Digital Library

São Bento Shipwreck (1554)

Chase Oswald, Filipe Castro


Country: South Africa
Place: Miskaba River
Coordinates: Lat. -31.3269; Long. 29.9692
Type: Nau
Identified: Yes
Dated: 1554 (Historical accounts)



S. Bento was a large ship built in Lisbon in 1551 for the India Route. It had a tonnage of around 900 tons, left Cochi in early February of 1554 and was lost on its first return voyage to Portugal, at the mouth of the Msikaba River.

Also known from Bernardo Gomes de Brito’s História Trágico-MarítimaRelação sumária da viagem que fez Fernão d’Alvares Cabral, desde que partio deste Reino por Capitão Mor da Armada que foi no ano de 1553 às partes da Índia até que se perdeu no cabo da Boa Esperança no ano de 1554, escrita por Manuel de Mesquita Perestrelo que se achou no dito naufrágio – the ship S. Bento was lost in 1554 during a violent storm off the coast of South Africa, with a load of pepper and the other typical merchandise brought from Asia to Portugal.

About 150 people died in the shipwreck. Once again like in the S. João story, the more than 300 persons that managed to make it to shore during the wrecking event were forced to walk to the mouth of the Maputo River. Only 62 people arrived in the Maputo River, two and a half months later.

The written account gave precious clues to the sport divers who found this shipwreck site in 1968, and recovered 18 bronze guns, and many artifacts from it.  Judging from the reports it seems that no hull remains were found.

Part of the collection of artifacts is in the Natal Museum and another part is in the Durban History Museum.


After leaving Cochin in February of 1554 the Indiaman São Bento wrecked upon the coast of Transkei on April 22nd, 1554, the following is based on the account of Manuel de Mesquita Perestrelo, one of the survivors of the São Bento (Duffy, 1955 27-28).

São Bento’s troubles began when the vessel was overtaken by a storm and its helm and rudder were damaged. At a latitude of 32 and 1/3° São Bento drifted toward the nearest land in sight, a sandy expanse of beach at the mouth of the Infante River. The ship grounded upon a rocky islet that lay a few hundred meters from the river’s mouth (Theal, 1898: 218). This seemingly fortunate landing allowed São Bento to narrowly avoid the rocky shore at their intended destination. However, the relief felt by the ship’s landing on this islet was short-lived, as waters around the island proved to be incredibly deep, reaching seven fathoms, a short distance offshore. Within moments of nearing the rocky island the ship struck bottom, splitting immediately, the lower portion of the hull stuck amongst the rocks while other portions broke off and washed ashore. The wreckage left the hulk of the ship level with the sea with only the ship’s castles being visible above the water (Theal, 1898: 218).

The sea was so violent that the people who sought refuge amongst the projecting portions of the ship found themselves partly underwater others attached themselves to nearby barrels and planks and attempted to swim ashore. Still others who were stranded on the sinking ship cut away the shrouds on the seaward side of the mast, allowing it to fall overboard. Crew and passengers attempted to use the mast as a bridge to shore. The efforts of this group were short-lived as a series of heavy waves struck the mast throwing everyone into the sea, all but one of which were drawn back by the waves until they were caught up in the sail, and drowned (Theal, 1898: 219).

The surface of the sea was now littered with boxes, barrels, lances, and a variety of ship parts and cargo. As the people tried to swim ashore, they found themselves battered by waves, rocks, and debris, and a great number of them drowned from injuries or exhaustion, turning the surf red with their blood. São Bento broke into two halves, with the forecastle comprising one portion and the stern the other. Those who could not swim sought shelter on the remaining pieces of the hull until the waves had cast them ashore (Theal, 1898: 220).

The survivors gave thanks to god for having endured the wrecking of São Bento, then gathered and treated the injured to the best of their abilities (Theal, 1898: 220). As soon as the tide receded these survivors waded from the rocky islet upon which they wrecked to the nearby mainland. A headcount determined that the number who perished included 44 Portuguese and over 106 slaves, leaving 322 survivors on the beach (Theal, 1898: 221; Duffy, 1955: 27).

The survivors retreated to the woods as the sun began to set, where they found a river which quenched their thirst for the first time since the wrecking. In the morning they searched the shore for clothes and supplies. The beach was covered with mangled corpses, appendages, and a variety of goods (Theal, 1898: 222). The abundance of materials that washed ashore supplied the survivors with food and goods which allowed them to regain strength and build shelters among the tree line. The survivors constructed a relatively comfortable refuge from of lavish carpets, gold cloth, and silk (Theal, 1898: 223). After this task was completed the captain commanded that the surrounding countryside be surveyed from a nearby mountain top, to discern any inhabitants and scout a safe path to cross the Infante River. Approximately a dozen men performed this task. While this scouting party was away a group of natives was seen overlooking the Portuguese encampment but they fled before any contact was made (Theal, 1898: 223).

The next day another group of natives was spotted on the other side of the Infante River, burning pieces of the ship to extract iron fastenings. A few of the natives swam across to greet the Portuguese and were given modest gifts of cloth, iron, and any spare provisions. The natives were friendly and talkative but none of the survivors understood their language, thus little knowledge was ascertained during this encounter (Theal, 1898: 224). The following afternoon approximately 100 more natives appeared nearby armed with wooden and iron spears. The Portuguese armed themselves in turn, but the native force proved peaceful. The Portuguese rejoiced to see that the leader of the native group wore red beads around his neck resembling cornelian beads. The natives’ possession of these beads meant that they must be near a river visited by merchant vessels, as this form of bead originated only from Cambay and was brought to Africa by Portuguese traders (Theal 1898: 225).

