The Hidden Museums of the Mediterranean

David Gibbins

This is an updated version of an article I wrote for New Scientist magazine in 1990, shortly after I had finished my PhD at Cambridge University. At the time, New Scientist sold more than half a million copies, making this my bestselling publication until my novel Atlantis fifteen years later! I was delighted when the editor of the European science annual Mannheimer Forum read this and asked me to write a much longer version for him; that eventually appeared as a 60-page chapter in the 1993 edition, translated into German (for the full reference, click here).

In revising this article, I have only updated factual matters and added a few essential developments. The article was written for a popular science readership who were neither archaeologists nor historians, and was copyedited in the house style. Most of the conclusions I drew then remain valid today. There have been further shipwreck excavations in the Mediterranean, some of major importance – for example, the 5th century BC wreck at Tektaş, off Turkey, which I helped to excavate in 1999-2000. That wreck now allows us to say more about classical Greek trade, and other excavations have revealed more about Roman trade in stone, for example, and Phoenician commerce. Each new wreck tells a rich and unique story in its own right, yet it is difficult to imagine new discoveries that could exceed some of the wrecks excavated in the 1980s – the fabulous Bronze Age wreck at Uluburun, for example, which you can read more about by clicking here.

One area of the Mediterranean that remains almost totally unexplored is the abyssal plain, at depths that can only be reached by submersibles and remote-operated vehicles. The average depth of the Mediterranean is over 1,500 metres, and the maximum recorded depth is a staggering 5,267 metres, in the Ionian Sea off Sicily. In several publications, including this one, I voiced doubts over whether these depths would produce the pristine wrecks that people imagined, or whether scientific excavation was possible. Some wonderful discoveries have now been made in abyssal depths, but those doubts remain. Far more exciting to me has been the increasing ability of divers to explore coastal waters at depths greater than 50 metres, down to about 100 metres, something I predicted would become the new frontier in underwater archaeology – yet the rapidity with which mixed gas and rebreather systems have become mainstream diving equipment took us all by surprise. This, and not abyssal exploration, could be the stepping stone to the great wreck excavations of the future.

In this article I described shipwrecks as ‘time capsules’, a term some archaeologists resist because time capsules contain material deliberately selected for future discovery, whereas wrecks almost all represent accidental loss. This is a valid point, but in other respects the term time capsule captures the essence and thrill of shipwrecks – slices through history that give an unparalleled glimpse of one instance in time, with all the technology, belongings and trade goods of people preserved with great immediacy, vivid and unique portholes into the ancient world. 

These two pictures of me were taken on the same exploratory dive off eastern Sicily, when we discovered two ancient shipwrecks. To the right, in more than 50 metres depth, I am holding the heavily encrusted top of a Byzantine amphora, dating from about the 7th century AD. To the left, I am raising the top of an amphora from one of the oldest sites discovered off Sicily - a ship carrying Corinthian Greek amphoras like this one, dating to the period when the Greeks were first colonising Sicily in the 7th and 6th century BC.  

Despite decades of plundering, the seabed of the Mediterranean remains the last great repository of archaeological knowledge in the ancient world. Now underwater archaeologists are revealing its secrets.

Shipwrecks are time capsules. A crew’s equipment, a passenger’s belongings, and the remains of cargoes from ancient merchant ships can give us a fleeting glimpse of ancient life and trade. Over the past fifty years, underwater archaeologists have uncovered fascinating details of Greek, Roman and even more ancient civilizations.

Mediterranean underwater archaeology really began with the invention of the aqualung in 1943. But in those early days, many land archaeologists were skeptical about its ability to meet their exacting standards. Then, in 1960, George Bass and a team from the University of Pennsylvania proved them wrong by meticulously excavating a Bronze Age wreck at Cape Gelidonya, off south-west Turkey. Now a staggering 100 or more pre-medieval wrecks are discovered every year, with the total running to well over two thousand.

Sports divers continue to discover many of them by chance. But the sheer number of sites means that less than half have been visited – often only briefly – by archaeologists, and just a small proportion of those wrecks are excavated (see Box 1, below). Another problem is that of clandestini, the Italian term for tomb and wreck robbers, who have plundered and destroyed many sites before they could be recorded.

The Mediterranean sea is the last great repository for undiscovered ancient works of art, especially bronzes. Even accessible, shallow waters continue to produce spectacular finds. In 1971, a spearfisherman discovered two 5th century BC lifesize bronze statues at Riace, off southern Italy; and a magnificent life-size bronze Poseidon – now in the Athens Archaeological Museum – was dredged from the sea years earlier near Boeotia in Greece.

