaped to receive the spindle of the the mast foot. This is the case on the Red Bay wreck, for example (LOEWEN, 2007, 162 to 167). This

morphology is found on many other wrecks of Atlantic building tradition. It is ‘trait’ number 7 of the Oertling table classification (OERTLING, 1998, 234). We do not know of any text that sheds light on the origins of these technical traditions and what may have motivated their adoption. However, some remarks can be made:

Figure 122. Diagram of the main mast step of the wreck of Red Bay (Courtesy of: Carol Pillar, Parks Canada).

99

The Mortella III Wreck • From a practical point of view, the Atlantic tradition requires very large timbers to be able to widen the keelson in sufficient proportions to carve a mortise capable of receiving the spindle of the mast foot. It therefore implies a supply likely to provide pieces of wood from trees of large diameter. The Mary-Rose (Portsmouth, 1545), for example, had a keelson bulge of about 80 cm to form the mast-step mortise (Fig. 123). The one for the Red Bay wreck came from a tree with a trunk greater than 60 cm in diameter at a height between 5 and 8 meters. Brad Loewen therefore sees this likely to be an indicator of the difference in supply possibilities between the Atlantic area, where forest resources were abundant, and the Mediterranean, where they were scarcer (LOEWEN, 2007, 167). The Atlantic technique also requires very intensive work on the timber in which the mast-step is literally sculpted. It finally involves a big loss of wood. However, in the end, it has the advantage of having a single timber mast-step which is an integral part of the ship’s structure and of maintaining the spindle which does not depend on parts fasten together. • The Mediterranean-type mast-step system to which the wreck of the Mortella III belongs, characterized by the enclosure of the keelson by two timbers, the sisterkeelsons, logically seem to have a lower mechanical strength than the ‘Atlantic’ system. In this case, the cohesion of the pieces depends on the quality of the fastening work.

and/or tree-nails to the floor-timbers on which they rest. In the Mediterranean case, the buttresses are longer (90 cm for the Mortella III) and are not fixed to the floors, which, moreover, are not necessarily located directly above them. They are simply arched between the timbers. In the case of the wreck of the Mortella III, the builder also took care to use a species other than oak, in this case juniper, a very specific and unusual species characterized—in addition to its good mechanical properties—by its high elasticity. In the end, everything seems to be aimed at organizing a system with a certain degree of flexibility, playing the role of a sort of transversal ‘damper’ as opposed to the model observed on wrecks of the Atlantic tradition, whose maststep system appears more rigid.

It should be stressed here that in both systems, sister-keelsons and mortised keelsons, a lateral reinforcement is placed on each side of the mast-step, arched between the keelson or sister-keelson and the first clamp called foot-wale. However, the size of lateral buttresses seems proportionally more important in the Mediterranean system, perhaps reflecting the greater need for this lateral reinforcement. Its ‘floating’ structure also seems to reflect a difference in nature: in the Atlantic case, the buttresses are relatively short (40 cm for Red Bay, 50 cm for Mary-Rose) and are fixed with nails

The components of the bilge pump system were located in the immediate vicinity of the main mast-step, as it is usual, on the port side. First of all, the remains of the pump well’s planks were discovered. It consisted of four planks of chestnut wood, 3 cm thick and approximately 30 cm wide (fig. 124).

4.2. The pump Elements of the pump and pump well participating in the drainage system were uncovered during the 2010 excavation campaign. As this device is closely linked to the structure— of which it is sometimes a part, as is the case for Atlantic tradition ships—it seemed necessary to associate it with the architectural study and include it in this work. There is rarely an opportunity to observe and study a bilge pump from this period. Only half a dozen wrecks, including the Mary-Rose (1545), Red Bay (1565) and Lomellina wrecks (1516), have brought this type of device to light.

These planks were apparently joined on their four corners by a squared piece of wood of approximately 5 cm in crosssection located in the inner side of each corner (fig. 125).

Figure 123. Diagram of the main mast step of the wreck of the Mary-Rose (Illustration: D.M. McEvolgue).

100

The attributes of the hull

Figure 124. Remains of the pump well where the debris of the upper deck laid (Photo: Arnaud Cazenave de la Roche).

Figure 125. Wooden pieces joining the end of the planks of the pump well (Photo: Arnaud Cazenave de la Roche).

101

The Mortella III Wreck and V22, was designed so that the lower part of the valve was located about 5 cm from the planking, thus allowing an abundant passage of water (fig. 131).

There were two components, a foot-valve and a pumptube, that have studied before putting them back in their original position. We found no trace of the internal pumping system, flaps, shaft, etc.

This process is therefore somewhat different from that of the Villefranche wreck (fig. 132 a) or Red Bay wreck (fig. 132 b), whose lower valves were equipped with lateral holes designed to allow water to pass through.

4.2.1. The foot-valve The pump-valve, called foot-valve, consisted of a single wooden piece of square shape with 15 to 23 cm sides and a central orifice of about 10 cm in diameter (fig. 129 and fig. 130). The angles of two vertical sides were chamfered over a width of 3.5 cm. It had a 4 cm thick collar that exceeded a little less than 4 cm on two sides, allowing it to be supported on two floor-timbers. In our case it was found inserted between M21 and M22 (fig. 126).

4.2.2. The pump-tube (fig.133, fig. 134, fig. 135) The second component, the pump-tube itself, was located in the continuity of the base. It was discovered inclined towards the stern, and levelled in its upper part, remaining from its height only a maximum of 62 cm. The tube was composed of three wooden pieces 7 cm thick, with a curved surface, assembled by a system of tenons and mortises located along the edges. The whole was apparently tied by a thick rope that was no longer in place. But traces of it could still be seen on the outside of the walls which undulating surfaces indicated a very strong belting. This was undoubtedly intended first of all to maintain the cohesion of the parts and the waterproofness of the tube, but also to avoid cracks of the wood when it dried: in his Dictionnaire historique de la Marine, Alexandre Savérien notes in this regard: ‘… they are surmounted (the pumps) with ropes to avoid that they dry too much and ensure like this they do not split.’ (SAVERIAN, 1758, vol. II, 240).

Three holes were counted, two 10 mm in diameter on the collar, in an inclined plane, suggesting attachment to the floor-timbers, and one 6 mm in an inclined plane on its lower part. In the pumping system we are dealing with, the footvalve had a dual function; the first was to collect water by means of lateral holes carved in its base. The second was its retention, when the piston was lowered, by means of a leather flap. In our case, the flap disappeared, but curiously enough, no orifice allowing the passage of water was visible. The morphology of the wooden piece, provided with a collar resting on two floor-timbers V21

Figure 126. Foot pump between V21 et V22 (port) (Photo: Arnaud Cazenave de la Roche).

102

The attributes of the hull

Figure 127. Lower end of the pump tube in situ (Photo: Arnaud Cazenave de la Roche).

Once assembled, the pump-tube had a diameter of about 30 cm and an inner diameter of 15 cm. A collar of about 22 cm in external diameter narrowed it at its base (fig.133, fig. 134. Fig. 135). A question remains as to how this pump-tube was fixed to the pump-valve whose shapes in diameters did not coincide? Was there a connection piece between them?

‘From the period of 1500-1840, the three main types of pumps used on ships were the burr pump, the suction or common pump, and the chain pump. The burr pump was in general use in the 16th century, but declined in the first quarter of the 17th century in favor of the commun pump.”…“The burr pump was a very simple machine that was used on ships on the early 16th century, and it almost entirely disappeared by the early 17th century.’

The pump device uncovered is, in its principle, similar to most of those found on wrecks of the same period. It probably corresponds to the ‘burr-pumps’ typology, ‘pompes soulevantes’ in French, currently in use in the 16th century. This French term was in used in 18th century technical dictionaries to describe the pumping device that raises water by the action of a piston, as opposed to the system that consists in causing a suction phenomenon.1 The water was drained by the action of a shaft moving inside the pump-tube, at the end of which were placed a series of leather discs whose diameter decreased in order to obtain a cone. The first disc, slightly larger in diameter than the core of the pump-tube, curled up as it descended, allowing water to pass through. Rigid when climbing, due to the leather discs that followed it, it retained water and allowed it to drain upwards. The characteristics of this device fall within the first of the three types of pumps described by Oertling in his thesis on bilge-pumps, which is described as follows (OERTLING, 1984, III, 30):

Beyond its chronological framework, the Mortella III’s pump has some particularities that deserve to be mentioned: The first, as we have already pointed out, is the singularity of the foot-valve, which was free of holes on its sides. The second concerns the pump-tube made up of an assembly of three curved wood pieces. However, Thomas Oertling notes that all the wrecks he has studied are equipped with a pump-tube made up of a single wooden timber whose core has been drilled.2 It is in this respect interesting to observe that the only exception he finds to this rule is the Spanish shipbuilding, as William Dampier explains in 1683 that its pumping devices are made of wooden assembled pieces (DAMPIER, 1729, 443): ‘And our pumps being faulty, and not serviceable, they did cut a tree to make a pump. They first squared it, then

These two types of pumps are described in the Grand dictionnaire des Arts et des Sciences, vol.4, Amsterdam, 1696, p.144.

In some cases, such as Red Bay, the pump-tube consists of two single pieces that adjust lengthwise.

1

2

103

The Mortella III Wreck

Figure 128. Location of the pump (Illustration: Arnaud Cazenave de la Roche).

sawed it in the middle, and then hollowed each side exactly. The two hollow sides were made big enough to contain a pump-box in the midst of them both, when they were joined together: and it required their utmost skill to close them exactly to the making a tight cylinder for the pump-box; being unaccustomed to such work. We learnt this way of pump making from the Spaniards; who make their pumps that they use in their ships in the South-seas after this manner…’

Finally, he gave the example of a Spanish ship discovered in the early 1970s near Port Royal, Honduras3, where an assembled pump-tube was found4. However, it should be stressed that the pump-tube of the Red Bay wreck escapes to this construction technique. The third observation concerns the way the bilge pump is attached to its foot-valve. Brad Loewen notes that: ‘The technique of seating the foot-valve in a sump carved into the keelson and a floor timber has not only been observed on each of the 16th-century Iberian

Oertling adds: ‘The Spanish, therefore, were using this type of pumptube in the West-Indies and possibly in Europe, as well as, in the pacific in the 17th and 18th centuries.’

3 4

104

Thomas Oertling probably refers to the island of Roatán. Oertling, Thomas James, op. cit., p.21

The attributes of the hull

Figure 129. Drawing of the foot-valve (Illustration: Samantha Heitzmann).

105

The Mortella III Wreck

Figure 130. Photography of the foot-valve (Photo: Arnaud Cazenave de la Roche).

Figure 131. Location of the foot-valve upon the floor-timbers (Illustration: Arnaud Cazenave de la Roche).

Figure 132. a. Foot-valve of the Villefranche wreck b. Foot-valve of the Red Bay wreck (Illustration: Courtesy of: Max Guérout, GRAN/Laurie-Bourque, Parks-Canada).

106

The attributes of the hull

Figure 133. Inner part of the pump (Photo: Arnaud Cazenave de la Roche).

Figure 134. External part of the pump (Photo: Arnaud Cazenave de la Roche).

Figure 135. Drawing of the pump-tube (Illustration: Samantha Heitzmann).

wrecks investigated in the New World, but also on contemporaneous shipwrecks found in England, including the Mary Rose, the (possibly Iberian) Cattewater wreck, and the Rye A vessel.’5

was to support the pump on two frames, on Mortella III at least. Therefore, one can legitimately wonder whether these characteristics could be part of a Mediterranean technical tradition? 4.3. The rudder

This remark is of interest since the fastening of Mortella III’s foot-valve does not comply with this rule. Like at Villefranche, it was provided of a collar. Its function 5

During the 2015 excavation campaign was uncovered of a wooden panel located immediately behind the keel heel (square E5 of the grid of the general two-dimensional

Loewen, op. cit., 167.

107

The Mortella III Wreck

Figure 136. Excavation area AF 15/1 located at the fore end of the wreck (Illustration: Arnaud Cazenave de la Roche).

survey of the site) and identified as the remains of the ship’s rudder. It consisted of an assembly of 8 highly eroded pieces of wood forming a panel about 2.20 meters long, about 1 meter wide and 18 cm thick. It was laying flat on a muddy bottom oriented in a North/South axis (fig.136). This wooden panel had been identified during the 2007 appraisal but could not be identified then. Its proximity to the keel heel, its south-western end being located about thirty centimeters away, combined with the presence of large concretions fixed on its side enclosing hinges remains, finally made it possible to reach its interpretation. 4.3.1. The rudder device: definitions and vocabulary Before continuing their description, a review of the vocabulary related to the components of the steering system that we will discuss may be useful: The rudder is an assembly of timbers fixed to the stern by iron brackets (hinges called ‘gudgeons’ fixed to the stern and ‘pintles’ fixed to the rudder stock) which allows the ship to adjust its steering. The rudder blade usually composed of three pieces of wood assembled vertically. The first part is called the stock (or mainpiece), then comes the middle-piece. The later part is called the after-piece. In the 16th century, the rudder was currently called ‘gouvernal’ in French (NICOT, 1606).

Figure 137. Rudder parts (BONNEFOUX, 1848).

4.3.2. Morphology of the rudder blade

To the east, a timber (P1) 1.80 m long and 20 cm wide vertically formed the outer part of the panel (fig. 140). However, this piece didn’t reach the end of the panel, but was followed to the south by two square pieces of wood (P2 and P3) placed perpendicularly on a horizontal plane.

It is made up of an unusual assembly both by the shape of its wooden pieces and by their arrangement. However, the description of this remarkable architectural remain has been made difficult due to its poor condition. 108

The attributes of the hull centimeters in a notch in the lower end of the P1 timber. The assembly was held by 5 metal nails of a generally circular shape, 11 to 12 mm in diameter. They were driven in with a lost point from the outer part of the board to the heart of the P1 to P4 pieces. The pieces adjacent to P1 timber, namely P6, P7 and P8, were placed in a vertical plane and juxtaposed perpendicular to P1. However, they were very deteriorated and had many gaps. P2, P3, P4 and P5 timbers were placed in a horizontal plane forming an angle of 260° with P1. Morphologically, while P2 and P3 were straight and quadrangular in shape, P4 and P5 were characterized by a complex cut-out: P4 was embedded with P5 with a particular oblique assembly perfectly adjusted (fig. 140). The south edge of P5 was provided with two notches in which a piece of wood, which has now disappeared, was embedded. Apart from the nails that fastened the reinforcement piece we have mentioned, the system for fastening the timbers of the blade was provided by iron nails, bolts and brackets. Finally, it should be noted that the wooden pieces were made of oak, except the shole (P3) which was of a softer species (fir).6 4.3.3. The nailing The vertical wooden pieces were too eroded and incomplete to be able to certify the presence of nails. Nevertheless, horizontal timbers retained the trace of a strong nailing. The low piece P3 was fastened to P2 with six nails with a circular section of 11 to 13mm in diameter. In addition, as mentioned above, a reinforcement timber was fastened with five nails increased the connection between the pieces P1 to P4.

Figure 138. Photomosaic of the remains of the rudder (Photo: Arnaud Cazenave de la Roche).

4.3.4. The bolts In addition to the nailing of the pieces together, there was a horizontal and vertical bolting system that reinforced the strength of the whole: Despite their poor state of conservation, we could observe the presence of an orifice passing through pieces P1, P6, P7 and P8, revealing the initial presence of a bolt about 4 to 5 centimeters in diameter that joined these four vertical pieces together at a height of about 1.40 meters from the lower part of the rudder (fig.140 a).

Figure 139. Edge of the end of the rudder (Photo: Arnaud Cazenave de la Roche).

Vertically, a four-centimeter diameter bolt, located on the western side of the rudder, also crossed the horizontal pieces (fig. 140 b).

An assembly reinforcement between these three pieces was provided by a timber about 50 cm long by 25 cm wide cut in the shape of a whistle: its thickness was 8 cm on its outer edge and only 1 to 3 cm on its inner side (fig.139 and 141). Its thinnest part was finally embedded for 15

In his Elémens de l’architecture navale, Duhamel du Monceau writes this about the species to be used in the manufacture of the rudder: ‘The part of the rudder that touches the stern is of oak. The rest, called blade, is made of a lighter wood like fir.’ (Chapter II, art. 19).

6

109

The Mortella III Wreck

Figure 140. Layout of the remains of the rudder (Illustration: Arnaud Cazenave de la Roche).

110

The attributes of the hull

Figure 141. Bottom view: A diagonal cut plank holds the two timbers together (Photo: Arnaud Cazenave de la Roche).

4.3.5. The iron-works

concretions containing the hinges remains were present in the immediate vicinity of the panel. However, only one— the largest—was still in place on its west side (fig. 140, fig. 143 and fig. 144), located in the axis of the bolt previously mentioned. The inform nature of the concretion does not

A set of iron-works was, in addition to be the attaching device of the rudder to the stern, the third system for fastening the wooden pieces to each other. Several large

Figure 142. Outboard edge of the lower timber of the rudder (P3) (Photo: Arnaud Cazenave de la Roche).

111

The Mortella III Wreck

Figure 143. Concretions located on the West edge of the rudder (Photo: Arnaud Cazenave de la Roche).

Figure 144. Concretion enclosing a rudder pintle (pintle a. of fig.140) (Photo: Arnaud Cazenave de la Roche).

112

The attributes of the hull

Figure 145. Broken concretion enclosing a second rudder pintle (pintle b.) (Photo: Arnaud Cazenave de la Roche).

allow a precise description of this iron work. However we were able to interpret it as the remnant of a pintle: we note the presence of two horizontal metal strips more than 40 cm long—but probably initially longer—and 4 to 5 cm wide and one cm thick which enclosed the two sides of the panel and joined on its edge forming a loop. The concretion of a second iron-work that was no longer in place had similar morphological characteristics (fig. 145): In both cases, at the end of the loop, there was an orifice, oriented towards the South in the case of the concretion still in place, which probably constitutes the remains of the base of the hook7 intended to be inserted into the eye of the googin. Immediately behind this iron mass, there was a gap of about 20 cm in diameter in which we think a vertical piece of wood was passing through. Initially, we were surprised by the closing of the back part of this ‘eye’ which seemed to isolate the main piece of the rudder blade from the other. In fact, it seems that this stay corresponds to the nail that ran through the piece of wood and joined

Figure 146. 18th c. rudder pintle (Photo: Arnaud Cazenave de la Roche).

the two branches of the iron-work. Fig. 146 shows an 18th century spindle exhibited at the Madrid Museum, which is representative of this device as we believe it is inside the concretion (with the hook pointing downwards in this case). The concretion still in place on the rudder blade also seemed to indicate that a vertical hoop existed in its lower part. There is indeed a vertical iron-work branch running through the pieces P4 and P5 (fig. 143). This broken concretion was interrupted at P4, but we believe that it was initially continuing its course, also enclosing parts P3 and P2. This hypothesis is based on the presence of a notch located on the edge of piece P3 about twenty centimeters from its eastern end through. An iron blade 35 mm wide fastened by a nail of square section was likely to have been in place there (fig. 142). Everything suggests, therefore, that at least two iron-works placed vertically held the horizontal wooden pieces of the rudder.

7 In the 16th century, the French term reported by Jacques Dupuys (DUPUYS, NICOT, 1573) to designate the hook of the pintle that fits into the eye of the googin was ‘vit’: ‘The rudder vits are six large iron hooks in the form of one-and-a-half foot long staples attached to the ship’s rudder, with large iron strips and ‘caravelle’ nails [hooks], which enter their females, which are six large iron rings sewn to the stern panel, …, that carry and hold the rudder. The Spaniard calls them Machos, it is to say Males, as well as the rings, Hembras, it is to say females, whose reason is quite obvious.’ In French, in the 18th century, the terms ‘croc’ and ‘gond’ were used by Duhamel du Monceau (DUHAMEL DU MONCEAU, 1758, Chap. II-19).

113

The Mortella III Wreck 4.4. Analysis and interpretation of the remains uncovered

base of the rudder. As such, one may wonder whether this assembly, which brings together pieces cut according to a curious geometry, is not the testimony of one or more repairs to a rudder whose lower part has been damaged? • The rudder made up with a significant number of wooden pieces. While in the case of the Mary Rose wreckage, the rudder consists of two pieces and only one in the case of the Red Bay wreckage (in Red Bay, the rudder blade was made up with a single plank), in Villefranche, we are dealing with five vertical pieces, a scenario similar to that of the Mortella III whose initial number of vertical pieces was probably five. However, it is difficult to know here whether this characteristic is related to a particular construction method or more simply to the available wood supply. • Finally, the method of fastening the wooden pieces, the cohesion of which is in both cases mainly ensured by bolts.

To our knowledge, the only remains of 16th century ship rudders that have been studied to date are those of the Mary Rose (1545), the presumed San Juan (1565) for the Atlantic area and the wreck of Villefranche-sur-mer (1512) for the Mediterranean area (fig.147 and 148). From the point of view of its morphology, it is naturally from the latter that the rudder of the Mortella III wreck is closest. The common points that can be mentioned are: • The presence of horizontal wooden pieces. In Villefranche, there is only one. It is a 13 cm high and 26 cm thick shole. It should be pointed out that in the case of Mortella III, the succession of four horizontal pieces of wood seems unusual. The orientation of the ‘hook’ of the pintle to the south suggests that they form the lower part of the rudder. Although strongly nailed, bolted, secured by a reinforcing plank and initially enclosed by two vertical straps, this assembly appears to have a weakness that would have been avoided with uninterrupted vertical wooden pieces all the way to the

The reinforcement plank that increases the fastening of the rudder elements is a part that is also found on the rudder of the Villefranche-sur-mer wreck. The authors of the ‘Navire génois de Villefranche’ quote in a note n°34 several

Figure 147. Rudder of the wreck of Villefranche-sur-mer (Illustration: Max Guérout, GRAN).

114

The attributes of the hull

Figure 148. 1 (Left). One-piece construction of the Red Bay wreck 2 (Right) The rudder of the Mary Rose (Courtesy of: D. Kappler, Parks-Canada).

examples of wooden reinforcements visible on Flemish engravings of the 15th century (GUEROUT, RIETH, 1989, 34). This process would therefore have been generalized. The special feature of the reinforcement of the Mortella III wreck is that the plank is embedded in the rudder pieces, so its surface remains flat. As far as iron-works are concerned, the two pintles of the Mortella III rudder are too concretioned to be able to compare their morphology in detail with those found on the Red Bay and Mary Rose wrecks. However, we can make one observation: the iron straps of their port and starboard branches do not encircle the entire rudder because no trace is visible on its outer edge. It is concluded that these branches were interrupted before the end of the aft part of the rudder, as is the case for the Villefranche wreck. The authors of the Navire génois de Villefranche noted in this regard that ‘it [the iron-work] seems to stop 8 cm from the aft end.’ From this point of view, there is a significant difference here with the Mary Rose (MARDSEN et al., 2009: 273) and Red Bay (LOEWEN, 2007, 153) iron-works, which enclosed the entire rudders. In the case of Red Bay, the port and starboard branches meet at the aft end of the rudder and cross on its edge, one passing upwards, the other just below, and their ends fold back a few centimeters to the opposite edge of the rudder (fig. 149).

Figure 149. Rudder pintle of Red Bay (Courtesy of: D. Kappler, Parks-Canada).

115

The Mortella III Wreck

Figure 150. Rudder pintle hook of the Mary Rose (Illustration: D. M. McEvolgue).

The thickness of the rudder: as we have said, the thickness of the rudder, measured on its rear part, is 18 cm. The distance between the two branches of the pintles points at the main-piece (which no longer exists) suggests that the thickness at this level was identical.

pieces is therefore the opposite of what it was in the case of the Villefranche-sur-mer wreck. Apart from the hypothesis of a trapezoidal morphology of the rudder, we do not know how to interpret this fact. In any case, our remains were in too deteriorated a state, especially on the western side, to formulate a solid hypothesis on this subject.

In the 16th century, however, there was a rule that seemed quite generalized, which was that the aft piece of the rudder should be thicker than the fore one. The Spanish authors Diego García de Palacio (GARCÍA DE PALACIO, 1587) and Tome Cano (CANO, 1610) express this.8 This rule seems to have lasted over time since several authors refer to it in the 17th century and it is still formulated by the French writer Duhamel du Monceau in the 18th century (DUHAMEL DU MONCEAU, 1752 : chap.I-5).9 Although formulated in the writings, it must be noted that 16th century naval archaeology does not allow us to illustrate it10 for the time being, since none of the wrecks we have mentioned had this particularity.11 The inclination of the rudder pieces: finally, a curious point concerns the orientation of the horizontal pieces, which had a degree of inclination of about 10° upwards on the stern side in relation to the plane. The inclination of the 8 Diego de García Palacio writes that the outer part of the rudder must be half as thick as the one next to the stern. For its part, Tome Cano recommends that the width of the outer part be twice as wide as that of the stern. 9 In Chapter I, Article 5 it is written: ‘Several [rules followed by the Brest builders] are that the horizontal part of the submerged part [of the rudder] increases in width as they move away from the ship: thus they shape it in a dovetail shape, in the view that its angle with the keel is less obtuse.’ 10 In the 17th century, an illustration of this principle was given by the Vasa rudder (1628), whose aft part of the rudder blade is 20% thicker than the fore part of the main piece. 11 It should be noted that Cayetano Hormaechea (HORMAECHEA, 2012, vol. II, 86) cites two Spanish authors from the late 17th / early 18th century who question this custom, which they consider unnecessary and even harmful to the ship’s steering. They are Francisco Garrote (GARROTE, 1691) and Antonio Gaztañeta (GAZTAÑETA, 1712).

116

5 The architectural profile of the Mortella III wreck: Shapes and proportions 5.1. Attempt to restore the shape of the master-frame M27

measured in 2013 to be 25 meters, the master-frame would have been implanted 40 cm forward of the middle of the keel. It should be taken into account, however, that in order to make an exact calculation of the keel length, it would be necessary to add the flat part of the stem ‘laying on ground’—now disappeared—which was attached to the end of the keel timber. It is reasonable to assume that the keel ‘laying on ground’ measured at least one additional meter, or about 26 meters in total. In this way, if we retain this value, we can therefore conclude that the main mast of the wreck of the Mortella III was located in the middle of the keel.

In 2013, a frame whose floor-timber had two first-futtocks attached to its fore and aft faces was uncovered at midship. This frame, which inverted the frames components assembly sequence, could be identified as the masterframe of the ship, the twenty-seventh starting from the aft. The importance of the information this frame could impart necessitated its study on land. Its study was carried out with the aim of setting out its morphological features, as its shape or ‘figure’ in old French, is a decisive factor in the geometry of the ship. In order to maintain the methodological coherence of our work, after having described the constructive characteristics of this frame in the previous chapter, addressing its dimensions, types of scarfs and fastening methods, we will focus now on its shape and place in the architectural framework.

On Red Bay wreck, presumed to be the San Juan (1565), the master-frame was measured to be about 1.30 meters forward of the middle of the keel (LOEWEN 2007, 161). The situation is identical in the case of the wreck of Villefranche-sur-Mer, and although the total length of the keel could only be estimated, it was clearly shown that the master-frame was located well ahead of the middle of the keel (GUÉROUT, RIETH et al., 1989, 39).

As a reminder, the remains of the M27 frame in tumulus A were composed of three parts: a floor-timber, two firstfuttocks and part of the second-futtock whose end was burned. These timbers were arranged as follows: the two first-futtocks G27 A and B were fastened on the ends of the fore and aft faces of the floor-timber V27. Then, the aft face of G27 A was fixed to the fore face of the secondfuttock A27 meanwhile G27 B lacked a second-futtock. Near A27 was located the second-futtock A28, which was attached to the first-futtock G28.

The place of the master-frame on the keel has been the subject of several discussions in the history of shipbuilding, but in general, it seems that there was a consensus amongst builders to locate it between the middle and a variable distance forward of the middle of the keel. In the modern period, one of the first authors to mention this question was Father Fernando Oliveira, who in his treatise recommended that the master-frame should be located at a distance 1/8 of the length of the keel forward of its middle (OLIVEIRA, 1570, chap. 8).

The total length of this set, which had lost its initial curve with the rupture of floor-timber V27´s head, was 5.20 meters in a 2D measurement from the centre of the floortimber to the end of the second-futtock.

In the mid-18th century, J.L. Duhamel du Monceau wrote a chapter of his Elemens d’architecture naval which— although written at a much later time and for vessels with very different designs from those we are dealing with—provides a didactic insight into the consequences of the master-frame’s position on the keel for her nautical features (DUHAMEL DU MONCEAU, 1752, chap. I, art. 14, 115). Thus, according to him, the reasons why builders tended to place it forward on the keel were as follows:

5.1.1. Position of the master-frame on the keel The M27 frame was attached to the keel 12.90 meters from its heel. Given that the total length of the keel was

Figure 151. Forward side face of the master-frame M27 after reassembling the parts on land (Photo: Arnaud Cazenave de la Roche).

117

The Mortella III Wreck

Figure 152. Drawing of the frame M27 (Illustration: Arnaud Cazenave de la Roche).

‘1°. When such a vessel has once opened the water column, it no longer experiences any resistance; and the water lines at the stern being very sharp, they are more favourable to the reunion of the water flows on the rudder, which has more power as the centre of gravity of the vessel is closer to the bow. 2°. The water flows that gather at the stern of the vessel, must push it in this part and make it go forward. 3°. This imitates the fish shape, which must be more advantageous to divide the fluids properly.’ The third argument is inspired by a school of ‘biomimetic’ thought that has been popular throughout the modern period. For example, it was argued by the Portuguese author F. Oliveira in Chapter VII of his ‘Livro da fábrica das naos’, entitled: ‘de como a arte na fabrica das naos imita o arremeda a natureza dalgus peyxes & animaes’ or in Mathew Baker’s in his ‘Fragment…’ (fig.153).

Figure 153. The ideal hull of a ship according to Mathew Baker (Courtesy of: Magdalene College, Pepys Library).

attempt to reproduce the design of the master-frame was important as all the other frames located between the tailframes mirrored its design model.

Duhamel du Monceau concluded:

5.1.2.1. The rising and narrowing of the master floortimber

‘although some mechanical considerations and the builders’ use require placing the master-frame a little forward, nevertheless we would incline to place it almost in the middle, to have softer water lines, and to swing the ship with more ease.’

The rising. The study of the master-frame started with the observation of its floor-timber V27. It made it possible to measure its rising, which is a fundamental element of the ship’s architecture. The rising of the floor-timbers is set out as the distance measured between the line passing by the bilge points and the upper face of the keel. The precise measurement of this rising was not easy because the regular curve of the branches and the absence of a

5.1.2. Characteristics and shape of the frame The choice to undertake a detailed study of the M27 frame was due to the potential to recover its original shape. This 118

The architectural profile of the Mortella III wreck breaking point at the beginning of the floor-timber head prevented the identification of these bilge points. No mark was present to help either.

little water. Moreover, it reduces the draught… and the flat floor-timbers support the standing better than those which are strongly rised: that is why the floor-timbers of the merchant vessels are made with a higher rising in the Mediterranean, where the vessels do not ground, than in the Ocean where, because of the tide, they often remain out of water.’ (DUHAMEL DU MONCEAU, 1752, chap. I, art. 17).

To calculate the value of the rising, the position of the bilge point has been defined under the clamp n°2—its conventional location—which allowed us to measure a 33 cm rising (fig. 154). Nevertheless, we will retain this figure as an approximate value.1

In view of the above, it is interesting to take into account that this morphological difference appears as an ‘architectural trait’ that could mark a distinction between the Mediterranean and Atlantic architectural traditions. Archaeological studies also seem to be moving in this direction. For example, the W54 floor-timbers of Villefranche-sur-Mer wreck and C22 of Calvi I, both close to the midship2 had also a high rising value, 35 cm and 39 cm, respectively (fig. 156 and 157).

The high rising value of the master floor-timber of the Mortella III wreck is a singular characteristic that fundamentally distinguishes it from the master floor-timber of the Red-Bay wreck, which was practically nil (Fig. 155). From the point of view of its nautical characteristics, its practical consequence is a high draught, in contrast to Red Bay, which is lower due to its flat bottom. It also reflects a series of nautical characteristics listed by H.L. Duhamel du Monceau. He outlined the advantages and disadvantages of building a ship whose master-frame is flat or rised. Among the instructive remarks mentioned in the ‘Elemens d’architecture navale’, we can read:

On the other hand, as the Red Bay wreck shows, the Atlantic or Ibero-Atlantic wrecks we know have—at midship—flat floor-timbers with little or no rising. In the case of the Red-Bay, it was only from the 7th floor-timbers on either side of the master-frame that a rising began to appear (LOEWEN, 2007, 60).

‘The rising of the master floor-timber prevents the vessels from drifting, because it increases their draught,…, this is why we give more rising to small vessels than to large ones, because they are more prone to drift; but the decrease in rising is favourable to sail downwind, and the elevation of the gun battery, because the depth of hold is measured from above the keel, where the rising of the master floor-timber makes an angle that moves

This approach is also documented in Iberian texts. If we look at the master-frame recommended by Diego García de Palacio in his Instruccion naútica (1587), for example, or by Father F. Oliveira in his Libro da fabrica das naus (1580), we will see that they are flat and devoid of rising.

Figure 154. Drawing of the floor-timber V27 and representation of its rising (Illustration: Arnaud Cazenave de la Roche). Often, as was the case with the Red Bay wreck, marks are incised in the wood to indicate the location of the bilge points. On the Mortella III wreck, for the moment, none of these marks have been found.

1

2

119

In Villefranche, the W59 frame was identified as the master-frame.

The Mortella III Wreck

Figure 155. Comparison of the master floor-timbers of the Mortella III and Red-Bay wrecks (Illustration: a. Arnaud Cazenave de la Roche; b. Courtesy of: Carol Pillar, Parks Canada).

Figure 156. Morphology and rising of the floor-timber W54 of the Villefranche wreck (Illustration: Arnaud Cazenave de la Roche).

Figure 157. Morphology and rising of the floor-timber C22 of the Calvi 1 wreck (Illustration: Arnaud Cazenave de la Roche).