The following day the Portuguese ventured back to the shore in search of sustenance as they had been eating nothing more than coconuts. Not many provisions had washed ashore, and they found only a barrel of biscuits, 42 pounds of rice, and a few pieces of meat, all of which were soaked. In addition to the lack of supplies, the weather also tormented the Portuguese; it was now winter in the region and if they remained on the shore they would not survive (Theal, 1898: 225-226). After much debate, the survivors decided to trek north towards Sofala, following the route previously taken by Manuel de Sousa after the wreck of São João. To prepare for their journey the Portuguese gathered as many provisions and iron nails for trade that they could carry (Theal, 1898: 226).

The following day, April 27th, 98 Portuguese and 224 slaves set out on an interior journey towards Sofala. One-quarter of them were still seriously injured from the wreck, using sticks as crutches. They left a ship’s boy and slave at the original encampment, as both suffered from broken legs and were not expected to have survived the trek. The company was armed with lances, swords, shields, and one musket with 12 charges of water damaged powder (Theal, 1898: 227). Putting themselves in single line marching order, the ship’s crew carried a crucifix affixed upon a lance and a blessed banner in the front, and in the rear, a religious icon was carried by the captain along with the passengers and slaves. In the middle of the formation, unarmed members of the company helped to carry the sick and wounded (Theal, 1898: 227). After 12 days of traveling through the hills, mountains, and rivers of the region the company stumbled upon an abandoned village that contained pieces of china and other western artifacts that they were certain originated from the wreck of São João (Theal, 1898: 233-234). The following day, the company reached the spot where the galleon São João came ashore, whereupon they discovered the capstan and other timbers of the ship cast onto the rocky reef that stretched along the coast. Here the company realized that they were mistaken in attempting an interior route north and that by following the coast like the survivors of the São João wrecking they would have a greater chance of success. The coast had gentler terrain, less affected by the rivers of the interior, and provided resources such as oysters, which when gathered at low tide supplied the company with moderate sustenance (Theal, 1898: 234).

Months later, after enduring the grueling afflictions that accompanied treks along this treacherous coast, and with aid from friendly native guides, the remaining survivors found themselves in a native village (Theal, 1898: 227-280). They stayed here for five months, dealing with immense daily anguish, suffering from starvation, illness, and exposure as the winter storms had depleted the area of resources. However, on November 3rd, a Portuguese ivory merchant ship entered the bay in which the village lay (Theal 1898: 280-281). Nine days later the captain of this ship landed in a port owned by Inhaca, the king of the village in which the São Bento survivors took refuge, and upon the king’s instructions, the pilot of this ship was made aware of their presence. Three days after this the survivors were greeted with open arms by their fellow countrymen and taken into their vessel. In total, only 20 Portuguese and three slaves survived the trek to board the ship, the rest having perished along the way (Theal, 1898: 281). However, due to the harsh easterly winds, the Portuguese were not able to continue their journey home for the next five months, in the meantime, the merchant ship landed at various villages on the river attempting to trade with the locals, but was instead met with increasing violence (Theal, 1898: 282). The Portuguese raided one of the villages and captured several women, including the chief’s daughters; these captives were eventually returned to maintain peace with Tembe, the king of this land (Theal, 1898: 283). When it was time for the Portuguese ship to make its way back to Mozambique, the pilot received word that two more survivors of the São Bento were being aided by his people and were on their way. Two days later, a Portuguese man and a slave who had been separated from the company joined the ship (Theal, 1898: 283). On March 20th, 1555, the Portuguese set sail for Mozambique. Aside from temporarily running the ship aground on a sandbank, they experienced no more tragedies along their journey, coming to port on April 2nd (Theal, 1898: 284).


In 1968 several bronze cannons were discovered by G. N. Harris near a rocky islet 400 meters from the mouth of the Miskaba River on the Pondolond coast of Transkei. Previously, this area yielded sherds of Ming porcelain and carnelian beads, which were well known for washing up on the surrounding shore. This site was investigated by a research and salvage group from Kokstad, who eventually recovered additional cannons and artifacts from the sea around the island (Auret and Maggs, 1982: 1). The resulting report was eventually written up by archaeologists Chris Auret and Tim Maggs. Based on the archaeological evidence discussed below the wreck is clearly of the 16th century and with this information, researchers conducted a cross-historical analysis of potential ships to have wrecked in the vicinity. Only two Portuguese ships were known to have wrecked near this location during the 16th century, São João (1552) and São Bento (1554) (Auret and Maggs, 1982: 36). Using the account of survivor Manuel de Mesquita Perestrello,  translated by George McCall Theal from Bernardo Gomes de Brito’s, História Trágico-Marítima, the researchers compared the description of the landscape from the historical account to the modern scene (de Brito, 1735; Theal, 1898; Auret and Maggs, 1982: 36). With very little trouble, Auret and Maggs were able to draw clear parallels between the modern landscape and the one presented in the survivor’s account, thus identifying the Miskaba site as the location of the wreck of São Bento (fig 12) (Auret and Maggs, 1982: 37-38).


Since the initial discovery of the wreck 18 bronze cannons were recovered from the seabed around the islet. Auret and Maggs categorized the cannons into four types:  small breech-loading cannons, large breech-loading cannons, medium-sized muzzle-loading cannons, and a large muzzle-loading cannon (Auret and Maggs, 1982: 3).