Wrecks also provide remarkable evidence for ancient technology. Back in 1900, helmeted sponge divers salvaged the ‘Antikythera computer’ from a first century BC wreck off the southern Greek islands after which it is named. The device is 14 centimetres across, with 30 interconnected bronze-toothed cogs, and it is astonishingly sophisticated. Crafted with the precision of a clock, it was probably used as a navigational calculator.

The range of materials that can be preserved underwater is enormous. Organic remains, such as timbers, will survive if buried in a relatively anoxic environment where there are no borer worms (Teredo navalis)  or marine gribble (Limnoria lignorum). Outside the Mediterranean, the excavated wreck of the English warship Mary Rose, sunk and buried in the Spithead channel off Portsmouth in 1545, is a familiar example. Ancient hulls can be preserved extraordinarily well: archaeologists know of at least 500 wrecks in the Mediterranean with some timber remains.

Objects found on board are just as revealing. Delicate organic remains found recently include a small folding wooden writing board, from a 14th century BC wreck at Uluburun, Turkey; wooden calipers from a 6th century BC Etruscan wreck off Tuscany; and, from the same region, many dozens of sealed glass phials, their contents still intact, from an apothecary’s medicine chest discovered inside a Roman wreck of the 1st century AD. Amphoras, the large, two-handled pottery jars of antiquity, are often found stoppered with food remains preserved inside. Such preservation does not depend on a wreck being intact. In 1982 I was one of a team of divers who excavated a late Roman wreck only four metres deep off Randello, in southern Sicily, which held unbroken cylindrical amphoras over a metre high filled with the remains of pickled fish, buried in sand.

Two images of the incidence of ancient shipwrecks, from my chapter in Mannheimer Forum. The map of the Mediterranean shows incidence within degree squares of latitude and longitude; the histogram shows incidence through time, with numbers of wrecks on the right and years BC and AD below. The data are derived from a catalogue of known wrecks in the early 1990s, amounting to some 1200 sites; the figure is now larger, but the overall patterns remain. The map shows areas of Mediterranean coastline which have seen less exploration and reporting of wrecks, such as north Africa, and the chart reveals the huge proportion of wrecks that date from the Roman period.

Shipwrecks and ancient trade

But perhaps the most interesting artefacts are the cargo goods carried by the ancient merchant ships. Almost all wrecks are of merchant ships. Divers have excavated only one ancient warship, a 3rd century BC Punic ship grounded off western Sicily. Most warships wrecked in action would have perished as flotsam, but the weight and durability of cargo often meant that wrecked merchant ships survived; in some cases, the cargo pinned down and preserved the lower hull.

Merchant ships transported an immense variety and quantity of goods. The most common cargo is pottery amphoras, which are easily spotted by divers. Other goods were often carried alongside. A spectacular 11th century AD Arab cargo at Serçe Limani, off Turkey, contained glass as raw cutlets and as blown vessels, and Italian archaeologists excavating a 3rd century AD Roman wreck near Trieste, in north-east Italy, have found glass vases. Stone cargoes include columns, blocks, millstones and sarcophagi.

Wrecks also reveal trade in raw materials and consumables, such as unworked ivory tusk from the Bronze Age wreck at Uluburun, kernels of grain from a 6th century AD wreck off southern France, and even pig bones, the remnants of a consignment of pork, from a 3rd century AD wreck at Giglio Porto in Tuscany. Ships also transported several types of raw metal, from the 2nd millennium BC bronze and copper ingots of Uluburun and Cape Gelidonya to Roman cargoes of lead and iron.

More than two-thirds of all pre-medieval wrecks discovered in the Mediterranean so far date from between the 4th century BC and the 5th century AD. In the first half of this period the Romans rapidly expanded their control over much of Italy and Sicily, Gaul, Spain, north Africa, and the old Greek lands to the east. From the 1st to the 4th century AD, the Mediterranean was in many respects a unified market, free from piracy and warfare at sea, and long-distance trade in exotic goods, from eastern spices to African circus animals, was commonplace. In addition, Sicily, Spain, Egypt and north Africa provided staple foodstuffs for Rome’s million-strong population, for other large cities and for the frontier armies, while the emperor’s political security depended partly on handouts of grain, olive oil and other staples to the poor in Rome.

The largest number of wrecks have been found in the west Mediterranean - off Provence in France, Tuscany in Italy, Bonifacio in Corsica, the Balearic islands, the Aeolian islands, and south-east Sicily - all places where sport diving and archaeological research have been intensive. Elsewhere, exploration has been more sporadic, as in the Adriatic, or very limited, as off north Africa. Only a few dozen wrecks have been investigated along the entire coast between Morocco and Egypt. The same is true of the Black Sea, although more finds are now being reported from the Bulgarian and northern Turkish coasts. In the Aegean, spectacular finds have now been made in Greek waters, though the best-known discoveries have been made off south-west Turkey, where local sponge divers have led teams from the Institute of Nautical Archaeology – based at Bodrum and at Texas A&M University – to dozens of well-preserved wrecks. American and British teams have carried out similar investigations off Cyprus. Further east, another important concentration of sites is known off Israel.