120

The architectural profile of the Mortella III wreck flat, is 1.72 meters. Therefore, the width of the flat of the Mortella III wreck, at the level of its maximum breadth, was 3.44 meters. This width steadily decreased on either side of the master-frame. Towards the stern, at the level of the floor-timber V9, 6 meters fore of the master-frame, the width of the flat was still 3.00 meters. After this point it the reduction accentuated towards the stern. 5.1.2.2. The shape of the master-frame M27 One of the difficulties in restoring the shape of the masterframe was the breaking of the floor-timbers heads, which had caused a general collapse of the frames downwards. During the excavation in 2013, the first theoretical approach was attempted to restore the original shape of the frame M27. A drawing of the head of the floor-timber raised at three different heights was done under three hypotheses— high, medium and low. The uncertainty remaining at the end of this approach led us to take a more pragmatic approach, dismantling frame M27 and its reassembling it once ashore. Nevertheless, for its ‘rebuilding’ to be conclusive, two conditions had to be met:

Figure 158. Profile of the master-frame according to the Ordinances of 1613 (Illustration: Cayetano Hormaechea).

1. First of all, it was essential to ensure that the head of the floor-timber was correctly replaced to find the exact position of the timbers at their scarf. This was made possible by repositioning the ends of the two firstfuttocks G27 A and B on the fore and aft faces of the head of the floor-timber V27 by matching the marks and holes of the nailing, achieved by inserting metal rods (fig. 160). 2. The second condition was to ensure that the remains which were to be returned to their original position had not been deformed during their stay in the water, and that the shape we obtained was therefore really that of the frame before the ship sank. After confiding in him our doubts on this point, Brad Loewen told us that his experience with waterlogged squared timber, particularly that of the Red Bay wreck, made him think it unlikely that the oak of the Mortella III frame had suffered any deformation. This was confirmed by the study of Fabien Langenegger (dendrochronologist of the Archaeology Department of the Canton of Neuchâtel—OPAN) who studied the wood cells of the frame M27 and observed that they were indeed intact and free from any deformation.4

Figure 159. Dead flat surface of the Mortella III wreck (Illustration: Arnaud Cazenave de la Roche).

Finally, it should be emphasized that the technical tradition of giving rising to the floor-timbers at midship was adopted in the Atlantic from the beginning of the 17th century. This fact is documented in the Spanish Ordinances of 1613. In Spain, the term ‘astilla’ was used to describe rising. From 1601 onwards, according to C. Fernandez Duro and Monleón, ‘in the time of D. Diego Brochero’, the masterframe rising would have been introduced . (FERNANDEZ DURO and MONLEON, 1892).3 From 1613 onwards. the Ordinances imposed it under the name of ‘astilla muerta’ (‘dead rising’). The flat: The location of the bilge point also enables us to set out the value of the ‘flat’, or ‘dead flat’. This is the second important data revealed with the study of the master-frame. The distance between the middle of the master floor-timber and its starboard bilge, i.e. the half3

Figure 160. Reconstitution of M27, essential cohesion and adjustment of the timbers (Photo: Arnaud Cazenave de la Roche).

Quoted by C. Hormaechea, (HORMAECHEA, 2012, vol.1, 22 and 153).

4

121

See the study of the master-frame wood cells in Annex IV.

The Mortella III Wreck

Figure 161. Survey of the master-frame on land (Photo: Arnaud Cazenave de la Roche).

As the two conditions mentioned above had been met, it was then possible to restore the shape of the frame M27 remains with a satisfactory degree of fidelity (figs. 162 and 163).

relayed by the first-futtocks without any breaking points marking the bilge. 5.1.2.3. The design of the master-frame

The conclusions of this restoration were as follows:

The ‘figure’ (shape) that has emerged from the study of the remains of the frame M27 is a useful milestone in the debate over how the master-frame was designed in the 16th century, the terms of which were set out in Chapter I. However, in the case of the Mortella III wreck, the characteristics of this morphology must be interpreted with caution, since we are dealing with a limited portion of the original hull. Nevertheless, they allow some remarks on the shape of its lower part at mid-ship:

• Once the frame was rebuilt, in 2 D view, the distance between the burned end of the second-futtock A27 and the middle of the keel was 4.60 meters. • The height of the end of the burned frame could be measured accurately to 2.10 meters. • The frame M27 showed a regular curvature whose shape was very close to a circular arc and had a radius of 5.80 meters. The floor-timber followed this curve 122

The architectural profile of the Mortella III wreck

Figure 162. Record of the master-frame on land (Illustration: Arnaud Cazenave de la Roche).

Figure 163. Adaptation of the frame M27 underwater record to the shape of land record (Illustration: Arnaud Cazenave de la Roche).

• The frame M27—the remains of its part which sat below the water line—had a shape that is difficult to ensure it was designed by a simple circle arc. This is because its first-futtock followed a line slightly set back from such a circular arc (fig. 164). However, we can see that it is extremely close to it, the distance separating them being only ten centimeters at the furthest point. • As such, this shape seems closer to that recommended by the Iberian authors of shipbuilding treatises in the late 16th and early 17th centuries, whose master-frame design uses a single arc, than that recommended by the Venetian treatises of the 15th and 16th centuries, whose

spindle-shaped geometry is similar to an ellipsoid.5 Comparing the master-frame of the Zorzi Trombetta da Modon’s 700-botte nave (1445) with that of the Mortella III -whose dimensions are similar- makes it possible to clearly visualize these shapes differences (fig.165). In the Venetian model, the lower part of the hull bulges more due to flatter bottoms. Then the first-futtock marks a break with a clearly visible bilge generated by a sharp curve over the first third first-futtock. A second pronounced 5

123

See Chapter 1

The Mortella III Wreck

Figure 164. Superposition of the underwater and land records of M27 (Illustration: Arnaud Cazenave de la Roche).

Figure 165. Superposition of the profile of a nave de 700 botte, according to Z. Trombetta (1445) and that of Mortella III wreck (Illustration: Arnaud Cazenave de la Roche).

124

The architectural profile of the Mortella III wreck

Figure 166. Design of the shape of the master-frame of a merchant ‘nave’, according to the prescriptions of Pré Theodoro de Nicoló (1550) (Illustration: Arnaud Cazenave de la Roche).

inward turn in the lines occurs at the beam level. In the case of the Mortella III, on the other hand, the bottom line of the hull was regular and did not tail off. Additionally, the first-futtock followed the curvature of the floor-timber quite closely.

• The Villefranche-sur-Mer wreck (1516): It was built in a similar period as the Mortella III and shares many similar design features s with it. The shape of its frame at mid-ship followed a circular arch, more rigorously than Mortella III’s. Its radius was 6.30 meters at the level of the frame W54 which was taken as a reference (GUÉROUT, RIETH and GASSEND, 1989, 95), and was located five frames after the master-frame, identified as W59 (fig.167).

More contemporary with the period of the Mortella III’s ship is the merchant nave of Pre Theodoro de Nicoló (1550), whose profile of the master-frame has been graphically reconstructed (fig. 166), still confirms the difference between the Venetian model and that highlighted by the Mortella III. However, the smaller size of the Theodoro de Nicolò’s ship accentuates this difference.

The Calvi I wreck (late 16th century): in the same way, the frame C21, part of the first twenty-two frames discovered at the end of the 1980s and located at mid-ship, had a floortimber and a first-futtock with a perfectly circular shape (fig.168). They followed a line forming an arc 2.81 meters in radius.

The Atlantic model represented by Mathew Baker’s English shape in his ‘Fragments…’ (BAKER, 1570), upon which Basque naval architecture would be based (LOEWEN, 2007, 97)6, is similar to the Venetian typology and remains very different from the shape of Mortella III’s master-frame.

5.1.2. Restoration of the master-frame shape Once the shape of the remains had been examined, the delicate question arose as to how the frames continued their trajectory upwards, beyond their current charred ends. As we have seen, the preserved part followed a 5.80 meter radius circular arc quite closely. Did the upper part of the second-futtock A27 and the third-futtock follow this line, or did they have a point at which they turned inwards?

On the other hand, archaeological work enables a comparison between the Mortella III and those of the two Mediterranean shipwrecks that have served as a reference throughout this work. It is indeed the frames of Villefranche-sur-Mer and Calvi I wrecks that have the closest shapes to that of the Mortella III:

6

It is impossible to answer this question with certainty. However, two scenarios can be considered:

See Chapter 1

125

The Mortella III Wreck

Figure 167. Shape of the frames at midship of the Villefranche-sur-Mer wreck (Drawing: a. Courtesy of J.M. Gassend, GRAN; b. Adaptation by the author).

Hypothesis 1. First, if we strictly follow the design rule put forward by the Iberian authors, using a single arc and extend the M27 frame by following a circular arc of 5.80 m radius, we obtain the transversal shape of a ship very close to that of F. Oliveira’s 48 palmos ship (OLIVEIRA, 1580, fº112, see fig.3 and fig.169) or that of Manoel Fernandes (MANOEL FERNANDES, 1616, Fº88, see fig.4). Its maximum breadth beam would have reached 11.60 meters and the ratio of its keel length / beam would then have been in the order of 1: 2.24, proportions not far from the rule ‘As, Dos, Tres’.7

that is missing. This hypothesis is based on the example of the wreck of Villefranche-sur-Mer, which shares many constructive similarities with the Mortella III.

Hypothesis 2. Alternatively, we can consider the possibility of an a break of the arc’s curve parallel to the second half of the second-futtock A27, which is precisely the part

The case of the wreck of Villefranche-sur-Mer is indeed interesting insofar as, in a manner similar to that observed on the Mortella III wreck, up to half of the second-futtock, the frame C54 which was located close to mid-ship, followed a regular curved line. This line formed a circular arc with a radius of 6.30 meters. Then, a little over half of the second-futtock, in the area below the maximum breadth, to a level probably close to the waterline, there was a change of the curve of this circle whose radius decreased by half (fig. 171). In this way, the hull straightened up and quickly reached a vertical position.

To calculate this ratio, we considered a length of keel ‘laying on ground’ of 26 meters, i.e. the 25 meters measured, plus 1 meter estimated for the part of the stern piece that joined the keel.

To sum up, the Mortella III master-frame had a single arc, but only for the part of the hull located under the waterline.

7

126

The architectural profile of the Mortella III wreck

Figure 168. Shape of the frame C20 of the Calvi I wreck at midship (Illustration: a. Pierre Villié; b. Adaptation by Arnaud Cazenave de la Roche).

Beyond it, the degree of the curve changed and the frame line straightened. To our knowledge, the design of this master-frame was not described—in any treatise of the period. We can only note the similarities with the case of the Villefranchesur-Mer wreck and hypothesize that the Mortella III’s master-frame could have been designed in the same way. Indeed, if we consider the many commonalities shared by these two wrecks, in particular their dating, origin and building processes, this second hypothesis is more plausible, implying that their master-frames shared a similar shape. In this case, the shape of the master-frame of Mortella III wreck would be as shown in fig. 172. We then have a maximum breadth line that stands at around 10.50 meters and more slender proportions than in previous estimations, with a ratio of beam / keel length of 1 : 2.48.

Figure 169. Shape of the master-frame and of the ‘almogamas’ or tails frame, according to Fernando Oliveira (fº114) (Courtesy of: Biblioteca Nacional de Portugal).

127

The Mortella III Wreck

Figure 170. Hypothesis of the shape of the master-frame of the Mortella III wreck designed with a single arc (hypothesis 1) (Illustration: Arnaud Cazenave de la Roche).

Figure 171. Evolution of the shape of the frame W54 of the wreck of Villefranche-sur-Mer with two arcs (Illustration: Arnaud Cazenave de la Roche).

128

The architectural profile of the Mortella III wreck

Figure 172. Hypothesis of the design of the master frame master-frame of the Mortella III wreck based on the model of the Villefranche wreck (hypothesis 2) (Illustration: Arnaud Cazenave de la Roche).

considerations that primarily affect an architectural issue related to the proportions of the ship, but also practical and commercial considerations.

These proportions are almost identical to that of the wreck of Villefranche-sur-Mer. 5.2. The deck and depth of hold

There is little evidence to be able to assess their original position with certainty. In fact, observations made on the structure of the Mortella III wreck have so far not revealed any trace of the overlop, nor any remains or hanging features on the clamps. In our attempt to establish the position of the decks of the Mortella III’s ship, the overlop was therefore located about halfway down the secondfuttock, a little after its charred end. In this case, its height would be 2.25 meters if measured from the upper face of the keel.

5.2.1. The number of decks The dimensions of Mortella III, with a keel of about 26 meters ‘laying on ground’, and 10.40 to 11.50 meters wide at beam, depending on the hypotheses, makes the presence of two or three decks a possibility. The breadth at the beam of a three-deck ship commonly reaches 10 meters, but little is known about the organization of the frames and the position of the decks. Nevertheless, the construction characteristics of the Mortella III wreck and—as we will see in the following chapter—the conclusions of the historical study lead us to point towards it being an Italian nave. We now know she is a commercial ship provided with two decks (GATTI, 1999, 145). This typology is also shared by the Villefranchesur-Mer wreck. As a result, a ship with two decks is the favoured model for recovering the characteristics of the Mortella III wreck. 5.2.2. Decks height

Looking to the wreck of Villefranche-sur-Mer, archaeology offers the example of the construction of a Genoese ship whose overlop height was 2.20 meters. Her deck knees were hooked onto the shelf clamp upholding the assembly between the first and second futtocks. The space between the decks was a little less than 2 meters. The height of the first deck was 4.30 meters. This value, measured from the upper face of the keel, was chosen as the reference for determining the depth of hold.8 The height of the second deck was 6.40 meters, half of the breadth. In this arrangement of decks, the beam line passed a little above

The decks height, in other words their position on the framing, is an important question. It involves technical

8 We will when dealing with the depth of hold that in the 16th century most builders took its measurement at the level of the second deck.

129

The Mortella III Wreck

Figure 173. Height of the decks of the Villefranche-sur-Mer wreck (Illustration: J.M. Gassend, GRAN/Arnaud Cazenave de la Roche).

the first deck, a common configuration in the 16th century before the introduction of the battery (fig.173): In the case of the Mortella III wreck—taking into account the shape of its master-frame suggested by the first hypothesis (H1), its depth of hold, measured according to the same criterion as at Villefranche, would be around 4.12 m. This is equivalent to 16½ palmi of Genoa9, a value similar to that of the Villefranche wreck. This height would also correspond to the position of its second deck. In this diagram, it would be located at the level of the beam line, in a configuration close to that recommended by Manoel Fernandes (1616) (fig.174). The height of the overlop could reach 2.25 meters, or 9 palmi. Therefore, the height between the decks would be 1.72 meters, or 7 palmi. Finally, if we use the shape of the master-frame of the second hypothesis as a reference (H2), the favoured one, the first deck would be located at the same level as the beam (fig. 176):

Figure 174. Arrangements of the decks of the 500 toneladas ship according to Manoel Fernandes (Courtesy of: Academia de Marinha de Portugal).

used in sixteenth century Basque Country. It was likely to be relatively standardised, regardless of the size of the ship, in order to optimise barrel cargoes: Barkham listed seventeen shipbuilding contracts between 180 and 300 toneladas in Guipúzcoa, dated from 1545 to 1590. They show that, regardless of their tonnage, the two decks ships had an overlop located at a height of 4 codos (2.30

Michael Barkham’s research on Basque shipbuilding provides an opportunity to highlight a deck arrangement One unit of length used in Genoese shipbuilding was the palmo (24.8 cm). 3 palmi were equivalent to a goa or gua (74.4 cm), a unit used exclusively in the maritime context. 10 palmi were equivalent to a canna (2.48 meters). And 12 palmi were equivalent to a canella (2.98 meters).

9

130

The architectural profile of the Mortella III wreck

Figure 175. Hypothesis of the height of the Mortella III wreck (h1) (Illustration: Arnaud Cazenave de la Roche).

Figure 176. Hypothesis of the height of the Mortella III wreck (h2) (Illustration: Arnaud Cazenave de la Roche).

131

The Mortella III Wreck

Table 12. Measurement of the depth of hold in various authors of the 16th and early 17th centuries Date

Starting point of the measure

Arrival point of the measure

Rodrigo Vargas

1570

 

To the second deck

1/2

AGI, Real Patronato, leg. 260, 2º, rº 35: Text published by J.L. Casado (CASADO SOTO, 1988)

Fernando Oliveira

1580

on ‘the upper face of the keel’

‘until the deck, i. e. second cover’

3/4

O livro da fábrica das naos… , 1570, Cap. V, fº120

Diego Garcia de Palacio

1587

From the keel…

…to the second deck.

2/3

Instrucion náutica,… 1587

Escalante de Mendoza

1575

‘The true depth of hold’ to ‘the first deck’ =’flush with the keel, at the level of the mast-foot”

Thomé Cano

1611

Nicoló Sagri

1570

Spanish Ordinances of 1590

1590

Anonymous Spanish Ordinances of 1613

 

Itinerario de Navegación, Libro Primero.

To the second deck

2/3

Arte para fabricar, fortificar y aparejar naos…, 1611, Dialogo segundo, p.67

 

‘The pontalle to the second deck ‘

1/2

“Il carteggiatore”, 1570, fº13v

‘from the soler, on the upper face of the floortimbers”

‘to the beam’

 

M.N.M. Colección Navarrete, Nº de catálogo 789: Arqueamiento de navíos, San Lorenzo el 20 /8/ 1590

‘to the beam’

 

Cristobal de Barros, MNM Vargas Ponce, T.XXV B, doc.19 f. 42-43

‘until the upper face of the main deck’

 

‘Ordenanzas de 1613 para la fábrica de navíos de guerra y mercantes’

Early ’From the soler’ 17th c. 1613

Ratio Texts depth of hold / breadth

‘From the upper face of the floor timbers’ (soler)

the beam. Although the Ragusan Nicolò Sagri also set out the value of the depth of hold to the second deck, it seems that the Venetian builders measured it at the first one. To our knowledge, the only Iberian shipwright to express the idea that the ‘true depth of hold’ must be measured to the first deck (‘primera cubierta fija’) was the Spanish Escalante de Mendoza. Measuring at the level of the first deck was developed in Spain from the 17th century onwards, but also in France where Duhamel du Monceau, in the 18th century, gave the following definition:

meters).10 This arrangement enabled three rows of barrels to be stowed. The following decks had a height of more or less 3 codos (1.72 meters), allowing two rows of barrels to be stowed. Brad Loewen stated that the Red-Bay ship ‘fit to this rule confirmed by the Basque shipbuilding contracts of the 16th century, since its decks were at heights of 4, 7 and 10 codos, measured from the top of the master frame-timber to the top of the beams.’ (LOEWEN, 2007, 160). 5.2.3. The depth of hold

‘The depth of hold is the distance between the top of the keel and the top of the beam of the first deck, not including the camber of this beam.’ (DUHAMEL DU MONCEAU, 1752, 37).

It is the height of the decks (or sometimes the height of the beam) that determines the size of the depth of hold. This is an important measure in the ship’s architecture because it often informs the proportions that set out her shape. It is also essential when calculating the gauge. The depth of hold is called a pontalle in Italian and a puntal in Spanish.

Another difficulty lies in determining the starting point of this measure. Some suggest it starts from the upper face of the keel, others from the upper face of the floor-timber (soler, in Spanish). The end point, on the other hand, generally seems to be determined on the upper surface of the beam. Iberian authors sometimes specify this point, Venetian authors also show it on their drawings. An offer of charters from the city of Marseille to the Count of Clermont for a crusade project entitled ‘Informationes Civitatis Massiliae pro passagio trasmarino’ dated 1318 also seems to attest to this practice in Provence during the Middle Ages as the value depth of hold was ‘tabulam in tabulam’ (FOURQUIN, 1990, 182). This approach is the opposite of the contemporary measurement done ‘under the beam’.

The problem faced by the researcher when studying the question of the depth of hold is the imprecision that surrounds the definition of this architectural concept, as many others at the time we are discussing. In the 16th century, there were builders who measured it at the height of the second deck, the majority of them of Iberian origin. Others measured it at 10 The measurement was taken from the soler (ceiling above the floortimbers) to above the deck, a method officially recommended by the Spanish Ordinances in the 17th century.

132

The architectural profile of the Mortella III wreck from the upper face of the keel, but from the soler (the ceiling covering the floor-timbers). Then, from 1613 onwards, the legislator required that the measure be taken from the upper part of the main deck because of the inaccuracy caused by the measurement of the breadth ‘in the air’, without any specific reference point. From this date and throughout the 17th century, this was the way the Spanish builders proceeded, in line with the rule stated by Duhamel du Monceau.

The above table shows that, as a general rule in the 16th century, the depth of hold was, as F. Oliveira says, a value setting out the total height of the ship. This is therefore why it was mostly measured up to the second deck. This point is important when considering this dimension in the proportions mentioned by builders. The proportion of the depth of hold measured at the second deck in relation to the beam goes from ½ to ¾. As a result, in the case of the Mortella III wreck, if we take into account a breadth measuring 10.50 meters (H2) and a depth of hold when measured to the second deck of 6.15 meters, we have a proportion ratio slightly higher than ½. It is almost exactly ½ in the case of the assumption that the breadth would be 11.50 meters (H1). This proportion of ½ is that recommended by the Spaniard Rodrigo Vargas (1570) and the Ragusan Nicoló Sagri (1570). In the case of the Villefranche-sur-Mer wreck, this proportion was also established at ½, with a breadth of 12.60 and a depth of hold up to the second deck of 6.70 meters.

5.3. Rake of the sternpost and shape of the stem 5.3.1. The rake of the sternpost and the overhang Despite the absence of the sternpost on the Mortella III wreck, the slope of the notch, located on the aft face of the upper part of the keel heel, into which it was fit, made it possible to estimate its value (fig. 177). This was of 76º 11, a value close to the general slope of the outer face of the keel heel. The ability to estimate this value is important for restoring the shape of the hull.

Finally, Table 11 shows that in Spain, the concept of measuring the depth of hold to the second deck was established at the beam line towards the end of the 16th century, to allow gauging calculations. From 1590 onwards, however, the measurement was no longer taken

The rake of the stern of the wreck of Villefranche-sur-Mer was very similar or identical to the Mortella III wreck: it was estimated between 75º and 78º (GUEROUT, RIETH, GASSEND, 1989, 29). In L’archéologie Subaquatique

Figure 177. 3D image of the upper part of the keel heel (Photo: Bérenger Debrand). 11 The mentioned degrees measure the angle formed by the horizontal line running in the prolongation of the keel and the line set out by the axis of the stern.

133

The Mortella III Wreck

Figure 178. The rake of stern posts of the wrecks of Mortella III, Calvi 1 and Red-Bay (Illustration: Arnaud Cazenave de la Roche).

de Red Bay, Brad Loewen presents a table in which he compares the slope of various sterns (LOEWEN, 2007, 53). He concluded: ‘the greater the number of decks, the closer to vertical was the sternpost.’ Two wrecks with similar dimensions are the Red Bay and the Calvi I. The calculated angle of Red Bay stern is 69º, quite close to Calvi I’s (66º), s (fig.178).

Figure 180. Rake and height of the stem post of Z. Trombetta nave (Courtesy of: British Library).

Whatever the hypothesis proposed, the height of the Mortella III’s second deck was most probably around 6 meters. This height, combined with the stern rake, sitting at 76º, makes it possible to calculate an aft overhang length of 2.14 meters when measured at the height of the second deck (fig. 179). This overhang value was about 1/12th of the keel length. This proportion is close to the 700 botte nave of Zorzi Trombetta da Modon. On the other hand, it differs from that given by Pre Theodoro de Nicolò for his merchant

Figure 179. Hypothesis of the rake of the stern of the Mortella III wreck (Illustration: Arnaud Cazenave de la Roche).

134

The architectural profile of the Mortella III wreck nave. Her dimensions being smaller (10 pas = 25 meters length), her stern rake and overhang are therefore greater (Overhang = 1/8th of the keel length).

panel make it unfortunately impossible to determine the initial curvature of the stern. From this point of view, we must rely on a theoretical approach to estimate the value of its radius and shape. Based on the methods used by builders in the 16th century, we are able to conclude that the Mortella III stem overhang was between 8 and 10 meters.

5.3.2. The stern and its overhang There was no trace of the stempost. Towards the front end, the only remains were those of the end of the planks which were inserted into the rabbet. The wooden panel formed by these remains has been described in Chapter III. It should only be added here that the shape of the line established by the end of the planking was not straight but slightly curved. Its radius reflects the initial curve of the stem but is difficult to evaluate (fig. 181). The poor remains of this

• The first case from Venetian sources is the 700 botte nave by Zorzi Trombetta, whose dimensions are close to those of the Mortella III. In this case, we are dealing with a very high stem of 12.50 meters (36 Venetian feet) corresponding to half the keel length. It had a

Figure 181. Fore end of the strakes of the planking (Photo: Christoph Gerigk).

135

The Mortella III Wreck great overhang exceeding 10 meters (29 Venetian feet). It contrasts with the weak 1.40 meter overhang of the stern (4 Venetian feet). Similar proportions can be found on the other models of navi proposed by Zorzi Trombetta as well as in the nave quadra of the Fabrica di galere, whose stem overhang was ½ of the keel length.

the overhang length of the stempost, which was between 36 and 30 Venetian feet, respectively, in the Libro de Zorzi Trombetta da Modon. The resulting hypotenuse, 46 feet in our case (16 meters), is drawn by the line that connects the outer face of the upper part of the stem to its lower part at the point of tangency with its face on laying on ground (fig. 182).

A rectangular triangle was used as a means to design these stems. The two basic measurements were the height and

The same method of designing the stem was used a century later in Pre Theodoro de Nicolò’s Instructione.

Figure 182. The stem post of Zorzi Trombetta’s 700-botte nave (Courtesy of: British Library).

136

The architectural profile of the Mortella III wreck maximum breadth (bocha or regia in Italian, lo más ancho in Spanish), the length of their keel and their overall length (longhezca in Italian, esloria in ancient Spanish). The value of the depth of hold, (pontalle in Italian, puntal in Spanish) and that of the dead flat (piana in Italian, plan en Spanish) were also involved in the establishing these proportions. The excavation of the Mortella III ship led to an analysis of the relationships between the dimensions of these architectural concepts, comparing them with those used in the 16th century Mediterranean. This is an important subject of this book which contributes to developing a typological approach to the ship, just as it is likely to provide key information on its nautical characteristics.

• Pre Theodoro de Nicolò recommends a smaller stem overhang for his merchant ship than Zorzi Trombetta. Its length is a little more than 1/3 of the keel length (15 feet for a keel length of 50 feet). Its height is also proportionally much lower and corresponds to a little more than a third of the length of the keel. The lower part of the stem, laying on ground, is 3 feet, or just over one meter. The proportions advocated by Theodoro de Nicolò (1550) for his merchant ship, the text of which is chronologically close to the sinking of the Mortella III ship, were used to restore its characteristics presented at the end of this chapter. The Iberian authors used different stem design methods to the Venetians, but the proportions are generally close to those given by Pre Theodoro de Nicolò. Most of them recommend an overhang of 1/3 of the keel length:

Nevertheless, it is necessary to underscore the two major limitations facing the implementation of this exercise:

• Father Fernando Oliveira, for example, outlined a design method based on a circular arc whose radius is 1/3 of the keel length. If we apply Father Oliveira’s method (OLIVEIRA, 1590) to Mortella III stem design, we obtain an overhang of 8.6 meters (fig. 183).

• First, the absence of stern and stem posts limit the calculations of actual measurements—i. e. not estimated—of the overall length of the ship and of the overall length of the keel, including the now-absent part of the stem laying on ground, • Secondly, since the end of the remaining frames at midship only reach a portion of the second-futtock, an estimate of the maximum breadth is necessary.

5.4. Proportions of the ship, the attempt to restore her shape and longitudinal dimensions 5.4.1. The main proportion ratios

Once these caveats have been made, care should be taken to specify what we are measuring, i.e. what is exactly meant by ‘maximum breadth’, ‘keel length’ and ‘overall length’, since the interpretation of how these measurements should be recorded has regularly led to confusion.

As we saw in Chapter I, in the traditional way, the proportions that shipbuilders gave to their ships were set out by the relationship between their maximum width or

Figure 183. The stern and stem posts of the Mortella III wreck, according of the method of F. Oliveira (Illustration: Arnaud Cazenave de la Roche).

137

The Mortella III Wreck 5.4.1.1. The keel width/length ratio

toward the aft and the beginning of the stempost toward the fore, it is difficult to set out its delimitation points. This caused confusion in the history of shipbuilding. The timbers that make up the keel of a ship, excluding the heel and stem, are what Spanish builders called the ‘quilla rigorosa’ (‘rigorous keel’). When the authors of the L’archéologie subaquatique de Red Bay mention that the length of the wreck’s keel timber is 14.20 meters, they were referring to this ‘rigorous keel’ which, in this case, was a single piece.

The notion of ‘maximum breadth’ or ‘main breadth.’ This is part of the ratio used for calculating the ship proportions. It must be deemed as her maximum width. Traditionally, texts agree on this point: it is recorded inside the hull at the level of the master-frame. It should not be confused with the width at beam, sometime simply called ‘beam’, which is measured at the main deck and is used for gauging, even if sometimes the measurement of the beam and the maximum breadth merge.

However, in the 16th and 17th centuries, the concept of keel length evolved to include part of the heel piece and the stem foot. In fact, in Spain, for example, the 1613 Ordinance explicitly established the measurement of the length of the keel between the aft face of the stern (‘cara de proa del codaste’) and the beginning of the stem curve (‘punto de tangencia o corte del arco que traza el interior de la roda con la cara alta de la quilla’). Under this rule, written in 1613, but probably applied well before, the keel of the Red Bay wreck would measure approximately 14.80 meters. Based on the distances measured in the Guipuzcoa shipbuilding contracts, Brad Loewen reaches the same conclusion with a total keel length measured at 14.75 meters (LOEWEN, 2007, 40-41), which is about 4% longer than the keel piece itself.

In the 16 century, builders often sought to match the waterline with that of the maximum breadth (LOEWEN, 2007, 323 and HORMAECHEA, 2012, 97 and 207). It was common to set it a little above the main deck (1/2 codo to 1 codo), so that once the ship was loaded, the maximum breadth line lowered down to the waterline. The question of the height of the waterline in relation to the maximum breadth and the location of the deck in relation to it will become very important in the 16th century given the emergence of artillery on board and the creation of the battery. In 15th century Venetian texts, the maximum breadth line appears under the term of ‘bocha’. It coincided with the main deck in the Fabrica de galere and in the Libro by Z. Trombetta. In the middle of the 16th century, however, in the Instructione de Pre Theodoro de Nicolò, the ‘bocha’ became the measure of the width measured nine Venetian feet above the keel. It no longer corresponded to the line of maximum breadth which was called ‘regia’ and which was located—in the example he gives of his merchant ship— two Venetian feet (69.5 cm) above the main deck. th

In France, the most distant information we have been able to find on how to measure the keel dates back to the 17th century with the notion of ‘keel on ground’. It starts from the end of the heel and ends at the point where the stem curve leaves the ground. It can therefore be concluded that for French builders too, the keel measurement included a part of the stempost. This concept was found in Spain under the term ‘quilla derecha’ (“straight keel”).

The first rules written about the location of the maximum breadth line are the Spanish gauging rules enacted by the 1613 Ordinances. They clearly make the distinction between ‘lo mas ancho’, the maximum or main breadth, and the ‘manga’, the width at deck level, or beam, used to calculate the gauge. Nevertheless, in 1613, the construction rules for merchant ships were putting these at the same level and thus merging the lines of the maximum breadth and the beam. But this was not always the case, nor will it be in 1618 when the new Ordinances set a line of the maximum breadth half a codo below the beam (HORMAECHEA, 2012, V.2, 147).

In the case of the Mortella III wreck, the keel measurement, exactly 25 meters, refers to the keel timbers and the lower part of its heel. In other words, to perform an accurate analysis of the ship’s proportions, the measurement of the total length of the keel ‘on ground’ is required. Therefore, it is necessary to account for the flat lower part of the stem, which has now disappeared. It is for this reason that we have used a keel length of 26 meters for our calculations. This includes the length of the keel piece, which was measured at 25 meters plus one meter, which corresponds to an additional 4% over the initial length measured.

It should be added that the notion of breadth—more than any other dimension—is fundamental in the architecture of modern ships. Indeed, this dimension remained the keystone of the ship on which all the others depend, by a proportion. It is symptomatic in this respect that in Spain, for example, in the 16th and 17th centuries the size of ships was characterised by their width and not by their length.

The ratio: the ratio of the length of the keel ‘on ground’ and the width of the maximum breadth was as follows: 26 / 10.5 = 2.48 => ratio = 1 : 2.48. The overall length. This is generally measured at the level of the second deck and on the outer part of the stern and stem posts. We have described above how the dimensions used to restore these two pieces make it possible to estimate the aft and fore overhang at 2.14 meters and 8.66 meters (1/15th and 1/3 of the keel length, respectively). These two measurements, when added to the keel length allow the

In the reconstruction of the master-frame diagram of the Mortella III wreck (H2, fig. 176), the maximum breadth line was? 10.50 meters. The length of the keel. This is a concept that also needs to be clarified. Since its ends are made up of the heel piece 138

The architectural profile of the Mortella III wreck Table 13. Evaluation of the proportions of the Mortella III ship in comparison with those of the Villefranche-sur-Mer and Calvi I. Main measures :

Maximum breadth

Keel length

Ratio to: Mortella III

Measures(m) Measures(m) Ratios

Calvi I

Depth of hold* to second deck

Depth of hold* to first deck

Flat

Maximum breadth

Ratios Villefranche

Overall length

Measures(m) Ratios

Stem overhang

Stern overhang

Keel length

10.5

26

36.80

6.15

4.27

3.45

8.66

1.65

1

2.48

3.50

0.58

0.40

0.33

0.33

0.06

12.5

32

44

6.70

4.55

 

10

2

0.36

1

2.56

3.52

0.53

 

0.31

0.06

7.80

17

24.90

 

 

5.10

2.80

1

2.18

3.19

 

 

0.3

0.16

Note*: the depth of hold is measured from the upper face of the keel to the top of the beam.

overall length to be calculated: 2.14 + 8.66 + 26 = 36.80 meters, or 146 palmi or just less than 50 goe12. From this value, we can deduce the ratio between the breadth and length which is as follows:

its main dimensions and shapes—have been outlined. Some of them were recorded on direct field surveys and others were estimated. After having proposed a transverse reconstruction of the ship at midship, here is an attempt to restore its longitudinal shape: Fig.184

36.80 / 10.50 => ratio 1 : 3.50

5.5. The gauging

5.4.1.2. Summary of proportions

5.5.1. An essential unit of volume in 16th century shipbuilding

To summarize the proportions of the maximum breadth, keel length and / overall length result in the following ratio:

Establishing the main dimensions of the Mortella III wreck allows us to evaluate its tonnage.