Ten of the breech-loading cannons were classified within the smaller grouping and five placed into the larger group. These cannons, defined as falconets by N.V. dos Santos, of the Museu de Marinha, Lisbon, are described as having:

A primitive breech directly to the rear of the first reinforcement ring. In the breech there are three rectangular slots while the cascabel has a circular hole through the center. This was designed to take a simple form of butt used for training and elevating the weapon. The cannon itself was not mounted on a carriage but, by means of a wishbone structure, directly onto the gunwales, lookout platforms, and other vantage points of the ship. The breech-loaders were designed to take a separately charged chamber which may best be described as looking like a large beer tankard. This was inserted into the breech and locked into position by means of a tapered metal bar driven through the two rear slots. (Auret and Maggs, 1982: 3)


The description above identifies falconets as a type of swivel-gun similar to that of the English ‘murderers’ or the Spanish versos found on the 1554 Padre Island wrecks (Arnold and Weddle, 1978: 240-243). Manwayring’s dictionary defines murderers as:

A small iron or brass piece with chambers: In Merchant-men they are most used at the Bulk-heads of the forecastle, half-deck, or steerage; and they have a Pintell, which is put into a stock, and so they stand and are traversed, out of which they use Murdering-shot, to shower the Decks, when men enter, but iron Murderers ae dangerous for them which discharge them, for they will scale extremely, and endanger their eyes much with them, I have known divers hurt shooting them off. (Manwayring, 1644: 69)


While definitions of other similar swivel-guns are useful in determining classifications and giving a general description of the different types of swivel-guns recovered from wreck locations, it should be noted that swivel-guns can be vastly morphologically distinct from one another, such as falconets, versos, murderers, and espingardas. In R. D. Smith’s 1988 publication Towards a new typology for wrought iron ordnance, he defined different forms of swivel guns, which will be discussed briefly to highlight their (aside from basic differentiations between wrought-iron and bronze). As a base definition, Smith defines a swivel gun as “… any gun which can be maneuvered by means of a swivel. All swivel guns have separate chambers” (fig 13) (Smith, 1988: 7). In his later chapter titled Wrought-iron swivel guns (in Mensun Bound’s The Archaeology of Ships of war), Smith denotes the common names used in association with his original classifications and expands on additional typologies not included in his original publication with archaeological examples (Smith, 1995: 104-113). As such, Smith originally classifies swivel-guns into five major classes, SW1 through SW5, based on their general morphology (Smith, 1988: 8).

Smith defines SW1 as the most common type of swivel-gun, comprised of a long narrow barrel forged using stave and hoop construction, having longitudinally running staves that are bound together using alternating bands and hoops, varying from two to three hoops on the muzzle (Smith, 1988: 6, 8). The chamber holders on these swivel-guns are open and connected to the barrel of the swivel-gun via a pair of crossways lugs and pegs or hammered lugs laid out like rivets. From here, the support bar for the chamber is either a separate segment that is placed through openings on each side of the chamber holder or is welded across the bottom of the chamber holder. The wedge for the swivel-gun was then placed into holes created on the sides of the chamber holder. The trunnions are adjoined to a hoop in front of the chamber holder and are inserted into the eyes of the swivel (Smith, 1988: 8). The swivel-gun’s rear projecting aiming handle, known as a tiller, is typically constructed of a simple bar with an ornamental apex (Smith, 1988: 6, 8). It should be noted that the term chamber refers to an independent portion of the gun that is separate from the barrel into which powder was loaded (Smith, 1988: 7). SW1 is later classified in Smith’s 1995 publication as a ‘sling’, citing an example of this type of gun on the early 18th century ‘Bronze Bell’ wreck, in North Wales (Wignall, 1982: 228-229, as cited in Smith, 1995: 107).

SW2 is the second most common form of swivel-gun, according to Smith. This type is forged using a similar stave and hoop construction but with a flatter, shorter barrel than the previous type of swivel-gun. The chamber holders for SW2 are less open than in SW1, and thus lacks a support bar, and are instead welded to the barrel’s rear band. The holes for the wedges are formed into the sides of the chamber holder projecting just above the holder. The swivel-gun’s tiller is usually short and forged in an upturned fashion. Typically, one of the hoops on the barrels of these types of swivel-guns is decorated in some sort of fashion, such as in a twisted formation. Additionally, the tiller and chamber holder are also decorated in some form or another (Smith, 1988: 8, 11). SW2 is classified as a murderer in his 1995 publication, this type of gun was referenced earlier in a quote from Manwayring’s Seaman’s Dictionary and has been recovered from the 1609 Mauritius wreck (L’Hour et al., 1989: 102-104, as cited in Smith, 1995: 110).

SW3 is a rare form of swivel-gun with only a few known examples. The barrel is a single piece along with the chamber holder and trunnions, the rear section of which is removed to hold the chamber, concealed by a hinged cover. The tiller is a straight continuation of the barrel with a plain apex (Smith, 1988: 11). SW4 consists of a “one-piece” or “forged-gun’” barrel containing an ordinary muzzle band and chamber holder that is integrated into the barrel and is entirely encased (Smith, 1988: 11; Roth, 1989: 194). The chamber holders are described as having a tiny iron loop at their back end, in which a wedge chain was fastened, while the holes for the wedge are forged on either side of the chamber holder (Smith, 1988: 11). Wedge chains were used to keep hold of the wedge while it was not inserted into the swivel-gun (Smith, 1988: 6). The trunnions were connected to the barrel in front of the chamber holder while the tiller is described as a simple bar originating out of the back end of the chamber holder and ending in a triangular arc (Smith, 1988: 11). SW5 is defined as being like SW4 but differing in that the trunnions are connected on the forward portion of the chamber holder, instead of the barrel. The bottom of the chamber holder is also slit in SW5, which does not occur on the others (Smith, 1988: 11). SW4 and SW5 type swivel guns do not use hoop and stave construction techniques and are classified as ‘base’ in Smith’s later publication. Such examples have been found in a series of 16th century wrecks sites, such as on the Padre Island wrecks, Highborn Cay, Bahia Mujeres, and Molasses Reef sites (Arnold and Weddle, 1978: 240-243; Bass, 1988: 56-65, as cited in Smith, 1995: 106).

