Building a Roman Pila

The International Journal of Nautical Archaeology

(2005)

34

.1: 123137doi: 10.1111/j.1095-9270.2005.00049.x

Notes

2005 The Nautical Archaeology Society. Published by Blackwell Publishing Ltd. 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA.

Blackwell Publishing, Ltd.NOTESNotes

Building a Roman

Pila

in the SeaExperimental Archaeology at Brindisi, Italy, September 2004

T

he basic texts for our understanding ofthe construction of Roman harbours arefound in Vitruviuss

De Architectura

2.6.1,and 5.12.26, written

c

.25 BC. A significantportion of these passages concerns the technologyof building harbour structures in hydraulicconcrete, a mortar of pozzolana, lime and watercombined with aggregate, that could be placedin inundated wooden formwork in the sea. Theextraordinary strength and durability of thismaterial, along with its ability to set underwater,made it crucial to the infrastructure of the RomanMediterranean. This passage in Vitruvius was thestarting point for the Roman Maritime ConcreteStudy (ROMACONS), co-directed by the authors,which has begun to bring new procedures to theanalysis of Roman hydraulic concrete in marinestructures throughout the Mediterranean. Anintroductory report on the background to thestudy of Roman concrete, the ROMACONSresearch design and procedures for collectingand testing concrete cores, along with the firstscientific results of the physical and mechanicaltesting of our samples, recently appeared in thisjournal (Oleson

et al

., 2004).The first two campaigns of the ROMACONS

project in 2002 and 2003 focused on the collec-tion and analysis of cores of Roman hydraulicconcrete taken from structures in the sea. In 2004we attempted a reconstruction of the proceduresused by the Romans to build inundated formworkin the sea and to prepare and place mortarand aggregate in the forms. Although Vitruviusdiscusses the design and construction offormwork in the sea as well as the preparationof mortar, many details are missing from hispresentation. Precisely how was the mortarplaced in the forms? Did the formwork itself haveto be constructed with tight seams or evencaulked in some way to prevent seepage of themortar before it had set? How were the formsactually constructed in the sea? Was aggregateadded to the mortar before it was placed within

the forms, or was it added later? Was the mortarand aggregate tamped, and if so, how? Why dothere appear to be such variations in the formworkthat has survived at various archaeological sites?Why does overlapping of boards appear in someforms and not others? How many man-hourswere required to create a cubic metre of Romanconcrete? How long would it take for Romanhydraulic concrete to set in sea-water and how longto reach its maximum strength? Our understandingof Roman imperial construction along withobservation of Roman harbour works suggestedthat a fixed building protocol did emerge for theuse of Roman hydraulic concrete, while at thesame time analysis of the evidence for ancientformwork indicates that many variables influencedits physical configuration (Oleson

et al

., 2004:216).

Given the paucity of textual evidence and thedifficulty of interpreting the archaeological data,the authors decided that the best way to resolvesome of these issues was to build somewherealong the Italian coast, using materials andtools available to Roman builders, a free-standingconcrete block. The Romans called this, a frequentfeature of maritime structures, a

pila

(e.g. Vitruvius,

De Architectura

5.12.4; C

orpus InscriptionumLatinarum

10.1640)

.

For financial, logisticaland practical considerations it had to be smallby Roman standards (approximately 8 m

3

), andconstructed in a location that permitted the bulkof this block to be under water while it cured, butwith its surface always above the high tide mark.

Finding a suitable site was easy; gainingpermission from the Italian authorities toconstruct a concrete block that might welllast for centuries was a far more difficult task.Fortunately, the primary sponsor of ROMACONS,the Italcementi Group, maintains a marinetesting station at the Brindisi branch of the LegaNavale (an Italian yachting association) in theSino de Ponente of the harbour of Brindisi, Italy.At this location Italcementi scientists analyse the

NAUTICAL ARCHAEOLOGY,

34

.1

124 2005 The Nautical Archaeology Society

mechanical and physical properties of variousmixes of modern hydraulic concrete. Throughthe efforts of Dario Belottii and Massimo Borsaof Italcementi, permission was obtained fromthe clubs president, Admiral P. Fadda, for ourexperimental archaeological project to proceed atthis facility. The work was carried out from 13 to21 September 2004.

The location was ideal for our test

pila

. Theapproved site was near a concrete quay leading toa local rowing club and within 1 m of a floatingdock of the Lega Navale in a depth of

c

.1.7 m athigh tide. The concrete quay provided a workspacefor our mortar trough and the storage of ourbuilding tools during the day. About 3 m abovethe site, the bulk quantities of lime, pozzolana,tuff and sand were stored in a section of theyacht clubs parking lot. Using a rope and abasket, quantities of the building materials wereeasily lowered to the quay as required. Transportof the mortar from the trough to the formwork(once it had been constructed) required only ashort walk. Moreover, the

pila

can be routinelychecked over time by personnel from Italcementiin the course of their work on their ownexperiments. Our site was also deep in the well-protected harbour itself and thus not vulnerableto sea conditions, and the tidal variation also wasideal,

c

.50 cm between low and high tides.Italcementi generously supplied the materials

necessary for our experimenta special orderof slaked lime putty (

grasello di calce

) selected tomatch best the composition of our lime samplestaken from Roman structures, pozzolana fromBacoli in the Bay of Naples, tuff from the samearea (to be used as aggregate), beach sand(to seal the modern contaminated mud and toprovide the appropriate Vitruvian foundationfor the first mortar layer), 0.3 m-wide planks (

c.