1 : 2.48 : 3.50 This ratio—accounting for a margin of error resulting from the imprecision of certain measurements—can be compared to that of the nave quadra of the Fabrica di galere (1 : 2.45 : 3.58) or the 700-bot nave of Zorzi Trombetta da Modon, although the latter has a slightly more stretched shape (1 : 2.56 : 3.80). It instead appears to be higher than that prescribed by the Iberian authors for their merchant ships in the first two thirds of the 16th century. This was mostly in accordance with the ‘As-DosTres’ rule, with the maximum breadth/keel length/overall length ratio being close to 1: 2: 3.

In the 16th century, the gauging of ships was intended to measure their load capacity. It was therefore a question of assessing volume and not displacement, a concept that would appear much later.13 The notion of contemporary gauging, which evaluates tonnage, the capacity of the ship, is based on the same concern (1 tonne is equivalent to 100 cubic feet, or 2.83 m3 according to current international conventions). During the Middle Ages, the regional diversity of units of measurement and the multiple ways of taking them, made for a complex and slow process when many nations opted to centralise their measurement system. However, the ability to measure a ship’s loading capacity was of crucial importance for shipbuilding contracts, usually based on the tonnage. It was also used for the assessment of taxes or, as in Spain, for payments for leases or embargoes on ships requisitioned by the Crown for war.

Recent archaeology studies also offer two examples of Italian ships which can be compared: The first is the Villefranche-sur-Mer wreck whose chronology and Genoese origin aligns it to the Mortella III wreck. Its proportions are very similar to those of the Mortella III. The other is the Calvi I wreck, further away chronologically (late 16th century) whose proportions are, on the other hand, closer to the rule ‘As-Dos-Tres’:

In Spain, royal ordinances were promulgated between 1590 and 1613 to unify a way of establishing a universal and centralised measuring system throughout the kingdom. The 1590 Ordinance officially regulated the calculation of the gauge (ORDENANZA, 1590, fº168r at 170v). It did so in two ways. First, generalizing the use of a single unit of measurement throughout the country. The codo de ribera,

5.4.2. Attempt to restore the longitudinal shape and the dimensions of the ship Throughout the preceding pages, the architectural characteristics of the remains of the Mortella III wreck— The Goa was used in Genoa as a unit of measurement exclusively in shipbuilding in the 16th century. Its value was 74.4 cm.

Displacement is the measurement of the mass of water displaced by a ship in different loading situations.

12

13

139

The Mortella III Wreck

140 Figure 184. Longitudinal shape of the Mortella III wreck and main measurements (Illustration: Arnaud Cazenave de la Roche).

The architectural profile of the Mortella III wreck which was equivalent to 33 dedos, or 57.47 cm, replaced the codo castellano, which was equivalent to 32 dedos (55.72 cm) and was in use in several regions, including Andalusia. Then the crown imposed a generalised system for gauging a ship on the basis calculating the value of the maximum breadth, the value of the depth of hold and the overall length. At the end of this calculation, a value in toneladas (barrels) was obtained. Each toneladas was equivalent to 8 codos per cube, or 1.5185 m3, or the equivalent of 2 barrels called pipas14.

According to this formula, the calculation of the tonnage of the nave cuadra, , is as follows:

Initially the gauge was designed to determine the number of barrels a ship was capable of carrying. The etymology of the units of measurement used remind us of this: the barrique in France, but also the tonel or tonelada in Spain, the botta in Venice and Genoa.

Maximum breadth = 10.50 m, or 30.17 Venetian feet (pie)15

[(26.5 × 13) × 13] / 6 = 746 botte. We saw that the number of botte had to be divided by 1.66 to obtain the equivalent of tonnage: 746 / 1.66 = 449 tonnes. If we apply the Venetian gauging rule to the estimated dimensions of the Mortella III wreck under our H2 hypothesis, the calculation is as follows:

Depth of hold at the first deck: 4.27 m, or 12.27 Venetian feet (pie) Keel length: 26 m, or 14.9 steps (pas)

In Genoa, the multiplicity of units used for gauging during the modern period is symptomatic of the problem raised by the lack of a centralised measurement system: The capacity of a ship was measured in salme de Sicilia, botti, mine or cantari, the latter also being a unit of mass (1 cantaro = 47.6 kg). All of these different units of measurement were used in Genoa to determine the tonnage of ships in the 16th century. L. Gatti states that during this period the cantaro predominated in notary contracts. But she adds that in the following century, it was the salma, a unit originating from Sicily, initially used to measure wheat cargo volumes, that was preferred (GATTI, 1999, 75-86). According to L. Gatti, 21 cantari were equivalent to 1 ton. And 4 cantari to 1 salma, from which it can be deduced that 5.25 salme was equivalent to 1 ton.

(30.17 × 12.27) × 14.9] / 6 = 919 botte, or 553 tonnes 5.5.2.2. The Spanish method It should be pointed out here that although it was officialised in 1590, the Spanish rule was in force throughout the 16th century and its trace can be found in 1523 (GUIARD, 1917, 77) and 1568 in a text by Domingo de Busturia (BUSTURIA, 1568)16. The Spanish gauging model has been studied by several researchers. First by Pierre Chaunu in the late 1950s (CHAUNU, 1959) which contained an unfortunate miscalculation that caused much confusion. Then it was reconsidered from the late 1980s onward (CASADO SOTO, 1989, 62, 63, LOEWEN, 2007, 322 to 338, HORMAECHEA, 2012 vol. II, 107 à 137 et CASTRO, 2013, 1138 à 1140). This method is based on three simple measures used in the last decade of the 16th century. It was pronounced in the Ordinances of 1613, which introduced a consideration of the flat and the depth of hold.

5.5.2. Calculation of tonnage in the 16th century As texts consulted do not provide information on the methods of gauging ships in Genoa in the 16th century, it is on the Venetian method of the 15th century and on the Spanish method, formalized by the text of 1590, inspired by Diego Brochero that we must rely upon for an evaluation of the tonnage of the Mortella III ship.

The Spanish method enabled a calculation of the tonelada or the tonel macho which were both gauge values, called arqueo in Spanish. The tonelada de carga was used in Andalusia before 1590 (8 codos castellanos per cube = 1.38 m3) and the tonel macho (8 codos de ribera per cube = 1.52 m3) was in force in northern Spain throughout the country from 1590 under the name tonelada. Independently of these gauging units, from the middle of the 16th century onwards, the concept of the tonelada de cuenta, 20% higher than the tonelada, appeared. It arose from the practice of refacción. It was a bonus intended to compensate owners whose ships had been requisitioned (embargo) for the risks of war. It is also on the basis of this tonelada de cuenta that the crew’s wages were calculated, hence it was also known as tonelada de sueldo (‘wage tonne’).

5.5.2.1. The Venetian method Folios 88v and 50r of La fabrica di galere inform us of the Venetian gauging method at the beginning of the 15th century . It delivers a very simple formula which—like the Spanish rule—involves three dimensions, the bocha (maximum breadth), the pontalle (depth of hold), measured at the first deck, and the keel length. The calculation to be carried out to obtain the tonnage expressed in botte is as follows: (BOCHA in pie × PONTALLE in pie) × BELT LENGTH in pas -----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------6 The equivalence of tonelada to two pipas has been documented from the end of the 15th century (LOEWEN, 2007, 324); in 1575, Escalante de Mendoza stated that it is equivalent to two pipas of 27 ½ arrobas each. In the second half of the 17th century. This equivalence to two pipas was still in force during the 17th century (VEITIA LINAJE, 1671, l. II, cap. XV, 2). 14

As a reminder, 1 Venetian foot = 34.8 cm and 1 step = 5 feet. These documents are cited by J. L. Casado Soto (CASADO SOTO, 1989, 63). Domingo Busturia’s work is incorrectly attributed to his brother Pedro with whom he is commonly confused.

15 16

141

Once these clarifications made, let us now return to the calculation of the tonnage of the Mortella III ship. The gauging formula that appeared in the 1590 text can be summarized as follows:

[(

)]

F L ------------ × C – 5% 2 T = -------------------------------------------------------------- 8 L = Length F = Maximum breadth C = Depth of hold T = tonelada In the case of the Mortella III wreck, the codo’s values of these dimensions are as follows: L = 36.80/0.5747= 64 F = 10.50/0.5747= 18.27 C = 3.92/0.5747= 6.82 So, 3987 – 5% T = -------------------------------------------- = 473 8 Conclusion: With regard to the Spanish calculation method, the Mortella III ship would therefore have had a tonnage of 473 toneladas that can be compared to contemporary registry tonnage (LOEWEN, 2007, 322) and is equivalent to nearly 10,000 cantars. It was by using this method with the 1590 Ordinance that the archaeologists of the Red-Bay wreck assessed it amount to 202 toneladas. However, this figure does not correspond to the ship’s carrying capacity because it does not take into account the space above the first deck. According to Brad Loewen, the additional 20% of refacción made it possible to arrive at the actual capacity of the ship, currently called tonne burthen (tonne de port in French) (LOEWEN, 2007, 322). Applying this to the Mortella III ship, we calculate a burthen of 568 toneladas, or about 11,900 cantars. Finally, it can be seen that the calculation of the Venetian tonnage applied to the Mortella III wreck resulted in a very similar result to that obtained using the Spanish method with 919 botte, or 553 tonnes burthen which was equivalent to 11,600 cantars.

142

6 Historical study: Issues, references, and identification attempt of the Mortella wrecks 6.1. Hypothesis and historical issues

of Naples in 1497, which he considered to be part of his Angevin heritage, itself a region that had been annexed in 1447 by Aragon. This conflict ceased in 1559, after more than 60 years of wars that had ravaged Europe with the peace of Cateau-Cambrésis signed by the heirs of François I and Charles Quint, Philippe II and Henri II. Whatever the proposed date of their sinking may be, the Mortella shipwrecks must be linked to this turbulent political context.

The discovery of the Mortella wrecks immediately raised the question of their origin and identity. This questioning quickly led to researching literature held in French, Italian and then Spanish archives and libraries. In the search for an historical episode likely to explain the presence of the Mortella wrecks, the oldest chronological information relating to the sinking of ships in the Bay of Saint-Florent was reported to us by Marie Antoine Graziani, historian and specialist of Corsican history. There are brief handwritten mentions found in documents from the Genoese archives that refer to two naval battles in the 15th century. The first refers to the sinking of Genoese ships in 1460, the second refers to the sinking of Catalan privateer ships in 1490. While the first episode seems to have been clearly established in the Bay of St Florent, the location of the second is more evasive, but can potentially be traced to the region of the ‘Cap de Saint-Florent’ (Cape of Saint-Florent).

In this chapter, we present the three episodes mentioned above (1555, 1526 and 1527)., Each of them shall be briefly placed in the historical context of their time. We then present extracts from the main texts collected during our research on the shipwrecks themselves. Unless otherwise indicated, these have been translated into French, then into English by us. The original versions of the main texts can be found in Annex VI. Finally, for each episode, a comparison between the historical events and the archaeological details is undertaken. 6.2. The naval combat and shipwrecks of 1555

The dating of Mortella wrecks, however, has oriented us towards the 16th century. Consequently, another event caught our attention and led us to put forward the first hypothesis for the identification of the wrecks. It was suggested that they could correspond to two Spanish ships sunk in December 1555 by Antoine Escalin, Baron Paulin de la Garde, General of the King of France’s galleys. This episode is indeed the only one that we initially managed to locate in documentary sources of Corsican history which relates to the sinking of seagoing ships in the Gulf of Saint-Florent during the 16th century.

6.2.1. The historical context of the period In 1555, the date of the first hypothesis of shipwreck, the ships reported to have been lost were part of Alonso Pimentel’s Spanish fleet. At that time, the admiral of the Spanish fleet in the Mediterranean was the Genoese Andrea Doria, who had been in the Emperor’s service since 1528 after turning against Francis I, his first ‘employer’. As a result, from that date onwards, Genoese vessels had been a regular part of the Spanish fleet. This is clearly seen in the expeditions that Charles V led to Tunis in 1535 or Algiers in 1541, for example.1 In these circumstances, it does not seem shocking that the Mortella wrecks, whose ballast and artefacts point towards a Genoese origin, could be ships that belonged to the Spanish fleet on that date.

However, this hypothesis has been frustrated by the Mortella III wreck’s dendrochronological dating. Indeed, as we have seen, it dated the shipwrecks to the first third of the 16th century. For this reason, we continued our literature research, aiming to find a maritime casualty in the Bay of Saint-Florent that matched the verdict of the dendrochronological dating. We identified of several texts revealing other shipwrecks earlier in time, first in 1526 and then in 1527. It is this last episode that seems to offer the best explanation for the presence of the Mortella wrecks.

6.2.2. Related Texts Arthur Filimon carried out an initial study in the archives and libraries of Paris, which located six instances of naval combat that could be related to the wrecks. The closest to the events of 1555 was written by Pierre de Bourdeille, known as Brantôme (1537-1614) in his complete works

First, we must place these events into the complex Mediterranean geopolitical context of their time. This period was particularly marked by the rivalry between the French and Spanish, as evidenced in the wars between Francis I and Charles V in Italy. The first half of the 16th century saw a succession of eleven wars. The starting point was Charles VIII’s attempt to recover the kingdom

1 For example, Charles V’s expedition against Algiers in November 1541 included 65 galleys, almost half of which, 27, were of Italian origin. Among them were four from Sicily, five from Naples, twelve belonging to Andrea Doria and six to Antonio Doria. The preparation of Andréa Doria’s fleet for this expedition is visible in a handwritten document from the Archivo General de Simancas (VILLAFRANCA, 1541, f°176).

143

The Mortella III Wreck But it was later, with the discovery of an account of these events by Marc-Antonio Ceccaldi (1520-1560), a chronicler of Corsica who was a contemporary of the shipwrecks, that we obtained a primary source that was the most precise. This text was published and translated into French by Antoine-Marie Graziani. He recounts the episode in some detail, dating it to mid-December. It coincides with Brantôme’s account in that the Baron de la Garde was described to have been coming from Civitavecchia. However, he estimated the number of French galleys to be fourteen and the number of Spanish ships to be eleven, which suggests a balance of power that seems more equitable and credible. The Baron de la Garde allegedly ‘vigorously’ attacked two Spanish ships that had arrived in the bay ahead of their fleet. Surprised, and seeking to rejoin their fleet still on the high seas, the two ships would have entangled their rigging in preparation for the manoeuvre, before hitting a reef near the coast. The Baron de la Garde would then have been able to capture more than 900 elite soldiers whereas the ships would have been looted by the inhabitants of Nebbio region. Ceccaldi pointed out that the villagers of Farinole would have been enriched by the thousands of ecus they recovered there (CECCALDI, 2007, 491): ‘While these things were happening in Corsica, the war in Italy was turning more and more to the King’s advantage. The Imperial Officers were therefore sent to Spain to seek infantry to distribute them among the various places where their presence would be necessary. About six thousand Spaniards were gathered for this purpose: they embarked on eleven ships and sailed for Italy. In the middle of December a north north-westerly wind pushed them towards the coasts of Corsica in the Gulf of Saint-Florent.

Figure 185. Captain Paulin de la Garde (Illustration: Public).

first published in 1655 (BOURDEILLE, 1868, 139-150). Brantôme recounted the confrontation between eleven Spanish round ships that had taken refuge in the Bay of Saint-Florent following a storm and two galleys of Baron Paulin (or Polin) de la Garde2. The combat that followed this meeting resulted in the wreck of two Spanish ships along with 1500 people, most of who died in the sinking. Brantôme’s description is unfortunately brief and not very detailed. The other five stories taken from secondary sources are as follows: Adrien Richer in Vies du capitaine Cassard et du capitaine Paulin (connu sous le nom de baron de la Garde), whose text dated from 1789 and related an epic version of events to which we can only give limited credit (RICHER, 1789, 183). The episode was also reported by Jean Gaudin, in a thesis from the Ecole des Chartes entitled Le capitaine Paulin (GAUDIN, 1900). Jacques-Auguste Thou in his Universal History (THOU, 1734, vol. II, 622) also discussed the events. Finally, we have chosen to ignore the accounts of Léon Guérin Les marins illustres de la France (GUERIN, 1861, 113) and of Commander Vivielle La vie tumultueuse du capitaine Polin baron de la Garde (VIVIELLE, 1935) are not mentioned

The ships, once they entered the Gulf, wanted to dock? Only two had reached the coast, when the Baron de La Garde arrived with fourteen French [galleys], coming from Civitavecchia on the way to Provence, having carried some French Cardinals who were going to Rome. As soon as he arrived, he vigorously attacked the two ships by surprise, which (…) seeking to join the others (who were still on the high seas) manoeuvred too hastily, in such a way they entangled in each other and struck a rock where they were crushed. When they saw this accident, the other ships that were on the open sea turned their bow with much difficulty in another direction. After a very laborious crossing, they arrived in Genoa where they landed about five thousand Spaniards, who went to the points where the emperor needed them most. The Baron de La Garde took from the two ships that had broken up about nine hundred elite soldiers: and since he could not (because the weight was too heavy) take more on his galleys, he was obliged to leave more than a hundred of them on land. They were disarmed by the Nebbio people (who had gone down to loot the

Antoine Escalin, Baron Paulin de la Garde (1498? - 1578) was from the Garde Adhémar. Of modest origin, his exceptional qualities made him noticeable early on and he quickly climbed the social ladders of his time. He served in the army and navy under the reigns of Francis I, Henry II, Charles IX and Henry III. He was knighted and played a leading political role as ambassador and General of the king’s galleys. 2

144

Historical study ships) and were sent to Bastia. From there, some were sent to Calvi and enlisted in the companies, others to the mainland, where they were most needed. The Baron de la Garde, bringing with him the Spaniards he had kept to chain up, returned to Provence with fourteen galleys and two others which, on their side, had captured at sea a Genoese ship loaded with wheat. It was claimed that when the two Spanish ships sank, in addition to the considerable spoil made by the people of the Nebbio, the people of Farinole found several thousand crowns (?) hidden there and a large number of them became rich in this way.’

misfortune, especially in the current economic climate. In view of the above, I have sent fifteen galleys to tow and escort ships that—due to bad weather—have not yet arrived. I have received the information that [the troops] have been landed and are arriving by land, which I sincerely hope for as they are arriving at the right time , although from what I understand, they are penniless, naked and mistreated by the sea.’ 6.2.3. Analysis of the hypothesis of the shipwrecks of 1555 The 1555 episode, which we had initially favoured, was weakened by the dating delivered by the dendrochronological study which placed the sinking too far away from the event.5 In fact, even if the construction of the Mortella’s ship could have taken place several years after the felling of the trees used for its framework6, if it took place between 1517 and 1520, it can reasonably be estimated that the ship would have been between 29 and 34 years old at the time of its sinking. However, even if the possibility of a ship of this age is not unthinkable, this lifetime is about three times the average lifetime of a large ship in the 16th century in the Mediterranean.7

Handwritten letters from the Ambassador to Genoa written to Joana of Austria, held in the Archivo General de Simancas (AGS), are a useful addition to the Ceccaldi version. These are two letters dated 21 December 1555 and 2 March 1556. These documents are cited by Rafael Vargas-Hidalgo (VARGAS-HIDALGO, 2002, 15 - 16). From these we haved learned that the Spanish fleet was commanded by Don Alonso Pimentel3. According to this Spanish version, the French fleet was composed of twelve vessels and the Imperial fleet of ten. One of the Spanish vessels was hit by a French artillery shot that broke her tiller. Then we find what is mentioned in Ceccaldi’s text: as she couldn’t manoeuvre, the ship would have unintentionally collided with another Spanish ship and, together, would have fallen into the hands of the French. This text addresses the great loss that these catches caused Spain However, at no time is there any question of sinking:

If we now examine the facts reported by the various primary source documents, other problems arise and put into question the adequacy between the event as described and the presence of the wrecks at the site where they were discovered:

‘The twelve galleys of France that had gone to Civitavecchia began the journey back and arrived in the Gulf of Saint-Florent on the island of Corsica on the 4th of this month[of December 1555] in the morning where they stayed all day. At nightfall, Don Alonso Pimentel arrived with ten ships [naos in the text] carrying infantry which, not detecting the presence of French galleys, dropped anchor because they were in great need of water. Two hours later, the said French galleys attacked the [Spanish] ships, which broke their moorings and set sail. But the galleys gave artillery and a shot hit the tiller of a ship that could not set sail because its rigging became entangled with that of another, and the two ships in which there were three companies of the best Spanish soldiers were captured by the French, except for a hundred of them who escaped by swimming. Of the other eight ships, five went to Diana’s pond4 with Alonso Pimentel, and two others to Monaco, and the last one came here to Genoa. This event was a great

•

First, Ceccaldi’s chronicle mentions a shipwreck on a reef near the coast in a north-north westerly wind. The inhabitants who looted the wrecks were described as those of Farinole, a village in Cap Corse region, at the entrance to the Bay. These seem to refer to rocks located near the shore, on its eastern side. This location appears inconsistent with that of the Mortella wrecks which lie in the middle of the bay of Saint-Florent, more than 2000 meters from the nearest coast and in the absence of any shoal in the vicinity. • At no time is there any mention that these ships were burned. It is only specified that they ‘hit a shoal where 5 The dendrochronological study presented in Annex III establishes that the wood used in the construction of the Mortella III ship was felled no earlier than 1517 and no later than 1520. 6 In the 16th century, the period that can elapse between the felling of trees intended for the construction of a ship and its launch may vary by several years. However, it can be estimated that between the cutting of the logs, the transport of the wood and its air drying, two to three years may have passed. Thereafter, we know that the construction of a large ship could have also taken up to two to three years, sometimes longer if funding is lacking. In total, an average period of between 4 and 6 years can be estimated to evaluate this period. 7 In the book La Grande Maîtresse, nave of François 1er, Recherches et documents d’archives, 2001, by Max Guérout and Bernard Liou, several documents from the Genoa archives show that the useful life of a nave in the 16th century can be estimated between 6 and 14 years. (GUEROUT, LIOU, 2001, chap.9). The book by Fréderic C. Lane Navires et constructeurs à Venise pendant la Renaissance, Paris, 1965 is also cited: the study of ship chronology leads his author to conclude : ‘All these indications confirm the traditional opinion that the average life span of a ship was ten years.’ (LANE, 1965, 259).

Antonio Bernardino Alonso Pimentel y Herrera de Velasco (Benavente 1514-Valladolid 1575), the sixth Count of Benavente, comes from a Portuguese family settled in Castile (Zamora) in the 14th century. He was a military leader and friend of Charles V, whom he accompanied on many of his expeditions in the Mediterranean. Godfather and tutor of Philip II, he was appointed Governor of La Goulette in 1565 and Viceroy of Valencia in 1566. 4 Diana’s pond is what we interpret from the French text that mentions ‘la fosse de Diane’. The word ‘fosse’ which means ‘pit’ in French is probably a bad translation of the old Italian word ‘foce’ which means ‘estuary’. In Corsica, Diana’s pond is also an estuary. 3

145

The Mortella III Wreck they would have crashed.’ Again, the description does not seem to coincide well with the archaeological facts.

by the action of Andrea Doria’s own Genoese galleys in the Gulf of Saint-Florent.

The problem raised by the excessive chronological distance between the event and the felling of the trees intended for the building, added to those related to its description, naturally led us to seek a similar naval episode earlier in the 16th century that would better match with the archaeological evidences. It should be recalled that these have a twofold particularity that the history of their sinking should be able to explain:

The historical circumstances of this event are as follows: Genoa was then governed by Doge Antoniotto Adorno on behalf of Charles V. At the initiative of Louise de Savoie, mother of Francis I, then held prisoner in Madrid by his cousin Charles V, the ‘Cognac League’ was formed. Signed in May, shortly after the liberation of the King of France, it brought together France, England, Pope Clement VII, Venice, Florence, Milan and the German princes against the Emperor. It gathered a fleet of 37 ships that threatened the Spanish. Savona was taken over by the French fleet, which then joined the Venetian and Pontifical fleets. The rivers that supplied Genoa also fell into League hands and the city, starving, was in danger. This situation prompted Charles V to send a rescue fleet at the head of which he placed Charles de Lannoy, a trusted man and viceroy of the Kingdom of Naples. Gathered in haste in Cartagena, it was composed of either 22 square-sail vessels, as according to Agostino Giustiniani (GIUSTINIANI, 1537, 278 and 278v), or thirty-two, as according to Francesco Guicciardini (GUICCIARDINI, 1738, 254).9

• The first is that the two ships of the Mortella were engulfed by flames. The fire burned the Mortella III’s frames to the first third of its second-futtock and consumed almost all its dead works. • The second is that the ships have probably been emptied of their contents, except from the artillery. The very small amount of artefacts found in Mortella III wreck shows that the ship was practically empty at the time it sank. These two facts suggest that either the ships were captured, emptied and burned, or voluntarily abandoned after being emptied and burned. The fact that both of them burnt makes it more likely a voluntary than a fortuitous act. But knowing the great value of the artillery at the time, how can we explain its lasting presence on board together with the anchors? Indeed, it is doubtful that all nine cannons from the wreck of Mortella III were out of order. Their presence is more akin to ships that were disarmed in a hurry and whose wrought iron artefacts, too heavy to be quickly unloaded, were left on board. But here we stray into the realm of assumption, let us now look at the information provided by the two episodes of 1526 and 1527, the latter being the one we favour as an explanation for the presence of the Mortella wrecks.

Several texts report that shipwrecks occurred when this fleet passed through the Bay of Saint-Florent. We present these texts here. 6.3.2. Related texts A version written in close proximity to the events is that of Francesco Guicciardini (1483-1540), Florentine historian, philosopher and diplomat of the time, who recounts the episode of the journey of the imperial fleet commanded by Charles de Lannoy in the Gulf of Saint-Florent in 1526 in his book History of the Italian Wars (GUICCIARDINI, 1738, 254, 257):

6.3. The shipwrecks of 1526

‘… [the pope] was advised that the viceroy [of Naples] was in the Gulf of St. Florent in Corsica with thirty-two ships, with three thousand horses, two-thousand and five hundred lansquenets, and three or four thousand Spaniards.’

6.3.1. The historical context of the period The continuation of historical research led us to then consider the events occurred in 1526. Although it is true that in 1526 and 1527, Mediterranean history was still dominated by the wars of Italy—at that time a seventh war was raging between France and Spain (1526-1529)—the geopolitical context was nevertheless very different from that of 1555, given that Admiral Andréa Doria of Genoa was still in the service of Francis I. It was an unusual situation. The Republic of Genoa was under Spanish control and therefore an Imperial Genoese fleet existed whose main enemy was the so-called ‘Cognac League´ fleet composed of the Genoese galleys of Andrea Doria and the Pope, those of the King of France and those of the Venetians.8 It is in this particular political context that a Genoese navi from the imperial fleet may have been sunk

…’In the meantime, the viceroy left Corsica with only twenty-five ships, because before anchoring at SaintFlorent, he had lost two in a storm, and the strength of the wind had carried five others to the open sea. At Sestri di Levante he met a squadron composed of six galleys from France, five from André Doria and an equal number of Venetians who fought him from four o’clock in the evening until nightfall. Doria wrote that he had sunk a ship of more than three hundred men and attacked the others with cannon shot; that the bad weather had forced him to release under the Cape of

9 These documentary sources are given in a doctoral thesis defended by Damien Broc in 2014 at the University of Corsica Pasquale Paoli, available online (BROC, 2014, 142).

The exact composition of this fleet is given by Francesco Guicciardini (GUICCIARDINI, 1738:190).

8

146

Historical study Portofino, where he was to be joined that same night by the rest of the Naval Army that was in Porto Venere.’

A fourth text that shares Guicciardini’s description is to be found in the account provided by the 17th century historian of 16th century Genoa, Filippo Casoni, in his book entitled Annali della Republica de Genova del secolo decimo sesto. Like Guicciardini, he placed the fleet in Sestri di Levante. The significance of his account lies in an additional fact that he reported. After the departure of the Imperial fleet from the Gulf of Saint-Florent, two ships of that fleet in poor condition would have turned back to the port of Saint-Florent where they would have been abandoned (CASONI, 1708, 91):

Moreover, two other texts place the naval battle between the Imperial fleet and Andrea Doria’s one in October/ November 1526 in the bay of Saint-Florent. They report that during the confrontation two Spanish ships would have been sunk. The episode is first mentioned by a contemporary author, Paolo Lingua (LINGUA, 2004, 94): ‘Clement VII, weak and undecided, enacted his intention to withdraw from the League, then he changed his mind again and decided to face Charles V again. The fleets of Venetians, French and Andrea [Doria] met again. Command was given to Doria and Peter of Navarre, of whom seventeen galleys inflicted on the thirty-six imperial ships commanded by Antonio de Lannoy10, Ferrante Gonzaga and Ferdinand d’Alarcon, the only resounding defeat suffered by the Spanish army… Andrea Doria, in turn, had a personal achievement: his galley sank two Spanish ships, he demolished the first with his falcons and bombards from a close range shot and he sank the second with a masterly shot fired from the castle. The scene of the clash is the bay of San Lorenzo (Saint-Florent), in Corsica. A thousand soldiers and sailors from Charles V’s fleet would lose their lives, but this victory would not be significant for the fate of the League.’

‘But the imperial army, pursuing its journey and fearing the wind, divided into several squadrons. A part, where Ferrante Gonzaga was, went to Sicily, then joined Gaeta. Two other vessels among those of poorest condition turned back to Corsica, and arrived at Saint Florent, where the crews abandoned the unusable vessels, and then passed through Sardinia.’ Finally, we had confirmation of this event through Agostino Giustiniani, bishop of Nebbio11 (GIUSTINIANI, 1537, 278 and 278v): ‘From Cartagena, the Spanish fleet arrived in Corsica in the Gulf of St. Florent with a fleet of twenty-two square sails, including the Viceroy of Naples and Captain Alarcon with about fourteen thousand infantrymen. The aim was to make a five- to six-day break for a much-needed rest. Then, the fleet sailed towards Genoa and arrived at Capo di Monte, where the wind ran out. And of the entire League fleet, only Captain Andrea Doria fought the Spanish fleet with his six galleys; sent a ship to the bottom and held the entire enemy army in check, which made no further attempt to reach Genoa, but headed for Naples. Two ships that were in a terrible state returned to the Gulf of St. Florent and it was by land that the infantry on board went to Bonifacio, and from Bonifacio went to Sardinia.’

However, we have not yet been able to trace Lingua’s information back to its source. A very similar text can be found on the Internet, published by a website dedicated to Italian history (DAMIANI, 2012, [online]). But here again, however, Mr Roberto Damiani, author of these uploads, could not give us the references of the document which—if we believe the style—seems old, and is perhaps Lingua’s source. Here is this text: ‘Oct/Nov. 1526. [Andrea Doria] reached Navarro’s army which was anchored in Portofino. He made this place his operational base, which he left to his nephew Filippino with five hundred infantrymen. With the arrival of the opponents into Corsican waters, he was soon joined by Navarro in the Gulf of La Spezia. Supported by six French galleys, five pontifical and five Venetian, he intercepted twenty-four enemy ships at Chiappa Point, near Portofino. He faced Navarro…. Doria threw himself between two ships with his galley, demolished the first with his falcons and bombards located in the bow and sunk the second (on which three hundred men were stowed) with a shot from the forecastle. In the battle many Imperial sailors died, and a thousand soldiers commanded by Charles de Lannoy, Ferrante Gonzaga and Ferdinand d’Alarcon. According to some sources, the shock occurred, however, in the bay of Saint-Florent, in Corsica. After the battle, which lasted four to five hours, two galleys from SaintBlancard came to assist the Doria ships.’ 10

6.3.3. Analysis of the texts on shipwrecks of 1526 As the excerpt from Guicciardini’s Wars of Italy shows, the Spanish fleet, commanded by the Viceroy of Naples, Charles de Lannoy, was hit by bad weather to the point that two ships sank. However, no details are given on the site of this shipwreck, it is only said that it occurred ‘before arriving in the Gulf of Saint-Florent’, which seems to exclude that potential that it could explain the presence of the Mortella ships. Guicciardini, on the other hand, does not mention a shipwreck in the bay of Saint-Florent, but just mentions the loss of a Spanish ship sunk later by 11 Agostino Giustiniani (1470-1536), Pantaleone by his baptismal name, was a Dominican priest and Genoese writer to whom we owe the Dialogo nominato Corsica (or Description of Corsica) and the Descrittione della Lyguria (1537). Appointed Bishop of Nebbio (Upper Corsica), he spent nine years in his diocese in Saint-Florent between 1521 and 1530 where he was a direct witness to the events that took place there. He died in 1536 in the sinking of the ship that brought him back from Italy to his diocese during a storm between the island of Capraia and Cap Corse.

It probably refers to Charles de Lannoy

147

The Mortella III Wreck Andrea Doria’s galleys during a clash between the League and Imperial fleets south of Genoa, at Sestri di Levante.

the Marshal of France, Theodore de Trivulce, who was subsequently appointed governor.

Conversely, Paolo Lingua and Roberto Damiani’s chronicle located the scene of the confrontation between the two fleets in the bay of Saint-Florent. According to them, the Doria’s galley sank two Genoese navi. The first ‘with his falcons and bombarding’, then the second ‘sinks too with a masterful shot fired from the castle’. So we have here two ships that could be those of the Mortella. With one caveat, however, as we have said, since the dead works of the Mortella ship were burned down, it probably spent some time before sinking. This does not coincide well with the description that the ships were sunk ‘straight down’ by guns. It could of course be assumed that the ships may have caught fire as a result of the battle, but the texts of Lingua and Damiani suggest a rapid shipwreck caused by enemy artillery. They are also the only ones to locate a naval battle between the Spanish fleet and the league fleet in Saint-Florent. Here arises the question of the extent to which these stories can be trusted, given the sources on which they are based are unknown and the primary sources that we have located agree on the fact that the clash between the Imperial and League fleets did not take place in Corsica, but off the Italian coast.