Two additional swivel guns were discussed in Smith’s 1995 publication: the serpentine, and the petriera a braga. The serpentine is described as a gun with a wrought-iron barrel fastened to a wooden base. The gun’s chamber is placed at the rear of the barrel and is secured by an iron wedge; the entire gun is placed on an iron swivel that is secured via a single perpendicular bolt. Smith also denotes that this is a fairly small gun with a bore measuring about five to 10 centimeters (Smith, 1995: 104). An example of this type of gun has been found on the 16th century Cattewater shipwreck (Redknap, 1984: 49-66, as cited in Smith, 1995: 105).

The petriera a braga is described as being a swivel-gun cast in bronze, including the trunnions, with two rear lugs to aid in securing the chamber holder. The chamber holder is rectangular and constructed of wrought iron, the front of the holder is made to be positioned around the rear of the barrel with two holes that fit over the bronze lugs on the rear barrel, the two pieces are then hammered together at this junction. The chamber holder is open on its rear and top and has a support strap welded to its bottom side. An iron tiller is also welded to the rear of the wrought iron chamber holder to aid in the gun’s aim. To secure the chamber in place a wedge is driven between the rear of the chamber and the rear inner side of the chamber holder (Smith, 1995: 108-109). Examples of this type of gun were found on the, 1588 Spanish Armada wreck Trinidad Valencera (Flanagan, 1988: 78; Martin and Parker, 1988: 220, as cited in Smith 1995: 109).

Chambers can be grouped into two separate typologies, those constructed using staves and hoops and those made from a single piece of metal. To differentiate between chamber types Smith created two major classifications for once piece chambers denoted as CH1 and CH2. CH1 is described as a one-piece chamber that has a slight taper from the rear to the front, with a stepped neck and a handle that has been welded on to the right side. When in use the touch hole was in a vertical position near the handle. CH2 is similar to CH1 except that the neck tapers away from the body and the handle is occasionally placed on the left-hand side (Smith, 1988: 12).

In their publication on naval guns Ian Hogg and John Batchelor display a table published in 1547 giving the estimated weight and caliber dimensions of a falconet: Weight 500 lb, Caliber 2.0 in, Shot Weight 2 lb, and Charge Weight 2.5 lb (Hogg and Batchelor, 1978: 12). The larger five cannons in this group from São Bento measure 264 cm in length with bores measuring 17 cm (Auret and Maggs, 1982: 4).

The falconets contained some evidence of foundry markings; however, because they had undergone a substantial amount of abrasion the markings could not be conclusively identified. Luckily, the markings on the larger cannons were still relatively discernable and were used to determine the markings found on the falconets by analogy (fig 14). The markings on both the smaller and larger falconets were uniform, bearing a crest near the muzzle which is trailed by an ‘armillary sphere’, which is then followed by a shield holding the monogrammed letters CFRO (fig 15). The armillary sphere was assumed as the mark of King Manuel of Portugal (1495-1521). The armillary sphere was also used by his successors, John III (1521-1557) and Sebastian (1557-1578), to a limited degree (the armillary sphere was not used alongside the royal coat of arms during these later rules) (Do Valle 1963, as cited in Auret and Maggs, 1982: 6). One of the explanations given for the CFRO monogram is that it is the marking of a foreign gun founder who made the cannons for several countries, including Portugal; in part because the royal Portuguese coat of arms is inaccurately depicted (Do Valle 1963, as cited in Auret and Maggs, 1982: 5-6).

Medium-Sized Muzzle-Loaders

Two identical medium-sized muzzle-loading cannons, were found on São Bento’s wreck site. Each measured 236 cm in length and a bore measuring 17 cm in diameter (fig 16). Both cannons had four lifting rings, which as Auret and Maggs point out, differs from later designs with only two lifting handles, often referred to as dolphins, located above the trunnions at the balancing point for the cannon (fig 17) (Auret and Maggs, 1982: 3). The trunnions were the two cylindrical projections from the barrel at the cannon’s balancing point, used for elevating the muzzle and mounting a cannon on a carriage (Hogg and Batchelor, 1978: 10). The cannons also have a single reinforce running from the front of the trunnions to the base. The base is described as being flat, and containing a loop styled in a ‘twisted rope’ motif, similar to the lugs holding the lifting rings (Auret and Maggs, 1982: 4). This flat base is without a cascabel, the rounded projecting knob cast into the base of a cannon (Auret and Maggs, 1982: 4; Mauncy, 1985: 83).