The distribution of these wrecks is not just a function of modern exploration, but also tells us a lot about patterns of ancient shipping. Mediterranean seafaring for colonization and trade was widespread before 3000 BC; the distribution of obsidian, from the island of Melos – used for stone tools – and early pottery, reveals a flourishing Aegean network during the Neolithic. We have some idea of these early boats from pottery models and depictions. Most were probably paddled and lightly built, not much different from a modern dory or large canoe, so they would be unlikely to leave a visible wreck. We know of only one wreck dated with certainty before 2000 BC; it lay in shallow water off the Greek island of Dhokos, and contained many hundreds of small pottery vessels - from the late 3rd millennium BC - as well as a stone anchor. A similar anchor, perhaps of the early 2nd millennium BC, found on land in Lebanon, weighs as estimated 700 kilograms, so some early vessels at least must have been quite large.

After about 1500 BC –  the late Bronze Age – wrecks begin to give a more detailed picture of the civilizations they represent. Excavation of the Cape Gelidonya wreck off Turkey in 1960 revealed evidence for a ship that may have been about 10 metres long, dated by radiocarbon and by stylistic analysis of the pottery to the 13th century BC. Its main cargo was 34 four-handled copper ingots of distinctive ‘oxhide’ shape, each weighing about 25 kilograms. There was also residue from ingots of tin, which would have been mixed with copper to make bronze. Interspersed among these ingots were baskets of bronze tools, most broken before the wrecking and apparently carried as scrap. The ingots were probably cast in Cyprus, and the ship may have belonged to an itinerant Syro-Palestinian metal merchant.

In 1983 George Bass and his team were rewarded with a second Bronze Age wreck, the marvelously opulent site at nearby Uluburun. The wreck was about a century older than the one he had excavated more than two decades previously, at Cape Gelidonya. The Uluburun ship, too, appears to have been traveling from the east, perhaps under the control of a Canaanite rather than a Myceanean Greek merchant. Substantial sections of hull timber were preserved, and teams from the Institute of Nautical Archaeology under project director Cemal Pulak have excavated a dazzling cargo: Mycenaean, Cypriot and Near Eastern pottery; copper and tin ingots; bronze tools, including axes, adzes, drill bits and tongs; cylinder seals, with merchants’ markings; bronze swords, daggers and arrowheads; a unique wooden writing tablet; objects in amber, shell, bone and faience (a molten silicate similar to glass); gold jewelry; and a magnificent gold chalice. The wreck proves that seaborne transport prospered more than a thousand years before Graeco-Roman trade reached its peak.

The Bronze Age Cape Gelidonya and Uluburun wrecks mark the establishment of a Mediterranean nautical tradition. Balance-pan weights from Uluburun resemble those found on Byzantine ships almost 2000 years later, suggesting a similar type of commercial activity. The Uluburun ship carried a cargo of so-called Canaanite jars, forerunners of the ubiquitous amphora (see Box 2, below). The modest size of the two ships seems typical of later amphora-carrying ships, which appear in pictures propelled by a single adjustable square sail, and steered by double oars fixed at the stern. From their kitchen and table utensils, it seems that crews were often small, sometimes only four or five men.

About 1100 BC the Mycenaean Greek world collapsed, and the Aegean plunged into a dark age for which no wrecks are recorded. Then from the 8th to the 6th century BC, expanding Greek city-states such as Corinth began colonizing the rich lands of the western Mediterranean, alongside Phoenicians from the old Canaanite lands who went west to found Carthage and Cadiz. Early on, the colonizing Greeks established contacts with the remarkable Etruscan civilization of central Tyrrhenian Italy. Iron, not bronze, now fuelled the spread of civilization, and the Etruscans controlled the rich iron mines of Elba and the Tuscan coast; in return for iron, the Greeks supplied manufactured goods, including fine painted pottery.

The wreck that best illustrates this Etruscan link lay off the island of Giglio in the Tuscan archipelago. In 1984 I joined a team under Mensun Bound from Oxford University to help excavate the site, which had been extensively looted over the years by divers but still contained a rich diversity of artefacts. The finds, of the early 6th century BC, included a collection of Etruscan amphoras – some probably carrying pitch – as well as ingots of copper and lead. There were also wine and olive oil amphoras from the Greek island of Samos and from Asia Minor, modern Turkey, as well as painted finewares, including 20 or more ‘aryballoi,’ beautiful round-bodied vessels about the size of inkpots used for carrying unguents and perfumes. Many of these came originally from Corinth, reflecting that city’s commercial and political prominence.