1Roman foot) and beams of varying sizes for theexterior frame of the formwork. Bacoli tuff wasselected since Roman harbour projects as farafield as Caesarea Palestinae imported it for usein the concrete. Unfortunately, we were unableto obtain the raw or uncured wood that mostprobably was used for formwork in the sea, sinceit was cheap and heavy. The planks and beamswe used were made of a laminated kiln-driedreconstituted wooden material with a very lowspecific gravity, which complicated our efforts atplacement in the sea.

The design of the formwork was based onan interpretation of the method described byVitruvius (5.12.2). Piled corner-posts were intended

to support edge-beams onto which vertically-driven sheet pile planks could be fixed (cf. Oleson

et al

., 2004: 201, fig. 1). But due to the excessivebuoyancy of the planks and beams availableto us, it was impossible to drive the piles intothe harbour mud. Changing the design strategysurmounted this problem. The planks werepounded in the harbour floor first, and thehorizontal frames were then secured to them,forming a box with sides each 2 m long. Thusthe formwork was constructed shell-first with aframe added later, in the sequence followed byRoman shipwrights. Although this modificationworked in this instance, it is unlikely thatformwork erected in this fashion could havewithstood the force of the sea in an exposed siteduring the time it took to build it.

The seams between the planks were generallytight except in the east wall, since our first effortsat sheet piling were the least successful. Gaps alsoexisted in the north-east and south-west corners.We realised how difficult it was to keep the plankstightly together as we drove them into the bottom.Sub-bottom obstacles such as rocks and miscella-neous harbour debris sometimes made thetask very difficult, as it surely was for the Romanbuilders. We realised quickly that such unseendebris constitutes one reason why ancientformwork is often very irregular in design, inaddition it became obvious that features such asoverlapping planks on ancient formwork are aresponse both to the difficulty of determiningthe exact dimensions of a form prior to itscompletion and to the need to ensure a tight jointbetween all the planks. These ancient designanomalies represent efforts to deal with unexpectedproblems that arose during construction. If ashort gap or a poor joint required patching, aplank could simply be hammered into the seabedalong the inside face of the existing wall. Evenwith repairs, our seams were hardly watertightand would not become so even if swelling of thewooden planks occurred. We were concernedabout the possible seepage of mortar throughthe seams in the walls of the form. Constructionof the form required 37.5 man-hours (Fig. 1).

Using the Vitruvian formula of two partspozzolana to one part lime, we mixed the mortarby hand in a quayside trough using mattocks,shovels and rakesvariations of which were inthe kit of Roman harbour-builders. We made adecision to use sea-water for our mix ratherthan fresh, since we believe that was standardoperating procedure in Roman times. Vitruvius

NOTES

2005 The Nautical Archaeology Society 125

was fastidious in his specifications of thecharacter and quality of the pozzolana and limethat went to make up hydraulic concrete foruse in the sea. If fresh water had been essentialto the mix rather than easily obtained salt water,we believe he would have specified its use. Weprepared a very stiff mortar that would notdissipate when dropped into the inundatedform, and one that potentially would be stronger.Such mortar is far too viscous to poured down atremie-tube like modern concrete. A solution toplacing such mortar into the water was to usewicker baskets with two drop-lines on the handlesand a trip-line attached to the base. The designof the baskets was based on those illustratedin Roman construction scenes and on the basketsrecently found at Pisa in the excavations of the

Portus Pisanus

. Once a basket had been loweredinto the form, it was easily tipped and emptied bypulling on the trip line (Oleson, 1985; illustratedin Hohlfelder, 1987, pp. 26465).

We discovered it was possible to lower thebasket to a desired location (after manoeuvring itin the water, taking advantage of the buoyancy ofthe wicker) and then deposit the mortar precisely(Fig. 2). The mortar came out of the wickerbasket with a gentle tug, while efforts to empty aplastic bucket failed, since the impermeable walls

of the container created too much suction. Oneman could lift, transport and dump a basket fullof mortar with considerable difficulty. Two menaccomplished the task easily. Three-man teamsadded no efficiency and merely involved a redundantworker.

Once a layer of mortar reached a thickness ofapproximately 0.2 m, the surface was tamped byindividuals using rakes and standing on planks

Figure 1. C. Brandon examines the completed form before filling. (R. L. Hohlfelder)

Figure 2. Mortar is lowered into the form using a wickerbasket with two drop lines along with a trip line. (R. L.Hohlfelder)


34

.1


placed on the external crossbeams of the form.The

caementa

, irregular pieces of aggregatebroken from blocks of Bacoli tuff with a sledge-hammer, were then thrown into the form piece bypiece in an effort to cover the entire surface of themortar unseen beneath the murky water within theform. Once distributed in this random manner,the aggregate was then tamped in the same waythe mortar had been. This protocol was repeateduntil the concrete mix finally breached thesurface. The concrete became more visible inthe murky water as the form began to fill, allow-ing us to verify that the practice of randomlythrowing in aggregate piece by piece did result inthe even distribution pattern we had hoped toachieve. By the end of 20 September (day 8), wehad placed the last layer of mortar and aggregateand tamped it in place by walking on the softmixture. The following day the concrete hadset sufficiently to walk on without any of theaggregate shifting or sinking deeper into themortar (Fig. 3). A final layer of mortar was placedover the aggregate, levelled using trowels, andthen paved with local tuff blocks. On Tuesday 21September our

pila

of

c

.8 m

3

was finished after273 man-hours of work, perhaps the first suchstructure built with these materials and techniquesin the last 1600 years or longer (Fig. 4).