This group of cannons have crests similar to those found near the muzzles of the falconets, also accompanied by an armillary sphere; however, the third marking on these canons consists of a rectangle carrying the letters ‘OC’ (fig 18) (Auret and Maggs, 1982: 5). The OC marks belong to Portuguese cannon founder Sebastio Cobris, who cast cannons in Portugal for 50 years, beginning in 1514 (Do Valle, 1963, as cited in Auret and Maggs, 1982: 6). It should be noted that each group of cannon markings has a different version of the armillary sphere. This difference suggests that only the falconets were manufactured by the same source (Auret and Maggs, 1982: 6).

Both the medium and the large-sized muzzle-loaders contain chapelet holes, square holes in front of the base rings, these features are described below (Auret and Maggs, 1982: 6-7):

Until the early years of the 18th century cannon were commonly cast on a core: to form the bore, a rod was placed centrally in the hollow clay gun mould and after the molten bronze had been poured and had cooled, this rod was removed, leaving a central passage which was then cleaned up by means of revolving boring tools to form the final bore. The all-important central rod was secured in the breech end of the gun mould by means of an ‘armature’ made of iron and called a ‘crown-iron’ or ‘chapelet’. This had either 3 or 4 arms which were used to fix the rod firmly in the clay mould. The ends of these arms would reach nearly to the outer surface of the mould. When the molten bronze filled the mould, the rod could be removed, while the armature remained forever embedded in the breech end of the cannon. Prolonged immersion of a bronze cannon in sea water would corrode the iron away leaving as in this case, holes where the chapelet was. (Auret and Maggs, 1982: 6-7)

Large Muzzle-Loader

The large muzzle-loader measures 301 cm in length with a bore measuring 23 cm in diameter. Auret and Maggs note that this cannon differs from standard designs in two aspects. Like the medium-sized muzzle-loaders from the São Bento site, this cannon contains four lifting rings in place of the standard two dolphin handles; however, this latter cannon deviates from the former design in that the second reinforce is thicker than the initial reinforce and the chase, while also running from behind the after lifting rings past the trunnions (fig 19). Unlike the smaller cannons, the larger one does have a cascabel which although damaged does resemble a more traditional design (Auret and Maggs, 1982: 4-5). During the cleaning of this gun, a series of tightly packed materials were retrieved from its barrel including stones, 50 grams of peppercorns, money cowries, a piece of oak, two pieces of fine cane, three possible sheep rib fragments, two heavily chewed bone shards, and a portion of a coconut shell. This odd collection of objects suggest that this gun was stored in the hold when the ship sank (Auret and Maggs, 1982: 8, 11).

The large muzzle-loader exhibited only two marks, a crest followed by an armillary sphere similar to the ones found on the other cannons (Auret and Maggs, 1982: 5). This cannon, however, had one more possible mark on the second reinforcement. Auret and Maggs suggested that this cannon is a 16th century type referred to as ‘camel’, possibly derived from its humped shape. They noted that the unidentifiable mark on the second reinforce of the cannon looked like a camel (Auret and Maggs, 1982: 5). The morphological characteristics of the cannons and their markings allowed Auret and Maggs to definitively label this wreck site as 16th century Portuguese (Auret and Maggs, 1982: 7).

Cannon Balls

All the cannonballs recovered from the site came in two separate sizes, the smallest being 3.5 cm in diameter and the larger being approximately 5 cm in diameter, meaning they fit the falconets. The cannonballs were of cast iron and covered in a layer of lead, and also contained either a slot or square hole penetrating through the lead sheathing. These holes indicate that the balls were part of either bar shot or chain shot, in which, before being encased in lead, two iron balls were joined with one another by a bar or a chain. At the time of the original publication no other shot sizes fitting the muzzle-loading cannons had been recovered, to which Auret and Maggs assert may indicate that they were made completely of iron and have since corroded away (Auret and Maggs, 1982: 7). Another reasoning was also put forward by A. N. Kennard, in that the large muzzle-loading cannon may have used stone shot, to which some irregular stones had been recovered in the barrel of the gun during cleaning. However, Auret and Maggs claim that it is also possible that these stones were either a part of the ship’s ballast or native to the area (Auret and Maggs, 1982: 8-9).

True Porcelain

Porcelain made up a significant portion of the ship’s cargo, as a great quantity of it has been recovered from the site even after over four centuries since the wreck; using both museums and private collections Auret and Maggs made a wide-ranging comparison between porcelain from the wreck of São Bento and other well-known assemblages. J. Ayers of the Victoria and Albert Museum dated the porcelain recovered from the São Bento site to approximately 1530-1560. Researchers described the assemblage as consistent and probably representing cargo rather than personal belongings, as commonplace dishes and bowls appeared in large numbers. Because exported porcelain was likely acquired by Portuguese merchants within a few years of its manufacturing, and that the ship sank two months after leaving port in February of 1554, it is believed that the porcelain dates to somewhere between 1550 and 1554. Auret and Maggs support this argument by drawing attention to the ‘emperor marks’ found on the porcelain. While most of the porcelain pieces contained the ‘emperor marks’ of “Xuande” (1425-1435), the motifs of the porcelain, however, did not fit the designs of that period (fig 20). Auret and Maggs point out that using earlier ‘emperor marks’ such as the marks of “Xuande” on porcelain is a well-known practice of Chinese potters. A few other pieces contained the ‘emperor marks’ of “Zhengde” (1506-1521) (fig 21), while a single piece contained the mark of “Jiajing” (1522-1566) (fig 22). The inclusion of the Jiajing piece places the porcelain assembly after 1522, at the earliest. Additionally, the pieces are commonly depicted with a typical 16th century inscription, ‘long life, honor and riches’, usually in the form of a ‘cash coin’ (Auret and Maggs, 1982: 12).