Other objects found at the site approach those of the Bronze Age wreck of Uluburun in diversity and wealth, and suggest what may have been in the wreck before it was looted: a magnificent bronze helmet, and arrowheads; an elaborately carved wooden lid, and gaming pieces; small copper nuggets that may have been pre-coinage currency; and fragments of ornate inlaid furniture. Delicate wooden pan pipes recall the Etruscan love of wind music noted by Roman writers, and are another case of unique and unexpected finds in a shipwreck.

Shipwrecks are not the only evidence for ancient trade. The photo to the left is from Ostia, the port of Rome at the entrance to the river Tiber, and shows a mosaic in the 'Square of the Merchants'. Above the ship - which provides valuable evidence for ancient rigging, rarely found in wrecks - you can see the words NAVICULI KARTHAG, showing that this merchant represented trade from Carthage, in north Africa. The image to the right is a photograph I took on Monte Testacchio, 'Mountain of potsherds', in Rome. Monte Testacchio is one of the most extraordinary monuments of ancient Rome - a mound 30 metres high and half a kilometre in circumference made up entirely of smashed amphoras, discarded at the wharfside after their contents - mainly olive oil - had been offloaded from boats on the Tiber. The mound contains at least 50 million amphoras, dating mainly from the first two centuries AD, a staggering number that reveals the huge scale of Roman amphora production and trade, and the importance of olive oil in the ancient diet. Many more smashed amphoras were reused as temper in Roman concrete, and you can see them all over Rome if you look closely - in the concrete that made up the huge circuit of late Rome's walls, and in Hadrians's great Pantheon in the heart of the Imperial city.

Roman control of the seas

From the 4th century BC, the number of known wrecks increases sharply, reflecting Hellenistic Greek trade from the Aegean and a boom in amphora production in the Roman west. Parts of Sicily and Italy began producing wine for export, packaged in local ‘Graeco-Italic’ amphoras that mimic Greek ones; the scale of this transport is seen in the number of wrecks with these amphoras in the Aeolian Islands, north of Sicily. By the 1st century BC, flourishing Italian vineyards were exporting vast quantities of wine. Wreck evidence then tracks a reversal in this trade, as wine producers in southern Gaul and the Roman province of Tarraconensis – modern Catalonia – began to establish a place in the Italian market.

In the 1st century AD, the focus again shifts to the ancient southern Spanish provinces of Baetica and Lusitania. Their export of olive oil, fish produce and wine dominated the western Mediterranean amphora trade for two centuries. The scale of Baetican olive oil export was immense: In Rome, a fifty-metre high wharfside mound of discarded amphoras, Monte Testacchio – literally, 'mountain of potsherds' – contains at least fifty million amphoras of this period, representing at least six million tonnes of oil. And there are dozens of wrecks known with cargoes of Spanish amphoras, many of them off the Balearic Islands and in the Strait of Bonifacio, on the direct route from Spain to Rome.

By the early 3rd century AD this traffic was being ousted by export from the newly prosperous provinces of Proconsulis and Tripolitania in north Africa, modern Tunisia and Libya. At least one hundred wrecks are known with the distinctive African cylindrical amphoras, a type that was still being exported two centuries later when the Roman empire in the west had been fragmented by barbarian invasions and had largely ceased to exist.

Underwater archaeology fills in the details of this picture. Among the most fascinating wrecks of this period are the massive wine carriers of the later Roman Republic, from the century or so before the first Roman emperor Augustus established his rule in 27 BC.  A remarkably well-preserved example, at Madrague de Giens in southern France, was excavated over many seasons under the direction of Patrice Pomey and André Tchernia from the University of Aix-en-Provence. The ship was huge, of a size only commonly seen otherwise among stone carriers – at least 35 metres long and able to carry 400 tonnes of cargo, equivalent to at least 6,000 amphoras. Some ships of this trade were even larger: one salvaged in the 1950s near Albenga in Italy had as many as 10,000 amphoras on board. Amphora ships of this size were exceptional: more often then are about 16 to 20 metres long, with about a 70 tonne capacity, and even smaller vessels of 12 to 16 metres carrying about 200 amphoras are also quite common. The discovery of these smaller ships has added a new dimension to our understanding of ancient economics, as the few Roman writers who mentioned trade tended to focus on the larger, more ostentatious vessels; they would suggest a minimum size for seaborne trade considerably larger than the majority of excavated wrecks.