Many of the questions we had hoped to answerwith this experiment have been addressed.Trip-baskets for the delivery of the mortar workefficiently and easily. The Vitruvian mortar retainsits integrity in the sea with very little seepage

between the planks of the form. The work ofbuilding a form and filling the form went smoothlyand quickly even for a small team of completelyunskilled labourers encountering a range ofproblems and finding their solutions for the firsttime. There is much more to add to the story ofthe Brindisi

pila

, but we shall wait until wehave taken our first core and tested it in thelaboratories of Italcementi before publishing afuller, more scientific account. The

pila

willremain in its form until March 2005, when theformwork will be partly removed (if that provesto be possible), and it will be cored for the firsttime. Another core will be taken in September2005, and probably one the following Septemberas well to determine the rate of curing. The

Figure 3. R. L. Hohlfelder stands on the concrete on the morning after the last mortar and aggregate had been placed inthe form. (J. P. Oleson)

Figure 4. The finished pila with pavers in place. (R. L.Hohlfelder)

NOTES


experiment appears to have been successful,but we shall have to wait for some time to learnits complete results.

Robert L. Hohlfelder

Department of History, University of ColoradoBoulder, CO 80309-0234 USA

Christopher Brandon

Pringle Brandon Architects, 10 Bonhill St.London EC2A 4QJ, UK

John Peter Oleson

Department of Greek and Roman StudiesUniversity of Victoria

Victoria BC V8W 3P4, Canada

Acknowledgements

We remain indebted to Dr Luigi Cassar, Mr Dario Belotti, Ms Isabella Mazza and Mr Massimo Borsa of of ItalcementiGroup for their support of ROMACONS and their invaluable assistance in surmounting numerous logistical problems forthis project. The Lega Navale and president Admiral P. Fadda provided every possible courtesy. The Italcementi Group, the LoebClassical Library Foundation (Harvard University), the University of Victoria and the University of Colorado, Boulder,provided generous financial support. Lastly we thank Mr Francesco Retta, Mr Fabrizio Orlandino, Mr Peppino Brescia, andMr Mario Colucci, of Italcementi, Brindisi, who worked with us to construct the

pila.

References

Hohlfelder, R. L., 1987, Caesarea Maritima, Herod the Greats City on the Sea,

National Geographic

171

.2: 26079.Oleson, J. P., 1985, Herod and Vitruvius: Preliminary Thoughts on Harbour Engineering at Sebastos, the Harbour of Caesarea

Maritima, in A. Raban (ed.),

Harbour Archaeology. Proceedings of the First International Workshop on Ancient MediterraneanHarbours, Caesarea Maritima, 24-28.6.83

, BAR Int. Series 257: 16572. Oxford.Oleson, J. P., Brandon, C., Cramer, S. M., Cucitore, R., Gotti, E., and Hohlfelder, R. L., 2004, The ROMACONS Project: a

Contribution to the Historical and Engineering Analysis of Hydraulic Concrete in Maritime Structures,

IJNA

33

.2: 199239.

WEAVING and ANCIENT SAILSNOTE

Weaving and Ancient Sails: Structural Changes to Ships as a Consequence of New Weaving Technology in the Mediterranean Late Bronze Age

D

uring the Late Bronze Age in Europe anew form of weaving developed. Insteadof the simple 1:1 scheme, we find a new

2:2 scheme that is typical of twill. This new clothcould be used in various fields including theproduction of sails, for which it was very suited. Itmay not be a coincidence that in the Late BronzeAge we can see a great development in sailing allaround the Mediterranean basin.

The introduction of sails for propulsion as analternative to the mechanical forces of rowing,paddling or towing was an epoch-making idea(Casson, 1971: 12). From at least the second halfof the 4th millennium BC Egyptian river-boatsseem to have been able to use simple square sails togo up the Nile. The first visual evidence for sailingcomes from a pre-dynastic vase-painting, dating to3500 BC and preserved in the British Museum,London (Casson, 1971: 14, fig. 10). This showsa rectangular sail fixed to a mast positioned atthe bow of the vessel (Fig. 1). A terracotta model

found in Eridu in southern Iraq, from the Tel elUbaid culture, also dating back to the 4th millennium,shows that in the Middle-East sailing propulsionwas used in river and lagoon transport at such anearly date (Bass, 1974: 12, fig. 3) (Fig. 2).

Iconographic sources shows that up to the 6thcentury BC at least, ships were propelled by amixed system of oars and sail, and ships withtwo masts appear by this date. Merchant ships,however, preferred to use sails alone, therebykeeping the deck clear, increasing stowage areaand reducing the number of crew needed, eventhough this might mean a reduction in speed andmanoeuvrability (Casson, 1994: 3646). Betweenthe 4th millennium and the 6th century BC, sailsseem to have maintained their standard rectangularshape, with only slight local or regional variations(Bass, 1974: 1136).