Auret and Maggs denote the cobalt pigment used on the porcelain as striking while also varying in color. The porcelain is typically covered in a transparent glaze that applied to the body of the porcelain quite well, except for in a few fluted dishes where there was modest ‘crawling’ and ‘pinholing’ (Auret and Maggs,1982: 12). Crawling occurs during the firing process of ceramics when the glaze separates into clumps, leaving behind patches of clay between said clumps (Hansen, 2020). Pinholing, a common surface defect, occurs erratically within the production of ceramics, in which gas-producing particles are present within glaze during the firing process, causing dimples and tiny holes to appear on the surface of the glaze (Hansen, 2020).

Hemispherical bowls varying in size were frequently recovered from the São Bento, their insides contain centrally circular floral motifs and a ‘diamond diaper border’ known as a ‘trellis’ or ‘classical scroll’, just below the rim. On their outsides, these bowls contain a plethora of different designs, including landscapes with vegetation, floral scrolls, the Taoist trigram and crane motif, waves and rocks, phoenix, and chrysanthemum scroll, and others (Auret and Maggs, 1982: 12-13). One of the trigrams and crane motifs contained the inscription ‘riches, honor and enduring spring’, on the foot of the dish (Auret and Maggs, 1982: 15). Auret and Maggs denote that the trellis and Taoist motifs are distinctive of porcelain made during the Jiajing rule (Auret and Maggs, 1982: 12). A base inscription on one of the bowls denotes ‘made in the Great Ming period’, while another reads, ‘made in the Xuande period’ (Auret and Maggs, 1982: 14).

Another typical bowl in the São Bento assemblage is smaller than the previously mentioned bowl, albeit proportionally wider and containing an inverted rim. Typically, inside of this second type of bowl, a circular motif depicting a dragon was displayed towards the center, while the common, diamond diaper border or trellis was set below the rim. On the outside, most of these bowls contained motifs of a dragon and floral scroll (Auret and Maggs, 1982: 12). A few contained other exterior motifs such as a lotus and other floral scroll designs (Auret and Maggs, 1982: 18).

Another porcelain item commonly found in the São Bento collection is a type of small fluted dish containing a rim decorated in leaflike motifs or foliate. Auret and Maggs note that the flutes or grooves were press-molded, though some appear to have been cut into the porcelain as well. Most of these fluted dishes contain central motifs referred to as ‘parrot in fruit tree’ (fig 23), in which the fruit tree is believed to be a pomegranate tree; as with the other dishes, a few alternative decorations occur, which were not directly defined by Auret and Maggs (Auret and Maggs, 1982: 12). The inside rims of these dishes are typically stylized with either trellis patterns or what is believed to be a possible cherry motif. The outside rims of these dishes contain motifs of waves and fungus with the classical scroll design (Auret and Maggs, 1982: 16). Auret and Maggs note that complete versions of these dishes have been retrieved from the Islamic cemetery of Vohemar in northern Madagascar, which contained numerous examples of well-preserved Chinese porcelain, and whose coast had a history of Portuguese raiding during the 16th century (Vernier and Millot, 1971: plates 88 and 89, as cited in Auret and Maggs, 1982: 16; Molet, 1972: 149-150).  Some of the dishes recovered from São Bento are similar to these fluted dishes but are either without blue decorated underglaze and were fired with a non-transparent white glaze instead, or were decorated with blue underglaze but are missing fluting (Auret and Maggs, 1982: 16).

Several larger dishes contained foliate rims, which were theorized by Auret and Maggs to have been used to serve food.  The insides of these dishes contained the Taoist trigrams alongside clouds and decorative borders, while the outsides were decorated with the classic scroll border which was accompanied by waves, rocks, dragons, and fire motifs. Other artifacts of interest included an item referred to by the researchers as a “candlestick”, with a broken candle holder portion (Auret and Maggs, 1982: 18). This porcelain candlestick was decorated with the diamond diaper border or trellis, along with fish and water plants placed above rocks and waves (Auret and Maggs, 1982: 23). Vase fragments include the neck of a “bottle-shaped” vase, pieces referred to as “hexagonal forms”, one of which contained an inscription that read ‘made in the Great Ming period’, and a potential “mei-p’ing” vase neck (fig 24) (Auret and Maggs, 1982: 18, 24). Mei-p’ing vases are a distinct type of Chinese ceramic vase that are traditionally used to display the branches of plum blossoms; the vase is tall, originating from a narrow base that gradually spreads into a wider body, which abruptly transitions into a rounded shoulder that then converges into a narrow and short neck, containing a small opening (Welch, 2013: 17).

Variously shaped ceramic boxes were recovered from the wreck, including round boxes, boxes with bracket-shaped sides, and square-shaped boxes, the latter being the most common. The boxes had lids and stand on feet shaped like the heads of Lions of Fo or Dogs of Fo (fig 25) (Auret and Maggs, 1982: 18). Lion of Fo, also known as Dog of fo, Lion of Buddha, or Shishi, is a decorative figure typically used as a guardian symbol for Buddhist temples and is found on a variety of Chinese pottery. The Lion of Fo is usually depicted as a snarling lion and accompanied by a mate, the male traditionally holds an orb, while the female is depicted with a cub (Encyclopædia Britannica, 2015). Auret and Maggs note that these legs were cast separately from their respective boxes and were attached after bevels had been cut into the boxes. The square boxes were decorated with lotus scrolls or floral motifs on their sides and had lids decorated with floral scrolls. Circular boxes have floral motifs on their sides and contain lids with repeating border patterns (Auret and Maggs, 1982: 18, 24). Lastly, a few pieces of ‘spouted ewers’ (jugs) were also recovered from the São Bento site (Auret and Maggs, 1982: 18).