It is in this context that the Genoese saw two ships that had been sent to relieve the starving city with grain chased by French galleys and scuttled in the bay of Saint-Florent. 6.4.2. Related texts It is therefore within this turbulent context that a new shipwreck episode was identified in the Bay of SaintFlorent. We had first noted this event in the text uploaded online by R. Damiani. It evoked an attack on four galleys that were set on fire in August by Andrea Doria’s fleet (DAMIANI, 2012, [online]): ‘August 1527. [Andrea Doria] is in the pay of Francis I with a charter contract agreed for eight galleys for 38,000 crowns per year. In the middle [of August] he reached Portofino with his galleys that of Saint-Blancard and Morette’s ships. Genoese and imperial ships were congested in the port, without any room for manoeuvre. The galleys crew screamed with terror under artillery fire and sent a wave of panic that spread throughout the crews. Thirty-two ships were destroyed or fell into the hands of the French, among them were Spanish galleys, which were captured along with their weapons; some galleons loaded with Sicilian wheat were also taken, as well as the large carrack of the Giustiniani family, coming from the Levant with a cargo of spices. In the aftermath, four Genoese galleys were attacked and burned in the Gulf of Saint-Florent. Genoa was now cornered. Doria and the Saint Blancard entered this port and bombarded the city defended by Brizio Giustiniani with twenty-three galleys. Genoa was taken over by the led by the Venetian Cesare Fregoso on land and at sea by the Venetian fleet and that of the French ships. Doria prevented Fregoso from becoming governor of the city and favoured Teodoro Trivulzio.’

Moreover, Agostino Giustiniani, who must be recalled as a contemporary observer of the facts, points out that after the departure of the Imperial fleet led by Charles de Lannoy from the Gulf of Saint Florent, two ships that were in too poor condition to continue their navigation turned back and returned to Saint-Florent. Filippo Casoni who also reports the event states that after the landing of the ‘people’, the ships were abandoned for being ‘useless’. So we have here information that could coincide with the presence of the wrecks; if the ships had been abandoned, it can be assumed that they have been emptied of their contents beforehand. However, the information on this episode remains tenuous, we do not know where and how the ships were abandoned, nothing is said about any fire on board the ships, nor do we understand why the artillery and anchors would have been abandoned with them.

But later, we spotted several descriptions of the event mentioning that these lost boats were in fact large ships and not galleys, information that aroused our interest. The episode seemed to have attracted a certain notoriety because it is reported by Charles de la Roncière in his Histoire de la Marine Française (LA RONCIERE, 1906, vol. III, 204) where Genoese ‘carracks’ are mentioned, specifically the Ferrara and the Boscaina which were ‘kidnapped’ by a fleet of eight French galleys. But the main interest of this text, which misdates the event in 152612, lies in the references it mentions. It led to contemporary

6.4. The shipwrecks of 1527 6.4.1. The historical context of the period In 1527, confrontation arose between the Cognac League and Spain. In reaction to Pope Clement VII’s alliance with the King of France in May, Charles III of Bourbon ordered the imperial army to attack Rome, which, after falling after a few hours was the object of looting lasting several days. The news of the sack of Rome had an enormous impact throughout Christianity.

12 The text of La Roncière places the event in September 1526 by associating sources referring to two different episodes: it first cites the ‘Comptes de Ragueneau’ (RAGUENEAU, 1601-1700, f°188) which set out a patent letter of September 1526 concerning the prize of ‘two carracks and four ships loaded with wheat belonging to the Genoese.’ He then cites the texts of A. Giustiniani and P. Giovio which report another event that took place in mid-August 1527. The description of La Roncière therefore shows that there is confusion between the two events.

At sea, the increase in military pressure from the League against Genoa resulted in a blockade that suffocated the city. In mid-August, the city finally fell under the pressure of the French fleet commanded by Andrea Doria and 148

Historical study authors who provide a further account of the facts and in particular two quite detailed recount of the episode.

having evacuated their crews. This famine resulted in that the Genoese sent merchant ships everywhere to buy wheat & that day after day, they waited for their arrival…’

The first was found in Agostino Giustiniani’s book, published in 1537 (GIUSTINIANI, 1537, 278v). This text is of great interest because it is written by a direct witness to the event and confirms that it was not galleys that were set on fire, but in fact two large ships. These were two navi which, pursued by the French galleys, were abandoned and set alight by their crews in the bay of Saint-Florent:

6.4.3. Analysis of the texts on shipwrecks of 1527 The first text posted online by R. Damiani describes the terrible famine experienced by the city of Genoa, due to the effective blockade organised League fleet. In midAugust, almost all of the Imperial fleet was destroyed: 35 ships were reportedly destroyed or taken in the port of Genoa, the starting point for the League’s takeover of the city. At the same time, the above story posted online by Mr. Damiani and that of Agostino Giustiniani describe four ships being sent to Sicily to look for grain. The first text we initially located—Mr Damiani’s—did not attract our attention because, as we have said, it refers to four galleys and not to large ships.

‘In the course of 1527, the city of [Genoa], then under the ‘dogato’ of Antiniotto Adorno, was in the grip of a terrible famine, the grain was so exhausted that the bread was rationed per capita and everyone received only three buns. And in the city a pound of grain was worth up to fourteen lira, and beyond its gates it was worth from eighteen to twenty-five lira. Four navi were armed with the support of ships from Sicily and elsewhere to transport grain to the city, including the Ferrara and Boscaina de Rapallo in the Gulf of SaintFlorent in Corsica, which were pursued by the French galleys, and were forced by lack of wind to touch land, the crews [‘chiourme’] were saved, but the hulls of the navi were burnt. The city had recovered Côte du Levant and Captain Andrea Doria, again in the pay of France, had been appointed General Captain and Admiral of the French fleet. He caused a lot of damage to the city, while it was holding the Portofino site, which had been fortified with bastions and other shelters…’

Agostino Giustiniani and Paolo Giovio’s accounts are more precise and detailed. They provide a series of corroborating information on the events. Of the two texts, however, it is Giustiniani’s that seems the most accurate, especially in its expression. It should be noted in this regard that Giustiniani writes in Italian while Giovio’s text is in French, translated from his original language, the Latin, which has led to inaccuracies. The texts of Giustiniani and Giovio provide information which makes the shipwrecks of 1527 the most suitable to match archaeological evidences with historical facts:

The other reference leads to a text written by Paolo Giovio, another Italian author also writing during the period. It corroborates A. Giustiniani’s text, except that it seems to indicate that the navi were burned by the French (GIOVIO, 1555, vol. II, 94):

• First, we would have to deal with ‘navi’, a term used by Giustiniani. Four in all, two of which were reported to have been shipwrecked in the Bay of Saint-Florent. These are ships of a very particular typology in the Genoese ship nomenclature of the time, which covers a precise architectural meaning. We will return to this point. • The circumstances of these shipwrecks are also particularly noteworthy: pursued by the French galleys, the two Genoese navi had to make landfall because of the lack of wind. The crews landed and probably unloaded everything on board at the same time. According to Giustiniani, the fire was deliberately set before they were abandoned, while according to Giovio, it is understood that they were captured and burned by the French after the crews had fled. This is the only significant discrepancy between the two texts.

‘Of some things that happened in France in 1527. So that we can better understand the causes of [Andrea Doria] breakdown with the King: Thus, about the time that Lautrec; having passed over the Alps; expelled Bosco, Alexandria, & Pavia, by the means that we have said, François; King of France evict an army of twenty-one galleys from the port of Marseilles. This army had been under the charge of André Doria whom he had recently appointed Admiral of the Mediterranean Sea, the most honourable title of all the Navy. Leading the siege of Genoa, ran by all the coasts above and below Liguria, and was diligently doing his duty in this charge, which was only that wheat shouldn’t be carried to the city. For then, there was a great scarcity of wheat in Genoa and the pestilence had begun to strike the inhabitants.: Antoniot Adorne, Prince of the city and of the Emperor’s party, did not have a marine army firm enough to face that of the François, who had previously pursued the Ferrara & Rapallina, two Genoese merchant ships, without favourable wind on the port of Saint-Florent in Corsica. He forced them to leave the shipboard and then there he burned them,

In fact, the descriptions made by Giovio and Giustiniani, even more so, coincide particularly well with the scenario suggested by the archaeological evidence. We are dealing with two Genoa merchant ships intended for a transport of wheat which, because of the lack of wind, they were facing the danger of being caught by a fleet of French galleys chasing them. They decided to unload the ships in a hurry and—in these circumstances – it is understandable that there was not enough time to land anchors and guns. 149

The Mortella III Wreck The ships were then set on fire voluntarily, according to Giustiniani’s version, to prevent them from falling into enemy hands. The only caveat we have regarding Giustiniani’s account is that he uses the term ‘chiourma’, to refer to the crews that managed to flee. In fact, this word is generally associated with galley rowers. It seems, therefore, in conflict with the ships normally referred to as navi.

the boat herself (GUAZZO, 1537, 305). This fact is further reinforced by the name reported by Giovio who indicates that the second ship was called the ‘Rapallina’. Again the name is obviously linked to Rapallo, the supposed port of origin of the ship.

Giustiniani’s text provides us with important information about the episode. First of all about the ships’ mission and route: they came from Sicily where they had loaded grain and went to Genoa where they intended to land it. Precious information is also provided: the names of the ships and their ports of origin. One was called Ferrara and the other Boscaina; they were from Rapallo, a port located about twenty kilometres southeast of Genoa. Giovio is less detailed, but he also gives the names of the two ships, the second one is said to have been named Rapallina, the second point of divergence with Giustiniani.

6.5.1. The architectural portrait of a Genoese nave

6.5. Provisional lessons drawn from historical research

Whatever the hypothesis, historical studies lead us to the sinking of two Genoese navi. We have so far taken care to put the word in italics because it is an Italian term which has a precise meaning. We will now examine it because the architectural characteristics associated with this terminology are a useful complement to the archaeological study, which was undertaken for a typological approach to these ships. In the first third of the 16th century, the Genoese built a type of ship called a ‘nave’ whose main characteristics were reported to us by Italian authors:

The two names of Ferrara and Boscaina can be associated, as it was the custom in Genoa in the 16th century, with surnames that generally refer to their owners. They differ from their baptismal name, which was often associated with a saint (CALEGARI, M, 1970, 21).

First, in her book Navi e cantieri della Republica di Genova, Luciana Gatti states that the nave was in Genoa ‘the largest commercial unit’ (GATTI, 1999, 145). This is important information; we are dealing with a type of ships designed for commercial use.

Ferrara was a very common surname in Italy, especially in Lombardy and Campania. It is a variant of the Ferrari family name, whose etymology evokes the blacksmith or farrier. This name is also—as is commonly the case in Italy—associated with a toponym: the city of Ferrara and the province of the same name located in the Po delta. For comparison, we are in a similar situation to the Villefranche-sur-Mer wreck, whose customary name is the Lomellina, named after her owner who was of important family in Genoa, the Lomellini (GUEROUT, RIETH et al., 1989, p. 145). Moreover, like Ferrara, Lomellina is also associated with a toponym located in the southwest of Lombardy. As far as Boscaina is concerned, the situation is the same. The surname Boscaina belongs to a noble Italian family from Piedmont. And as for Lomellina and Ferrara, there is also a toponymic correspondence, a small village in the province of Bergamo, in Lombardy, which now has 32 inhabitants. However, in the case of Boscaina, we have an additional similarity that links the word to the province of Vizcaya, in Spain. Etymologically, in fact, the word ‘Boscaina’ comes from ‘Biscaina’ which itself has its origin in ‘Viscaya’ or ‘Biscaya’: the Basque Country. As a result, it is difficult to say today whether the Boscaina nave owes her name to her owner or to her origin. It appears that the ships usernames was sometimes associated with the origin of its owner. In fact, Mr. Guazzo in his Historia di tutte le cose degne de memoria q vai del anno 1524, published in 1537, quotes a nave ‘Boscaina’ and another ‘Ragusea’ (from Ragusa, former Dubrovnik) showing that the word ‘Boscaina’ could refer to a Basque origin of the owner or

Figure 186. Fresco of the Palace of Viso del Marqués Álvaro de Bazán. Nave during the expedition led by Charles V against Algiers (1541) (Photo: Arnaud Cazenave de la Roche).

150

Historical study Dimensions/port

required mass transport, particularly because the city specialized in trading alum in the 15th century, a very dense and cheap raw material that required transport by large and solid seagoing vessels for it to be profitable (HEERS, 1954, 31-53 and 1958, 111, 112). This necessity will have led Genoa to construct increasingly large ships, of which the carrack will be the best representative. As a result it stood as the builder of the largest ships sailing across the Mediterranean for nearly a century. From the 16th century onwards, with the decline of the alum trade, Genoese manufacturers returned to the production of smaller nave sizes.

In terms of size, Gatti points out that it is the ‘larger’ commercial ship. Useful details are provided in a study carried out by Manlio Calegari where it appears that in 1509, the average tonnage of the nave in Genoa was 14,000 cantars, or 700 tonnes of registry and that the smallest units had a tonnage exceeding 8,000 cantars, or more than 400 tonnes of port (CALEGARI, 1970, 15 and 16). Features and rigging According to Gatti, the nave is a ship with three masts. It is provided with square sails on the foremast and mainmast, and a Latin sail on the mizzen.

The documentary study carried out in the Genoese archives by Manlio Calegari shows that between 11 and 16 navi with more than 8000 cantars were in service in Genoa each year during the first decade of the 16th century. The Genoese navi totalled 225,000 cantars in 1509, an average tonnage of 14,000 cantars per unit (669 tonnes of port). Over the period 1537-1539, there was an average of 11 units in service per year (CALEGARI, 1970, 26 et 31).

Another important information from the point of view of its architecture: it is a ship that most often has two decks. Military armament Gatti reports that—although navies were merchant ships—they were most often armed with several cannons to defend themselves against privateers. 6.5.2. Historical landmarks on the development of the Genoese nave According to Furio Ciciliot, who explored the Genoese archives in search of a chronological framework specific to each type of ship, the first mention of nave construction in Liguria could be found at the end of the 12th century, in 1190 (CICILIOT, 2005, 182). In his study on Italian shipping, Jacques Heers states that this type of vessel was introduced in Genoa in the 12th or 13th century by Basque sailors (HEERS, 1958, 109). The great Mediterranean trade that developed during the modern period was with the help of two types of ships, first of all the nave. It was the merchant ship favoured by Genoa, while Venice remains loyal to the galley. Nevertheless, for its trade, a new type of galley will appear, the ‘galee grosse da merchato’, a large merchant galley, which will replace the ‘galea sottile’, or light galley. On this subject, J. Heers writes: ‘Certainly in Venice, both [the galleys and the navi] are privately owned but the state monopoly applies to galleys with the whole system of regular auctions and convoys, so often described, while the navi were left to the individuals who had them of their own. And it is certain that, for most Mediterranean nations, the galley still prevails by far.’ This is the case in Florence, where, as in Venice, the galley predominated, while in Genoa, it was practically absent. As we can see, Genoa’s transport policy differed radically from its two rivals. According to J. Heers, who has carefully studied Genoa’s trade, the Genoese preference for the nave should be due to its very nature, which 151

Conclusion At the end of this study and the analysis of the information gathered during the five excavation campaigns carried out between 2010 and 2019, the wreck of the Mortella III stands as a precious illustration of Mediterranean shipbuilding in the first third of the 16th century. This work is the result of a convergence of archaeological evidence and historical documents that has led to the conclusion that the origin of the ship was probably Genoese. Her portrait, which has gradually been confirmed and clarified throughout these pages, is that of a nave, a merchant ship that contributed to the development of trade in the Mediterranean during the 16th century. This trade, controlled by the Italian states, was an important driving force in the European economy of the time, to which the nave contributed as the preferred Genoese merchant ship (HEERS, 1958, 110).

drift than another built with a flat floor. This characteristic, combined with the stretched longitudinal shape, would also provide an advantage when going upwind. On the other hand, the regular curvature of the sides and the absence of visible turn of the bilge can suggest that she must have been more sensitive to rolling. The historical portrait of the Mortella III ship The historical information gathered during the literature research (Chapter 6) furthers our knowledge of the Mortella’s ships in two ways: First, it provides a description of the events which led to their sinking plunging us into the heart of a turbulent historical period.

The architectural portrait of the Mortella III ship

Secondly, they lead us to favour a type of ship, the nave, coinciding with archaeological research and allows us to gain a better understanding of the ship we are dealing with and its characteristics.

The architectural portrait of the Mortella III ship can be summarized as follows: • It would have been a nave 550 to 570 tonnes burthen, a low value considering the average tonnage of this model of ship which was near 700 tonnes in the first third of the 16th century. This nave had two decks, three masts, and was nearly 37 meters long. Her maximum breadth was about 10.50 meters, her keel length 26 meters and her depth of hold 4.30 meters.

As we have seen, of the three historical episodes related to shipwrecks in the Bay of Saint-Florent that have been studied, that of 1527 is most likely to explain the presence of the wrecks of the Mortella. The historical and archaeological evidence corroborates the events of 1527 and therefore leads us to favour this hypothesis. In this hypothesis, the Mortella III wreck could therefore be the Boscaina or the Ferrara, a nave built in the Rapallo1 shipyard in the 1520s, a period suggested by the dendrochronological study (Chapter 2). As it was blocked by the fleet of the League of Cognac, composed of the French and Papal fleets alongside, Andrea Doria and Venice, Genoa would have sent the ship to Sicily with another nave. The aim of this expedition was to load wheat there and to try to supply the city, which was being starved by the blockade. Caught in Corsica by a French fleet of galleys in August, the two ships were trapped in the bay of Saint-Florent by the lack of wind. For this reason, they would have been hastily unloaded and deliberately set on fire to prevent them from falling into enemy hands.

The architectural study (Chapter 5) highlighted three major characteristics: 1. First, the ratio between the main breadth / keel length / overall length was of 1: 2.48: 3.50. It sketches the picture of a ship with particularly stretched shapes for a merchant vessel. 2. Her transverse shape was particularly round, as seen in the study of the ‘figure’ (shape) of its master-frame. This profile is closer to the shape recommended by the Iberian shipbuilders than that of the Venetians. 3. The pronounced rising of her master floor-timber is the third special feature of the Mortella III ship. In view of these architectural traits, what can we say about the nautical characteristics of the ship? This is a question that is difficult to answer precisely, especially considering that only a small portion of the remains have been preserved and the total absence of dead works. Nevertheless, it can be argued that the rather slender longitudinal shape will have limited the loading capacity in comparison to one built on the basis of a ratio of 1:2:3, exchanging this for higher nautical qualities. As the rising of the master-floortimber induced a strong draught, the ship was less prone to

The study of the Mortella III wreck provided an opportunity to look briefly at the historical context at the time of her 1 Rapallo is a port located about 25 km south-east of Genoa. It is a medieval fortified village linked to Genoa by an act of allegiance dated March 7, 1229 (DIZIONARIO COROGRAFICO, 1868, 742). The city now has a population of 30,000. Rapallo Bay is a strategic shelter on the Ligurian coast and its shipyard, renowned for the quality of its construction, has been a source of Genoa’s commercial ship production throughout the modern period.

153

The Mortella III Wreck has appeared throughout these pages as a reoccurring reference, with a lot of commonalities, starting with its origin and chronology. This does not mean that either of the Mortella III wrecks could be considered to be a ‘clone’ of Villefranche-sur-Mer’s one. The technical culture that links them was not static, and the several technical options and choices are specific to their shipwrights and it is this that makes them unique. Moreover, from the point of view of the construction techniques used, the richness of the comparative analysis comes as much from what separates the two ships from what brings them together.

sinking. It notes the rivalry between France and Spain within the arena of the Mediterranean of which the sinking of the Mortella III ship was an illustration (chapter 6). It highlights a complex political situation and the division of Italian cities stirred up by this Franco-Spanish rivalry. On the one hand, we see Venice united with the Vatican in the French camp; rising up against Genoa. On the other hand, we see how the alliance between Genoa and Spain and Genoa’s General Andrea Doria with France resulted in the merciless war that he waged against his own hometown. The texts show that the shipwrecks of the Mortella’s naves are the direct consequence of this thorny political situation.

What brings them together makes it possible to clearly distinguish and specify what common technical heritage they share. This is primarily a part of an Italian tradition, but with its regional specificities, which in this case is probably Genoese. Only once these Italian specificities have been extolled can it be then described as Mediterranean in a second place.

From an economic point of view, historians who have studied this period, notably Fernand Braudel and Jacques Heers, described it as being dominated by the Italian cities of Venice, Genoa and Florence. Of course, they were not the only ones in the structure of Mediterranean trade in the 16th century. We know, for example, that Venice has never managed to completely dominate ‘its Gulf’, the Adriatic Sea, where Ragusa was always overshadowing it (CHALINE, 2010). Barcelona was also an important player in the Mediterranean market. Nevertheless, it is true that the three Italian cities monopolized a large part of it. The archaeology in collaboration with the written documentation makes this picture perceptible. A result, the large majority of Italian shipwrecks in the Mediterranean in the modern era is an illustration of this fact.

1. Construction processes. They are part of a Mediterranean tradition whose origin has been highlighted by a series of markers consisting of technical ‘fingerprints.’ These were described in Chapters 3 and 4. In summary, and in connection with the Mediterranean technical model mentioned in Part 1.2 of Chapter 1, the main technical fingerprints identified on the wreck are:

The contribution of the wreck to a Mediterranean technical model

1. The nature of the fastening, attaching the planking to the frames by means of two iron nails placed near the edges, and the lack of use of treenails. 2. The use of ‘hook scarfs’, or ‘empatures à cadeau’ in French, for assembling the frame pieces in the central area of the ship, 3. The use of iron nails to fasten the frames timbers, and the absence of treenails, 4. The shape of the main-mast step device, arranged with a structure composed of two sister-keelsons, joined by dovetail keys and laterally braced by buttresses.

In Chapter 1, we have seen the limitations of the Mediterranean maritime space concept for establishing a culturally homogeneous and coherent geographical area. The diversity of the ethnic and cultural components of this immense territory makes it difficult to understand it other than through the prism of geography; several regions and peoples united by their shared access to a sea. In the field of naval archaeology, the concept of ‘Mediterranean technical culture’ has been developed in recent years, mainly in contrast to an ‘Atlantic’ model highlighted by shared building features common to some fifteen examined wrecks. However the three main representatives -Mortella III, included- of this ‘Mediterranean’ category are instead representative of a specifically western and northern Mediterranean building, and more explicitly Italian. In fact, both written documentation and archaeological discovery have confirmed the preponderance of the Italian states in the western Mediterranean during the 16th century.

Ultimately, these construction processes are shared with the Villefranche-sur-Mer wreck and partially shared with Calvi I. They show that the Mortella III ship closely matched the Mediterranean technical environment as it has been described in recent years. This study also makes it possible to complete the list of known technical fingerprints by identifying possible new ‘markers’ of Mediterranean technical culture. Five additional constructive characteristics must be added, the first of which is certainly Mediterranean in origin and the other four still requiring confirmation:

Once these caveats have been made, what assessment can be made of the contribution of the Mortella III wreck to the knowledge and definition of this Western Mediterranean technical model? This contribution cannot be evaluated without comparing its construction and architectural characteristics with those of its ‘older sister’, the Villefranche-sur-Mer wreck (1516). This wreck

1. The use of circular nails for joining the timbers of the frame together, and for fastening the planking to the frame. 2. The fastening of the planking to the frames with nails passing through the frames, and clenched on their inner faces. 154

Conclusion

Shipwreck / shipbuilding technique Cala Culip

Period

Table 14. Comparison of technical ‘fingerprints’ on some Mediterranean wrecks of Mediterranean building tradition Probable Origin

Scarf type floor/ first-futtock between tailframes

Fastening of the planking to the frames

Genoa

Hook scarfs. 2 iron nails Iron nails with circular bend tips

2 sister keelsons. Dual. 2 keys Butt scarfs

Pitch

Villefranche

Genoa

Hook scarfs, dovetail scarfs. 2 iron nails

Iron nails

circular

2 sister keelsons. Single 2 keys

Lead

Italy

Dovetail scarfs

Iron nails

Circular on part of the garboard

Genoa

Hook scarfs between Iron nails the crutches and the first-futtocks. Iron nails

16th century

14th century

Mortella III

3. A keel made up with two overlapped timbers. 4. The use of butt scarfs to join the keel timbers. 5. A water removal device whose pump-valve is located between two frames on which it rests.

 

2 sister keelsons

Keel Outer Morpho treatment logy of the hull

Hook scarfs

Delta II2

 

Type of maststep

Catalonia

Calvi I

 

Shape of the nails

 

  2 sister keelsons. Dual 2 keys

 

  Lead

Mortella III (1: 2.48 : 3.50) and Villefranche-sur-Mer (1 : 2.56 : 3.52) ships were closer to those of the Spanish ‘nueva fabrica’ of the late 16th and early 17th centuries (1 : 3 : 3.75) than those of the other 16th century merchant ships we know. As we saw in Chapter 1, the latter were mainly built using a ratio close to 1:2:3, in accordance with the very generalized ‘As-Dos-Tres’ rule. The proportions of the Atlantic tradition wrecks of the 16th century that have been studied fit particularly well with this rule. For instance, the Red-Bay wreck is such an example.

As shown in Table 14, most technical fingerprints are clues that, alone, do not have the power to establish an indisputable technical origin, except perhaps the typology of the mast-step. It is in fact the convergence of these markers that will make it possible to identify a technical tradition. 2

What conclusions can be drawn from the above? Accepting the Genoese origin of the Mortella III and Villefranchesur-Mer wrecks, and considering that their proportions coincide with those recommended by Venetian authors of the 15th century, such as Zorzi Trombetta da Modon or Fabrica di galere, we can deduce that the slender shape of the Italian navi could be a Mediterranean and more specifically an Italian architectural trait. However, this by no means implies that the As-Dos-Tres rule was unique to the Atlantic. Archaeological work has demonstrated its presence in the Mediterranean with the wreck of Calvi I, whose proportions were close to 1:2:3 ratio. Certain texts do the same (chapter 1, part 1.3.1), notably those of the Ragusan Nicolò Sagri (1550), the Roman Bartolomeo Crescentio (1601) or the construction contract for the Illyric fleet of Pedro de Ivella (1570-1580), to name but a few.

Nevertheless, the list that has been drawn up includes more or less strong markers. Some of them, such as the mast-step arrangement with two sister-keelsons, or the fastening of the planking to the framing with iron nails, without treenails, are very strong clues. In fact, to our knowledge, these technical processes are not found anywhere in the Atlantic tradition. Others, such as the use of dovetail scarfs to join the floor-timbers to the first-futtocks between the tailframes, are weaker indicators because, although associated with the Atlantic tradition, this technique is sometimes also found in the Mediterranean shipbuilding tradition (in Calvi I, for example). We could therefore establish a sort of hierarchy of the value of these technical fingerprints. The contribution of the wreck to the definition of a Mediterranean architectural model But the construction characteristics of a ship are not limited to the carpentry technics. They also take into account her architectural characteristics, in other words her forms and proportions. They must be examined in order to determine how they fit into the Mediterranean shipbuilding tradition.

The second particular architectural feature of the Mortella III wreck is the significant rising of its master floor-timber, a characteristic also found on the Villefranche-sur-Mer and Calvi I wrecks. This trait is important, because it gave a round transverse shape to the hull which contrasts with the tradition of flat floor-timbers of Atlantic ships observed on sixteenth-century shipwrecks.

First of all, with regard to the ratio between the maximum breadth / keel length / overall length, the ratios of the

In support of Duhamel du Monceau’s remarks (see Chapter 5), it seems that the rising of the master floor-timber can be considered to be an architectural trait linked to the

This information on this wreck is based on: (HIGUERAS-MILENA, GALLARDO, 2016).

2

155

The Mortella III Wreck So far, naval archaeology has been mainly devoted to the study of construction techniques. It is symptomatic in this respect that T. Oertling’s Ibero-Atlantic model is based only on technical ‘traits’ and that any architectural considerations related to shape and proportions are absent. However, from our point of view, their consideration is essential in the comparative analysis of Atlantic and Mediterranean technical cultures. Without them our vision would remain fragmentary and incomplete. It is this dimension that our excavation programme strives to take into account.

Mediterranean tradition, and more specifically, Italian. But this hypothesis needs to be qualified: It should be noted that not all the wrecks of Mediterranean tradition studied in recent years are built with a risen master floor-timber. Two examples can be mentioned here. The Cala Culip VI boat (RIETH 1998, 207) and the Ottoman wreck of Yassi-Ada (LABBE, 2010, 154), both had flat floor-timbers. In fact, since the rising of the master floor-timber directly influences the draught, it is understood that regional navigation conditions and/or the commercial need to navigate in shallow water (e. g. around port or river areas) are of crucial importance in this architectural choice, regardless of the nautical space we are dealing with. And it is perhaps this requirement that led the shipwrights of the small coastal navigation boats of Cala Culip and Yassi-Ada to opt for flat bottoms.

The exploration of the Venetian treatises in relation to the Mortella III wreck. We cannot conclude this work without saying a few words about the Venetian treatises. The study of the remains of the Mortella III wreck led us to explore them, since they were of such great help for the analysis. These texts, which are partly dated to the fifteenth century, are difficult to access because they are difficult to read; they were written in the old Venetian vernacular. They are all based on the same model, that of a monotonous enumeration of dimensions and values. This has sometimes led them being referred to as ‘technical recipe books’. But this pejorative assessment is probably a result of having been understudied. Actually, Venetian treatises teach us a series of rules of proportions that don’t discuss techniques, but architectural geometry and design processes. In this sense, they reveal a sophisticated naval architecture.

The third architectural feature that brings the wrecks of Calvi I, Villefranche and Mortella III together is the shape of their master frame. This followed a line of a perfect circular arc in the case of the first two and very closes in the third. This characteristic documented by archaeologists brings us to a contrasted situation in the Mediterranean since the Venetian builders advocate more complex fusiform shapes with two breaks in the lines, one above the bilge and another at the main breadth. The likely Genoese origin of the ships of the Mortella III and Villefranche, prompts the question as to whether this represents significant regional architectural difference in navi design? This conclusion, which still requires confirmation, would likely show that there was not one homogenous concept of Italian shipbuilding, but several. There is a clearly common technical heritage, but also there would have been significant regional differences. In the meantime, the master-frame design of the three wrecks should be compared to the Iberian conception. However, a significant exception is the Basque tradition highlighted by the Red-Bay wreck. This could have been inspired by the English method, itself probably influenced by the Venetian (Chapter 1, section 1.3.2 and 5, section 5.1.2.3).

Indeed, if we follow the suggestions of the Fabrica di galere (1410) or the Libro de Zorzi Trombetta da Modon (1445) in the description of their navi, we reach a precise restoration of their shapes and dimensions, each of them being linked to the previous by a rule of proportionality. This level of dimensional interrelation does not exist in Iberian shipbuilding, even in the 16th century. On the other hand, in the Iberian treatises of the 16th and early 17th centuries, there are many details regarding technical construction methods (methods of assembly, fastening of the timbers, etc.) that were missing from the Venetian texts. These also refer to the design process: the moulding process is mentioned and also innovative techniques, such as the ‘trébuchement’ or the ‘haul down the futtock’. This has been called ‘partisone del ramo’ since the beginning of the 15th century, whereas it will only be introduced in Iberian shipbuilding in the early 17th century under the term ‘joba’ as one of the great innovations of the ‘nueva fabrica’.

To complete this overview of the architectural study, it should be pointed out that so far, it has been based essentially on the study of the master-frame. But this alone can only provide a partial knowledge of the geometry of the ship. It may suggest it in some ways, since we know that the frames located between the tailframes were probably designed on this model. But this does not remove the need to study the frames forward of the master-frame.3 Only their study will make it possible to assess the real evolution of the shapes, the rising of the floors and the narrowing of the flat, parameters that cannot be neglected to restore the correct shape of the hull. This work was scheduled during the sixth and final excavation of the wreck, an essential task in bringing the excavation programme to an end.

The Venetian treatises, the first of which we know was written at least 150 years before the first known Iberian equivalents, are therefore surprising in their early foresight. In our opinion they testify to the high degree of development in Italian shipbuilding during this period, which itself is probably a mirror of the splendour of the arts of the Italian Renaissance. For this reason they would deserve to be studied in depth in the future. Do the visible remains of the Mortella III wreck give us a glimpse of this degree of achievement? Yes indeed, in

The frames after the master-frame were too degraded to perform this task satisfactorily.

3

156

Conclusion some aspects of the ship’s architecture. In the field of construction processes, for example, a device such as the mast-step is impressive for its high degree of technicality. However, it contrasts with the heterogeneous nature of the framework’s construction. The lack of adjustment of the pieces had constantly to be made up for by the addition of shims and wedges (Chapter III). This characteristic was also found in Villefranche-sur-Mer. From this point of view, the Red-Bay ship, for example, seems to be of a more neat construction.

origin Organization meaning ‘Collaborative Research Project’) on ships of the 16th century.6 Finally, as acknowledged, however influential Italy may have been in the modern Mediterranean, shipbuilding in this period cannot be reduced to the Italian states, nor to Ragusa. As its history clearly shows, the weight of the Ottoman Empire in the Mediterranean was so considerable throughout the 16th century that the development of the shipbuilding model in this nautical space would not give us an accurate picture without studying wrecks from other regions—in the South and East—under Ottoman influence. Researchers interested in shipbuilding and navigation during this period are aware of this. In this respect, they observe with interest the recent developments in nautical archaeological research, particularly in Turkey, Egypt, Algeria, and Morocco.