The immense size of the late Republican merchant ships reflects the wealth of the estate owners who produced the wine, and their willingness to invest heavily in the relatively safe, short route between central Tyrrhenian Italy and southern Gaul. Most of these wrecks date from before Julius Caesar’s conquest of Gaul in 55 BC, and the wine they carried was destined for the warrior tribes of central Gaul and Britain – where both the demand for alcohol and the profit margins were huge, with an amphora of wine trading for a slave. The wine merchant could also be the ship owner: at another wreck off south France, divers found the name of the merchant Sextus Arrius stamped on both the cargo amphoras and the lead anchor stocks. Shippers also traded in other goods. At the Madrague de Giens wreck, gaps between the wine amphoras, stacked four high, were filled with hundreds of attractive ‘Campanian’ black-glazed bowls and plates from southern Italy. This way of packing helps to explain how cheap, everyday items were dispersed widely around the Roman world.

The photo above shows the treacherous shore of Capo Murro di Porco off eastern Sicily, with the city of Syracuse in the background. Our expedition boat is visible just off the Cape, and the photo to the right shows two of us ascending from an exploratory dive. Below us in the blackness, the slope drops off to immense depths; the sea visible on the right of the picture above is over 3,000 metres deep. Syracuse was a hub of ancient trade, and many ancient wrecks must lie undiscovered in the waters visible in this picture, entombed in the pitch blackness of the abyss.

If we move on several centuries, and go south, wreck archaeology off Sicily gives us a detailed picture of amphora trade under the later Roman empire. The coast of Sicily between Gela on the south and the Strait of Messina to the north-east was a busy crossroads of ancient trade, where many ships sailing from Africa or Greece foundered on the dangerous rocky reefs, or were taken unawares by the unpredictable winds. A pioneer underwater archaeologist in the area was Dr Gerhard Kapitän, whose work laid the foundations for expeditions from the late 1960s from the Universities of Oxford and Bristol under Dr Toby Parker, the foremost British scholar of ancient wrecks, and for a further series of expeditions from 1983 under my direction, from the Universities of Bristol and Cambridge.

Several of these wrecks are of great interest. Around AD 200, a small merchant ship, probably sailing from Tunisia to Rome, sank in 30 metres off Plemmirio in south-east Sicily. The 200 or so cylindrical amphoras on board mark the beginning of large-scale export of amphoras from north Africa to Rome. Some were lined with resin, and probably contained fish produce; other, unlined one may have carried olive oil (see Box 2, below). Iron bars stacked among them were perhaps for sale speculatively as the ship passed through various ports, or were carried under special contract.

One end of the ship, including the galley, sank intact. Its contents give a vivid picture of life on board. The galley house was roofed with pottery tiles; inside was a brick-and-stone hearth, surrounded by pottery dishes, jugs, and bowls, a glass bottle, four pottery oil lamps, small amphoras – probably for storing water – and a stone mortarium, for grinding grain. Lead fishing weights suggest how the crew supplemented their diet. But perhaps most interesting was the discovery, unique on a shipwreck, of three bronze scalpel handles: they probably belonged to a passenger, who may have been a specialist eye surgeon. (To read more about these scalpels, click here).

A few kilometers away, in only five metres of water off Terrauzza, was a scattered wreck of about the same date containing distinctive high-handled Aegean amphoras. The resurgence of Greek wine export to the west coincided with a period of massive stone export, of columns and blocks quarried in Greece and Asia Minor for use in building projects in Rome and other western cities. The second to third century AD stone wrecks off Sicily – at Taormina, Marzamemi and Isola della Correnti – are incredibly impressive: one column we measured on the seabed at Marzamemi is estimated to have weighed more than 40 tonnes alone. Another, extraordinary wreck on the reef at Marzamemi contained the prefabricated stone facing of a 7th century AD Byzantine church; it was probably intended for one of the Byzantine communities of Sicily or north Africa, at a time when the eastern Roman empire, based at Constantinople – ancient Byzantium – had reconquered parts of the west from the Vandals and Goths, who had swept away the old Roman order in the 5th century AD.

Two scattered, shallow wrecks investigated by us off Sicily show the nature of amphora trade by the 4th century AD, when Rome was on the verge of dramatic decline. At Femmina Morta, a wreck of about AD 300 contained African amphoras of a greater variety of shapes than at Plemmirio, as well as Spanish amphoras. At this period, north African ports such as Carthage acted as transshipment centres where all manner of goods could be laden into one cargo; the Femmina Morta wreck also contained red-gloss fineware, and pottery ‘vaulting tubes’ used by the Romans as an ingenious way of strengthening and lightening concrete vaults and walls.