Coarse Porcelain

According to J. Ayers of the Victoria and Albert Museum, the coarse porcelain among the collection represents a regional product most likely from southern China, an idea which coincides with claims made by E. Vernier and J. Millot of the porcelain from the Vohemar site (Vernier and Millot, 1971: 83, as cited in  Molet, 1972: 149-150). The coarse porcelain is neither true porcelain nor translucent, but instead is a ‘porcellanous’ form of stoneware, according to C. Woodward. The coarse porcelain is described as having a porous body and ranging in color from white to a light brownish yellow, the underglaze varied from black to an ‘inky’ blue. The glaze was denoted as semi-transparent with a ‘pale grey-green’ to brown coloring. Auret and Maggs point out that many of these pieces suffer from pinholing and crazing (Auret and Maggs, 1982: 18). Crazing refers to a flaw in ceramic glazing where small hairline cracks appear in the glazed surface and are caused by a discrepancy in thermal expansion in the body of the ceramic and the glaze. Crazing typically appears directly after firing, but can also appear many years later depending on the glaze (Hansen 2020).

The coarse porcelain is also much softer than the true porcelain, it therefore suffers from chemical weathering and has the proclivity to be broken into tinier pieces. The majority of the coarse porcelain within the São Bento collection is made up of fluted dishes with foliate rims, similar to the ones found among the true porcelain collection. These dishes also have motifs akin to those on the true porcelain versions, containing similar scrolls and diamond diaper border decorations. Although these pieces do contain their own unique style of border motif consisting of a row of nesting arcs, occasionally broken up with a crosshatch design at certain intervals (Auret and Maggs, 1982: 18). Because most of these porcelain shards were too broken up to have their central motifs readily identifiable Auret and Maggs compared them to examples from the Vohemar cemetery in Madagascar. Of the central motifs, the most frequently depicted the ‘ch’i-lin’ (Qilin), surrounded by fire forms (Auret and Maggs, 1982: 30). The Qilin is a mythical creature known in Chinese and other East Asian cultures. It is often depicted as a hooved chimera, sporting a single horn on its forehead, a deer body, an oxtail, and a series of dragon-like features, such as manes, beards, scales, and whiskers (Encyclopædia Britannica, 2020). The best example of a Qilin among the coarse porcelain in the collection contained the head, a fire form, part of the mane, and a shaggy tail. Most of the other sherds only depicted the tails and the front hoof. Another identifiable motif among the assemblage was of two lions with ribboned balls. There were other examples among the coarse porcelain persisting of both larger and smaller bowls, however, none of these pieces were complete enough to allow for a reconstruction (Auret and Maggs, 1982: 30).


Only one piece of earthenware from the São Bento collection contained decorations. This piece is believed to be of provincial Chinese origin and is described as a large dish having an indented pattern on its rim and a ‘sgraffito’ motif on its inside (Auret and Maggs, 1982: 30). Sgraffito refers to a decoration technique found in either pottery, produced by applying two layers of contrasting glaze or slip to an unfired ceramic body, or on wall décor, by applying layers of contrastingly colored plaster to a moist surface, and then, in either case, scratching the surface to reveal the underlying layer(s) (Weyer et al., 2015: 102). The glaze of this piece is a transparent olive-green color over a grey-buff colored body (Auret and Magss, 1982: 30). Other earthenware pieces from the collection primarily consist of jars, mainly made up of large containers with four strap-like handles and ‘rolled’ rims. Most of these jars are unglazed, aside from a few with a flakey dark greenish-brown glaze and are believed to most likely be of Chinese origin (Auret and Maggs, 1982: 36).

Gold Rings

14 rings of two different varieties were recovered from the area around the São Bento wreck site, one of the rings fits into a more basic category while the other 13 rings fit into a more intricate grouping. According to Auret and Maggs, the basic ring holds a ‘cabochon’ ruby while the 13 intricate rings contain a “U-shaped panel or ‘quatrefoil bezel’ (fig 26) on each of the four sides of the ring setting”; while this first ring holds a simple “table-cut” ruby, the other rings lack their gemstones or instead contain “table-cut” diamonds (Godfrey, 1969: 109, as cited in Auret and Maggs, 1982: 8, 10). Before the invention of gemstone cutting, cabochon was the default method in preparing gemstones. Cabochon refers to a gemstone that has been shaped and polished instead of faceted, resulting in a rounded obverse form with a flat reverse (Bonewitz, 2005). A quatrefoil refers to a decorative component that forms a symmetrical outline of four partially overlapping and equal circles, resembling a four-leafed clover, and is traditionally used in Christian symbolism and believed to symbolize something in fours (Rest, 1956: 36). These 13 complex rings were identified by A. Somer-Cox of the Victoria and Albert Museum as being either common “early or mid-sixteenth century Renaissance jewelry”; with the simpler example potentially being crafted as early as the 14th century (Auret and Maggs, 1982: 8). Auret and Maggs point out that these intricate rings were likely been coveted by the European elite of the mid-16th century, as images of the rings appear in several portraits from this period; including that of Lady Jane Grey by Master John, which has recently been reidentified to be a painting of Katherine Parr (approx. 1545), the English Court and Henry VIII by Hans Holbein (1533-1543), Dr. Léon Braunsberg by Cranach (1472-1553), and Steven van Herwijck by Mor (1519-1576) (Auret and Maggs, 1982: 8-9; James 2018). Auret and Maggs also point out that such rings do not show up in later portraits meaning they most like went out of fashion toward the end of the 16th century (Auret and Maggs, 1982: 9).