Future prospects for knowledge of Mediterranean shipbuilding. The Mortella III wreck undoubtedly contributes to the still limited information documenting shipbuilding in the Mediterranean at the time of the Renaissance. Its architectural testimony consolidates that provided by the wreck of Villefranche-sur-Mer and Calvi I. It helps to clarify the dominance of Italian building amongst the western Mediterranean in the 16th century. However, there is still a long way to go to shed light on the various questions raised by the study of the remains of the wreck’s hull. While those of Calvi I and Villefranche-sur-Mer are important references for archaeological analysis, many of the answers to the problems addressed in our work require the documentation of further wrecks built within the Mediterranean construction tradition. They are also necessary for the emergence of statistical guidelines that can be confidently relied on. The excavation programmes recently organised on Mediterranean-built shipwrecks from the modern period, particularly in Croatia and Spain4, combined with the Mortella II wreck, are likely to change this situation in the coming years. We now have the opportunity to build a documentary corpus that—for the first time—opens the prospect of laying the foundations for a Mediterranean model, or more precisely an ‘Italo-Mediterranean’ model, as was the case for the ‘Ibero-Atlantic’ model, who’s foundations were set out in the 1980s.

Shipbuilding in the French and Spanish ‘Levant’ also has a lot to teach us. It is hoped that future discoveries from these regions and from the eastern and southern sides of the Mediterranean will lead to a comprehensive understanding of Mediterranean shipbuilding in all its complexity and diversity. In the meantime, the panel of wrecks at our disposal make it possible to undertake a foundational work by paving the way of a model that we could call ‘Italo-Mediterranean’. This is a first step. It will undoubtedly allow us to develop our knowledge of this extraordinary floating machine that the boat represented during the Renaissance period.

This project, entitled ‘ModernShip project’ was awarded by the European Union’s Horizon 2020 programme for Research and Innovation (Marie Sklodowska-Curie Actions nº843337). It includes the participation of a group of researchers from the ForSEAdiscovery project consortium led by Prof. Ana Crespo Solana (Spanish National Research Council - CSIC).5 It also collaborates with an international group of institutions and researchers organized into a ‘Projet Commun de Recherche’ (French 4 Spain is making a major contribution to the documentation of Mediterranean building tradition with the wreck of the San Giacomo de Galizia (which appears in Spanish texts as Santiago de Galicia) located in Ribadeo (Galicia, Spain). The ship was built in Naples in the last third of the 16th century and the wreck remains in a remarkable state of conservation. The excavation programme is led by Dr. Miguel San Claudio with the support of the ForSEAdiscovery Consortium (MarieCurie Actions, PITN-GA-2013-607545)) and the Institute of Nautical Archaeology (INA, Texas). 5 The ForSEAdiscovery project involves, among others, the Spanish CSIC (Spanish National Research Council), the University of Texas A&M, the University of Wales and the Maritime Archaeolgy Trust, MAT, U.K.

6 This ‘PCR’ was founded by the NGOs Groupe de Recherche en Archéologie navale -GRAN- and Centre d’Etudes en Archéologie Nautique -CEAN. It brings together the following research institutions: the University of Paris-Sorbonne and the Musée National de la Marine for France; the Spanish National Research Council (CSIC) for Spain; the Università degli Studi di Genova, the Università Ca’ Foscari Venezia and the Università degli Studi di Sassari for Italy; the Croatian Conservation Institute - CCI for Croatia.

157

Sources and Bibliography 1. SOURCES

Diaz Pimienta, Manuscript of the Colección Vargas Ponce, 1645, Archivo del Museo Naval de Madrid – MNM, T.3, doc.102.

1.1. Handwritten sources Anon., c.1410, ‘Fabrica di galere’, Biblioteca Nazionale Centrale di Firenze, codex Magliabecchiano, XIX.7. Manuscript partially transcribed and translated by A. Jal in Archéologie Navale, vol II, Mémoire nº5, p.1 à 106, Paris, 1840. Extracts not published by Jal were published by Anderson, R.C., 1945, ‘Jal’s Memoire no. 5 and the Manuscript ‘Fabrica di Galere’ ’, in Mariner’s Mirror, 31:160-167. The manuscript is also commented in Bellabarba, S. 1988. ‘The Square-Rigged Ship of the ‘Fabrica di Galere’ ’. The Mariner’s Mirror, 74.2: 113130, 225-239.

Dracchio, Baldissera Quinto, c.1594, Visione, Venice, ASV Archivio di Stato di Venezia, fondos Contarini, ms 19, Arsenal, b.1. English translation of Louis T. Lehmann, 1992, Amsterdam Echeverri, J. A., betweeen 1648 et 1666. Manuscript of the Archivo del Museo Naval de Madrid –MNM- attributed to J.A. Echeverri by Fernández Duro. Colección Vargas Ponce, T.3ª, Doc. 108 fol. 391-395. Garrote, Francisco, 1691, ‘Nueva fábrica de baxeles españoles’, Archivo del Museo Naval de Madrid, MNM, Madrid.

Anon., 15th c., ‘Ragioni antique spettanti all’arte del mare et fabriche de vasselli’, Venetian manuscript kept at the Greenwich National Maritime Museum, ms NVT 19. Published under the direction of de Bonfiglio Dosio, G., 1987, with a study of Pieter Van Der Merwe, Alvise Chiggiato, David V. Proctor, Venice.

Lasalde, Juan de, 1581, ‘Juan de Lasalde al rey ofreciéndose para la fabricación de ocho galeones, mayo 1581. Reales cedulas expedidas por el rey a Cristóbal de Barros y siete pareceres…’, MNM, Colección Fernández Navarrete, T.XXII, doc.76, f°299-301

Anon., c.1570–1580 – ‘Relación de la fábrica de doce galeones de guerra de la Escuadra Yllirica de Pedro de Ivella y Estéfano Dolisti’. MNM, Colección Navarrete, Tomo IX, doc. 27.

Madeleine, Jean-Baptiste de la, 1712, ‘Tablettes de marine [manuscrit]: proposées par M. de la Madeleine pour l’instruction de ses neveux’, BMM, côte R711, Paris. Michele da Rodi, early 15th c. ‘Libro’. Collection privée. The text of this manuscript is largely copied in ‘Fabrica di galere’. The facsimile of the manuscript was published in Long, P.O., Mac Gee, D., Stahl, A. (ed), 2009, The Book of Michael of Rhodes, a fifteenthcentury Maritime Manuscript, 3 vol., Cambridge MA. A digitized copy of the manuscript is available on the website: http://brunelleschi.imss.fi.it/michaelofrhodes/ manuscript.html

Anon., 17th c., Document probably earlier than 1613. ‘El arqueo de Cristóbal de Barros’. MNM, Colección Vargas Ponce, T.XXV B, doc.19 f. 42-43 Anon., 1691, ‘Traitté de la construction des galères’, 1691, SHM, Service Historique de la Marine, Vincennes, ms SH 134) ; commented edition in Fennis, J., 1983, Un manuel de construction des galères de 1691, Amsterdam.

Pre Theodoro de Nicolò, c.1550, ‘Instructione sul modo di fabricare galere’, Biblioteca Nazionale Marciana di Venezia, Manoscritti italiani, cl. IV cod. XXVI (5131).

Baker, Mathew, c. 1580, ‘Fragments of Ancient English Shipwrightry’. Cambridge, Magdalene College, Pepysian Library, Ms. 2820.

Sagri, Nicolò, 1570, ‘Il carteggiatore’ manuscrit n°31951SA0111372C, TC Wilson Library, U. de Minneapolis. Transcription in Dell’Osa, D., 2010, Il carteggiatore di Nicolò Sagri, Commented transcription of the manuscript. Ed. Francoangeli, Milan.

Busturia, Domingo de, 1568, ‘Relación del maestre Domingo de Busturia en lo tocante a los arqueamientos de las naos que se toman para armada en esta costa de Biscaya por mandado de su Magestad.’ AGS, Guerra Antigua, Leg. 347, nº 23. 1568. Published by Casado Soto, J.L. in ‘Flota atlántica y tecnología naval hispana en tiempos de Felipe II’, ‘’ – Vol. 2. In Congreso Internacional sobre las sociedades ibéricas y el mar a finales del siglo XVI, 1998, Comisaría General de España, Expo de Lisboa ‘98.

Vargas, Rodrigo, c.1570, ‘Apuntamientos de Rodrigo de Vargas.’ AGI, Real Patronato, leg. 260, 2º, rº 35. Published in Casado Soto, 1988, Los barcos españoles del siglo XVI y La Gran Armada de 1588, Madrid. Villafranca, Marques de, 1541 ‘Correspondencia del Marques de Villafranca’, preparativos en la Armada de Andrea Doria para la empresa de Argel. AGS: ESTADO 1033 fol.176.

‘Comptes de Ragueneau, année 1526’ in Mélanges sur les Finances, la Marine, les Pèches et le Jansénisme, handwritten and printed documents of the BNF, ed. 1601 to 1700, mss. f°188. 159

The Mortella III Wreck construction des vaisseaux, Ed. Charles-Antoine Jombert, Paris.

Zorzi Trombetta da Modon, c.1444, ‘Libro’, British Library, London, Cotton ms. Titus A XXVI, Manuscript partially transcribed by R. Anderson in ’Italian Naval Architecture about 15th century’, Mariner’s Mirror, T.11, 1925, p.135-163.

Dupuys, Jacques and Nicot, Jean, 1573, Dictionnaire français-latin, Paris J. Dupuys Escalante de Mendoza, Juan, 1575, ‘Ytinerario de navegación de los mares y tierras occiden-Tales’, ed. Cesáreo Fernández Duro, in Disquisiciones náuticas (Madrid: Aribau, 1880), vol. 5, p. 413-515 (republishing under the title Itinerario de navegación de los mares y tierras occidentales, Madrid: Museo Naval, 1985).

1.2. Printed sources Anon., 18th c., ‘Tratado de galafatería.’ Text from the 17th c., in Fernández Duro, Cesáreo,1888, Disquiciones Náuticas, Ed. del Ministerio de defensa, Instituto de Historia y Cultura Naval, Madrid, 1996, vol. VI, p.243.

Fernandes, Manoel, 1616, ‘Livro de traças de carpintaria’, ed. Manuel Leitão under the title Livro de traças de carpintaria de Manuel Fernandes, Academia de Marinha, 1989, Lisbon.

Bonfadio, Jacopo, 1586, Gli Annali di Genova. Dal 1528, che recuperò la libertà, fino al 1550, ed. Bartoli, Genova. Bonnefoux, P.M.-J., 1848, Dictionnaire de la Marine à voile, Paris

Foglietta, Oberto, 1575, Delle cose della Repubblica di Genova, Ed. Antonii, Genova.

Bouguer, Pierre, 1746, Traité du navire, de sa construction et de ses mouvemens, ed. Jombert, Paris.

Fournier, Georges, 1643, Hydrographie, contenant la théorie et la practique de toutes les parties de la navigation…, Paris.

Bourdeille, Pierre de, said Brantôme, 1868, Œuvres complètes, Vol. IV, Ed. Mme Ve Jules Renouard, Paris.

Furttenbach, Joseph, 1629, Architectura navalis, dast ist von dem Schiff, Gebau auff dem Meer und Seekusten zugebrauchen, Published at Ulm (Germany).

Cano, Thomé, 1611, ‘Arte para fabricar, fortificar y apareiar naos de guerra merchante, con las reglas de arquearlas reduzido a toda cuenta y medida, y en grande utilidad de la navegación’, 1611, Luys Estupiñan, Seville.Transcription of the manuscript in Duro, Cesário Fernandez, 1996, Disquisiciones nauticas, vol. V, 1880, Madrid: Instituto de Historia y Cultura Naval, p.36-97.

García de Palacio, Diego, 1587, Instrucion náutica, para el buen uso y regimiento de las naos Mexico: Pedro Ocharte. Republished under the title: Instrucción náutica para navegar, Ed. Cultura Hispánica, 1944, Madrid. Republished under the title Nautical Instruction, 1587, translation. J. Bankston, Bisbee,Arizona : Terrenate Associates, 1986.

Capelloni, Lorenzo, 1562, La vita, e gesti di Andrea D’Oria. di Lorenzo Capelloni – In Vinegia : appresso Gabriel Giolito de Ferrari et fratelli, Venice.

Gaztañeta, José Antonio de, 1696, ‘Proposiciones de las medidas arregladas a la construcción de un Bajel de Guerra…’ in Grand dictionnaire des Arts et des Sciences, 1712, Amsterdam.

Casoni, Filippo, 1708, Annali della Republica de Genova del secolo decimo sesto, ed. Antonio Casamara, Genova.

Giovio, Paulo, 1555, Histoires sur les choses faictes et avenues de son temps en toutes les parties du monde, vol.II, ed. Guillaume Roville, Lyon.

Ceccaldi, Marc-Antonio, 2007, Histoire de la Corse – 1464 - 1560, Ed. Alain Piazzola.

Guicciardini, Francesco, 1738, Histoire des guerres d’Italie, vol. III, Ed. P. and I. Vaillant, London, p.190, p.254 et 257.

Colomb, C., 1979, ‘Lettre sur le quatrième et dernier voyage, said ‘lettre rarissime’ (7 juillet 1503)’ in La découverte de l’Amérique II Relations de Voyage. 1493-1504. Ed. La Découverte, Maspero.

Giustiniani, Agostino, 1537, Castigatissimi annali con la loro copiosa tavola della Eccelsa et Illustrissima Repubblica di Genova, Genoa. Ed. Cambridge (Mass.): Omnisys, 1990. Online on Gallica: http://gallica.bnf.fr/ ark:/12148/bpt6k58804h

Crescentio Romano, Bartolomeo, 1607, Nautica Mediterranea, Editions Bartolomeo Bonfadino, Rome. Dampier, William, 1729A new voyage round the world, 7th ed. James & John Knapton, London.

Guérin, Léon, 1861, Les marins illustres de la France, Ed. Morizot.

‘Decreto del Consejo de guerra sobre los inclusos papeles que trajo el Señor Diego Brochero Anaya tocantes a la nueva ordenanza de navíos.’ AGS – Guerra y Marina, legajo 776 published by Rodriguez Mendoza, B. M., 2008, Standardization of spanish shipbuilding: ordenanzas para la fábrica de navíos de guerra y mercante – 1607, 1613, 1618, Texas A&M University.

Jal, Auguste, 1840, L’archéologie navale, 2 volumes, Arthus Bertrand, Paris. Jal, Auguste, 1842, Documents inédit pour l’histoire de la Marine au XVIème siècle. Imprimerie royale, Paris.

Duhamel du Monceau, Henri Louis, 1752, Elémens de l’architecture navale ou traité pratique de la 160

Sources and Bibliography Vial de Clairbois, H.S., 1787, Traité élémentaire de la construction des vaisseaux, edition of 1805, Paris.

Juan, Jorge, 1757, ‘Compendio de navegación para el uso de los cavalleros Guardias Marinas, en Cádiz’, in Academia de los mismos cavalleros.

2. BIBLIOGRAPHY AND REFERENCES

Lavanha, João Baptista, 1610 ‘Livro primeiro da arquitectura naval’, c. 1598-1620, ed. Joào da Gama Pimentel Barata, Ethnos: Revista do Instituto portugués de arqueología, historia e etnografía, vol. 4 (1965), p. 221-298 (réedition sous le titre Livro primeiro da arquitectura naval, 1996, Academia de Marinha, Lisbon.

Agosto, Aldo, 1971, Origini ed evoluzione storica degli stemmi dei capoluoghi delle quattro province liguri. Ed.A. Compagna, Genoa. Alves, F., Rieth, E., Rodrigues, P., Aleluia, M., Rodrigo, R., Garcia, C., Riccardi, E., 2001, ‘The hull remains of Ria de Aveiro A, a mid-15th century shipwreck from Portugal: a preliminary analysis’, In Proceedings, International Symposium on Archaeology of Medieval and Modern Ships of Iberian-Atlantic Tradition: Hull Remains, Manuscripts and Ethnographic Sources: A Comparative Approach, Ed. Francisco Alves. Trabalhos de Arqueologia 18. Instituto Portugués de Arqueologia, Lisbon, p.317 to 345.

Nicot, Jean, 1606, Thresor de la langue françoyse tant ancienne que moderne, edition David Douceur, Paris Oliveira, Fernando, 1570, ‘Livro da fábrica das naos’, Ed. Manuel Leitão, 1991, under the title O livro da fábrica das naos do Padre Fernando Oliveira, Academia de Marinha, Lisbon. Ordenanzas de 1590. In ‘Arqueamiento de navíos’ M.N.M. Colección Navarrete, vol.I, nº de catálogo 789, Orden dada en San Lorenzo el 20 de agosto de 1590, doc. 13, fº168r à 170v.

Anderson, R. C, 1925, ‘Italian Naval Architecture about 1445’. The Mariner’s Mirror, vol. 11, London, p. 136168. Apestegui, Cruz, 1998, ‘Arquitectura y construcción navales en la España Atlántica, el siglo XVII y primera mitad del XVIII. Una nueva sistematización’ in Proceedings International symposium on archaeology of medieval and modern ships of Iberian-Atlantic tradition, Centro Nacional de Arqueologia Náutica e Subaquática, Academia de Marinha, Lisbon 7 to 9 sept. 1998, Lisbon, p.163 à 212.

Ordenanzas de 1607, ‘para la fábrica de navíos de guerra y mercantes’, In Colección de documentos y manuscriptos compilados. M. Fernandez Navarrete, Kraus-Thomson Organization Limited, Nendeln, 1971, Liechtenstein. Ordenanzas de 1613, ‘para la fábrica de navíos de guerra y mercantes’. Appendix of Función y evolución del galeón en la carrera de Indias. F. Serrano Mangas, 1992, 211-36. Colección Mar y América, 9. Ed MAPFRE, Madrid.

Arnold, J. Barto, III et Robert S. Weddle, 1978, The Nautical Archeology of Padre Island: The Spanish Shipwrecks of 1554. Academic Press, New York.

Ordenanza de 1618, ‘Recopilación de Leyes de Indias mandadas imprimir y publicar por la Magestad Católica del Rey don Cárlos II. Nuestro Señor.’ 4. Impresión 1943, 340-362. Gráficas Ultra, Madrid.

Aroztegui, Martin de, 1920, ‘Spanish Shipbuilding Ordinance, 1613’, Cambridge, Harvard University, Houghton Library, Palha Manuscripts, Ms. 4794, vol. 2. Published in La arquitectura naval española (en madera) bosquejo de sus condiciones y rasgos de su evolución, Ed. Gervasio de Artíñano y de Galdácano, Ap. 9.

Pantero Pantera, 1614, L’Armata navale, E. Spada, Roma. Petit, Edouard, 1887, André Doria, un amiral condottiere au XVIe siècle (1466-1560), Thesis defended at the Faculty of Arts of Aix, Quentin, Paris.

Assereto, Giovanni, 2000, Le metamorfosi della Repubblica : saggi di storia genovese tra il XVI e il XIX secolo, Ed. Daner Elio Ferraris, Savona.

Richer, Adrien, 1789, Vies du capitaine Cassard et du capitaine Paulin (connu sous le nom de baron de la Garde), Belin, Paris.

Aufan, Robert et Thierry François, 1990, Histoire des produits résineux landais: Goudrons, poix et brays gras depuis l’Antiquité, en Buch, Born et Marensin. Société historique et archéologique d’Arcachon et du pays de Buch, Arcachon.

Richer, Adrien, 1783, Vie d’André Doria, prince de Melfi, général des armées navales de France sous François Ier, ensuite de celles de l’empereur Charles Quint, Belin, Paris. Sigonio, G., 1586, De vita et rebus gestis Andrea Auriae Melphiae principis, Genoa.

Baker, W. A., 1978, ‘Comments on the Padre Island ship.’ Appendix F, The nautical archeology of Padre Island: the Spanish shipwrecks of 1554, Academic Press, New York, 385-389.

Thou, Jacques Auguste, 1734, Histoire universelle depuis 1543 jusqu’en 1607, Vol II, London. Veitia Linage, J., 1671, Norte de la Contratación de las Indias Occidentales. Publicaciones de la Comisión Argentina de Fomento Interamericano, 1945. Buenos Aires.

Barker, Richard A., 2003, ‘Whole-moulding: a preliminary study of early English and other sources’ in Shipbuilding Practice and Ship Design Methods from the Renaissance to the 18th Century, Max Planck Institute for the History of Science. 161

The Mortella III Wreck Barker, Richard A., 1998, ‘Sources for Lusitanian shipbuilding’, Lisbon Round Table, September 1998: Archaeology of Medieval and Modern Ships of IberianAtlantic Tradition. Online: http://home.clara.net/ rabarker/LISBON98.htm

Beltrame, Carlo, Gelichi, Sauro, Miholjek, Igor, 2014, Sveti Pavao shipwreck, a XVIth century Ventian merchantman from Mljet, Croatia, Oxbow Book, Oxford. Benvenuti, Gino, 1977, Storia della Repubblica di Genova, Ed. Mursia, Milan.

Barker, Richard A., 1991, ‘Design in the Dockyards, about 1600.’ In Reinders, T.; Paul, K., Ed. Carvel Construction Technique. Oxford: Oxbow, p. 61-69.

Bernard, Hélène, 2008, Expertises corses 2007, Golfe de Saint-Florent, Haute-Corse, épaves de la Mortella 2 et 3.  DRASSM – DAPA – Ministère de la Culture et de la Communication, No published report.

Barker, Richard A., 1990 ‘What Fernando Oliveira Did Not Say about Cork Oak’, in International Reunion for the History of Nautical Science and Hydrography, Aveiro, 15-19 sept. 1998. Published under the title ‘Fernando Oliveira e o seu tempo: Humanismo e arte de navegar no renascimento europeu (1450-1650) in Actas da IX Reunido internacional de historia da nautical e hidrografía, Ed. Inácio Guerreiro and Francisco Contente Domingues, p. 163-175, Patrimonia, Cascáis.

Bernier, M.-A-, Grenier, R. et al., 2007, L’archéologie subaquatique de Red Bay, la construction navale et la pêche de la baleine basques au XVIème siècle, 5 volumes, Parcs Canada, Ed. Canadien Museum of Nature, Ottawa. Bettencourt, J. 2011, ‘Angra B e Angra F (Terceira, Açores): dois navios ibéricos modernos para a navegação oceânica’, in Actas do Colóquio Internacional ‘A Herança do Infante’, Câmara Municipal de LagosCentro de Estudos dos Povos e das Culturas de Expressão Portuguesa FCH/UCP, Centro de História de Além- Mar, Lagos, 217-235.

Barker, Richard A., 1988, ‘Many may peruse us’: Ribbands, Moulds and Models in the Dockyards, Revista da Universidade de Coimbra, 34, 539–59. Barkham, Michael, 1985, ‘Sixteenth Century Spanish Basque Ships and Shipbuilding: The Multipurpose Nao’. In Postmedieval Boat and Ship Archaeology: Papers Based on Those. Presented to the International Symposium on Boat and Ship Archaeology in Stockholm,1982, Ed. Carl Olof Cederlund, p. 113-135. Swedish National Maritime Museum Report N. 20, BAR Publishing International Series 256, Oxford.

Bettencourt, J. 2014, ‘Angra B, un pecio español del siglo XVI en la bahía de Angra (isla Tercera, Azores, Portugal): resultados de una investigación en curso’, I Congreso de Arqueología Náutica y Subacuática Española, Cartagena 14, 15 y 16 de Marzo de 2013, Ministerio de Educación, Cultura y Deporte, Madrid.

Barkham, Michael, 1985, Spanish Basque shipbuilding: the port of Zumaya, 1560-1600 ; an historical economic geography of a merchant capitalist industry, Master’s thesis of Newfoundland University.

Biancotti, A., 1936, Andrea Doria. Turin. Blake W. et Green J., 1986, ‘A Mid XVI century Portuguese wreck in the Seychelles.’ International Journal of Nautical archaeology. Volume 15, Issue 1, fev. 1986.

Barkham, Michael, 2007, ‘Etude sur la construction basque espagnole c. 1550-C.1600’, N°422 (Ottawa : Parcs Canada, 1981), p. 8-12, published in Annex 1 of L’archéologie subaquatique de Red Bay, la construction navale et la pêche de la baleine basques au XVIème siècle, Parcs Canada, Ed. Canadien Museum of Nature, Ottawa, volume 5, p. V1 to V48.

Braudel, Fernand, 1977, La Mediterranée, les hommes et l’héritage (vol. 1), l’espace et l’histoire (vol.2), Paris. Braudel, Fernand, 1979, Civilisation matérielle, économie et capitalisme, XVe- XVIIIe siècle, Armand Colin, Paris. Braudel, Fernand, 1949, La Méditerranée et le monde méditerranéen à l’époque de Philippe II, Armand Colin, Paris.

Bass, George F., 1972. A History of Seafaring Based on Underwater Archaeology. Thames & Hudson, London. Republished in Book Club Associates, 1974, London.

Broc, Damien, 2014, Dynamiques politiques, économiques et sociales in la Corse mediévale: le Diocèse de Nebbio (XI siècle - c.1540). PHD thesis. Online on HAL: https://tel.archives-ouvertes.fr/tel-01258829/document

Basso, Enrico, 1994. Genova:un impero sul mare. CNR, Istituto sui rapporti italo-iberici, Cagliari. Bellabarba, Sergio, 1988, ‘The square-rigged ship of the ‘Fabrica di galere’ manuscript’ in The Mariner’s Mirror, t. 82, p.113 à 130.

Cabanes, Pierre, Chaline, Olivier et al., 2001, Histoire de l’Adriatique, Ed. du Seuil, Paris Calegari, Manlio, 1970, ‘Navi e barche a Genova tra il XV e il XVI secolo’ in Miscellanea storica Ligure, I, Guerra e commercio nell’evoluzione della marina genovese tra XV e XVII secolo, Ed. Universita di Genova, p.21 à 55.

Bellabarba, Sergio, 1996, ‘The Origins of the Ancient Methods of Designing Hulls: a Hypothesis ‘, in The Mariner’s Mirror, t. 74, p.259-268. Beltrame, Carlo, 2019, ‘Three Venetian ships’, in the Armada, Mediterranean Historical Review, 34:2, 195206.

Campodonico, Pierangelo, 1997, Andrea Doria, Tormena, Genoa.

162

Sources and Bibliography Candiani G., Lo Basso L., 2010 Mutazione e permanenze nella storia navale del Mediterraneo secc. XVI-XIX, Milano, Franco Angeli, 2010.

Castro, Filipe, Fonseca, N., Vacas, T. and Ciciliot, F. 2008, ‘A quantitative look at mediterranean lateenand Square-rigged ships (part 1), The International Journal of Nautical Archaeology, 37 (2), Oxford, 347-359.

Carbone, Giunio, 1867, Compendio della storia ligure : dall’origine fino a 1814, Ed. Carbone, Genoa.

Castro, Filipe, 2007, ‘Rising and Narrowing: 16th-century geometric algorithms used to design the bottom of ships in Portugal’, in International Journal of Nautical Archaeology –IJNA, Oxford, 36.1, 148-154.

Casaban, José-Luis, 2017, ‘Santiago de Galicia and the Illyrian squadron: Characteristics, dimensions and tonnages of Mediterranean-built galleons for Philip’s II Atlantic fleets (1593–1597)’. The International Journal of Maritime History, 29(2), 238- 260

Castro, Filipe, 2006, ‘The Arade 1 shipwreck. A small ship at the mouth of the Arade River, Portugal,’ In Blue, L., Hocker, F., and Englert, A., Eds. Connected by the Sea. Proceedings of the 10th Symposium on Boat and Ship Archaeology, Roskilde 2003 (ISBSA 10), Oxbow Books, Oxford, p.300-305.

Casado Soto, J. L. 2006, ‘Barcos para la guerra: soporte de la monarquía hispánica’, Cuadernos de Historia Moderna. Anejos 5, Tema monográfico: Armar y marear en los siglos modernos XV-XVIII, 15-53.

Castro, Filipe, 2005, The Pepper Wreck: A Portuguese Indiaman at the Mouth of the Tagus River, Texas A&M University Press.

Casado Soto, J. L. 2006, Entre el Mediterráneo y el Atlántico: los barcos de los Austrias, Guerra y sociedad en la monarquía hispánica: política, estrategia y cultura en la Europa moderna, Vol. 1, Madrid, 861890.

Cazenave de la Roche, 2019, ‘La construction navale en Méditerranée au XVIe siècle. L’apport de l’épave de La Mortella III (Saint-Florent, Haute-Corse)’, in Revue d’Histoire Maritime, nº26, June 2019, Presse Sorbonne Université, Paris.

Casado Soto, José Luis, 1998, ‘ Aproximación a la tipología naval cantábrica en la primera mitad del siglo XVI ‘. In Usas memoria: Revista de estudios marítimos del País Vasco, Donostia-San Sebastián, n° 2, 169-191.

Cazenave de la Roche, Arnaud, 2017, ‘Les épaves de la Mortella…’ in Catalogue de l’exposition Secrets d’épaves, 50 ans d’archéologie sous-marine en Corse. Ed. du Musée de Bastia, juillet 2017, p.54-59.

Casado Soto, J. L., 1998, ‘Flota atlántica y tecnología naval hispana en tiempos de Felipe II’, in Congreso Internacional Las sociedades ibéricas y el mar a finales del siglo XVI. Vol. 2., Comisaría General de España, Expo de Lisboa ‘98, Lisbon.

Cazenave de la Roche, Arnaud, 2016, ‘L’épave de la Mortella III (Saint-Florent, Haute-Corse): le témoignage d’un grand vaisseau méditerranéen du XVIème siècle naufragé en Corse ’ in Actes des XVIIèmes Journées Universitaires d’Histoire Maritime de Bonifacio (Corse du Sud).

Casado Soto, J. L., 1989, ‘La construcción naval atlántica española del siglo XVI y la armada de 1588’, in Simposio Hispano -Británica sobre la Gran Armada, London-Madrid 1988, 1989, Madrid, 51-85.

Cazenave de la Roche, Arnaud, 2016, ‘El pecio de la Mortella III (Córcega, Francia): un aporte al conocimiento de la arquitectura naval del siglo XVI en el Mediterráneo.’ In Proceedings of the Fifth Internationaler Kongreß für Unterwasserarchäologie, IKUWA V, (oct.2014), 863-870.

Casado Soto, J. L., 1988, Barcos españoles del siglo XVI y La Gran Armada de 1588. Ed. San Martín, Madrid. Castagna, Donemico, 1971, Genova nella Storia, Ed. Lala, Genoa. Castro, Filipe Vieira de, 2015, ‘Moulds, Graminhos and Ribbands: a pilot study of the construction of Saveiros in Valença and the Baía de Todos os Santos area, Brazil’, in International Journal of Nautical Archaeology, sept. 2015, Oxford, 44(2), 410-422.

Cazenave de la Roche, Arnaud, 2011, ‘The Renaissance shipwrecks of the Saint-Florent bay (Mortella II and III) two sites of a high archaeological potential in Corsica (France)’. International Journal of Nautical Archaeology, IJNA- sept. 2011, 40.1, 69-86.

Castro, Filipe Vieira de, 2013, ‘Tonnages and displacements in the 16th century’, in Journal of Archaeological Science nº40, 1136-1143.

Cazenave de la Roche et al., 2010-2015, ‘Rapports de fouille annuelle programmée du site de la Mortella III ’, CEAN, années : 2010, 2012, 2013, 2014 et 2015. No published Reports kept by the Département des Recherches Archéologiques Subaquatiques et SousMarines, DRASSM, Marseille.

Castro, Filipe Vieira de, 2008, In search of unique Iberian ship design concepts, Historical Archaeology 42 (2), Society for historical archaeology, Bethlehem, 63-87. Castro, Filipe Vieira de, 2008, ‘A group for the study of Iberian seafaring’, Edge of Empire: proceedings the Symposium ‘Edge of Empire’, Society for Historical Archaeology, SHA, 2006, Caleidoscópio, Casal cambra, 7-22.

Cazenave de la Roche, Arnaud, 2009, ‘Les épaves de la Mortella II et II : Observations Préliminaires sur des sites archéologiques de la période de la Renaissance découverts dans la baie de Saint-Florent (HauteCorse) ’ In Les Cahiers d’Archéologie Subaquatique. –CAS-. November/décember 2008, nºXVII, 5 – 53. 163

The Mortella III Wreck Cazenave de la Roche, Arnaud, 2008, Les épaves de la Mortella II & III, une contribution à l’expertise de deux sites, CEAN. No published study kept by the Département des Recherches Archéologiques Subaquatiques et Sous-Marines, DRASSM, Marseille.

Archaeology, Roskilde, 13th-15th March 1989, Roskilde, 41-54. Crumelin-Pedersen, 1991, ‘Ship types and sizes AD 8001400.’ Proceedings of the Nordic Seminar on Maritime Aspects and Archaeology, Roskilde, 13th-15th March 1989, Roskilde, 69-82.

Cerveira de Aguilar, M., 1640, Advertências de Navegantes, Lisbon.

Costantini, Claudio, 1977, La Repubblica di Genova nell’età moderna, UTET, Turin.

Chaline, Olivier, 2010, La Mer vénitienne, Actes Sud, coll. ‘Beaux Arts’, Arles

Daeffler, Michel, 2007, ‘L’épave des Marinières: un témoin des mutations technologiques in les chantiers navals du XVème siècle’ in la Revue d’Histoire Maritime, nº7. Ed des Presses de l’Université de Paris-Sorbonne, 9-56.

Chaunu, Pierre, 1957, ‘La tonelada espagnole aux XVIe et XVIIe siècles.’ In Le navire et l’économie maritime du XVe au XVIIIe siècles: Travaux du Colloque d’histoire maritime tenu, le 17 mai 1956, à l’Académie de Marine, Ed. Michel Mollat, 71-89. SEVPEN, Paris.

Daeffler M., 2005, ‘Portrait d’une victime : la ‘nave’ de commerce méditerranéenne du XVIe Siècle.’ In Course, corsaires et forbans en Méditerranée, XIVe-XXIe siècle, actes du colloque tenu au Musée de la Marine de Toulon du 15 au 17 septembre 2005. Paris : Riveneuve editions, 23-42. Online : http://hal.archives-ouvertes.fr/ hal-00475070

Chaunu, Huguette et Pierre, 1955-1960, Séville et l’Atlantique (1504-1650). Paris, S. E. V. P. E. N., 12 volumes. École pratique des Hautes-Études. VIe section. Centre de recherches historiques. Collection ‘Ports, routes, trafics’, n° 6.