The second wreck, at Randello, only held amphoras from southern Spain. They were intact and full of sardine bones, the remains of fish in brine being transported whole – perhaps for processing into garum, a favourite sauce of the Romans made from the effluvia of decomposed fish entrails. There were extensive sardine shoals around Gibraltar, and the ruins of Roman fish-salting installations – the nearest we have to ancient ‘factories’ – survive on both sides of the Strait. Analysis of the bones showed that the fish were graded and sorted carefully before packaging.

From subtle variations in the shape of the Randello amphoras, we think that the fish-salters were supplied by small workshops employing two or three master-potters. Their cylindrical form copied the efficient shape of north African amphoras, such as those dating over a century earlier from the Plemmirio wreck, and reflects the increasing standardization of amphora shapes across the Mediterranean regardless of their origin or contents. Indeed, by the 5th century AD, amphoras may commonly have been transported empty, as a commodity in their own right, and were more frequently reused than in earlier periods when the scale of production was much greater. These features are consistent with the changing economics of the late Roman period; on the other hand, wreck evidence also suggests that entrepreneurial commercial activity thrived at all periods, regardless of political change. This has been one of the most important conclusions about the ancient economy revealed by shipwrecks.

Future exploration

There is no doubt that tens of thousands of ancient shipwrecks remain to be discovered in the Mediterranean and Black Seas, one of the greatest archaeological resources of the future. Many archaeologists are now looking to the well-preserved, undisturbed sites that must lie beyond the coastal strip which has seen the most intensive exploration, below the safe air-diving depth of about 50 metres. Ancient wrecks have been discovered in abyssal depths – on the ocean plains, which lie over 5,000 metres deep in parts of the Mediterranean – during pipeline laying and submersible testing, as well as through archaeological survey. One of the most successful projects has been in the Strait of Sicily, where a team from Woods Hole Oceanographic Institute under Dr Robert Ballard found a group of ancient wrecks at 800 metres depth, and other wrecks have recently been discovered in great depths in the east Mediterranean.

Exploration at abyssal depths is costly, and excavation is logistically difficult; moreover, few of the wrecks that have been found in abyssal depths have been as well-preserved as shallower wrecks, a result of greater dispersal of materials from open-hold ships as they sank. A more realistic focus for most archaeological teams may be the extended coastal seabed, from depths of about 50 to 100 metres. Small submersibles and remote-operated vehicles to explore these depths are now more readily available, and closed-circuit rebreathers and mixed-gas diving – substituting a less toxic gas for nitrogen (see Box 1, below) – allow divers to carry out scientific work at depths too dangerous for compressed-air diving. Neverthless, deep diving will always pose special hazards, and only the most important wrecks will be excavated – those containing great works of art, perhaps, or a long-sought shipwreck of the Minoan civilization of Bronze Age Crete, yet to be found.

Meanwhile, archaeologists have important work to do at sites in more easily accessible depths. Many known wrecks, some found decades ago, are incompletely characterized or unpublished, their finds languishing in museum stores or private collections. And many more wrecks remain to be discovered in shallow waters where sport and archaeological diving has been limited, for example along the coast of north Africa and in the Greek islands of the Aegean. As our research off Sicily shows, even scattered wrecks in shallow water, easy for small teams with modest budgets to investigate, have a great deal to offer. As they piece together the evidence, underwater archaeologists will continue to reveal not only spectacular objects, but also more fascinating insights into the world’s great ancient civilizations.

Two images of excavation underwater, on Roman wrecks off Sicily. At the left, a diver uses a dredge to clear sediment; he is not driving the dredge into the seabed like a vacuum-cleaner, which would suck up and destroy artefacts, but instead gently wafts sediment towards the nozzle with his hand. At this wreck, in only 4 metres of water off Femmina Morta, careful excavation paid off - you can see a hull timber emerging from the seabed beside the broken amphora neck, and buried just below we found a beautifully preserved wooden comb. To the right, divers excavate amphoras in almost 50 metres depth at Plemmirio, the colours almost completely filtered out at this depth. At the shallow site, divers could spend several hours excavating, whereas at the deeper site they could only spend 10 or 15 minutes on the seabed, no more than twice a day - a severe limitation of deep-water excavation.

1. Techniques of underwater archaeology

Archaeological excavations, whether on land or underwater, should be of comparable quality. Underwater archaeologist may therefore spend as much time surveying the site, collecting information about its surroundings, and recording artefacts before disturbing them, as their land colleagues. Archaeologists traditionally use either grid or open-area excavation techniques, and both can be applied in modified form underwater. In the former, they cover the site with a metal or plastic grid of measuring squares, which are excavated one by one; in the latter, the site may be covered by several baselines and then excavated over a large area, taken down layer by layer. The locations of finds are mapped by measuring distances from two or more datum points, for example stakes in the seabed, whose positions are fixed on the site plan.