Because the intricately styled rings were standardized and regularly recovered from Miskaba near the wreck site, Auret and Maggs theorize that they were part of the ship’s cargo rather than personal effects. Additionally, all of the rubies from the São Bento were analyzed for their zircon impurities and found to have originated from Sri Lanka. Because India was the sole source of diamonds during this period, they most likely have a similar origin as well. This evidence indicates that jewelry makers in India were crafting rings intended for exportation in the European market (Auret and Maggs, 1982: 9).


Two gold pieces were identified by Skelton of the Victoria and Albert Museum, as the lower portion of an Indian “filagree” style earring, known as Jhumka. Auret and Maggs describe the ear-rings as “two hemispherical pieces worked in gold… The larger one is set with small cabochon rubies, many of poor quality; the smaller is made up of twisted gold threads” (fig 27) (Auret and Maggs, 1982: 9). The researchers point out that while this style of earring is primarily used in the 18th century, it is possible that such earrings were also found in the 16th century (Auret and Maggs, 1982: 9). Filigree refers to a method in jewelry crafting in which wires are soldered into patterns; the wires can also be organized into braids, twist, or singularly, they may also be either beaded or plain. The most common filigree patterns consist of straight lines, spirals, and circles (Higgins and Higgins, 1980: 21).


The intaglio is an anonymous bust or portrait engraved into a piece of carnelian. This artifact was discovered as early as 1934 and had since been set into a ring. Intaglios were a popular trinket during the Renaissance period (Auret and Maggs, 1982: 9-10). The term intaglio is often used to refer to an engraved gem, typically these contain depictions of portraits or important symbols, and were found throughout the ancient world (Middleton, 1891: 26).

Gold Coin (Cruzado)

A single gold cruzado was recovered from the São Bento site, found stuck between two rocks on the island near shallow water. The coin displays the Royal Portuguese coat of arms encompassed by the legend ‘I0A 111 POR ET AL R’, which according to Auret and Maggs refers to “John III King of Portugal and the Algarve”, who ruled from 1521 to 1557. Engraved on the reverse of the coin is a cross with the phrase ‘In hoc signo vincit’. This coin places the date of the wreck no earlier than 1521 (Auret, 1977: 231-235, as cited in Auret and Maggs, 1982: 9, 11).

Carnelian Beads

Carnelian beads from Cambay, India were among the most frequently recovered artifacts from the São Bento site. According to Auret and Maggs these beads were traded to colonial areas (such as West Africa), by several European maritime nations, including the Portuguese. Most of the carnelian beads from the São Bento site are shaped like lozenges, however, other examples have been recovered as well, including a faceted biconically shaped bead, an elongated hexagonally shaped bead, and a short cylindrically shaped bead measuring 7 mm long and 12 mm in diameter (Auret and Maggs, 1982: 9). As pointed out in the São João section, carnelian beads, also referred to as trade-wind beads, are indicators of a homeward-bound shipwreck, as they were only present on ships returning to Europe from India (Bell-Cross, 1987: 22; Boxer 1959).


Aside from the carnelian beads, there was a broken small faceted garnet measuring 4 mm wide and 7 mm long (Auret and Maggs, 1982: 10-11).

Money Cowries

An estimated 18.7 kg of money cowries were recovered from the barrel of the large muzzle-loading cannon, as mentioned previously. Auret and Maggs note that the cowries were packed incredibly tightly into the muzzle of the canon, which means that they had been purposely stored inside the canon, further alluding to the idea that the canon must have been stored away in the hull of the ship at the time of the wreck (Auret and Maggs, 1982: 11). As originally referenced by Auret and Maggs, money cowries originate from tropical areas in the Indian Ocean and were primarily exported to West Africa from the Maldive Islands, a trade which existed for centuries before Portuguese colonization. Money cowries were used in West Africa as a popular form of currency, hence their name; however, after the Portuguese began trading them with West Africa their value depreciated significantly (Mauny, 1961, as cited in Auret and Maggs, 1982: 11).

Size and scantlings

This ship is referred with a tonnage of 900 tons.


Brito, Bernardo Gomes de – História Trágico-Marítima, 2 volumes, Lisboa, Coleção Livros de Bolso nº. 275, Publicações Europa-América.

Kennedy, R.F. – Shipwrecks on and off the Coasts of Southern Africa, Johannesburg Public Library, Johannesburg, 1955.

Maggs, Tim e Auret, C. – The Great Ship S.Bento – remains from a mid-sixteenth century Portuguese wreck on the  Pondoland coast, in “Annals Natal Museum”, vol. 25 (1), Pietermaritzburg, October 1982, p.1-39.

Santos, Nuno Valdez dos – A Artilharia Naval e os canhões do Galeão Santiago, Lisboa, Academia da Marinha, 1986.

Warren, Blake & Green, Jeremy – A Mid-XVI century Portuguese wreck in the Seychelles, IJNA, 15.1, 1986, p 1-23.

Valle, Henrique Pereira do – Nomenclatura dos Bocas de Fogo Portuguesas do Século XVI, in, Revista de Artilharia nº439-440, Ano LVIII, 2ª Série, Lisboa, Março-Abril de 1962, p.381-390.