Beginning from the Old Kingdom in Egypt(25752134 BC), the sail was orientated with theshortest edge at the base (Baines and Malek,


34

.1


1985: 69). In this way the sail was better able toexploit the northern winds which allowedupstream sailing against the mechanical forceof the currents of the Nile. Images from the NewKingdom (15501070 BC) representing the SeaPeople invasion (1190 BC) on the other hand,show sails with numerous brails, orientated withthe wider edge at the base (Baines and Malek,1985: 689). This change could be accountedfor by the fact that by now ships were usuallyventuring out into the open sea, where theyconfronted stronger wind, for which this newshape is more adapted (Casson, 1994: 36). Forthe same reason from the 12th century BCall images of both merchant and naval shipsbelonging to Mediterranean nations show sails ofthe second type, in which resistance to the windmust mediate with wind variability, and thestructure of the sails is given greater stabilityand ease of manoeuvre by the presence of brailsattached to the yard (Casson, 1971: 3174).

Unfortunately such sails were made fromperishable animal or vegetable fibres, and despite

numerous finds from both land and marinecontexts, no direct evidence of such sails has beenfound. Consequently, for a critical study of thisparticular aspect of ancient marine technology, itis necessary to take clues from indirect archaeologicalsources, as for example, ship models, painted orcarved images, as well as the information whichcan be deduced from historical texts such asMiddle Eastern and Greek epic poetry. First ofall, images from the Egyptian era give us veryfew precise details about sails, other than theircolour, generally white, and their basic shape(Casson, 1971: fig. 19). Combining such icono-graphic information from the Egyptian periodwith details from written sources, however, allowsus to affirm that, originally, the selected materialfor making sails had to be linen (Homer,Odyssey: V. 234ff). We know that the cultivationand processing of domestic linen (

linum usitatis-simum

) was introduced into Syria in approx-imately the 8th millennium BC, as the findings ofTell Aswad demonstrate (Bazzanella, 1999: 199),and from there they became known very rapidlyin all of the Eastern and Mediterranean cultures.In Italy, for example, fibres have been recoveredfrom a Neolithic village submerged in LakeBracciano (Fugazzola Delpino, 2001: 1325). Thisvegetable fibre, easy to find and extremelyversatile, offered all the characteristics of resistanceand lightness necessary for the creation of a sailexposed to the pressure of the wind.

Initially, at least starting from the Neolithicperiod (9th to 8th millennium BC in Syria), thefibre was obtained directly by retting the plantand weaving was carried out with very elementaryvertical looms (Bazzanella

et al

., 2003: 8893),suitable only for weaving according to a

cloth

scheme, with which it is not possible to makecloths of large dimensions (Rast-Eicher, 1997: 547).Therefore, since the dawning of navigation, sailshad been produced by assembling differentpieces of cloth, sewn together with twisted fibresof linen to create vertical strips of significantdimensions. Such strips were then placed togetherand sewn until the desired sailing surface wasobtained. This is described by Homer duringthe departure of Odysseus from the Island ofCirce (V, 25860), and seems to have remainedsubstantially unaltered for the entire Mediterraneanprehistoric era, or at least until an advancedstage of the Bronze Age (15th to 14th centuryBC).

The introduction of the double-step loom datesback to this period and meant that clothes could

Figure 1. A rectangular sail on a mast at the bow of a boat,from a pre-dynastic vase-painting in the British Museum.(after Casson, 1971: fig. 10)

Figure 2. A terracotta model from southern Iraq, dating tothe 4th millennium BC. (after Bass, 1974: fig. 3)

NOTES


be made with a diagonal reinforcement, definedas twill. This innovation, apparently of limitedsignificance, gained popularity and characterisesthe entire textile production from the beginning ofthe European Iron Age (Rast-Eicher, 1997: 547;Bazzanella

et al

., 2003: 8892). This new weavemeant that the fabrics appear more elastic andresistant (Sievers

et al

., 1991: 462). In addition, andmore relevant of the production of fabric for sails,was the possibility of realising orientated fabrics,in which the different pieces are united to createdischarge lines for the different elastic stains:centrifuge or diagonal according to necessity (Fig. 3).

The 2/2 weave allowed the creation of differenttypes of sails: suitable for stern bearing, in whichthe elastic diagonals create a sort of X, orsuitable for bow bearing, in which instead of theX, the discharge lines follow course with paralleldiagonal lines. That technological innovation wasof importance in the maritime field is demon-strated, for example, by the rapidity with whichstructural variations and substantial innovationswere introduced in boat and shipbuilding. As wehave seen, in fact, from the 4th millennium BCand for approximately 5000 years, sailing shipshad only one mast, initially positioned in thebow of the ship, which over time gradually movedback until reaching the appropriate central position(Rast-Eicher, 1997: 547). The first evidence ofthis new position of the mast appears in Egyptduring the New Dynasty (15501070 BC), and isamply documented in the relief of the templeof Rameses III in Medinet Habu (Casson, 1994:24, fig. 19). In the representation of the battlebetween the fleet of the Pharaoh and the SeaPeoples, dated to 1190 BC, all the ships have amast, with the sails furled in a central position.

From this moment onwards, the presence of asingle mast continues at least until the 7th to6th century BC, as demonstrated by numerous

representations of ships from different geographicaland cultural ambits (Casson, 1971: figs 1100;Doumas, 1999). It is only in the 7th century BCthat we see the first representation of a two-mastedship (Casson, 1971: fig. 98). Another example ofa double-masted merchant ship, of the 5thcentury BC, is painted on one of the walls of theTomb of the Tarquinia Ship, with a main mast,and a smaller one in the bow (Casson, 1971:fig. 97) (Fig. 4). The appearance of the secondmast toward the bow involved some substantialmodifications in navigation methods and in thetechniques of reduction of the sails, though itwould have improved the sailing potential indifficult wind conditions (Medas, 2000: 209ff).