Range sightings can be made as on land: a measuring pole is sighted through a device which calculates distance from the observed angle. A wreck, which results from on brief event, does not have the layers of artefacts through time seen on a typical land site, but they can be excavated and recorded in a similar way: seabed sediment falling over the wreck may stabilize into distinctive layers according to coarseness of grain, often covering a grey-black anoxic layer underneath. Sometimes, at an exceptionally well-preserved wreck, divers can excavate through layers of cargo, ballast and hull timbers still in their shipboard configuration.

‘Mark 1 eyeball’ remains the best way of discovered ancient wrecks in the Mediterranean, where a scatter of pottery may only be seen close-up and not be revealed by remote sensing. Wrecks which contain iron may be revealed by magnetometer survey, and upstanding hulls by sonar. Remote sensing has been most successful in discovered more recent wrecks outside the Mediterranean, for example in the Goodwin Sands off the Thames Estuary in England where many well-preserved wrecks lie completely buried.

Underwater archaeologists have had to develop specialized tools and modify old techniques. For example, to help them measure differences in depth, they may use a standard diver’s oil-filled depth gauge. A length of clear garden hose, strung from a datum point, adds precision: because of water pressure, an air bubble introduced at one end will rise to a constant level, from which the relative depths of various points around a site can be plumbed.

Another advantage of excavating underwater lies in the removal of sediment. The underwater equivalent of troweling – the mainstay land technique – is to waft away sediment by gentle hand movements, either downcurrent or into a suction tube that disperses material some distance away. The most familiar suction device is the airlift – an angled tube through which air pumped from a surface compressor rises, forming a vacuum. Dredges, where water is pumped backwards through a tube to provide suction, are used at sites too shallow for an effective airlift.

Many of the apparent diffulties of the underwater environment are easily dealt with. A Perspex board with replaceable sheets of coarse drafting film and a pencil replaces a clipboard, the land archaeologist’s other basic tool. Simple ideas and improvisation are often best: at a Byzantine wreck at Yassi Ada, off south-west Turkey, George Bass’s team discovered that the most effective way to hold exposed wood fragments in place was to pinion them into the seabed with bicycle spokes.

Artefact conservation is a special bugbear of underwater archaeology, just as in the excavation of inundated wetlands sites. Organic materials that are well-preserved underwater may become extremely fragile on exposure to air. Divers pack such finds into sand-filled trays and raise them from a site using lifting bags, ventable air balloons that can be controlled by the diver on ascent. Then they transfer the finds to a conservation lab, where they add stabilizing solutions to the fresh water in which the finds are soaked.

Hull timbers are often raised individually. Waterlogged wood poses special problems, as it is usually the largest organic find from a wreck and may shrink and disintegrate if left to dry. The most successful treatment – used first on the 17^th century AD warship Vasa, raised intact in 1962 from Stockholm harbour – is to saturate the timber with polyethylene glycol (PEG), which eventually replaces the cellular structure of the wood. But it is a time-consuming and costly process, and hull timbers not destined for display are often left in situ and reburied after they have been recorded and sampled.

Divers working on archaeological sites underwater have to contend with special physiological problems. Even at a shallow Mediterranean site in the height of summer, a fully wetsuited diver will eventually be debilitated by cold after a hour or two; below about 30 metres the water can be as cold as off Britain. For divers using compressed air, either from tanks (SCUBA – Self Contained Underwater Breathing Apparatus) or supplied by a ‘hookah’, from a surface compressor, underwater time is limited by the danger of a nitrogen ‘bend.’ As external pressure increases with depth – by about one atmosphere every ten metres – the bloodstream absorbs ever-larger quantities of nitrogen, which then expands when the diver ascends and must be allowed to dissipate. A ‘bend’, often crippling or fatal, occurs when a diver decompresses too quickly and a bubble of nitrogen forms in the bloodstream, large enough to lodge in a joint or the brain. The only treatment is immediate recompression in a pressurized chamber, and then gradual decompression under medical supervision. A diver using compressed air can only remain at 30 metres depth for about 20 minutes without having to undergo decompression stops on ascent, and below 50 metres the risks become too great. The toxicity of compressed nitrogen also has an anaesthetic effect, increasingly apparent below about 40 metres, which may leave a diver incapable of performing quite simple tasks.