Daeffler, Michel, 2004, Formes de carène et navires de combat. L’invention du vaisseau de ligne en Angleterre (1560-1642), Centre de Recherche d’Histoire Quantitative, Histoire maritime, n°1, Caen.

Ciciliot, Furio, 1999, ‘Genoese Ship Carpenters in Lisbon.’ Published under the title ‘Genoese Shipbuilders in Portugal and in Asia (Early 16th Century)’, International Reunion for the History of Nautical Science and Hydrography, Fernando Oliveira e o seu tempo: Humanismo e arte de navegar no renascimento europeu (1450-1650), Aveiro, 1519 sept. 1998. In Actas da IX Reunido internacional de historia da nautical e hidrografía, Ed. Inácio Guerreiro and Francisco Contente Domingues, Patrimonia, Cascáis, 153-16.

Damiani, Roberto, 2016, ‘Corsari del Mediterraneo,’ Andrea Doria, ott./nov. 1526 et ago. 1527. Online since sept.2012: http://www.corsaridelmediterraneo.it/index. php?option=com_content&view=article&id=995:dor ia-andrea&catid=34&Itemid=142 Damianidis, K. A., 1998, ‘Methods used to Control the Form of the Vessels in the Greek Traditional Boatyards.’ In E. Rieth (ed.), Technologies / Ideés / Pratiques: Concevoir et Construire les Navires, 217–44

Ciciliot, Furio, 1999, ‘Garbo Timer: Oak Cultivation for Shipbuilding in Genoa in the Middle Ages’, handwritten report, Centro Studi Attività Marinare, Savone.

Delahaye, Marion, 2000 ‘L’apport de l’épave médiévale de Cavalaire (France). In L’étude de l’évolution de la construction navale au XVe siècle.’ 33th Conference on Historical and Underwater Archaeology, Society for Historical Archaeology, 4-9 January 2000, Quebec City.

Ciciliot, Furio, 2005, Le superbe navi. Cantieri e tipologie navali liguri medievali, Società savonese di storia patria, Attie memoroe, nuova serie – vol XLI, Savone. Crespo Solana, Ana et al., 2019, ‘Itinerarios de Globalización: Historia y Arqueología Marítima entre España y América’, Magallánica. Revista de Historia Moderna, vol. 6, nº11 (2019), Online access: https:// fh.mdp.edu.ar/revistas/index.php/magallanica/issue/ view/MAG11/showToc)

Doran, E. B. y Doran, M. F. 1978, ‘A reconstruction of the Padre Island ship.’ Appendix E, The nautical archeology of Padre Island: the Spanish shipwrecks of 1554, Academic Press, New York, 375-384.

Crespo Solana, Ana, 2019, “ForSEAdiscovery: la construcción naval y el comercio de la madera del siglo XVI al XVIII”, in revista PH Instituto Andaluz del Patrimonio Histórico, n-º 96 febrero 2019 pp. 114-141

Fernández Duro, Cesáreo., et Monleón, Rafael. 1892, ‘Respuesta dada por los Sres. Fernández Duro y Monleón ’ in La nao Santa María. Memoria de la Comisión Arqueológica ejecutiva. Ministerio de Marina, Madrid.

Crespo Solana, Ana, 2014, Spatio-Temporal Narratives. Historical GIS and the Study of Global Trading Networks (1500-1800), Cambridge Scholar Publishing, Newcastle upon Tyne.

Fernández González, F. 2010, The Spanish regulations for shipbuilding (Ordenanzas) of the Seventeenth Century. In International Journal of Naval History, 8 (3).

Crumelin-Pedersen, 1991, ‘Maritime aspects of the archaeology of Roman and Migration-Period. Aspects of Maritime Scandinavia AD 200-1200.’ Proceedings of the Nordic Seminar on Maritime Aspects and

Fernández González, L. 2002, La construcción naval en el Cantábrico a finales de la Edad Media: técnicas, modelos e innovaciones en el Puerto de Santander, Edades: revista de historia 10, Asociación Universitaria 164

Sources and Bibliography Guérout, Max, 2005, ‘Il rivestimento in piombo della nave Lomellina.’ In Pratiche e linguaggi – Contributi a una storia della cultura tecnica scientifica, Genoa, 9-26

de Jóvenes Historiadores de Cantabria, Santander, 97121. Fernández González, F. 2001, La construcción y la arquitectura naval, Historia de la tecnología en España. Vol. 2, Valatenea, Barcelona, 473-498.

Guérout, Max and Liou, Bernard, 2001, La Grande Maîtresse, nef de François 1er, Recherches et documents d’archives, Presse Paris-Sorbonne, Paris.

Fiorita, A. L., 1950, Andrea Doria. I Grandi Liguri – Collana di biografie storiche, Ceretti, Genoa.

Guérout, M, Gassend, J.M, et Rieth, E., 1989, ‘Le navire génois de Villefranche, un naufrage de 1516 ( ?)’, Archaeonautica n°9, Ed. of the CNRS, Paris

Fourquin, Noël, 1990, ‘Navires marseillais au MoyenAge.’ In Navigations et migrations en Méditerranée de la Préhistoire à nos jours, Ed. CNRS, 1990, 181 à 250.

Guerrazzi, F.D., 1864, Vita di Andrea Doria, Ed. Guigoni

Fourquin, Noël, 2001, ‘Un devis de construction navale de c.1273’ in Villain-Gandossi Christiane, Rieth, Eric, Pour une histoire du fait maritime, Ed. of the CTHS, 2001, Paris, 263-278.

Guiard, Teófilo, 1917, La industria naval vizcaína. Bilbao. Hasslöf, O., 1958, ‘Carvel construction technique. Nature and Origin.’ In Folk-Liv 1957, 58: 49-60. Stockholm.

Friel, Ian, 1994, ‘The Carrack: The Advent of the Full Rigged Ship’, in Gardiner, r., Unger, W., Cogs, Caravels and Galleons. The Sailing Ship 1000-1650. Conway Maritime Press, London, 77-90

Heers, Jacques, 1958, ‘Types de navires et spécialisation des trafics en Mediterranée à la fin du Moyen-âge.’ In Le Navire et l’économie maritime du moyen-âge au XVIIIe siècle principalement en Méditerranée, Colloque International d’Histoire Maritime, 17th and 18th May 1957, Académie de Marine, S.E.V.P.E.N., Paris, 107-118.

Gatti, Luciana, 1999, Navi e cantieri della Repubblica di Genova (secoli XVI - XVIII), Ed. Brigati, Genoa. Gaudin, Jean, 1900, Le baron de La Garde, dit le capitaine Polin (15??-1578). Thesis of the Ecole des Chartes, Paris.

Higueras-Milena Castellano, José Manuel and Gallardo Abárzuza, Mercedes, 2016, ‘Proyecto Delta: pecios localizados y excavados durante las obras de construcción de una nueva terminal de contenedores en el puerto de Cádiz’, In Proceedings of the V Congreso Internacional de Arqueología Subacuática (IKUWA V), 871-883.

Gendron, François et Gendron-Badou, Aïcha, 2009, ‘Rapports d’expertise d’échantillons de l’épave de la Mortella II et Mortella III. Baie de Saint-Florent (HauteCorse).’ In Les Cahiers d’Archéologie Subaquatique – CAS-. November/décember 2009, nºXVII, 55-62.

Hocker, Frederick M. and McManamon, John M., 2006, ‘Mediaeval Shipbuilding in the Mediterranean and Written Culture at Venice’, in Mediterranean Historical Review, 21:1, 1-37.

Gianfrotta, P.A., Pomey, P., 1981, Archeologia subacquea, Milan. Giudice, Giuseppe del, 1871, Diplomu iniditi di Carlo I d’Angió, Naples.

Hormaechea, Cayetano, Isidro Rivera, Manuel Derqui, 2018, Los barcos oceánicos del Atlántico ibérico en los siglo XVI y XVII, 3 vol. Published online: https://independent.academia.edu/ CayetanoHormaecheaArenaza

Glasgow, Tom,1970, ‘Maturing the Naval Administration.’ In The Mariner’s Mirror 56:3-26, 10-24. González Cruz, david, et al., 2018, Barcos y construcción naval entre Atlántico y el Mediterráneo en la epoca de los descubrimientos (siglo XV y XVI), Colección Biblioteca de Historia, Ed. Del CSIC, Madrid.

Hormaechea, Cayetano, Isidro Rivera, Manuel Derqui, 2012,  Los galeones del siglo XVII, 2 Vol., Ed. Associació d’Amics del Museu Maritim de Barcelona, Barcelona.

Gosse, Philip, 2008, Storia della pirateria, Ed. Odoya, Bologna.

Jézégou, Marie-Pierre, 1989, ‘L’épave II de l’anse SaintGervais à Fossur-Mer (Bouches-du-Rhône): Un navire du haut moyen-âge construit sur squelette.’ In Tropis I: 1st International Symposium on Ship Construction in Antiquity, Piraeus, 30 August-1 September 1985, Proceedings. Ed. Harry E. Tzalas, 139-146. Hellenic Institute for the Preservation of Nautical Tradition, Athens.

Graziani, Antoine, 2008, Andréa Doria, prince de la Renaissance, Tallandier, Paris. Guérout, Max, 2017, ‘L’épave de la Lomellina (1516), l’emplanture du grand-mât.’ Article published online on 11th February 2017, HAL archives ouvertes : https:// hal.archives-ouvertes.fr/hal-01465273 Guérout, Max, 2014, ‘L’épave de la Lomellina (1516), Le système d’épuisement des eaux de Cale.’ In Archeologia Postmedievale, 2014, Archeologia dei relitti postmedievali, 18, 49-61. Article published online on 21th February 2017, HAL archives ouvertes: https://hal.archives-ouvertes.fr/hal-01472921

Johnston, Stephen, 1994, ‘Making mathematical practice: gentlemen, practitioners and artisans in Elizabethan England.’ Ph.D. Cambridge, 1994. Published online : http://www.mhs.ox.ac.uk/staff/saj/thesis/

165

The Mortella III Wreck Joncheray, Jean-Pierre, 1988, ‘Un navire de commerce de la fin du XVIIème siècle, première partie: le navire et son mode de chargement.’ In Les Cahiers d’Archéologie Subaquatique, nº VII, 21-67.

Loewen, Brad, 1998, ‘The Red Bay vessel. An exemple of a 16th-century Biscayan ship’ in Revista de Estudios Maritimos del Pais Vasco, 2, Untzi Museoa-Museo Naval, Donostia-San Sebastian, 1998, 193-199.

Keith, Donald H. et Joe J. Simmons III, 1985, ‘Analysis of Hull Remains, Ballast, and Artifact Distribution of a 16th-Century Shipwreck, Molasses Reef, British West Indies.’ Journal of Field Archaeology, Boston, vol. 12, n° 4, 411-424.

Long, P.O., Mac Gee, D., Stahl, A. (ed), 2009, The Book of Michael of Rhodes, a fifteenth-century Maritime Manuscript, 3 vol., MA: MIT, Press Cambridge. Lopez, Roberto S., 1975, Su e giù per la storia di Genova, Università di Genova, Istituto di paleografia e storia medievale, Genoa.

Keith, D. H., J. A. Duff, S. R. James, T. J. Oertling and J. J. Simmons, 1984, ‘The Molasses Reef Wreck, Turks and Caicos Islands, B.W.I.: A Preliminary Report.’ In The International Journal of Nautical Archaeology and Underwater Exploration, London, vol. 13, n° 1, 45-63.

Loureiro, Vanessa, 2009, ‘ L’exemple de l’épave du xvie siècle Arade 1 (Algarve, Portugal) et des traditions architecturales Ibériques : une première approche ’, in Archeo Doct. Objets et symboles. De la culture matérielle à l’espace culturel. Proceedings of the Première journée doctorale d’archéologie, Paris 20 mai 2006. Publication de la Sorbonne, 115-130.

Labbe, M., 2010, ‘A preliminary reconstruction of the Yassiada sixteenth century ottoman wreck’, Master thesis of Texas A&M University. Not published. Online: http://nautarch.tamu.edu/Theses/pdf-files/ Labbe-MA2010.pdf

Maarleveld, Thyjs, 1992, ‘Archaeology and Early Modern Merchant Ships. Building Sequence and Consequences.’ In A. Carmiggelt (ed.), Rotterdam Papers VII, 155–73.

Lane, Frederic C., 1965, Navires et constructeurs à Venise pendant la Renaissance, Paris.

Marsano, Giacomo, 1985, Storia e tradizioni di Genova nelle vicende storiche italiane dalle origini al 1945, Ed. Valenti, Genoa.

Lane, Frederic C., 1934, ‘Naval Architecture about 1550.’ In The Mariner’s Mirror 20, 24-49.

Marsden, Peter et al., 2009, Your Noblest Shippe: Anatomy of a Tudor Warship, vol. 2, Mary Rose Trust, The Archaeology of the Mary Rose, Portsmouth.

L’Hour, Michel, Veyrat, Elisabeth, 1994, ‘The French Medieval Clinker Wreck from Aber Wrac’h’, in C. Westerdahl, Crossroads of Ancient Shipbuilding, Oxford, 165-180.

Martin, Colin J. M. 1979, ‘La Trinidad Valencera: An Armada Invasion Transport Lost Off Donegal, Interim Site Report, 1971-1976.’ The International Journal of Nautical Archaeology and Underwater Exploration, London, vol. 8, n° 1, 13-38.

Lingua, Paolo, 2006, Andrea Doria, principe e pirata nell’Italia del ‘500. Ed. Fratelli Frill. Lingua, Paolo, 2004, Breve storia dei genovesi, Ed. Laterza. Lo Basso, Luca, 2012, ‘Entre galères et vaisseaux. Armement et constructions navales en Ligurie au XVIIème siècle’, in Cahiers de la Mediterranée, nº84, 273.

Martinez Ruiz, E. 2008, ‘Un proyecto naval en perspectiva de la crisis de 1635’, in Revista de Historia Naval, n°100.

Lo Basso, L., 2004, A vela e a remi. Navigazione, guerra e schiavitù (secc. XVI-XVIII), Philobiblon Edizioni, Ventimiglia.

Mollat du Jourdin, Michel et Chillaud-Toutée, Florence, 1984, ‘Le livre des Faiz de la marine et navigaiges d’Antoine de Conflans, v. 1516-1520.’ 107th-108th Congrès nationaux des sociétés savantes, Brest 1982 and Grenoble 1983. Hist, mediévale et philologie – Hist. mod. et contemp. Etudes d’histoire maritime, C.T.H.S., Paris, pp. 9-44.

Loewen, Brad, 2007, ‘La carène du navire 24M’, in Bernier, M.-A-, Grenier, R. et al., L’archéologie subaquatique de Red Bay, la construction navale et la pêche de la baleine basques au XVIème siècle, 5 vol., Parcs Canada, Ed. Canadien Museum of Nature, Volume III, Ottawa.

Nayling, N. and Jones, T., 2012, ‘Three-Dimensional Recording and Hull Form Modelling of the Newport (Wales) Medieval Ship’, in N. Gunsenin (ed.), Between Continents. Proceedings of the Twelfth Symposium on Boat and Ship Archaeology Istanbul 2009, 319–24, Istanbul.

Loewen, Brad, 2001, ‘The structures of Atlantic shipbuilding in the XVIth century. An archaeological perspective’, in Proceedings International symposium on archaeology of medieval and modern ships of Iberian-Atlantic tradition, Centro Nacional de Arqueologia Náutica e Subaquática, Academia de Marinha Lisboa,7th to 9th Sept. 1998, Lisboa, 241-258.

Nayling, N. and Jones, T., 2013, ‘The Newport Medieval Ship, Wales, United Kingdom’, in International Journal of Nautical Archaeology 43(2)

Lo Basso, Luca, 2011, Capitani, corsari e armatori. I mestieri e le culture del mare dalla tratta degli schiavi a Garibaldi, Novi Ligure.

Nieto Prieto, J. and Raurich, X, 1998, Excavacions arqueològiques subaquàtiques a Cala Culip VI, II. Girona.

166

Sources and Bibliography Redknap, Mark, 1984, The Cattewater Wreck: The Investigation of an Armed Vessel of the Early Sixteenth Century. BAR British Series 131, National Maritime Museum Archaeological Series No. 8, Oxford.

Oertling, Thomas James, 1989, ‘The Few Remaining Clues’ in Underwater Archaeology Proceedings from the Society for Historical Archaeology Conference, 100-103. J. Barto Arnold III, Ed. Pleasant Hill, Ca, Society for Historical Archaeology.

Rieth, Eric, 2016, Navires et construction navale au Moyen Age. Archéologie nautique de la Baltique à la Mediterranée, Editions Picard, Paris.

Oertling, Thomas James, 1989, ‘The Molasses Reef Wreck Hull Analysis: Final Report.’ The International journal of Nautical Archaeology and Underwater Exploration, London, vol. 18, n° 3, p. 229-243.

Rieth, Eric, 2009, ‘L’architecture navale mediévale en mediterranée: quelques axes de réflexion’, Arqueología Mediterránea Náutica, monographies du CASC 8, Centre d’Arqueologia Subaquàtica de Catalunya, Barcelona.

Oertling, Thomas James, 1984, The history and development of ships bilges pumps, 1500 – 1840. These de Master, Texas University, published in 1996 par College Station : Texas A & M University Press, Texas.

Rieth, Eric, 2000, ‘ La méthode moderne de conception des carènes du whole-moulding, une mémoire des chantiers navals mediterranéens du Moyen-Age.’ Neptunia nº220, 199-202.

Oertling, T.J., 2001, ‘The Concept of the Atlantic Vessel.’ In Proceedings of the International Symposium on Archaeology of Medieval and Modern Ships of IberianAtlantic Tradition. F. J. S. Alves, Symposium de l’année 1998. Ed. Lisboa, Instituto Português de Arqueologia, 233-240.

Rieth, Eric, 1998, ‘L’épave du caboteur de Culip VI (Catalogne, Espagne), un témoignage du début du XIVème siècle d’une conception reposant sur la méthode du maître-gabarit et de la tablette’, in Archaeonautica 14, 205-212.

Padovano, Aldo et Volpe, Felice, 2008, La grande storia di Genova, Ed. Artlibri, Genoa.

Rieth, Eric, 1998, ‘Construction navale à Franc-Bord en Mediterranée et Atlantique (XVIe-XVIIe) et ‘Signatures Architecturales’: une première approche archéologique, in Mediterranée Antique. Pêche, navigation, commerce / Sous la direction d’Éric Rieth. Editions du CTHS, p. 177-187.

Palou H., Rieth, E., Izaguirre, M., Jover, A., Nieto, X., Pujol, M., Raurich, X., Apestegui, C., 1998, Excavacions Arqueològiques Subaquàtiques a Cala Culip. Culip VI. Museu d’Arqueologia de Catalunya / Generalitat de Catalunya, Barcelona. Paviot, J., 1994, ‘Aspects de la navigation et de la vie maritime génoises au XVe siècle, d’après les comptes des baillis de L’Écluse’, in La Storia dei Genovesi, Atti del Convegno …, 11-14th June1991, vol. XII, parte I, Genoa, 247-261.

Rieth, Eric, 1996, ‘Essai d’évaluation des savoirs des constructeurs de navires au XVIème siècle’, in Acera, Martine et Martinière, Guy, Coligny, les protestants et la mer, Presse de l’Université de Paris-Sorbonne. Rieth, Eric, 1996, Le maître-gabarit, la tablette et le trébuchet, essai sur la conception non-graphique des carènes du Moyen-Age au XXe siècle, Ed. of the CTHS, Paris.

Piane, Riccardo Delle and Giacomone Piana, Paolo, 2003, Militarium: fonti archivistiche e bibliografia per la storia militare della Repubblica di genova (15281797), della Repubblica Ligure (1797-1805) e della Liguria napoleonica (1805-1814), Ed. Daner Elio Ferraris, Savona.

Rodriguez Mendoza, M. B., 2008, Standardization of spanish shipbuilding: ordenanzas para la fábrica de navíos de guerra y mercante – 1607, 1613, 1618, Texas A&M University.

Prescott, WH. 1843, La mort de l’empire aztèque, Translation and adaptation of P. Guillot, 1965. Ed. de Saint-Clair, Paris.

Rodrigues, P, Alves F., Rieth, E., Castro, L. F., 1998, ‘L’épave d’un navire de la deuxième moitié du XVème siècle / début du XVIème, trouvée au Cais do Sodré (Lisbonne), Trabalhos de Arqueologia n°18, Lisbon.

Pujol, Marcel, 1990, ‘Culip VI: L’épave d’un navire long à rames du XIVe siècle.’ Neptunio vol. 179, n° 3, Paris, 37-40.

Roncière, Charles de la, 1906, Histoire de la Marine française, Ed. Plon-Nourrit, Paris.

Pujol i Hamelink, M., Rieth, E., 1994, ‘Arqueología naval medieval: Cala Culip VI y Les Sorres X’. Actas de las Aulas del Mar de Cartagena (20-24th September 1993), vol. I, Murcia, 179-190.

Rosloff, J., Barto Arnold III, J., 1984, ‘The keel of the San Esteban (1554): continued análisis’, in Internacional Journal of Archaeology, 13, 4, p.286-296.

Pulak, C., 2005, ‘A Rare Ottoman Wreck: Yassıada, Turkey’ in Beneath the Seven Seas,George Bass Ed., Thames & Hudson, London, 138-141.

Rubio Serrano, José Luis, 1991, Arquitectura de las naos y galeones de las flotas de Indias. 2 vol., Ed. Seyer, Malaga.

Rahn Philipps, Carla, 1986, Six Galeons for the King of Spain, Imperial Defense in the Early Seventheen Century, Johns Hopkins University Press.

Rule, Margaret, 1982, The Mary Rose: The Excavation and Raising of Henry VIU’s Flagship. Conway Maritime Press, London. 167

The Mortella III Wreck Remains of a 16th-Century Shipwreck in Bermuda.’ The International Journal of Nautical Archaeology, vol. 22, n° 2, 103-124. London

Sarsfield, John Patrick, 1988, ‘Survival of pre-sixteenth century Mediterranean lofting techniques in Bahia, Brasil’, in Filgueiras (ed.), BAR Publishing International Series 438, vol. I, 63-86. Savérien, Alexandre, 1758, Dictionnaire historique, théorique et pratique de la Marine, Paris.

Westerdhal, C., 1992, ‘The maritime cultural landscape.’ In International Journal of Nautical Archaeology, IJNA 21.1, 5

Smith, R. C, J. R. Bratten, J. Cozzi et K. Plaskett, 1998, The Emanuel Point Ship: Archaeological Investigations, 1997-1998. Report of Investigations nº 68. Archaeology Institute, University of West Florida, Pensacola.

Zwick, Daniel, 2016 ‘Bayonese cogs, Genoese carracks, English dromons and Iberian carvels: Tracing technology transfer in medieval Atlantic shipbuilding’, Itsas Memoria.

Smith, R. C, J. Spirek, J. Bratten and D. Scott-Ireton, 1999, The Emanuel Point Ship: Archaeological Investigations, 1992-1995, Preliminary Report. Bureau of Archaeological Research, Division of Historical Resources, Florida Department of State, Tallahassee. Smith, R. C, D. H. Keith et D. Lakey, 1985, ‘The Highborn Cay Wreck: Further Exploration of a 16th-Century Bahaman Shipwreck.’ The International Journal of Nautical Archaeology and Underwater Exploration, vol. 14, n° 1, p. 63-72. London. Spissu, Anna, 2008, Il pirata e il condottiero, romanzo storico, Corbaccio, Milan. Steffy, J. Richard, 1994, Wooden Shipbuilding and the Interpretation of Shipwrecks, College Station: Texas A&M University Press. Steffy, J. R., 1989, ‘Ancient scantlings: the projection and control of Mediterranean hull Shapes.’ Tropis. 111, 417-428. Stevens, Willis et Waddell, Peter J.A., 2007, ‘Les vestiges in situ du site 29M’ in Bernier, M.-A-, Grenier, R. et al., 2007, L’archéologie subaquatique de Red Bay, la construction navale et la pêche de la baleine basques au XVIème siècle, Volume I, Parcs Canada, ed. Canadien Museum of Nature, Ottawa, V.1-223. Vargas-Hidalgo, Rafael, 2002, Guerra y Diplomacia en el Mediterráneo: Correspondencia inedita de Felipe II con Andrea Doria y Juan Andrea Doria, Madrid. Vergé-Franceschi, 2001 ‘Andrea Doria, Amiral génois et ennemi des Corses’, in Les Doria, la Mediterranée et la Corse, Piazzolla, Ajaccio. Villié, Pierre, 1989, ‘L’épave Calvi 1’, in Cahiers d’Archéologie Subaquatique, CAS nº8. Villié, Pierre, 1990, ‘L’épave Calvi 1’, in Cahiers d’Archéologie Subaquatique, CAS nº9. Villié, Pierre, 1991, ‘L’épave Calvi 1’, in Cahiers d’Archéologie Subaquatique, CAS nº10. Waddell, Peter J. A., 1985, ‘The Pump and Pump Well of a 16th Century Galleon.’ The International Journal of Nautical Archaeology and Exploration, vol. 14, n° 3, 243-259. London. Watts, Gordon P., Jr., 1993, ‘The Western Ledge Reef Wreck: A Preliminary Report on Investigation of the 168

Glossary Definitions of technical words and expressions used in the text.

planks. It is also generally associated with covering the hull with tar or pitch.

Adze (Fr. Herminette n. f.): kind of axe with the difference that the edge is horizontal. The etymology of the French word comes from the ermine whose nose is curved, like the tool. It is traditionally an essential tool in marine carpentry.

Ceiling (Fr. Vaigrage, n. m.): Inner strakes covering longitudinally the bottom of the hold between the KEELSON and the SILL.

Beam (Fr. 1) Bau or Barrot n. m. 2) Fort, n. m.): 1) A transverse timber which, at the same time as it provides structural reinforcement for the hull in its upper part, contributes to the structure of the ship’s decks and supports their floor. 2) Measurement of breadth to the outside of the frames at the ship’s widest point at the level of the main deck. See BREADTH

Clamp (Fr. Serre n. f.): Thick ceiling notched on the frames used to reinforce the longitudinal structure of the hull in general. It also has a specific function depending on its position in the hull:

Ceiling Plank (Fr. Vaigre n. m.): See CEILING.

The Foot Wale (Fr. Serre D’empature n. f.) is the first fixed ceiling outboard of the keelson. It is located on the scarf of the floor-timber to the first futtocks to reinforce their junction. – The BILGE CLAMP (Fr. SERRE DE BOUCHAIN, n. f.) is the second clamp located on the scarf to the floor-timber to the first futtocks, just inboard of the sill. – These thick ceilings located in the turn of the bilge were also called thick stuff. The SHELF CLAMP (SERRE BAUQUIÈRE, n. f.) is a thick interior strake on which the deck beam and ledge ends rest.

Bilge or Turn of the Bilge (Fr. Bouchain, n. m. or Escoue, n. f.): Line formed by the overlapping of the floor timbers heads with the ends of the first futtocks (area called bilge points = ‘points d’escoue’). This line marks the end of the flat inside the hull and is generally characterized by an inflection of the line of the hull. Bilge Clamp (Fr. Serre de Bouchain, n. f.): See CLAMP.

Clinker Building (Fr. Construction à Clin): Construction method of the hull based on an overlapping of the strakes of the planking. It is associated with a ‘shellfirst’ construction method. It predominated in the Atlantic nautical space until the second half of the 15th century, after which the Mediterranean ‘carvel building’ gradually prevailed. See CARVEL BUILDING.

Breadth, Main or Maximum Breadth (Fr. Fort, n. m.): The measure of breadth to the outside of the frames at the ship’s widest point at the level of the main deck. It is one of the important measures of a ship linked to the others with a relation of proportion. Buttress (Fr. Taquet): Timber of the mainmast step arranged between the keelson (or the longitudinal timbers of the mast step arrangement) and the first clamp to contain the lateral forces of the mast foot. See MAST STEP.

Crutch (Fr. Fourcat, n. m.): Crutches are floor-timbers located beyond the tail-frames. They take a Y shaped morphology approaching the stern and a V shape towards the bow.

Camber, Rounding or Convexity (Fr. Bouge, n .f.): Applied to the ship, it corresponds to the transverse curvature given to the beams in order to obtain a convexity of the decks, useful for water flow.

Dead Flat: See FLAT Dead Work (Fr. Oeuvres Mortes): Part of the framing located above the floating line.

Carvel Building (Fr. Construction A Franc-Bord): Hull building technique of Mediterranean origin which is based on a layout of the planks edge to edge without connection or overlap between them. Waterproofing is achieved by the natural swelling of the wood in contact with water. The other hull building technique is the CLINKER BUILDING.

Depth of Hold (Fr. Creux, n. m.): The ship’s depth measured, for the purposes of tonnage calculations or GAUGE, in the pump well from the top of the ceiling to the top of the beam at the main deck. Diagonal Scarf (Écart En Sifflet): See SCARF.

Caulking (Fr. Calfatage n. m.): Action of sealing a hull carvel built. It is carried out with fibrous organic material, generally oakum, placed in force between the edges of the

Dovetail Scarf (Fr. Écart en Forme de Queue D’aronde): See SCARF.

169

The Mortella III Wreck False Keel (Fr. Fausse Quille): Timbers located to the bottom of the keel in order to protect it. When thick, its main function was to increase the size and strength of the keel.

Hauling Down/Up The Futtock (Fr. Recalement, n.m.): See MOULDING PROCESS. Hawse Hole (Fr. Écubier, n. m.): Orifices made at the front of the ship on either side of the bow to allow the passage of hawsers for mooring.

Flat Vertical Scarf (Fr. Ecart en Trait de Jupiter): See SCARF.

Heel of the Keel (Fr. Talon de Quille): See KEEL

Flat or Dead Flat (Plat, n. m.): The flat part of the floortimbers of the central area of the hull, located between the bilge points of the midship frames. This architectural concept set out a flat area between the tail-frames. Floor Timber (Fr. Varangue, n. f.): see FRAME

Hinges of the rudder (Fr. Gonds, n. m, Penture, n. f., or Charnière n. f.): Iron work attached to the stern post whose ring-shaped end (GUDGEON or GOOGIN) receives the hook of the PINTLE strap and thus supports the rudder.

Foot Wale (Fr. Première Serre n. f. D’empature): See CLAMP.

Hook Scarf (Fr. Écart à Croc or Écart à Adent or Écart à Cadeau): See SCARF

Foot Valve (Fr. Pied De Pompe / Valve de Pied de Pompe): Small component at the base of the pump mated with the tube that rested upon it.

Hull Planking: See PLANKING Keel (Fr. Quille, n. f.): In the CARVEL construction, the keel is the founding element of the FRAMEWORK. It is either single or made up of a set of timbers arranged in a longitudinal axis. The keel supports the frames that make up the transverse structure. It is extended in its fore part by the STEM and in its aft part, at the level of its HEEL, by the STERN. The stern, keel and stem together form the longitudinal axis of the framework. Along the two sides of the upper part of the keel, runs a groove cut called a RABBET in which fit the first planking strake, also called GARBOARD.

Frame (Fr. Membrure, n. f. or Couple n. m.): A set of longitudinally assembled timbers that constitute the transverse structure of a ship’s hull. They are arranged on the upper face of the keel on which they can be fastened with bolts. On their upper face, a longitudinal timber named KEELSON is notched on them. Each frame is composed of a central piece, the FLOORTIMBER, scarfed in the continuity of its port and starboard branches with pieces called FUTTOCKS. In France, the frames of a vessel of ‘Premier rang’ were currently composed of six futtocks. The last futtock was called top-timber. The largest frame located at mid-ship is called MASTER-FRAME. In the CARVEL construction, the TAIL-FRAMES are located on both side of the master-frame and are delimiting the central area of the ship. The frames located beyond the tailframes are called CRUTCHES.

Keelson (Carlingue n. f. or Contre-Quille, n. f.): Strong timbers set placed in the longitudinal axis of a ship, notched on the upper side of the floor-timbers. Limber Holes or Watercourses (Fr. Anguillers or Canal Des Anguilliers): Holes in the floor timbers allowing the water to flow towards the bilge pumps.

Futtock (Fr. First futtock = Genou, n., m. Second futtock = Première Allonge n. f., etc.): A timber that contributes to the composition of the frames. A frame is made of a floor and several extensions named futtocks attached to each other. For example, a ship of ‘Premier rang’ in France had up to six futtocks including the last one called top-timber. See FRAME.

Longitudinal Timbers of the Mast Step Arrangement or Sister Keelsons (Fr. Carlingots, n.m. or Escasses): Two strong timbers arranged longitudinally on either side of the keelson, notched on the floors, whose purpose is to maintain the mainmast foot. They were joined by two dovetail-shaped keys and held laterally by BUTTRESSES. They were part of the classic Mediterranean mainmast step.

Garboard (Fr. Galbord n. f.): First strake of the planking, whose lower edge is fitted into the RABBET of the keel.

Mast-Step (Fr. Emplanture du mât or Massif D’emplanture du mât): A mast step is a wooden device designed to hold the mast foot. For the mainmast, it consists of an orifice mortised in the keelson or a place between two strong timbers (named ‘ESCASSES’ in the maritime French Mediterranean language) where takes place the mast foot. Several buttresses are arched between the keelson and the first clamp to maintain it laterally.

Gauge (Fr. Jauge, n. f.): See DEPTH OF HOLD. Googin of the rudder (Fr.: Femelot n. m.): See GUDGEON. Gudgeon or Googin of the rudder (Fr. FÉMELOT, n. m.): Hinge attached to the stern post whose ring-shaped end receives the hook of the pintle strap and thus supports the rudder.

Master-Frame (Fr. Maître-Couple): See FRAME 170

Glossary Molded or Molded Dimension (Fr. Largeur Sur Le Droit): The sizes of the sections of timbers are expressed by their molded (width of their upper and bottom faces) and sided (vertical surface or height). Conversely, planks and wales are listed in thicknesses and widths.

Pump Tube (Fr. Corps De Pompe n. m.) : See PUMP Pump Valve (Fr. Valve De Pompe n. f.): See PUMP Pump Well (Fr. Archipompe n. f.): Quadrangular wooden box in which the bilge pump is located and whose function is to protect it.