In conclusion, it is suggested that a change inweaving technology at the end of the Late BronzeAge in the Mediterranean and Near East led toimproved sails and ships, as is demonstrated bythe introductions of brails in about 1190 BC.The chronological link between these structuralmodifications and the generalised increase inmerchant ships within the Mediterranean basin,demonstrated, among other things, by the fortunatediscovery of Uluburun and Cape Gelidoniya,may really have been a consequence of the newpossibilities of wind and therefore route exploit-ation. More than this, it is from this historicalmoment that the great marine civilisations set thefoundations for future colonial undertakings,intensifying not only trade, but also war contactsand piracy (Bondi, 1988; Hallager, 1988; Doumas,1996: 258). On the other hand, these sail changesset the foundations for a new way of ship-building, well demonstrated by the introductionof a foremast starting from the 6th

5th centuryBC, and even of the mobile mast in warships.

Francesco Tiboni

Via Posta 11, 25079 Vobarno (BS), Italy

Figure 3. Pieces of fabric can be joined together in different ways to give added strength. (Francesco Tiboni)


34

.1


References

Baines J. and Malek, J., 1985, Le imbarcazioni sul Nilo, in

Atlas of Ancient Egypt

. Novara.Bass G. F., 1974,

Navi e Civilt

. Milan.Bazzanella, M., 1999,

La Tessitura nel neolitico

. Udine.Bazzanella, M., Mayr, A., and Rast-Eicher A., 2003, I telai preistorici tra Neolitico ed et del Bronzo, in A. Rast-Eicher (ed.),

Textiles. Interlacement and fabrics from prehistory

. Trento.Bondi S. F., 1988, Problemi della precolonizzazione fenicia nel Mediterraneo centro occidentale, in S. Bondi (ed.),

Momentiprecoloniali nel Mediterraneo antico

. Roma.Casson L., 1971,

Ships and seamanship in the ancient world.

Baltimore and London.Casson, L., 1994,

Ships and seafaring in ancient times

. Austin, Texas.Dellamico P., 1997, Le attrezzature veliche,

Rivista Marittima

CXXX

.5.Dellamico P., 2002,

Costruzione navale antica

. Albenga.Doumas, C., 1996, La talassocrazia nel mar Egeo, in Pugliese Caratelli,

I Greci in Occidente

. Milan.Doumas, C., 1999,

The wall painting of Thera

. Athens.Fugazzola Delpino, M. A., 2001, Lo scavo subacqueo di un villaggio perilacustre del VI millennio a.C., in M. Giacobelli (ed.),

Lezioni Fabio Faccenna

. Bari.Grimal, N., 1988,

Storia dellantico Egitto

. Roma-Bari.Hallager, E., 1988, Aspects of Aegean long-distance trade in the second millenium BC, in S. Bondi (ed.),

Momenti precolonialinel Mediterraneo antico

. Rome.Homer,

Odyssey

.Johnstone, P., 1978, Alla ricerca delle navi scomparse, in P. Johnstone (ed.),

Civilt del passato

. Rome.Mark, S., 1998, The earliest Mast Step,

INA Quarterly

25

.

2

.Medas, S., 2000,

La Marineria Cartaginese

. Sassari.Rast-Eicher, A., 1997. Fabrics from the Bronze Age in Europe, in M. Cremaschi (ed.),

Le Terramare, the most ancientcivilisation of the Po

. Milan.Sievers, S., Pleiner, R., Venclova, N., and Geilengrugge, U., 1991, Lartigianato, in

I Celti

. Milan.

NOTE

Note

Figure 4. Ship painted on the wall of an Etruscan tomb at Tarquinia. (Francesco Tiboni)

NOTES


An Indo-Arabian Type of Stone Anchor from Kannur, Kerala, West Coast of India

T

he study of stone anchors can provideinformation on the size, date and pro-venance of the ships that carried them, as

they were provided by the boat- or ship-builder.In general terms the size of the ship can becorrelated with the anchor, but finds of stoneanchors from various sites both in India andabroad suggest that ships carried a number ofanchors of various sizes. This may have beenrelated to the type of journey, or expected weatherconditions. In storms or other unfavourableconditions sailors looked for a sheltered placeto anchor. The types of anchors depend on thenature of the seabed. Generally, anchors are foundin harbour sites, dockyards, shipwreck sites, andsheltered places along trade routes. Many exampleshave been found on rocky sea-beds and in shallowareas in Indian coastal waters (Sundaresh

et al

.,1999; Gaur

et al

., 2001; Sila Tripati

et al

., 2003).Over the years exploration and chance finds haveyielded 59 triangular anchors, 109 Indo-Arabianand 60 ringstone (Fig. 1) and these have helpedto reconstruct the history and development ofstone anchors of India.

There is also historical evidence for the use ofstone anchors. For instance, ancient Indian texts

such as the

Tilakamanjari

, the

Samaraiccakaha

and the

Milindapanho

mention the use of stoneanchors by Indian sailors (Gopal, 1999). However,the first foreign traveller to mention the useof stone anchors was Ludovico Varthema (AD14701510), who mentions that Calicut, on thecoast of Kerala, was a great shipbuilding centre.The shipwrights did not use oakum betweenplanks, but joined planks in such a way that waterdid not seep inside. Further, he mentions shipscarrying anchors made of a piece of marble eight

palmi

(spans) long and two

palmi

every other way.The said marble has two large ropes attachedto it and these are the anchors (Mookerji, 1912:203). According to Varthema Arab sailors usediron anchors on a particular occasion called

hadid

, while stone ones were in regular use, andwere referred to in the Persian glossary,

Miftah-ulfuzala

, compiled during the second half of the 15thcentury at Mandu (Malwa), which defines

langar

(anchor) as the stone of a

kishti

(boat) which isheavy. The use of iron anchors started in Indiain the 17th century under European influence(Qaisar, 1982).

