Nossa Senhora da Atalaia do Pinheiro, 1647

Country: South Africa

Place: Near the mouth of the Cefane River, Eastern Cape coast, South Africa

Type: Nau

Identified: Yes

Dated: 1647 (historical accounts)

History of the shipwreck

The following description of the Atalaia wreck site and its finds comes from personal communications between Professor C.R. Boxer and Graham Bell-Cross, which were originally published in Bell-Cross's chapter "Portuguese Shipwrecks and Identification of Their Sites." Mentioned as early as 1640, the full name of this vessel was Nossa Senhora da Atalaia do Pinheiro, referring to the sacred image of the Virgin Mary in a church in the "Pinhal of Santa Maria da Castelo de Torres Vedras." Although some texts refer to the vessel as a galleon, Bell-Cross's publication assures readers that Nossa Senhora da Atalaia do Pinheiro was a four-deck carrack of approximately 1,000 tons and most likely mounted about 30 guns, as most Indiamen on the Portuguese India Route were typically minimally equipped (Bell-Cross, 1988: 74).

After the discovery of the Atalaia's consort Santíssimo Sacramento in 1977, researchers decided to initiate a survey in 1978 to uncover the location of the contemporary wreck. Bell-Cross notes that available data at the time indicated that Portugal-bound ships from the East carried cargoes of blue-and-white porcelain. By locating deposits of said porcelain it was possible to establish the locations of Portuguese wrecks along the eastern South African coast. Researchers began an extended survey of beaches on the south-east coast of Africa, from the Mozambique border to Algoa Bay at the eastern Cape of South Africa. This included interviewing locals for archaeological evidence such as porcelain deposits and surveying already-known porcelain sites. The survey also promoted an intensive focus on inspecting the coast between Cape Padrone and the Kei River, with identification of the Atalaia wreck site as the main objective (Bell-Cross, 1988: 74-75). At the time of the survey's conception, a single site containing blue-and-white porcelain was known to be at Ivy Bay, Port Edward, on the Natal coast. Additionally, one porcelain site was also known on the Transkei coast at the mouth of the Misakaba river. Researchers were also aware of five sites on the eastern Cape coast at Cannon Rocks, Mtana River, Bonza Bay, Sunrise-on-Sea, and Double Mouth. It was assumed that most porcelain could be dated to within 20 years of its manufacture, and, with the aid of artifacts recovered from surveys and geographic descriptions found within the historical accounts of the wrecks, identification of the sites would be possible (Bell-Cross, 1988: 75).

The survey originally centered on the coast around 33 1/3° South latitude, which included the known porcelain site by the Mtana River, approximately located between the Keiskamma and Fish Rivers; this was based on a survivor's account of the area. However, nothing was found, and researchers soon focused on the entirety of the coast from Algoa Bay to the Kei River. In 1979 Bell-Cross found a large (12 by 12 cm) sherd of "martevan" located between the mouths of the Cinsta and Cefane Rivers (Bell-Cross, 1988: 75). Martevan, or "martaban" as it is more widely known, refers to large ceramic storage jars used in south-east Asia and originating from Martaban (Mottama) in Lower Burma. Martaban ware was commonly used during the 16th and 17th centuries and was primarily exported from Burma to Sri Lanka and India (Borell, 2014: 257; Castro, 2001: 16). Later investigations at this site provided additional pieces of martaban and finds of blue-and-white porcelain. The inclusion of this site meant there were six known porcelain sites between Algoa Bay and the Kei River (Bell-Cross, 1988: 75).

The key to determining which of these porcelain sites was that of Atalaia was found in Feyo's description of the coast seen just before the ship wrecked. Specifically, Feyo noted that the ship anchored in seven fathoms (27 m) off a long sandy beach. This description meant that only two known porcelain sites could potentially be that of Atalaia: the Mtana and the Cefane River sites. In January 1980, divers involved with the survey project, Peter Sachs and Sean Mitchely, dived at the Cefane River site. Searching in a grid pattern, the duo first came upon a muzzle-loading cannon, followed by a cluster of 10 bronze and eight iron cannons (fig. 33). The bronze cannons possessed cascabels shaped in the same fist-with-the-thumb-projecting-from-under-the-index-and-middle-fingers style found on the Bocarro cannons from Santíssimo Sacramento. The cannons were arranged in a way that suggested they had been stored in the hull at the time of the wrecking, as was the case with the cannons recovered from Santíssimo Sacramento. A later examination of an inscription on one of the cannon barrels established that it was a Bocarro cannon, of which some were known to have been transported aboard Atalaia (Bell-Cross, 1988: 76).