Mould (Fr. Gabarit n. m.): See MOULDING PROCESS. Mould to the Midship-Bend or Master Mould (Fr. Maître-Gabarit, n. m.): See MOULDING PROCESS.

Rake (Fr. Quête, n. f.): slope of the stern and stem posts. It sets out what are called OVERHANGS, i. e. the lengths that must be added to the keel length to estimate the length of a vessel.

Moulding Process / Method (Fr. Gabariage n. m. or Méthode du Maître-Gabarit): It is a non-graphic method largely used during the modern period to design boat’s hull. Likely to be of Mediterranean origin, the moulding process was first described in the Italian treaties of the 14th and 15th centuries. Three wooden tools were used for this operation: the mould to the midship-bend or master-mould, the rising-square and the “trébuchet” (French word), a wooden scale used to obtain a bulge of the breadth. They allowed four modifications to the frames located on each side of the master-frame between the tail-frames: the NARROWING (Fr. RÉDUCTION DU PLAT, It. fondo), the RISING of the floor-frames (Fr. ACCULEMENT, n. m., It.stella), the “TRÉBUCHEMENTt” (It. ramo) and the HAULING DOWN/UP the futtock (Fr. RECALEMENT, n.m., It. scorrer del sesto). A similar method named WHOLE-MOULDING is documented in the British shipyards from the end of the 18th century onwards.

Rabbet (Fr. Râblure n. f.): A groove cut running, to port and starboard, along the sides’ timbers pieces that form the longitudinal axis of a ship. It is intended to fit the end of the strakes from the planking on the stern and stem posts, and the lower edges of the garboards on the keel. Rising (Fr. Acculement n. m.): The rising is defined by the distance measured between the line formed by the two points of the bilge located at the end of a floor and the upper face of the keel. The rising given to the floors allows to gradually rise the flat from either side of the master frame to the ends of the ship. The “rising” combined with the “narrowing” and the “hauling down the futtock” were the main method used to give its form to the hull during the no-graphic period of shipbuilding. Room and Space (Fr. La Maille et le Plein): The interval from the centre of one floor-timber to the centre of the next measured alongside the keel.

Orlop Deck or Overlop or Overloop or Spare Deck (Fr. Faux-Pont, n. m.): the lowest deck in a line of battleship, being a platform laid over timbers of the hold.

Rudder (Fr. Gouvernail n. m.): The word rudder can first of all refer to the entire steering system of the ship. But in the strict meaning of the term, it describes the wooden blade fixed to the stern by iron works (hinges composed of a pin located on the rudder stock fitting in a gudgeon fixed to the stern) which allows the ship to adjust its direction. The rudder blade is usually composed of two or three pieces of wood assembled vertically: the main, middle and after pieces.

Overhang (Élancement, n. f.). See RAKE. Planking, Hull Planking (Fr. Bordage / Bordés n. m.): A timber or set of timbers that forms the hull of a ship. The strakes of planks are fixed on the outside of the frames. Pintle, rudder pintle or pintle strap (Fr.: Aiguillot, n., m.): Iron-work fixed on the main piece of the rudder at the end of which there is a hinge or “hook” fitting into the googin located on the stern post and allowing the rudder to rotate.

Scarf (Fr. Ecart, n. m.): Carpentry technique aiming to join two timbers. The scarfs can be of different typologies: FLAT VERTICAL SCARF (Fr. ÉCART EN TRAIT DE JUPITER), HOOK SCARF (Fr. ÉCART À CROC or ÉCART À ADENT or ÉCART À CADEAU), BUTT SCARF (Fr. ÉCART PLAT), DOVETAIL SCARF (Fr. ÉCART EN FORME DE QUEUE D’ARONDE), DIAGONAL SCARF (ÉCART EN SIFFLET), etc.

Pump or Bilge Pump (Fr. Pompe n. f.): Device designed to evacuate the water accumulated at the bottom of the hold. It consists of a PUMP VALVE (Fr. VALVE DE POMPE n. f., PIED DE POMPE n. m. ou CHOPINE, n. f.) provided with a leather flap to prevent the return of water. It was mated to a PUMP TUBE (Fr. CORPS DE POMPE n. m.) in which water is sucked upwards by means of a device of different typology, according to the period a SPEAR-LIKE PLUNGER (Fr. PISTON, n. m.) with a SHAFT (Fr. GAULE n. f.) and with a piston made of leather discs in the 16th century, for example. The pump is located in a receptacle called PUMP WELL or SUMP (Fr. ARCHIPOMPE, n. f.) to protect it.

Scarf of a Frame (Fr. Empature n. f. D’une Membrure): Assembly of two timbers part of a frame whose ends are placed side by side. See SCARF. Shelf Clamp (Fr. Serre Bauquière, n. f.): Thick ceiling strake whose function is to support the beams and ledges of a deck.

171

The Mortella III Wreck Shole or Shoe of the rudder (Fr. Semelle, n. f.):

Waist (Embelle, n. f.): Fore part of a boat located between the main mast and the bow

SIDED or SIDED DIMENSION (Fr. LARGEUR SUR LE DROIT): See MOLDED or MOLDED DIMENSION

Well: see PUMP WELL.

Sill (Fr. Accotards or Accotars n. m.): Thick ceiling planks notched on the frames located above the bilge clamp against which they rely on. Between each frame and the upper part of the sills are inserted FILLING PLANKS (Fr. PLANCHETTES DE REMPLISSAGE) designed to prevent the transit of objects and ensure the cleanliness of the central part of the hull.

Whole-Moulding (Fr. Gabariage, MOULDING PROCESS.

n.

m.):

See

Wrunghead (Fr. Tête de Varangue, n. f.): The extremity of a floor-timber.

Sister Keelson: See LONGITUDINAL TIMBERS OF THE MAST STEP ARRANGEMENT. Spear-Like Plunger (Fr. Piston, n. m.): See PUMP. Stanchion (Fr. Épontille, n. f.): Vertical timber or pillar used to support deck beams. Stem Post (Fr. Étrave, n. f.): A curving timber or a set of joined of timbers, attached to the keel at its fore end. Its slope is called RAKE and the lengths that must be added to the keel length to estimate the length of a vessel is called OVERHANG. Stern Post (Fr. Étambot, n. m.): A vertical timber or a set of joined of timbers, attached to the keel at its aft end, called heel of the keel. Its slope is called RAKE and the lengths that must be added to the keel length to estimate the length of a vessel is called OVERHANG. Strake or streak (Fr. Virure, n. f.): Line formed by a series of hull planks butt joined together, starting from the stern and ending at the stem. Stringer (Lisse, n. f.): A single inner strake running through the midship area between the lower and main decks. Surmark or Sirmark – 1: (Marque, n. f.): A mark made on the molds of a ship when building to show where the frames should be beveled, it is to say where the angles of the timbers are to be placed. – 2 (TAQUET, n. m.): a cleat temporarily placed on the side of a ship on the ways or in a ship dock to support the ribband against which the shores rest. Tailframes (Fr. Couples de Balancement): These are the two frames delimiting the frames of the central area of the ship, located on both side of the master-frame. Between the tailframes are located the frames of the moulding process. Taper, mast foot’s taper (Fr. Mèche n.f., Michon n.m.): The mast foot’s taper is a progressive reduction of its diameter in order to accommodate it in the room provided for this purpose in the mast step Tree Nail (Fr. Gounable, n. f.): big wooden dowel 172

Figure A1.1. General layout of the site and excavation areas. For a better understanding of the drawings and photomosaic, it is possible to access to a digital version online at barpublishing.com/additional-downloads.html.

Annex 1

General planimetry of the Mortella III wreck

173

Figure A1.2. General planimetry of the site. For a better understanding of the drawings and photomosaic, it is possible to access to a digital version online at barpublishing.com/ additional-downloads.html.

Annex 2 Photomosaic of the Mortella III wreck Christoph Gerigk Dipl. Photo-Designer FH [email protected]

175

Annex 3 Dendrochronology Study Fabien Langenegger Archaeologist and dendrochronologist, Cultural Affairs Department, Office and Museum of Archaeology, Republic of Neuchâtel, Switzerland. [email protected] Les échantillons

Tous les échantillons montrent un façonnage intensif et un seul élément conserve encore quelques cernes d’aubier, l’allonge n°20. Un bois plus jeune, donc un diamètre du tronc plus réduit, a été utilisé pour cette partie du bateau et un maximum de matière première a été gardé pour obtenir une section suffisamment importante. Ces quelques cernes de “bois vivant” sont indispensables pour estimer la date du début de la construction de la coque. La synchronisation de ces huit moyennes a permis de construire une moyenne longue de 206 ans (fig.3). A noter que pour le genou n°20, plusieurs échantillons ont été sciés, mais l’échantillon n° 28 ne provient pas du même arbre que l’échantillon n°22. De ce fait, soit ce genou a été constitué de deux parties différentes ou soit une erreur dans la numérotation des pièces à été faite et ces deux échantillons proviennent de deux genoux différents.

Lors de la campagne de 2010 de la Mortella 3 dans la baie de Saint-Florent (Corse), l’équipe d’Arnaud Delaroche a prélevé 12 échantillons de bois dans le but d’obtenir des datations dendrochronologiques précises (fig. 1, pl. 1-2). Ces pièces de bois proviennent toutes de l’abattage de chênes de type sessile, comme semble démontrer la structure des cellules et la vitesse de croissance particulièrement lente de ces arbres (en moyenne 1 et 2 mm par année). L’estimation du diamètre des chênes exploités varie entre 22 et 44 cm. Le trend des courbes de croissance est très différent d’un échantillon à l’autre et montre une provenance très hétérogène du bois de construction. La datation relative Lors de l’expertise dendrochronologique, un seul échantillon (n°24D) trop détérioré n’a pas pu être mesuré. Trente-cinq mesures ont été effectuées, totalisant 3074 cernes, ont permis de constituer une courbe moyenne la plus représentative possible par élément d’architecture (pl. 3-10). Ensuite, ces huit moyennes ont été synchronisées entre elles afin d’obtenir leur positionnement les unes par rapport aux autres. Une datation relative est ainsi proposée entre ces différents bois.

Malgré l’hétérogénéité des croissances des chênes, la courbe moyenne reste très représentative des différents échantillons (fig.4). La datation absolue Les différentes moyennes obtenues pour la Mortella 3 ont été testées sur les nombreux référentiels disponibles au laboratoire de dendrochronologie de Neuchâtel pour le Nord des Alpes. Les résultats montrent des bons indices avec la zone centrale de la France et notamment la Bourgogne. Le carlingot, le genou n°20 (éch.28b), la serre d’empature n°4, ainsi que la moyenne générale donnent des résultats concordants sur plusieurs courbes de référence. Les coefficients tournent autour de 4 pour le test d’Eckstein et de 5 à 7,3 pour le test de Student T (BP). La synchronisation visuelle des courbes à l’écran est très bonne pour l’échantillon prélevé dans le carlingot et permet de proposer une bonne datation absolue. La seule incertitude qui pèse sur ces résultats est une provenance des bois essentiellement du Sud des Alpes et qui permettrait peut-être d’obtenir des résultats différents. L’ensemble des courbes seront envoyées prochainement en Italie pour éliminer ce dernier doute. La demande en bois dur était très importante au 16e siècle, notamment pour la construction navale et le chêne manquait dans certaines régions d’Europe amenant à un essor du commerce du bois organisé sur de très longue distance. Le stock de bois, constitué pour la construction

Figure A3.1. Liste des échantillons prélevés sur l’épave de la Mortella 3

177

The Mortella III Wreck

Figure A3.2. Courbes moyennes en position relative.

Figure A3.3. Courbe moyenne du l’épave de la Mortella 3.

de la Mortella 3, a peut-être nécessité l’importation de chêne de régions très éloignées les unes des autres. Une étude dendrochronologique plus poussée avec l’apport de nouveaux échantillons permettrait peut-être de définir précisément la provenance du bois.

La date de construction peut être définie grâce à l’allonge n°20 et ces 4 cernes d’aubier. La largeur de croissance des derniers cernes de cet échantillon est particulièrement élevée et atteint 4 à 5 mm. Ainsi, la largeur conservée de l’aubier est déjà de 2,14 cm. Une largeur normale peut atteindre 3 à 3,5 cm et ainsi 3 cernes d’aubier manqueraient jusqu’au cambium. Une estimation de 3 à 6 cernes manquants a été notée pour ce prélèvement. Ainsi le début de la construction a pu commencer au plus tôt entre 1517 et 1520 (fig. 6).

Pour l’instant, les corrélations obtenues permettent de donner une date précise pour le début de la construction de la Mortella 3. La moyenne générale est longue de 206 ans et se situe chronologiquement entre 1309 et 1514 (fig.5). 178

Dendrochronology Study courbe moyenne de la Mortella 3 Coeffi cient de concordance

test d’Eckstein

test de Student T (BP)

longueur calculée (an)

carlingot

77,30%

5,96

9,05

161

bordé A

80,85%

5,98

9,98

132

bordé B

82,35%

4,62

5,15

66

genou n°20 (éch.22)

77,08%

5,31

7

125

genou n°20 (éch.28)

72,92%

4,49

6,53

139

serre d’empature n°4

88,63%

7,25

9,63

122

allonge n°20

96,08%

6,58

16,96

63

pièce sur le bordé

79,71%

4,94

5,17

99

Figure A3.4. Coefficients de corrélation des échantillons avec la courbe moyenne de la Mortella 3

Figure A3.5. blocs-diagramme des différents éléments d’architecture en datation absolue.

origine (ap. J.-C.)

terme (ap. J.-C.)

estimation de la date d’abattage

nb de cernes mesurés

carlingot

1309

1469

> 1489

161

bordé A

1330

1461

> 1481

132

bordé B

1324

1389

> 1409

66

genou n°20 (éch.22)

1313

1437

> 1457

125

genou n°20 (éch.28)

1352

1490

> 1510

139

serre d’empature n°4

1370

1491

> 1511

122

allonge n°20

1452

1514

1517 à 1520 ap. J.-C.

63

pièce sur le bordé

1336

1434

> 1454

99

Figure A3.6. Datation absolue des échantillons prélevés sur la Mortella 3

Planches

Ainsi grâce au positionnement de chaque élément d’architecture en datation relative, une date absolue peut être proposée pour tous les échantillons prélevés sur l’épave de la Mortella 3 (fig. 6). Lors de l’estimation de la date d’abattage, si un bois n’a pas d’aubier, 20 cernes sont ajoutés systématiquement qui donne un terminus post quem. Plus ce dernier est éloigné de la date probable de construction du bateau, plus le façonnage sur cet élément a été intensif. A moins qu’il ne s’agisse d’un bois de récupération, ce qui est peu envisageable pour une telle construction.

Photos des échantillons : E : ¼ Dessins des échantioons, E : ¼ Montages des courbes moyennes par élément d’architecture Tableau du protocole de mesure

179

The Mortella III Wreck

180

Dendrochronology Study

181

The Mortella III Wreck

182

Dendrochronology Study

183

The Mortella III Wreck

184

Dendrochronology Study

185

The Mortella III Wreck

186

Dendrochronology Study

187

The Mortella III Wreck

188

Dendrochronology Study

189

The Mortella III Wreck

Mortella 3 : protocole de mesure

aubier

optimal

maximal

cambium

 

 

 

 

 

 

 

 

 

 

 

QU

 

2

2

1

119

/

20

 

 

fin très serrée

20a / 02

S.E

QU

 

1

1

1

99

/

20

 

 

 

20b / 01

S.E

QU

 

1

1

3

116

/

20

 

 

 

20b / 02

S.E

QU

 

1

1

3

120

/

20

 

 

 

22b / 01

G20

QU

 

1

1

5

123

/

20

 

 

 

22b / 02

G20

QU

 

1

1

3

112

/

20

 

 

 

22b / 03

G20

QU

 

1

1

3

114

/

20

 

 

 

22b / 04

G20

QU

 

1

1

5

110

/

20

 

 

 

D

T

moelle

  S.E

objet

  20a / 01

nom

A

âge mesuré

estimation

espèce

croissance

remarques

22c / 01

G20

QU

 

1

1

5

101

/

20

 

 

 

23b / 01

A20

QU

2

2

2

5

59

4

3

6

 

 

23b / 02

A20

QU

2

2

2

5

59

4

3

6

 

 

23b / 03

A20

QU

1

2

2

1

63

4

5

10

 

début de croissance très perturbée

23b / 04

A20

QU

 

1

1

20

39

/

20

 

 

20 premiers cernes non mesurés

23c / 01

A20

QU

 

1

1

1

47

/

20

 

 

 

23d / 01

A20

QU

1

1

1

1

60

4

3

6

 

 

23d / 02

A20

QU

1

1

1

1

58

3

4

8

 

 

23d / 03

A20

QU

1

1

1

1

58

2

5

10

 

 

23d / 04

A20

QU

1

1

1

20

63

4

3

6

 

 

25b / 01

B.B

QU

 

1

1

1

61

/

20

 

 

blessure après 10 cernes

25b / 02

B.B

QU

 

1

1

1

66

/

20

 

 

blessure + cassure vers la fin

25b / 03

B.B

QU

 

1

1

1

30

/

20

 

 

 

25b / 04

B.B

QU

 

1

1

2

58

/

25

 

 

5 derniers cernes non mesurés

25b / 05

B.B

QU

 

1

1

1

30

/

20

 

 

 

26a / 01

CAR

QU

 

1

1

1

160

/

20

 

 

 

26a / 02

CAR

QU

 

1

1

1

117

/

20

 

 

 

26a / 03

CAR

QU

 

1

1

1

98

/

20

 

 

 

26b / 01

B.A

QU

 

1

1

5

107

/

20

 

 

 

26b / 02

B.A

QU

 

1

1

5

132

/

20

 

 

 

26b / 03

B.A

QU

 

1

1

5

128

/

20

 

 

 

26b / 04

B.A

QU

 

1

1

5

99

/

20

 

 

 

27b / 01

PSB

QU

 

2

2

4

92

/

20

 

 

croissance très perturbée

27b / 02

PSB

QU

 

1

1

1

99

/

20

 

 

 

28b / 01

G20

QU

 

1

1

10

73

/

20

 

 

 

28b / 02

G20

QU

 

1

1

1

78

/

20

 

 

 

28b / 03

G20

QU

 

1

1

12

126

/

20

 

 

190

Annex 4 Étude de la compression transversale des bois de l’épave de la Mortella III Fabien Langenegger Archaeologist and dendrochronologist, Cultural Affairs Department, Office and Museum of Archaeology, Republic of Neuchâtel, Switzerland. [email protected] Les échantillons

En fonction du type désiré de pièces d’architecture, on sélectionne un tronc parmi un stock de bois, de provenance très hétérogène, et on adapte le type de façonnage. Le but étant d’obtenir des sections avoisinant les 180 cm2 pour les varangues, les genoux et les allonges. La section du galbord est plus étroite (111cm2) et celle de la clé plus imposante (345 cm2).

Lors de l’opération 2014 de la fouille sous-marine de la Mortella III, six nouveaux échantillons de bois ont été prélevés. L’objectif pour cette année est d’étudier l’état sanitaire des bois et de déterminer si les sections transversales conservent encore leurs dimensions d’origine. En outre, de nouvelles mesures de la croissance des cernes ont été effectuées pour compléter les datations obtenues précédemment. Les six échantillons proviennent d’une varangue, d’un genou (2 prélèvements), d’une allonge, d’un galbord et d’une clé avant. Tous sont en chêne et cette diversité d’éléments d’architecture permet d’avoir une bonne représentativité pour quantifier d’éventuelles déformations dans les sections.

Premier groupe :

Les dessins des sections transversales permettent une reconstitution de la taille minimale des troncs des arbres abattus et de déterminer le ou les types de chênaies exploitées. Deux échantillons conservent de l’aubier, la partie “vivante” de l’arbre et permettent ainsi d’estimer avec précision le diamètre des grumes. Les représentations graphiques montrent clairement deux types différents de façonnage liés étroitement à la matière première à disposition :

Second groupe :

• Le premier groupe est composé de chênes âgés de 120 ans avec des troncs d’un diamètre compris entre 20 et 24 cm. Pour obtenir les pièces, les grumes sont équarries plus ou moins fortement sans aucun travail de refente. La section de la clé avant conserve encore l’essentiel de l’aubier. Pour le galbord, la date obtenue pour le dernier cerne mesuré montre que le façonnage périphérique a ôté une soixantaine de cernes de croissance et le diamètre minimal reconstitué ci-dessous est trop petit de trois centimètre environ. Il en va de même pour la section transversale du genou G27B. Nous sommes donc en présence de trois chênes de 24 cm de diamètre. En étudiant leur courbe de croissance, on remarque que ces trois arbres ne proviennent pas du tout du même terroir forestier. • Le deuxième groupe est constitué de chênes âgés de 170 ans environ avec des troncs d’un diamètre compris entre 34 et 40 cm. Pour obtenir les sections désirées, les billes sont refendues en quart. 191

The Mortella III Wreck

espèce

moelle

âge mesuré

pos. aubier

optimal

maximal

cambium

section en cm2

Galbord 25/27

galbord

QU

1

43

/

20

 

 

111

Galbord 25/27

galbord

QU

1

52

/

20

 

 

111

nom

objet

estimation

G27B

genou

QU

1

41

/

20

 

 

175

G27A

genou

QU

5

65

/

20

 

 

197

A27

allonge

QU

2

134

/

20

 

 

189

V27

varangue

QU

3

124

/

20

 

 

170

Clé avant

Clé avant

QU

1

116

101

2

4

 

345

Le protocole de mesure des bois L’analyse des échantillons Une première observation visuelle permet de déterminer l’état de conservation du bois et de définir les zones dégradées par les bactéries et sensibles à un écrasement important. La résistance du bois en compression transversale varie selon la position des couches annuelles. Elle sera bien meilleure lorsque la compression se fait dans le sens de la croissance que latéralement. Autrement dit, lorsque la compression se fait perpendiculairement au fil du bois. L’observation des vaisseaux et des pores du bois permettra de constater si une compression transversale existe sur les échantillons et de la quantifier. L’allonge A27 Pour réaliser cette pièce, l’aubier a été enlevé et les deux images faites en périphérie du duramen montrent que les vaisseaux du bois initial sont intacts et n’ont subi aucune déformation transversale (image 1 et 2). Dans le bois final, les pores sont également bien visibles et ne présentent aucune compression. En étudiant les courbes de croissances des échantillons, on devrait observer, lors d’un écrasement des fibres du bois, une diminution brusque de la largeur des cernes. Les courbes mesurées sur l’allonge 27 sont très régulières avec un cerne moyen de 0.88 mm. La varangue V27 Pas d’aubier sur cet échantillon. Le duramen est dégradé et les vaisseaux des trois derniers cernes sont entièrement fermés. La compression est estimée à 1mm. Le reste du bois est sain. Les courbes de croissance montrent une reprise dans les trente derniers cernes. Les 3 cernes comprimés ont été pris en compte pour déterminer l’âge de l’échantillon, mais n’ont évidemment pas été mesurés et n’apparaissent pas sur le graphe.

192

Étude de la compression transversale des bois de l’épave de la Mortella III

La clé avant

Le genou G27B

La section transversale de cette pièce de bois conserve encore une grande partie de l’aubier. La présence de l’aubier s’explique certainement par l’obligation d’obtenir, pour la clé avant, une section très importante, mesurée sur cet échantillon à 345 cm2. Le bois vivant est beaucoup plus fragile que le duramen et se déforme plus facilement.

Pas d’aubier sur cet échantillon. Le bord du duramen est fortement dégradé, mais la compression se limite au dernier cerne et la perte n’excède pas 1 mm. Au centre, le duramen est sain et les vaisseaux sont bien conservés. Les parois des cellules du bois initial sont intactes. Dans le bois final, on observe la disposition caractéristique des pores en groupes radiaux formant un décor flammé et aucun écrasement des fibres du bois n’est visible. La courbe montre une croissance du chêne qui diminue fortement avec l’âge. Mais la pente est régulière et cette diminution de l’accroissement est propre à l’arbre et non à la dégradation de l’échantillon.

Pourtant, la compression observée des cellules est très limitée dans l’aubier. Les trois premières illustrations concernent le bois vivant. En périphérie, dans certaines zones, on constate un écrasement des vaisseaux (image 1), dans d’autres, le bois est intact (image 2). A l’intérieur de l’aubier, les cellules sont partout très bien conservées et 193

The Mortella III Wreck ne présentent aucune compression (image 3), et il en va de même dans le duramen de la clé avant (image 4). La réduction du diamètre de l’échantillon due à l’altération du bois n’excède pas 1 mm. Le trend de la courbe de croissance de la clé avant est très comparable à l’échantillon de l’allonge 27, avec un cerne moyen très réduit de 0,84 mm et une courbe très plate sous le millimètre.

Le galbord C’est le prélèvement transversal qui présente la plus grande déformation des cellules du bois. Elle est observable sur une largeur d’un centimètre (image 1). Les vaisseaux du bois initial sont complètement écrasés, mais la différence avec la section d’origine reste limitée et se chiffre à environ 5 mm. La partie centrale n’est pas concernée par cet écrasement et les cellules sont bien conservées (image 2). Cette compression est bien visible sur la courbe de croissance. La courbe plonge subitement sous le millimètre au niveau de l’écrasement du bois.

194

Étude de la compression transversale des bois de l’épave de la Mortella III

Conclusion En analysant les échantillons prélevés lors de la campagne 2014, on constate que les compressions transversales remarquées sur les sections sont dues uniquement à l’altération de certaines parties du bois dans l’eau. Tant que le bois est gorgé d’eau, il conserve ses dimensions d’origine, les parois des cellules ne se déforment pas. En revanche, une rétraction du bois est déjà constatée après un bref séjour hors de l’eau (1 heure), mais tant que le bois est sous l’eau, il ne se rétracte pas. Par contre, une compression des cellules est observable sur plusieurs échantillons, mais elle est localisée uniquement à la partie périphérique des pièces de bois qui est la plus rapidement altérée. Les dimensions mesurées sur les sections de ces pièces sont équivalentes aux dimensions d’origine et les pertes se limitent à quelques millimètres au maximum. L’épave de la Mortella III est aujourd’hui complètement envasée et protégée des phénomènes extérieurs et une perte de matière due à l’érosion est peu probable. Aucun prélèvement ne présente une surface érodées, les traces de façonnage sont encore visibles.

195

Annex 5 Chemical analysis of caulking and sealing materials Dr. Carole Mathe Associated professor, IMBE UMR 7263/ IRD237, Avignon University/CNRS/ IRD/Aix-Marseille University, Restoration Engineering of Natural and Cultural Heritage, Faculty of Sciences, Campus Jean-Henri Fabre, 301 rue Baruch de Spinoza BP 21239, 84916 Avignon Cedex 9, France. [email protected] - Phone : +33(0)490 144 454 I. Objet de l’étude – Objet : demande d’analyses chimiques – Pièces : prélèvements provenant de l’épave la Mortella (fouilles 2010). – Localisation : baie de St Florent, Corse, France.

III. Analyse par Chromatographie en Phase Gazeuse couplée à une Spectromètre de Masse (CPG-SM) 1. Matériel et méthode Les analyses en chromatographie en phase gazeuse ont été réalisées à l’aide d’un chromatographe Varian Saturn 3900, équipé d’un injecteur Varian 1177 et couplé à un spectromètre de masse à ion trap, Varian 2100 T. La colonne capillaire utilisée possède une longueur de 30 m,

II. Description des échantillons Le tableau ci-dessous (tableau 1) donne la description des prélèvements analysés (crédits photos : ©IRPNC, IMBE).

Tableau A5.1. Description des échantillons étudiés Références R136 Ech 2010

R136-A MIII/10/Ech 06

Description macroscopique

Masse totale

Couche noire luisante entre deux couches de bois

82,2974 g

Substance noire hétérogène luisante avec quelques dépôts blanchâtres

4,0645 g

Substance noire hétérogène luisante avec des impuretés

4,4585 g

Substance noire hétérogène luisante collante au broyage

3,8685 g

R136-B Bordé G20 Face xet B MIII/10/Ech 0030.A R136-C Bordé G20 Face Ext A MIII/10/Ech 0031.C R136-D Etoupe A Boradé 620 milieu MIII/10/Ech 0029.B

197

Photo

The Mortella III Wreck un diamètre interne de 0,25 mm et une épaisseur de film de 0,25 mm de 5% phényl, 95% dimethylsiloxane : il s’agit d’une CP-Sil 8 CB Low Bleed/MS (Varian).

L’interprétation de ces chromatogrammes traduit de grandes similitudes entre eux notamment dans la zone d’élution comprise entre 22,5 et 33 min (figure 2).

Le voltage du multiplicateur d’électron est à 1400 V, le temps d’ionisation dure 25000 ms et il s’effectue par impact électronique. La ligne de transfert, la trappe à ions et l’enceinte de la trappe (“manifold”) sont respectivement maintenues à 300°C, 200°C et 50°C. Le détecteur scanne des masses comprises entre 40 et 650 (m/z) avec un voltage ionisant de 70 eV. Les échantillons sont injectés (1 mL) en mode splitless. Un débit continu de 1 mL/min d’hélium de grade analytique est utilisé.

Les chromatogrammes des échantillons R136 B et C sont très semblables, celui de R136 A est similaire aux deux précédents avec cependant des différences qualitatives et quantitatives. Par contre, le chromatogramme du prélèvement R136 D apparaît différent de ces homologues. Une interprétation moléculaire de chacun des pics obtenus a donc été réalisée à partir de l’étude des spectres de masse et en référence avec des molécules standards et/ou la banque de données NIST’08 (tableau 3).

2. Dérivation des échantillons via une triméthylsilylation

Les échantillons analysés possèdent tous une population en dérivés aromatiques appartenant à la famille des diterpènes. Tous les échantillons possèdent une population diterpénique commune qui se distingue cependant de par la proportion relative de chacun des pics présents, mais également sur le plan qualitatif. En effet, ces prélèvements traduisent la présence de molécules à squelettes abiétanes comme l’acide abiétique ou le méthyl abiétate qui sont des composés caractéristiques de la famille des Pinacées. A noter également, la présence de composés pimaranes indiquant l’appartenance de la résine à l’ordre des Conifères. La GC-SM a notamment mis en évidence l’acide déhydroabiétique et l’acide 15-hydroxy-déhydroabiétique dans la fraction acide et le rétène, tétrahydrorétène, le 18-Norabiéata-8,11,13-triène, le 19-Norabiéata-8,11,13triène, 10,18-bisnorabiéta-8,11,13-triène dans la fraction neutre. Le rétène est considéré comme un produit final stable de des différentes voies réactionnelles de l’oxydation des dérivés abiétanes acides et les norabiétatrienes et des tétrahydrorétène représente les intermédiaires de ces réactions (Pollard et al., 1996, Colombini et al., 2005).

Une quantité de 5 à 10 mg d’échantillons à analyser est triméthylsilylée avec 0,1 mL d’une solution constituée de 0,5 mL de pyridine anhydre, 0,45 mL d’hexaméthyldisilazane (HMDS) et 0,3 mL de triméthylchlorosilane (TMSCl). La réaction est effectuée à température ambiante pendant 30 min, temps au bout duquel la solution est évaporée à sec sous courant d’azote ou d’argon avec un chauffage inférieur à 40°C. Le résidu ainsi obtenu est alors solubilisé dans 0,6 mL d’éther éthylique de grade analytique (Merck), puis directement injectée en CPG/SM. 3. Gradient d’analyse La température initiale du four est de 50°C, pendant 2 min, puis il y a une augmentation de celle-ci de 8°C/min jusqu’à 250°C, suivi d’une deuxième élévation de température à 3°C/min jusqu’à 350°C. L’analyse s’effectue en mode split avec un rapport de 20. L’injecteur, la trappe ainsi que la ligne de transfert sont maintenus à respectivement 250, 200 et 300°C. Le temps d’analyse est d’environ une heure (tableau 2).

De plus, il est possible d’observer notamment sur les chromatogrammes l’occurrence de l’acide déhydroabiétique méthyl ester.

4. Résultats et discussion L’objectif de ce travail est de caractériser par CPG-SM quatre échantillons de nature archéologique référencés respectivement R136 A, B, C et D.

Les acides diterpéniques méthylés indiquent l’emploi d’une poix résultant d’un traitement par distillation destructive (pyrolyse) de bois résineux. En effet, durant ce procédé, du méthanol gazeux rejeté par le chauffage du bois réagit facilement avec les acides diterpéniques initialement présents pour produire du déhydroabiétate de méthyle (Pollard et al., 1996, Colombini et al., 2005), qui est absent de la poix produite par pyrolyse de la résine seule (Colombini et al., 2009 ; Izzo et al., 2012).

4.1 Chromatogrammes obtenus Après dérivation des échantillons par triméthylsilylation, les chromatogrammes obtenus sont présentés dans la figure suivante (figure 1).

Tableau A5.2. Gradient de température pour l’analyse en CPG/SM Température en °C

Montée de température (°C/min)

Temps de maintien de la température (min)

Durée totale (min)

50

–

2

10

250

8

0

27

350

3

0

60

Ttrappe = 200°C

Tligne de transfert = 300°C

T injecteur = 250°C

Rapport de split = 20

198

Chemical analysis of caulking and sealing materials

Figure A5.1. Chromatogrammes obtenus par CPG-SM

Figure A5.2. Agrandissement des chromatogrammes entre 22,5 et 33 min

199

The Mortella III Wreck Tableau A5.3. Terpènes identifiés par CPG-SM Echantillons tR (min)

R136 A

R136 B

R136 C

R136 D

1

N°

18-Norabiéata-8,11,13-triène

Composés

23,39

√

√

√

–

2

Dérivé de l’ac. déhydroabiétique méthylester

23,42

√

√

√

√

3

19-Norabiéata-8,11,13-triène

23,85

√

√

√

√

4

10,18-bisnorabiéta-8,11,13-triène

24,12

√

√

√

√

5

Méthylpodocarpa-8,11,13-trien-15-oate

24,71

√

–

–

–

6

1,2,3,4-tétrahydrorétène

24,81

√

√

√

√

7

Inconnu

25,78

√

√

√

√

8

Méthylabiétate

25,92

√

√

√

–

9

Inconnu

25,99

√

√

√

–

10

Rétène

26,43

√

√

√

√

11

Dérivé de l’ac. isopimarique

26,43

√

√

√

–

12

2-isopropyl-10-methylphénanthrène

26,75

tr

–

–

–

13

Acide sandaracopimarique

27,15

√

√

√

–

14

Acide pimarique

27,16

√

√

√

–

15

8-isopropyl-1,3-diméthylphenanthrène

27,23

√

√

√

√

16

Acide isopimarique

27,34

√

√

√

√

17

Acide déhydroabiétique méthylester

27,44

√

√

√

√

18

Acide palustrique

27,50

tr

–

√

√

19

Acide déhydrodéhydroabiétique

27,59

√

√

√

√

20

Acide déhydroabiétique

27,86

√

√

√

√

21

Δ

28,15

√

√

√

√

22

Acide abiétique

- Déhydroabiétate de méthyle

28,27

√

√

√

–

23

Acide 15-hydroxydéhydroabiétique

29,63

√

–

–

√

24

Acide 7-hydroxydéhydroabiétique

29,86

√

–

–

√

25

Acide 7-oxodéhydroabiétique méthyl ester

30,53

tr

–

–

–

6,15

√ : présence ; - : absence ; tr : traces

Tableau A5.4. Acides gras identifiés par CPG-SM

La figure 3 montre le schéma les voies de dégradation thermique menant à la formation des composés caractéristiques présents dans la poix (Colombini, et al., 2009).