But occasionally stone anchors are also reportedfrom inland sites, which may indicate re-use

Figure 1. Stone anchor find sites in India. (R. Uchil and S. Chitari)


34

.1


outside the maritime community. This paper isthe outcome of explorations carried out by theauthors in and around Kannur and adjoiningregions to trace the earliest archaeological remainsof maritime activity. One important find was anIndo-Arabian type of stone anchor, buried nearthe mosque close to the shore. Comparison withsimilar stone anchors from other parts of Indiaand beyond and its probable date are discussedbelow. The excavations at Kadakkarappally inAlappuzha of a medieval boat also recently yieldeda possible stone anchor (Nair

et al., 2004; Tomalinet al., 2004: fig. 12).

History of KannurLocal tradition states that during in antiquityKannur was known as the city of Maidans.The recorded history of Kannur dates back to theSangam Period (3rd century BC to 3rd centuryAD). Geographically, Kannur in the 1st to5th centuries AD extended from Mangalore inthe north to Badagara in the south, and duringlater periods the region was under the hegemonyof Haider Ali, Tipu Sultan, the Portuguese,the Dutch and lastly the British. Hemmedbetween the Western Ghats and the ArabianSea, the area is filled with forest, cash crops,coconut plants, spices, teak, rosewood, bamboo,and teak for the construction of ships. Thecoastline is interspersed with a number of rivers,lagoons and backwaters (Fig. 2), which haveprovided suitable bases for trading since ancienttimes. Marco Polo in the 13th century AD referredto Kannur as the emporium of the spice trade(Bhatt, 1997).

Maritime contact between Malabar and ArabiaThe Malabar region of Kerala has had trade andcultural contacts with Arabian and Mediterraneancountries since the earliest times. The findingof Indian teak logs in the temple of the Moon atMugheir and in the Palace of Nebuchadenezzarprovides evidence for contact between Malabarand the Mediterranean (Panikkar, 1997). Recentexcavations at Berenike and Myos Hormos inEgypt show archaeobotanical evidence of blackpepper, which might have been imported fromthe Malabar Coast of India (Tomber, 2002).The finding of large number of Roman coinsof Augustus, Tiberius, Claudius and Nero fromPudukkadu near Cannanore and Kottayam

in Travancore, Malabar, proves trade contactswith Rome over an extended period (Innes, 1951;Sathyamurthy, 1992), while a temple of Augustusat Muzuris in Malabar is evidence of a Romanmerchant settlement.

Ships were sailing between the Malabar Coastand the Red Sea from at least 120 BC if notearlier, first between the Mahrah coast andthe Indus delta, later direct from the mouth ofthe Red Sea to Malabar (Hourani, 1975). Shipsproceeding to India used to halt at Ocelis or Canefor replenishment, and it was 40 days sailing fromOcelis to Muzuris. Pliny mentions that archers

Figure 2. Location of stone anchor find in Kannur, Kerala.(R. Uchil and S. Chitari)

NOTES


were carried on board to protect ship and cargofrom pirates. Ships sailing from the Persian Gulfreached India by two routes. One was via Suhar,Masqat and ports on the coast of Oman wherethey took on water, then sailed straight across theIndian Ocean for Quilon in southern Malabar.The other route was coasting via Qays Island, oldHurmuz, al-Daybul, al-Mansurah and ports alongthe coast of al-Sind to the Gulf of Kutch andKathiawar. Malabar was an important centre forArabs because teakwood was the main sourcefor construction of houses and ships. Malabarsimportance increased in the medieval period as itlay on the trade route between Arabian countriesand China (Hourani, 1975). During this periodthe trade of Malabar passed into the hands ofMohammedan merchants. Sulaiman, the Arabtraveller, mentions that Chinese ships called atQuilon on their way to Arabian and Persian Gulfcountries during the 8th9th century AD (Panikkar,1997).

Indian goods were shipped directly to Yemenand from there to the Red Sea coast, Syria andthence to Europe, either directly or via Egypt.In ancient times the main items of export fromMalabar were pepper, spices, pearls and gems,while the main imports were coral, lead, tin, gold,wine and other finished articles. From literaryevidence it appears that trade contact betweenMalabar and Arabia was continuous from ancienttimes to the coming of the Portuguese (Mohamed,2001). The decline of Chola power created avacuum in overseas trade and commerce and thishelped the Arabs to reach the zenith of maritimecommerce (Sridharan, 1982). The relation ofZamorins of Calicut, Kerala with the Arab tradersbecame cordial in the 13th15th century AD andCalicut grew as an important centre of tradeof pepper and other spices. Subsequently Arabsmonopolised the trade on this coast. Pepper,cardamom and textiles from Malabar were ingreat demand and were exported from the port ofCalicut (Pearson, 1981).