After discovering the cannons, the divers obtained a salvage permit and made arrangements to raise them. For the salvaging operations, the team attached buoys to the Bocarro cannons and plotted their positions, then hired the same trawler that was used to lift the cannons of Santíssimo Sacramento. The team attached strops around the cannons, which were then connected to the five-ton lift winch of the trawler. By the end of the first week, 10 of the Bocarro cannons were raised and sent to the East London Museum. Recovery of the thirteen cannons closer to shore took the team more than eighteen months, even though they weighed significantly less than the Bocarro cannons (Bell-Cross, 1988: 76). Raising the cannons closer to shore was more dangerous; lifting bags were initially attached to them and they were repositioned in deeper water and later raised by the trawler. In addition to the 10 Bocarro cannons, the team also recovered sheets of lead, sherds of blue-and-white porcelain, glazed and unglazed martaban sherds, lead shot for muskets and pistols, two anchors, a crushed pewter jar, a lead sounding weight, and thirteen other cannons, one of which had the same A.G. Feyo founder's mark as a cannon found on Santíssimo Sacramento. One large iron cannon was also lifted. Notably, one of the bronze cannons had the date 1638 cast in relief on its upper surface, near the touchhole (Bell-Cross, 1988: 77-78). Wood most likely belonging to the hull was recovered from underneath one of the Bocarro cannons and later identified as evergreen oak, confirming that the ship was built in Portugal (Bell-Cross, 1988: 74).

Additionally, two bronze coaks weighing approximately 18 kilograms apiece were recovered from the site (fig. 34). Typically found on 16th-century wrecks, coaks are defined in Sir Henry Manwayring's Seaman's Dictionary as "little square things of brass with a hole in them, put in the middle of some of the greatest wooden sheaves, to keep them from splitting and gulling by the pin of the block whereon they turn" (Manwayring, 1644: 27). It is thought that coaks were mostly used within the larger halyard sheaves placed on the mastheads, as they are recovered very infrequently from wreck sites (Keith, 1987: 117). In Bell-Cross's text, it is suggested that the coaks were placed within the sheaves that would have loaded the Bocarro cannons onto Atalaia (Bell-Cross, 1988: 78). The sheaves and blocks in which coaks were placed were typically constructed of lignum vitae, ash, or elm. Eventually, multipart wooden sheaves would be replaced with sheaves made of solid bronze (Marquardt, 1992: 248; Horsley, 1978: 219-220). As alluded to by Manwayring, coaks supplied structural support to wooden sheaves and protected them from splitting; this was done by inserting the coak along the grain of the wood, while leaving the exposed side of the coak flush with the face of the sheave (Marquardt, 1992: 148). The pins on which the sheaves ran were usually cut from lignum vitae, cog wood, greenheart, or, in larger metal sheaves, iron (Marquardt, 1992: 248; Marsden, 2009: 267; Rule, 1984: 144).

According to an image in Gillian Vernon's book Even the Cows Were Amazed: Shipwreck Survivors in South-East Africa, at least one of the coaks from the Atalaia shipwreck was a rare variant subtype containing a pair of twin trapezoidal projections, commonly referred to as "keys." The keys protrude in a parallel manner from two of the outer sides of the coak and are flush with one of its faces (Vernon, 2013: 63). This unusual variation of coak has been found on Espíritu Santo, one of the 1554 Spanish shipwrecks from Padre Island, and on Bom Jesus, another Portuguese ship of the Carreira da Índia, lost off the Cape of Good Hope in 1533 (Olds, 1976: 43-45; Vernon, 2013: 63; Alves, 2011: 26; Werz, 2010: 434). The appearance of this form of coak on Atalaia is interesting, as its parallels are nearly a century older.

Ten of the Bocarro cannons measured between 350 and 355 centimeters in length, with an average of 349 centimeters, similar to those found on Santíssimo Sacramento. The largest cascabel of Atalaia's Bocarro cannons measured 14 centimeters, and, as mentioned previously, contained the same clenched-fist design as the Bocarro cannons recovered from Santíssimo Sacramento. Atalaia's Bocarro cannons bore the same inscriptions, including Macao's coat of arms and a rampant lion and crown — the insignia of the Governor of India, António Telles de Meneses (1639-1640), and the Viceroy João da Silva Telo de Meneses, Conde de Aveiras (1640-1646) — although the inscriptions on most of the cannons were partial at best, as the cannons were heavily eroded (Bell-Cross, 1988: 78).