Composés

D’une manière générale, le terme poix désigne le résidu obtenu après traitement par le feu d’une oléorésine ou de bois insoluble dans l’eau, mais soluble dans un solvant organique (Garnier, 2003 ; Colombini et al., 2009) La recherche d’acides gras libres a été également réalisée et les résultats obtenus sont présentés dans le tableau 4. Le prélèvement R136 A ne contient aucune trace d’acide gras libre. R136 B et C traduisent la présence d’acides gras saturés et insaturés comme les acides myristique (C14 :0), palmitique (C16 :0), margarique (C17 :0), oléique ((C18 :1) et stéarique) en proportion non négligeable.

tR (min)

Echantillons R136 A R136 B R136 C R136 D

Acide myristique C14 :0

21,47

–

tr

tr

–

Acide palmitique C16 :0

23,90

–

√

√

√

Acide margarique C17 :0

25,09

–

tr

tr

–

Acide oléique C18 :1

25,90

–

√

√

–

Acide stéarique C18 :0

26,15

–

tr

√

√

√ : présence ; - : absence ; tr : traces

D’une manière générale la proportion relative de ces acides gras est relativement importante au sein de l’échantillon R136 D et relativement modérée pour R136 B et C. A noter que ces acides gras sont absents du prélèvement R136 A.

Le prélèvement R136 D contient uniquement des acides palmitique et stéarique, mais en plus grande proportion comparativement aux deux autres échantillons. 200

Chemical analysis of caulking and sealing materials

Figure A5.3. Schéma de l’oxydation et de la dégradation thermique conduisant à la formation des composés caractéristiques de la poix de pin (Colombini, et al., 2009).

5. Conclusion

pas subi de dégradation naturelle (vieillissement) et/ ou anthropique (chauffage). Ceci indique que ces trois échantillons correspondent à un mélange de résine de pin et de poix. Par contre, cette molécule est absente dans R136 D ce qui exclut la présence de gemme (ou gemme-résine, c’est-à-dire l’exsudat récolté de l’arbre). L’utilisation de substances extraites de l’exsudat et/ou du bois des espèces arboricoles de la famille des Pinacées est en accord avec l’abondance des diverses espèces de Pinus spp. présentes sur le pourtour du bassin Méditerranéen.

L’étude analytique développée a permis la caractérisation et la comparaison entre les résidus organiques de trois échantillons récoltés sur l’épave La Mortella, dans la baie de St Florent en Corse, (France). Les échantillons R136 A, B et C possèdent une composition chimique semblable voire similaire pour R136 B et C. L’échantillon R136 D se distingue de ses homologues. La présence systématique de molécules à squelette abiétane et pimaranes indique la présence de résine appartenant à la famille des Pinacées et plus particulièrement à l’ordre des Conifères, c’est-à-dire de la résine de pin (type colophane).

A noter que les résultats obtenus par chromatographie en phase gazeuse sont en parfaite adéquation avec ceux décrits dans le précédent rapport d’analyse (juin, 2001), à savoir :

Le haut degré d’oxydation des diterpénoïdes présents dans les résidus organiques prélevés suggère que ces résines ont été chauffées dans une atmosphère oxydante ou ont fait l’objet d’un processus de vieillissement par oxydation. Tous les échantillons étudiés ont révélé la présence de poix ou, plus précisément, de goudrons de bois utilisés pour préparer la poix. Cette information est confirmée par ailleurs par la présence de rétène, et de méthylester d’acide déhydroabiétique, qui sont des produits issus de la réaction chimique entre les acides diterpéniques natifs et le méthanol gazeux formé au cours de la distillation du bois. L’utilisation de poix obtenue par combustion de résines végétales a été largement documentée dans le passé, notamment comme traitement d’imperméabilisation.

  i) tous les prélèvements contiennent de la matière organique,   ii) Répartition des échantillons en deux lots distincts : R136 A, B, C et R136 D, iii) Présence de molécules diterpéniques de type colophane pour R136 A, B, C. En ce qui concerne les matières grasses, la présence d’acide gras a été détectée dans R136B et D (population identique) et dans R136 D en proportion plus importante. De plus, R136 D contrairement aux autres prélèvements, traduirait la présence de triterpènes de type friedelan-3-one. Par conséquent, comme il a été proposé pour les esters mixtes des acides gras et le lupéol (Dudd et al., 1999) ou longibornéol (Charrié-Duhaut et a.l, 2007), ces résidus organiques correspondent par exemple au résultat du mixage ou du chauffage de graisses animales avec la poix

La présence de l’acide abiétique en proportion non négligeable au sein des échantillons R136 A, B, C indique la présence de résine pin non altérée c’est-à-dire n’ayant 201

The Mortella III Wreck Tableau A5.5. Résumé des résultats obtenus par CPG-SM Echantillon

Composition chimique

R136 A

Poix + résine de pin

R136 B

Poix + résine de pin +graisse

R136 C

Poix + résine de pin +graisse

R136 D

Poix + acides gras + graisse + résine triterpénique?

(Charrié-Duhaut et al., 2009). L’origine animale de la graisse employée reste cependant à être confortée. Dans un souci de clarté, le tableau 5 résume les résultats obtenus. Références bibiographiques Charrié-Duhaut, A., Connan, J., Rouquette, N., Adam, P., Barbotin, C., de Rozières, M.-F., Tchapla, A., Albrecht, P., 2007. The canopic jars of Rameses II: real use revealed by molecular study of organic residues. Journal of Archaeological Science 34, 957–967. Charrié-Duhaut, A. Connan, J., Darnel, M. Spangenberg, J., Szymczyk, E., Bissada, A., Albrecht, P., 2009. Molecular and isotopic characterization of organic samples from the wreck of the Saint-Etienne merchant ship (XVIIIth century): Identification of pitch, fat, hair and sulfur. Organic Geochemistry 40, 647–665 Colombini, M.P., Modugno, F., Ribechini, E. 2005. Direct exposure electron ionization mass spectrometry and gas chromatography/mass spectrometry techniques to study organic coatings on archaeological amphorae, Journal of Mass Spectrometry, 40, 675 687. Colombini, M.P, Modugno F., 2009. Organic Mass spectrometry in Art and Archaeology. Wiley & Sons, Chichester, p221. Dudd, S.N., Evershed, R.P., 1999. Unusual triterpenoid fatty acyl ester components of archaeological birch bark tars. Tetrahedron Letters 40, 359–362. Izzo, F.C., Zendri, E., Bernardi, A., Balliana, E., Sgobbi, M. 2012. The study of pitch via gas chromatography– mass spectrometry and Fourier-transformed infrared spectroscopy: the case of the Roman amphoras from Monte Poro, Calabria (Italy). Journal of Archaeological Science, in press. Garnier, N., 2003. Analyse structurale de matériaux organiques conservés dans des céramiques antiques, thèse de doctorat, Université de Paris VI. Pollard, A.M., Heron, C. 1996. Archaeological Chemistry; RSC Paperbacks, Cambridge.

202

Annex 6 Main original texts resulting from the literature research 1. 1527 Shipwrecks Giustiniani, Agostino, 1537, Castigatissimi annali con la loro copiosa tavola della Eccelsa et Illustrissima Repubblica di Genova, Gênes. Edité par Cambridge (Mass.): Omnisys, 1990, fº270v

203

The Mortella III Wreck Traduction “Dans le courant de l’année 1527, la ville de [Gênes] alors sous le dogat d’Antiniotto Adorno était en proie une terrible famine, le grain était épuisé à tel point que le pain était rationné par tête d’habitant et que chacun ne recevait plus que trois petits pains. Et dans la ville, la livre de grain valait jusqu’à quatorze lires, et au-delà de ses portes, elle valait de dix-huit jusqu’à vingt-cinq lires. On arma quatre nave avec l’appui des bateaux qui étaient en provenance de Sicile et d’ ailleurs, pour transporter du grain à la ville, parmi lesquelles les navi la Ferrara y la Boscaina de Rapallo, dans le golfe de Saint-Florent en Corse, qui furent poursuivies par les galères françaises, et furent par manque de vent contraintes de toucher terre, les équipages furent sauvés, mais les corps des navi furent brulés. La ville avait récupéré Côte du Levant, et le capitaine Andrea Doria, de nouveau à la solde de la France avait été nommé capitaine général et amiral de la flotte française. Il provoqua beaucoup de dégâts à la ville, alors qu’elle tenait (?) le site de Portofino qui avait été fortifié avec des bastions et d’autres abris. … » Broc, Damien, 2014, Dynamiques politiques, économiques et sociales dans la Corse médiévale : le Diocèse de Nebbio (XI siècle – c. 1540). Histoire. Université Pascal Paoli, p.143 https://tel.archives-ouvertes.fr/tel-01258829/document P.143. En 1527, le golfe de Saint-Florent était également le théâtre d’un affrontement naval entre Français et Génois pro-Impériaux. Gênes souffrait alors d’une grande pénurie de grain, à cause du blocus que la Ligue lui imposait. Ne pouvant plus s’approvisionner dans les Rivières, les Génois armaient quatre naves destinées à compléter une flotte partie chercher du grain en Sicile et en d’autres lieux. Or, deux de ces naves de Rapallo, la Ferrara et la Boscaina, étaient prises en chasse par les galères françaises jusque dans le golfe de Nebbio. Sans vent, ces navires étaient contraints de toucher terre. Les équipages se sauvaient mais les bâtiments étaient incendiés (Juillet 1527)

204

Main original texts resulting from the literature research 2. 1526 Shipwrecks Casoni, Filippo, Annali della Republica de Genova del secolo decimo sesto, ed. Antonio Casamara, Gênes 1708, p. 91 p.91, année 1526:

Lingua, Paolo, Breve storia dei genovesi, ed. Laterza, 2004, p.93

l’arrivo degli avversari nelle acque corse, viene raggiunto dal Navarro nel golfo di La Spezia; lo affianca con 6 galee francesi, 5 pontificie, 5 veneziane ed intercetta a punta Chiappa, presso Portofino, 24 navi nemiche. Le affronta con il Navarro. L’ammiraglia degli avversari, la “Portunda”, è rasa come un pontone dall’artiglieria del Navarro. Il Doria si getta con la sua galea in mezzo a 2 vascelli, demolisce il primo con i falconetti e le bombarde che operano ad alzo zero e cola a picco il secondo (sul quale sono imbarcati 300 uomini) con un colpo di rostro. Nella battaglia muoiono numerosi galeotti e marinai della flotta imperiale, nonché un migliaio di soldati dell’esercito comandato da Carlo di Lannoy, da Ferrante Gonzaga e da Ferdinando d’Alarcon. Per alcune fonti lo scontro si verifica, invece, nella baia di San Lorenzo, in Corsica. Al termine del combattimento, durato dalle quattro alle cinque ore, si congiungono alle navi del Doria, allo scopo di affiancarlo nelle operazioni, anche 2 galee del SaintBlancard.’

Clemente VII, fragile e indeciso, accenna a ritirarsi dalla Lega, poi cambia ancora idea e decide di affrontare nuovamente Carlo V. Si riuniscono le flotte dei Veneziani, dei Francesi e di Andrea. Il comando è affidato al Doria e a Pietro di Navarra, che con diciassette galee infliggono alle trentasei navi imperiali comandate da Antonio Lannoye, Ferrante Gonzaga e Ferdinando d’Alençon, l’unica sconfitta sonora alle armi spagnole. La galea ammiraglia è rasa come un pontone dall’artiglieria diretta dallo stesso Pietro di Navarra. Andrea Doria, a sua volta, compie una prodezza personale: la sua galea si getta in mezzo a due vascelli spagnoli, ne demolisce il primo con i falconetti e le bombarde che sparano ad alzo zero e ne cola a picco il secondo con un magistrale colpo di rostro. Teatro dello scontro è la Baia di San Lorenzo, in Corsica. Un migliaio di soldati e marinai della flotta di Carlo V vi perdono la vita: ma è una vittoria inutile per le sorti della Lega. Infatti, per chiudere definitivamente la partita, Carlo fa calare in Italia un’armata terrificante, i 12.000 lanzichenecchi comandati dal feroce Georg von Frundsberg, uno dei protagonisti della vittoria della Bicocca. Roberto Damiani https://corsaridelmediterraneo.it/doria-andrea/ ‘Ottobre / novembre Raggiunge l’armata del Navarro che è ormeggiata a Portofino. Pone in tale località la sua base operativa e vi lascia alla guardia il nipote Filippino con 500 fanti. Con 205

The Mortella III Wreck 3. 1555 Shipwrecks Ceccaldi, Marc-Antonio, 2007, Histoire de la Corse – 1464–1560, ed. A. Piazzola, p.491

Document issu de l’Archivo General de Simancas, cité par VARGAS-HIDALGO, Rafael, dans Guerra y Diplomacia en el Mediterráneo: Correspondencia inédita de Felipe II con Andrea Doria y Juan Andrea Doria, Madrid, 2002, pp.15-16 –

Gómez Suárez de Figueroa, embajador en Génova, a Juana de Austria (Génova, 21 de diciembre de 1555) “Las doze galeras de Françia que avian ydo a Çivitavieja (Civitavecchia) se bolvieron y llegando en la ysla de 206

Main original texts resulting from the literature research Corzega en el golfo de San Florençio (Saint-Florent) a los 4 del presente por la mañana estubieron alli todo el dia en el qual al anochezer llego Don Alonso Pimentel con diez naos cargadas de ynfanteria y sin poder descubrir las galeras dieron fondo porque trayan gran nezesidad de agua y a las dos oras las dichas galeras de Françia dieron sobre las naos y quando las descubrieron cortaron los cabos que tenian dado a la mar e hizieron vela pero las galeras desparando su artelleria dio a una nao y le rompio el timon y no podiendo salir enbarazandose con otra fueron todas dos tomadas de franzeses en las quales avia tres compañias de spañoles de las mejores y tomaron toda la gente eçeto asta cien hombres que se escaparon a nado en la ysla de Corzega. Las otras ocho naos binieron las cinco dellas con Don Alonso Pimentel a la fosa de Dian y otras dos a Monago y la otra vino aqui a Genova que ha sido una desgraçia grandisima en especial en esta coyuntura y visto esto despache luego a la buelta de las naos quinze galeras para que las remolcasen e hiziesen escolta las que por los malos tiempos aun no son llegadas. Tengo aviso que se an desembarcado y que bienen por tierra los quales deseo que lleguen porque vernan a muy buen tiempo aunque segun entiendo heran sin un real y desnudos y mal tractados de la mar; darsele el mejor remedio que se pudiera.” Gómez Suárez de Figueroa, embajador en Génova, a Juana de Austria (Génova, 2 de marzo de 1556) “Ya di aviso a Vuestra Alteza como en Corzega se perdieron dos naos de las que venian con Don Alonso Pimentel y que los franceses los avian tomado con doce galeras a donde avia tres compañias de infanteria spañola y la que despues partio de Malaga con otras compañias dio al traste en Zerdena y se salbaron todos pero quiso la desgracia que se tornaron a embarcar en una nao y un escorchapin que les dio el presidente de aquel reyno (Jerónimo de Aragal) y al segundo dia que se partieron toparon con cuatro galeras franzesas y las tomaron y el escorchapin se escapo y bino en esta ribera con çiento y veinte hombres de manera que en Françia ay mas de mil quinientos hombres presos de lo que traya a su cargo el dicho Alonso Pimentel….”

207

Annex 7 Selection of texts mentioning or inducing the As-Dos-Tres rule (16th and 17th centuries) Domingo de Busturia, 1568, “Relación del maestre Domingo de Busturia en lo tocante a los arqueamientos de las naos que se toman para armada en esta costa de Biscaya por mandado de su Magestad.” AGS, Guerra Antigua, Leg. 347, nº 23. 1568. Publié par Casado Soto, J.L. dans l’apendice de “Flota atlántica y tecnología naval hispana en tiempos de Felipe II”, dans Las sociedades ibéricas y el mar a finales del siglo XVI. – Tomo 2. (Textes du Congreso internacional du même nom, 1998 – Comisaría General de España en la Expo de Lisboa ‘98).

Fernando Oliveira, «Livro da fábrica das naos », 1570, éd. Manuel Leitão, , sous le titre O livro da fábrica das naos do Padre Fernando Oliveira, 1991, Academia de Marinha, Lisbonne. «… naos, caravelas, barcos, esquifes & todos os que tem proporçao de tres por hum, ou menos.” Capitulo V. Nicolò Sagri, , 1570, “Il carteggiatore” manuscrit n°31951SA0111372C, TC Wilson Library, U. de Minneapolis, transcrit et publié dans Dell’Osa, D., 2010, Il carteggiatore di Nicolò Sagri, Transcription commentée du manuscrit. Ed. Francoangeli, Milan.

“Las naos que en esta costa de Biscaya se fabrican para de mercancía son por la mayor parte de tres y a una (...) ha de ser tres veces de ancho de manga y una vez de largo de proa a popa por la esloría.”

(13R) « È notorio a ciaschuno che ogni chorpo [h]a tre principali mesure cioiè la longecza largecza et altecza osia profondità sencza le quah non saria corpo, e perciò hogni buon maestro in queste tre mesure deve essere pratichissimo aciò lui possa e sapia dare le giuste misure (13V) ciaschuna sortte del vascello e specialmentte de la nave de la qualle noi traiamo.

Rodrigo Vargas, c.1570, “Apuntamientos de Rodrigo de Vargas.” AGI, Real Patronato, leg. 260, 2º, rº 35. Publié dans Casado Soto, 1988, Los barcos españoles del siglo XVI y La Gran Armada de 1588, Madrid. “La orden que se a de tener en arquear qualquiera nabe española o llebantista y beneçianos es, y la nao a de ser e ir en perfeción al tres, dos y as, a saber:

...dire solo delle tre principalli misure di essa nave nel tutto onde dovette sapere che ogni nave dovria essere tre voltte tantto longa per il ventto cioiè da roda a roda nella sechonda copertta quanto è quella larga nella sua magior largecza in essa sechonda chopertta et la sua altecza osia profondittà che noi dicemo nel pontalle sino a essa sechonda choperta deve essere per la mittà di essa largecza e questo è la più giusta migliore et più proporcionatta missura che si possi inmaginare sebene poche navi ogidi nel paesse nostro cossi si fabrichano ma Ile antiche chossì si fabrichavano (14R) et hoggi dei genovessi anchora cossi mantengono e li bischaini et portogalessi al medessimo e perciò quelle loro navi sono miglior veliere e specialmente borinevolle e di miglior governo del timone...” (p.123)

A 30 codos de quilla, 15 de manga y de 7 1/2 a 8 la cuvierta y de largo de 45 a 46 de popa a proa en la cuvierta.” Juan Escalante de Mendoza, 1575, « Ytinerario de navegación de los mares y tierras occidentales », éd. Cesáreo Fernández Duro, dans Disquisiciones náuticas (Madrid: Aribau, 1880), vol. 5, p. 413-515 (réédition sous le titre itinerario de navegación de los mares y tierras occidentales, Madrid: Museo Naval, 1985). “La medida más perfecta que hasta ahora se sabe, que qualquier buque de nao puede tener en la quilla para que salga conforme a la manga es: por cada 5 codos de quilla derecha se han de dar dos codos y un quinto de codo de manga, por los cuales tenga, cinco codos de quilla derecha, y conforme a esta cuenta, más o menos, al respecto…” (p.39).

Tomé Cano, 1611, Arte para fabricar, fortificar y apareiar naos de guerra merchante, con las reglas de arquearlas reduzido a toda cuenta y medida, y en grande utilidad de la navegación, 1611, Luys Estupiñan, Seville.Transcription du manuscrit dans Duro, Cesário Fernandez, 1996, Disquisiciones nauticas, vol. V, 1880, Madrid: Instituto de Historia y Cultura Naval, Dialogo segundo:

…Esloria se llama lo que hay por encima de la primera cubierta desde el codaste al branque y tajamar de proa…, por cada cinco codos de la misma quilla derecha tenga dos codos más de lanzamiento, que por esta cuenta, cada cinco codos de quilla derecha serían siete de esloria…” (p.40).

« …todos los maestros españoles, italianos y de otras naciones que manejan estas fábricas de naos an tenido uso de les dar a un codo1 de manga dos de quilla; a otro Il s’agit ici du « codo de Ribera » qui a une valeur de 57,47 cm approximativement.

1

209

The Mortella III Wreck de manga, tres de esloría, y a tres codos de manga, uno de plan; y el puntal tres cuarto de manga. » Bartolomeo Crescentio Romano, 1607, Nautica Mediterranea, Roma: Bartolomeo Bonfadino. capítulo IX, fº 63. “... le misure del Galeone per l’ordinario si fanno in terzo. Verbigracia: se il Galeone far à lungo de 9 o in 93 piedi, la sua maggior larghezza fara da 30 in 32...” Diego Brochero, AGS – Guerra y Marina, legajo 776 publié par Rodriguez Mendoza, B. M. 2008, Standardization of spanish shipbuilding: ordenanzas para la fábrica de navíos de guerra y mercante – 1607, 1613, 1618, U. du Texas. “Manga = 2 veces el puntal en la segunda cubierta Eslora = 3 veces la manga” Anonyme. “El arqueo de Cristóbal de Barros.” MNM: Colección Vargas Ponce, Tomo XXV B, doc.19 fol. 42-43. XVIIème siècle, probablement antérieur à 1613. “Para estar bien proporcionado un navío para su tiempo de paz y guerra, y jugar bien la artillería y no hacer tanto daño el enemigo, ha de tener tres, dos, as, que quiere decir la manga dos partes, el puntal una y la esloría tres.” J. A. Echeverri, entre 1648 et 1666. Manuscrit du Museo Naval de Madrid –MNM- attribué à. J.A Echeverri par Fernández Duro. Colección Vargas Ponce, T 3A Doc. 108 fol. 391-395. “ Es de saber que a las medidas de los vajeles que navegaron entre los nuestros asta los años de 600 llamaron los prácticos de aquel tiempo 3, dos y has, y se formavan de esta manera: sobre 30 codos de largo o quilla, 15 de ancho ó manga, y 10 de profundidad o puntal, y secundariamente se les dava 5 de plan y 45 de esloría.” “... en la fábrica más antigua, a un galeón con 18 codos de manga se le daba 12 de puntal y 3 codos entre una cubierta y otra, porque navegavan en dos y una cámara, con que todo el contenido de la altitud no era más que 15 codos, y este galeón fondearía cargado 11 codos y medio.”

210

Annex 8 Study of the Mortella III West Anchor Fabrizio Ciacchella Independent researcher in naval history and underwater archaeology, associated to the NavLab – Laboratorio di Storia Marittima e Navale of the University of Genoa, Italy. [email protected] Tableau A8.1. Dimensions et mesures

L’ancre Ouest de l’épave Mortella III (M III-AW) est une ancre en fer à deux bras, pourvue à l’origine d’un jas en bois qui n’a pas survécu. Elle gît sur le fond en position de repos partiellement couverte de sédiment et présente une gangue de concrétions dont l’épaisseur a été estimée à 1-2 cm.

Mesure Mesure actuelle avec estimée hors concrétion concrétion (cm) (cm) Longueur de la verge et envergure

452 × 205

449 × 202

Longueur axiale du bras au diamant

120

117

Longueur interne2 du bras à l’aisselle

103

106

64 et 48 (moy.56)

61 et 51

Épaisseur de l’organeau

8

5

Largeur et épaisseur de la verge

 

 

  à la tête de la culasse

13 × 13

10 × 10

  près de sa moitié

18 × 13

15 × 10

L’épaisseur de la gangue augmente évidemment les dimensions, avec une erreur relative qui est négligeable pour les grandes valeurs (telles que la longueur de la verge,  l’envergure et la longueur des bras), mais importante pour les petites, telles que les mesures des sections.1 Les valeurs mentionnées ci-après (en cm) ont été prises au-dessus de la gangue. Les mesures entre parenthèses sont les dimensions originales estimées, sachant que la gangue formée par les concrétions est estimée à 1,5 cm d’épaisseur.

  près du diamant

19 × 14

16 × 11

 

 

  près du diamant

20 × 14

17 × 11

  près de son tiers interne

19 × 14

16 × 11

  à sa moitié (base des pattes)

13 × 13

13 × 13

Longueur et largeur des pattes

58 × 49

55 × 46

6

3

Angles et proportions

obtenue en calculant le volume total de l’ancre et en le multipliant par la densité du fer.3 On a répété le procédé pour des valeurs d’épaisseur des concrétions de 1 cm, 1,5

Morphologie La tête de culasse présente un renflement arrondi en forme de goutte avec le sommet plat et porte encore l’organeau. Juste au-dessous se trouvent deux tourillons (tenons) pour fixer le jas, situés de chaque côté, sur le même plan des bras. La verge a une section quadrangulaire ainsi que les bras, qui suivent deux courbures différentes et se rencontrent pour former un diamant arrondi; à leurs extrémités se trouvent deux pattes de forme triangulaire.

Diamètre externe et interne de l’organeau

Dimensions

Largeur et épaisseur des bras 

Épaisseur des pattes

2

Les angles ont été mesurés avec un rapporteur sur la photo et vérifiés trigonométriquement en utilisant les dimensions linéaires; les proportions ont été calculées par les valeurs estimées au-dessous des concrétions.

Pour les mesures internes (longueur interne du bras et diamètre interne de l’organeau), la dimension estimée au-dessous des concrétions est plus grande que celle prise au-dessus. 3 Pour trouver le volume total on a additionné (valeurs prises en exemple pour des concrétions de 1,5 cm d’épaisseur) : le volume de l’organeau = section par circonférence moyenne = 2,52π x 56π = 3454,5 cm3 ; le volume de la verge = moyenne de ses sections par sa longueur = 13 x 10,5 × 449 = 61288,5 cm3 ; le volume de chaque bras = moyenne de ses sections par sa longueur interne = 13,5 × 10,5 × 106 = 15025,5 cm3 ; le volume de chaque patte = épaisseur par base par hauteur divisé par deux = 3 × 46 × 55 : 2 = 3795 cm3 . Le résultat est un volume total de 102384 cm3 soit 102,384 dm3 qui, multiplié par la densité du fer 7,79 kg/dm3, donne le poids moyen estimé de 798 kg. Le même procédé calculé pour des concrétions d’épaisseur de 1 cm et 2 cm donne des poids estimés à 651 kg et 958 kg. 2

Poids estimé En partant des dimensions originales estimées de M IIIAW, il est possible d’arriver à une estimation de son poids, 1 Pour l’ancre M III-AW chaque centimètre de concrétion autour de la structure métallique comporte une augmentation de la mesure (erreur relative) de 0,4% pour la longueur de la verge, 1% pour l’envergure, 2% pour la longueur des bras, 3% et 4% pour la longueur et pour la largeur des pattes, 15% à 20% pour les mesures des sections de la vergue ainsi que des bras.

211

The Mortella III Wreck Tableau A8.2. Angles et rapports Angles

cm et 2 cm, en obtenant des poids estimés entre 651 e 958 kg (moyenne 798 kg).

Droit Gauche Moyenne

Angle interne (à l’aisselle) entre la verge et le bras

62°

64°

63°

Conclusions

Angle axial (au diamant) entre la verge et le bras

55°

57°

56°

L’ancre M III-AW est une des plus longues connues de son époque: en Méditerranée elle est dépassée seulement par l’ancre isolée de San Nicolò di Camogli (520 cm) et par la maîtresse-ancre de l’épave de Gnalic (486 cm), récemment identifiée comme étant le navire vénitien Gagliana grossa naufragé en 1583. L’ancre M III-AW présente de nombreux aspects propres aux ancres du XVI siècle (le renflement de la tête de culasse, la position latérale des tourillons, les bras qui forment deux différentes courbures, les pattes longues qui atteignent la moitié des bras), mais aussi des caractéristiques atypiques : la verge très longue par rapport aux bras (4,24 - 3,84 fois, suivant que l’on considère la longueur interne ou axiale des bras) et l’organeau très petit par rapport à la verge (0,125 fois, alors que pour les autres ancres méditerranéennes connues de cette époque il est compris entre 0,135 et 0,173). La forme de la tête de culasse, en goutte à sommet plat, est à rapprocher de celle de la maîtresse-ancre de l’épave de Gnalic (Gn-A1). Du point de vue des dimensions, M III-AW est un peu plus courte de Gn-A1 (– 8%), mais beaucoup plus fine et plus légère, la surface de sa section et le poids étant à peine plus que la moitié (Planche 1). Avec un poids estimé entre 650 et 950 kg environ (moyenne près de 800 kg), elle n’était pas assez lourde pour être la maîtresse-ancre d’un navire tel que celui de Mortella III, avec une quille de 25 m de long et une longueur de tête à tête estimée à 35 m, comparable à celle de la nommée Gagliana grossa de Gnalic, qui était 36 m et dont la maîtresse-ancre avait un poids estimé à 1,4 t.

Rapports Rapport entre l’envergure et la longueur de la verge 

0,45

Rapport entre la longueur de la verge et la longueur interne / axiale du bras

4,24 / 3,84

entre le diamètre moyen de l’organeau et la longueur de la verge

0,125 (1/8)

entre l‘épaisseur de l’organeau et la longueur de la verge

0,011

entre l’épaisseur de l’organeau et son diamètre moyen

0,089

Finesse de la verge (rapport longueur / largeur max)

28

Effilement de la verge (rapport largeur min / largeur max)

0,625

Rapport entre la longueur de la patte et la longueur interne du bras 

0,52

Tableau A8.3. Comparaison entre les ancres Gn-A1 et M III-AW ancre

Gn-A 1

longueur de la verge

486 cm

449 cm

–8%

larg. max / long. de la verge

1/24

1/28

– 14 %

section max de la verge

M III-AW comparaison

20 × 17 cm 16 × 11 cm

poids de l’ancre

1,4 t

0,8 t

longueur du navire

36 m

35 m

– 48% (surface) – 43%

212

Annex 9 Excavation team of the Mortella III excavation 2010-2019 MEMBRES DE L’EQUIPE DE FOUILLE BARBOT, Alexandra

DEBRAND, Bérenger

HEITZMANN, Samantha

NATER, Géraldine

BERTONCINI, Alain

DROGUE, Gilles

JAMME, Stéphane

NAYLING, Nigel

BOURDEAUD’HUI, Cédric

FILIPPI, Jean-José

JOYARD, Anne

PHILIPS, Agnès

C. DE LA ROCHE, Arnaud

GENDRON, François

KÜHN, Laurent

PINELLI, Charles

CASAMARTA, Dominique

GERIGK, Christoph

LANGENEGGER, Fabien

RIGO, Inmaculada

CATTEAU, Sidonie

GRIMOND, Jonathan

LANLEAU, Jacques

SANCHEZ, Didier

CLEMENT, Normann

GUESNON, Joë

MAGER, Patrick

SEGURA G., Maite

CIACCHELLA, Fabrizio

GUEVARA, Jesus

MARIE, Jehan

TER-JUNG, Marine

COQUOZ, Xavier

HARDY, Grégory

MARTINS, Guillaume

TOMAS, Emilie

COUPPEY, Antoine

HEAMAGI, Christin

MASON, Bradon

CRESPO S., Ana

HAUREZ, Olivier

MOMBER, Gary

213

Annex 10 Excavation photos

215

The Mortella III Wreck

216

Excavation photos

217

The Mortella III Wreck

218

Excavation photos

219

The Mortella III Wreck

220

Excavation photos

221

The Mortella III Wreck

222

BAR INTERNATIONAL SE RIE S 2976

‘The quality of both archaeological and historical data is excellent.’ Professor Sylviane Llinares, Université Bretagne Sud ‘Le travail présenté par Arnaud Cazenave de la Roche est vraiment original.’ Dr David Plouviez, Université de Nantes During the Renaissance period, Mediterranean shipbuilding—particularly Italian— was renowned for its quality. But it is largely unappreciated today due to the scarcity of written sources and the lack of archaeological documentation. The discovery of the Mortella wrecks in Saint-Florent, Corsica, in 2005–2006, and the 2010–2019 excavation of the 16th century Mortella III, helps to fill these gaps. The main objective of this archaeological study is to identify ‘technical fingerprints’ and ‘architectural traits’ that could contribute to the formulation of an Italo-Mediterranean shipbuilding model from the early modern period. The analysis is based on comparisons with archaeological data from other wrecks of the period as well as written sources. Finally, literature research allows us to link the Mortella wrecks to their history, that of Genoese ‘navis’ sunk during the Italian wars of 1527, complementing the archaeological study with historical research. Arnaud Cazenave de la Roche is a researcher at the Consejo Superior de Investigaciones Científicas (CSIC) with a research grant from Marie Sklodowsca-Curie Actions (Horizon 2020 EU Programme). He has been studying naval architecture from the Renaissance and modern periods for several years. He is an associate member of the Laboratoire d’Histoire et d’Archéologie Maritimes (FED 4124) at the University of Paris-Sorbonne where he defended his doctoral thesis.

Printed in England