The stone anchor from KannurThe Indo-Arabian stone anchor at Kannur islocated at the south-east corner of Hydross PalliMosque, near the beach. It is half-buried in theground (Fig. 3). The visible end has two squareholes for wooden pegs. The hawser hole, ifpresent, would be in the buried end. A thick coatof grey paint has been applied on its surface,so no marks are visible. The exposed portion is

890 mm long and 30 mm wide. The anchor is madeof granite, and the estimated weight of the exposedportion is 189.70 kg (1 m3 of granite weighs 2900 kg).The stone has been trimmed neatly by chiselleaving prominent straight lines all around on itssurface. A big piece has been chipped from oneside of the base, but there are no signs of wear. Itseems to have been embedded in the ground atthe same time as the construction of the mosque.

The date of the anchorNo Indo-Arabian anchors have so far been foundfrom before the 8th11th century AD. Twofragmentary examples from the stratified layersof Siraf on the Persian Gulf were datable to the8th11th century AD (Whitehouse, 1970). Indo-Arabian stone anchors with remains of woodenflukes were found in Galle harbour, Sri Lanka,and the wood radiocarbon dated to 430 80 yearsBP (Souter, 1998). On the basis of historicalevidences, Indo-Arabian stone anchors could

Figure 3. The Indo-Arabian type anchor from Kannur,Kerala. (Musthafa Mohamed)

NAUTICAL ARCHAEOLOGY, 34.1


be dated to the medieval period. The HydrossPalli mosque of Kannur has two lines of Arabicinscription on a wooden panel (Fig. 4). The secondline, translated, reads this mosque building wasrenovated in the month of Rabi-ul-Awwal of1126 of the Hijri calendar (about 300 years ago).Unfortunately the name of the donor is illegible.The mosque is claimed locally to have beenconstructed 300 years before its renovation. Fromthis, it is inferred that the anchor was embeddednear the mosque 600 years ago.

Another recent example of definite medievaldate was found in the excavation of a boat atKadakkarappally, in the Alappuzha district ofKerala (Tomalin et al., 2004: fig. 12). However,the identification is not definite as the stone doesnot have any hole to fasten a rope (Fig. 5).

DiscussionIn India maritime archaeology has not yet yieldedshipwrecks from the Arab period (7th14th centuriesAD), but direct evidence of trade contacts betweenIndia and Arabian countries is provided byIndo-Arabian stone anchors. More examplesare expected to be found in the future at or nearother Arab trade centres in India. Inscriptions,coins and other archaeological finds indicatemaritime contacts with other countries, butfinds of anchors provide firmer evidence for andinformation about visiting ships. Explorationsalong the coast of India have revealed triangular,Indo-Arabian, and ringstone anchors from Dwarka,Bet Dwarka, Somnath; triangular and Indo-Arabian stone anchors from Aramda (Sundareshet al., 1999; Gaur et al., 2001; Gaur et al., 2004)Sindhudurg Fort, Padmagad, Vijaydurg (Sila Tripatiand Gaur, 1997; Sila Tripati et al., 1998); and

Indo-Arabian types from Goa (Sila Tripati et al.,2003), Minicoy Island (Sila Tripati et al., 2004)and Tamil Nadu (Athiyaman and Jayakumar,2004). Recently, stone anchors have also beenreported from Puri on the Orissa coast (Fig. 1).None, however, has so far been reported from WestBengal, Andhra Pradesh, the Andaman andNicobar Islands and Karnataka. Furthermore,no wooden or lead stock anchors have so farbeen reported from Indian waters. The use ofiron anchors started in India under Europeaninfluence. But it is possible that the use of stoneanchors continued for some years. Stone anchorshave also been found re-used, as for example aslintels, paving, and mooring bits at Sindhudurg,Padmagad and Vijaydurg in India (17th centuryAD) and at Qalhat in Iran (10th16th centuryAD) (Vosmer, 1999).

None of the Indian stone anchors has beenfound in a stratified context or associated withany datable antiquities. All have been dated onthe basis of historical evidence and comparativestudies. The Indo-Arabian stone anchor fromKannur is the first example reported fromKerala, but it is also unstratified, and with noassociated finds. A lost stone anchor marks thepassage of a vessel: the Kannur anchor indicatesthat either an Arab ship visited Kannur or a shipusing the Indo-Arabian type of anchor inspiredby Arab mariners visited Kannur. As sailors beforesetting sail pray for a safe voyage, it is possible thatthe stone anchor may have a symbolic associationwith this coastal mosque.

One of the most important features of theKannur stone anchor is that the tool-marks bearsimilarities with the Indo-Arabian stone anchorfrom Grande Island, off Goa, and the anchorat the National Museum of Mogadishu (Fig. 6)

Figure 4. Arabic inscription on the wooden panel of the Hydross Palli mosque, Kannur, Kerala. (Musthafa Mohamed)

NOTES


Figure 5. Possible stone anchor from Kadakkarappally,Kerala

Figure 6. Stone anchors from (a) Grande Island off Goa;(b) Mogadishu Museum; (c) Kannur, Kerala. (R. Uchil andS. Chitari)

NAUTICAL ARCHAEOLOGY, 34.1


(Chittick, 1980). The stone anchors of GrandeIsland, Goa and Mogadishu do not have an upperhole for the hawser, and the Kannur anchor mayalso not have such a hole. These three anchorsmay perhaps have been manufactured by the samegroup, or even belong to the same owner. Furtherstudies on triangular and ringstone anchors wouldhelp to conclude whether these are indigenousor foreign introductions. However, triangularstone anchors have long been reported fromMediterranean waters, and ringstone anchorshave also been reported from Oman and Qalhat.