Not long after Atalaia was discovered, the remains of the survivors' campsite were also located. Less than a kilometer from the wreck site, researchers found a plateaued sand dune approximately 50 meters long and covered in brush. Crawling through the dune, the team initially located a sherd of blue-and-white porcelain measuring 7.5 by 6 centimeters, showing no signs of water damage. An excavation of the site was carried out by Simon Hall of the Albany Museum in 1980 (Bell-Cross, 1988: 77). According to personal communication between Hall and Bell-Cross, the following artifacts were recovered from the campsite: eight sherds of blue-and-white porcelain from the transitional Ming period, homogeneous in style with both the campsite and wreck collections (Bell-Cross, 1988: 77) — this form of porcelain was manufactured during the 17th century (Valenstein, 1989: 195); eight sherds of Blanc de Chine, or Dehua porcelain, an all-white ware originating in Dehua, Fujian province, China, and highly coveted by the 16th century (Yap and Hua, 1994: 65) (Bell-Cross, 1988: 77); and three stoneware sherds from three different vessels, with brown glaze, believed to have come from martaban jars (Bell-Cross, 1988: 77).

Eleven items of flint were also found at the campsite; nine of these pieces were knapped into level rectangles, most likely for use in firing flintlock firearms. Citing personal communication with Peterson Hall, Bell-Cross points out that the flintlock pieces ranged in size, and that this variation most likely indicates they were meant for different guns: the smallest pieces were most likely for pistols, the medium pieces for an arquebus (carbine), as mentioned in Feyo's account, and the largest for muskets (Bell-Cross, 1988: 77). A series of other, mainly ornamental artifacts were also recovered, most of them luxury items: a piece of marble, a "small cog-like ornamental piece of bronze," a portion of bronze sheet plated in gold, a tiny semicircular portion of silver, a small fragment of silver, and two pieces of mother-of-pearl believed to have originated from a woman's comb; 25 fragments of corroded iron nails and two pieces of lead were also recovered (Bell-Cross, 1988: 77). Some biological evidence was also found, possibly indicating what the survivors ate on the beach: 498 fragments of bone and shell belonging to the white sand mussel Donax serra. As pointed out by Bell-Cross, Feyo's account mentions the survivors consuming mussels on the beach for sustenance; compared to the intact Early Iron Age Donax mussels found near the mouth of the Cefane River, the heavy fragmentation of the mussels at the campsite indicates that they were broken for food by people unfamiliar with, and unaccustomed to, their consumption (Bell-Cross, 1988: 77).

Description of the site

The wreck lies near the mouth of the Cefane River, on the eastern Cape coast of South Africa, in an area where the ship apparently anchored in seven fathoms (27 m) off a long sandy beach before wrecking. The site included a cluster of bronze and iron cannons arranged as they would have been stored in the hull, along with sheets of lead, porcelain and martaban sherds, lead shot, two anchors, a crushed pewter jar, and a lead sounding weight. A related survivors' campsite was located less than 1 Km away, on a plateaued sand dune about 50 m long.

Cargo

Porcelain and stoneware sherds recovered from both the wreck site and the survivors' campsite — including transitional Ming blue-and-white porcelain, Blanc de Chine, and martaban storage-jar sherds — are described in detail in History of the shipwreck above.

Hull Remains

A single piece of wood, most likely belonging to the hull, was recovered from underneath one of the Bocarro cannons; it was later identified as evergreen oak, confirming that the ship was built in Portugal (Bell-Cross, 1988: 74).

Guns

The cannon assemblage recovered from the site consisted of at least 10 bronze Bocarro cannons and more than a dozen additional iron and bronze cannons; see History of the shipwreck above for details of their recovery, dimensions, inscriptions, and associated finds.

References

Alves (2011);

Bell-Cross (1988);

Borell (2014);

Castro (2001);

Horsley (1978);

Keith (1987);

Manwayring (1644);

Marquardt (1992);

Marsden (2009);

Olds (1976);

Rule (1984);

Valenstein (1989);

Vernon (2013);

Werz (2010);

Yap and Hua (1994).