Since the mid 1990s, the widespread availability and acceptance of mixed-gas diving, as well as the use of rebreathers, has extended the time divers can spend at shallow depths and increased the depth range of sport divers to below 50 metres. Mixed gas diving involves the substitution of a less toxic gas for nitrogen, and rebreathers allow extended bottom time by recycling expelled gas. Nevertheless, it remains the case that excavation at the deep end of the compressed air range, below about 40 metres, is only justifiable on a large scale if the site is exceptional and there is a recompression chamber on site.

Few institutions have the resources to mount projects on this scale, and to maintain the funding necessary for many seasons of excavation and post-excavation work. Many projects are thus on a smaller scale, for example conducting a shallow-water survey or limited excavation, and might involve a dozen or so divers operating from a shore base with inflatable boats and a compressor. However, providing that rigorous techniques of recording are maintained, the quality of data collected by a small-scale expedition can equal that of a more lavish project.

Amphoras from a 4th century AD shipwreck at Randello, south Sicily. When we found them, the amphoras were still filled with fish bones - the remains of a cargo of pickled sardines. I spent many hours measuring and drawing these amphoras, and as a result we were able to discern the hands of two or three master potters, probably working for a Roman fish-processing factory in south Spain. To the right, some experimental archaeology. I am often asked why amphoras were shaped that way, and this shows why: they were perfectly suited for carrying around, as well as for stowing in a ship's hold. Amphoras were among the most efficient containers ever invented, with a good volume to weight ratio, and made from a raw material so abundant that there was no need to clean out and reuse them - providing expert potters were available, fresh containers could be made for each new shipment.

2: Clues from the pottery of the past

Amphoras are among the most familiar of artefacts from the ancient Mediterranean – cheaply produced, disposable pottery receptacles manufactured in vast quantities for the transport of everything from wine, olive oil and fish sauce to nuts, meat, lime and pitch. They found their way to the furthest outposts of the Roman world, to Britain and to the trading colonies of southern India, but they are associated above all with seaborne trade in the Mediterranean. Their pointed bases meant that they could be stacked in interlocking layers, and also made carrying and pouring easier.

Amphoras varied considerable in size, typically holding from 20 to 60 litres, but were always easily manageable by one or two men. They are seen in their earliest form in the 2nd millennium BC, in the ‘Canaanite jars’ from the Uluburun Bronze Age wreck, and Byzantine and Arab merchants three thousand years later still used then – though by then less pottery was being made, and reusable wooden barrels were becoming more popular. Nevertheless, the olive jars of 16th to 18th century Spain, found as cargo on treasure ships of the Spanish Main, reflect a persistent tradition, and the two-handled shape survives today in north African and Egyptian water jars.

Amphoras can often be dated, if only broadly, and their place of origin identified. Because of this, they provide the nearest that archaeology has to offer for a statistics of trade, and their study has become a flourishing sub-discipline of archaeology. The first to put the study of amphoras on a scholarly footing was a 19th century German antiquarian in Rome, Heinrich Dressel, who nearly went blind in his fastidious recording of their epigraphic markings – impressed stamps and painted inscriptions that can indicate the place of origin, the name of the producer or shipper, and the contents.

Dressel’s typological guide has been greatly expanded and revised, but it remains a useful image of the range of Roman amphora shapes. He was the first to recognize that particular shapes might be tied to a region, date of manufacture, and sometimes specific contents; thus, for example, the long-handled Dressel 1 is an Italian wine amphora of the 2nd to 1st century BC, the globular Dressel 20 is a south Spanish olive oil amphora of the 1st to 3rd century AD, and the cylindrical Dressel 27 is a later form typical of north African production.

On land, amphora fragments  are often found in rubbish deposits of pottery discarded over many years near a wharf or a market. Because of the problems dating such deposits, a typology derived from them can be confused, whereas shipwrecks with coins or other closely datable material can provide exact dating for the amphora forms in the cargo, which are far more likely to be well-preserved with contents than on a land site. It is therefore no surprise that the advances made in amphora research have gone hand in hand with underwater archaeology. Characterisation techniques have also become more sophisticated; as well as details metrical study of nuances in shape, thin-section petrology, the microscopic study of pottery fabric, is now standard. The particular mixture of clay and its mineral inclusions may be typical of a manufacturing area, such as coastal Tunisia. Even more precision can be gained from neutron activation analysis, used to study trace elements in the clay, which can even tie an amphora to a particular kiln complex if comparative pottery samples are available. Another important area of scientific research, still being developed, is the study of amphora contents from minute residues, using gas and high-pressure liquid chromatography to identify olive oil lipids impregnated in sherds, for example.

----------------------------------------

This is an updated version of Gibbins, D.  1990. The hidden museums of the Mediterranean. New Scientist 1739: 35-40.

Text and all pictures copyright © 2006 D J L Gibbins

 TOP