ConclusionIt is generally agreed that Arab mariners broughtthe Indo-Arabian type of stone anchor toIndia, and it was subsequently adopted by Indianmariners. This type of stone anchor was widelyused in the Indian Ocean region until the

introduction of iron anchor from Europe, thoughstone anchors almost certainly continued in usefor centuries afterwards. The find-spots ofIndo-Arabian stone anchor do seem to correlatewith known areas of trade and cultural contactsbetween the East African coast, Arabian Sea, RedSea and India. The finding of such a stone anchorat Kannur clearly validates the statement byLudovico Varthema that the people of Calicut usedstone anchors. On the basis of historical evidencethis anchor is datable to the medieval period, butonly excavation will yield more information.

Sila Tripati,Marine Archaeology Centre, National Institute of

Oceanography Dona Paula, Goa 403 004, India

Ali Manikfan and Musthafa MohamedExpress Housing Complex, Kallai Road, Calicut,

Kerala 673 002, India

AcknowledgementsThe authors express their deep gratitude to Dr Satish Shetye, Director of the National Institute of Oceanography and ShriK. H. Vora, Scientist-in-Charge, Marine Archaeology Centre, for their encouragement. We thank Dr V. Selvakumar forinformation about the stone anchor from Kadakkarappally, and Dr Hisham Nagi, Ebtisam Shamsan and M. S. Hussain fortranslating the Arabic inscription. We also acknowledge Shri Ravindra Uchil and S. B. Chitari for computer tracing andpreparing the figures. This is NIOs contribution No: 3930.

References

Athiyaman, N. and Jayakumar, P., 2004, Ancient anchors off Tamil Nadu coast and ship tonnage analysis, Current Science86.9: 12617.

Bhatt, S. C. (ed.), 1997, The Encyclopaedic District Gazetteers of India, 2. New Delhi.Chittick, H. N., 1980, Stone anchor shanks in the Western Indian Ocean, IJNA 9.1: 736.Gaur, A. S., Sundaresh, Sila Tripati, Gudigar, P., Vora, K. H., and Bandodker, S. N., 2001, A Group of 20 Stone anchors

from the waters of Dwarka on the Gujarat Coast, India, IJNA 30.1: 95108.Gaur, A. S., Sundaresh, and Sila Tripati, 2004, Grapnel stone anchors from Saurashtra: Remnants of Indo-Arab trade on the

Indian coast, Mariners Mirror 90.2: 13451.Gopal, L., 1999, Indian shipping in early medieval period, in K. S. Behera (ed.), Maritime Heritage of India, 90106. New Delhi.Hourani, G. F., 1975, Arab Seafaring in the Indian Ocean in ancient and early medieval times. New York.Innes, C. A., 1951, Malabar District Gazetteer 1. Madras.Mohamed, K. M., 2001, Arab relation with Malabar Coast from 9th to 16th centuries, The Malabar 1.1: 715.Mookerji, R. K., 1912, Indian Shipping. Bombay.Nair, M. V., Selvakumar, V., and Gopi, P. K., 2004, Excavation of a unique sailboat at Kadakkarapally, Kerala,

Current Science 86.5: 70912.Panikkar, K. M., 1997, Malabar and the Portuguese. New Delhi (reprint).Pearson, M. N., 1981, Coastal Western India. New Delhi.Puech, A., 1984, The Use of anchors in offshore petroleum operations. Houston.Qaisar, A. J., 1982, The Indian response to European Technology and Culture (AD 14981707). New Delhi.Sathyamurthy, J., 1992, Catalogue of Roman Gold Coins. Department of Archaeology, Government of Kerala, Thiruvananthapuram.Sila Tripati and Gaur, A. S., 1997, Stone anchors from Sindhudurg Fort on the west coast of India, IJNA 26.1: 517.Sila Tripati, Gaur, A. S., Sundaresh, and Bandodker, S. N., 1998, Historical period Stone anchors from Vijaydurg on the west

coast of India, Bulletin Australian Institute of Maritime Archaeology 22: 18.Sila Tripati, Gaur, A. S., and Sundaresh, 2003, Anchors from Goa waters, Central West Coast of India: Remains of Goas

overseas trade contacts with Arabian countries and Portugal, Bulletin of the Australasian Institute for Maritime Archaeology27: 97106.

Sila Tripati, Gaur, A. S., and Sundaresh, 2004, Marine Archaeology in India, Man and Environment 29.1: 2841.Souter, C., 1998, Stone anchors near Black Fort, Galle, Sri Lanka, IJNA 27.4: 33142.

NOTES


Sridharan, K., 1982, A Maritime History of India. New Delhi.Sundaresh, Gaur, A. S., Gudigar, P., Sila Tripati, Vora, K. H., and Bandodker, S. N., 1999, Stone anchors from the Okha-

mandala Region, Gujarat Coast, India, IJNA 28.3: 22952.Tomalin, V., Selvakumar, V., Nair, M. V., and Gopi, P. K., 2004, The Thaikkal-Kadakkarappally Boat: an Archaeological

Example of Medieval Shipbuilding in the Western Indian Ocean, IJNA 33.2: 25363.Tomber, R., 2002, Indian fine wares from the Red Sea coast of Egypt, Man and Environment 27.1: 2531.Vosmer, T., 1999, Indo-Arabian stone anchors in the western Indian Ocean and Arabian Sea, Arabian Archaeology and Epig-

raphy 10: 24863.Whitehouse, D., 1970, Excavations at Siraf. Third Interim Report. Iran 8: 117.

Building a Roman Pila

Documents

Transcript of Building a Roman Pila