SCOTT SEMANS WORLD COINS

MANILLA:

MONEY OF THE SLAVE TRADE

Scroll down for price list


    Copper was the "red gold" of Africa and had been both mined there and traded across the Sahara by Italian and Arab merchants. The early Portuguese explorers of the 1470s observed that copper bracelets and legbands were the principal money all along the west African coast. They were usually worn by women to display their husband's wealth. The Portuguese crown contracted with manufacturers in Antwerp and elsewhere to produce crescent rings with flared ends of wearable size which came to be called "manilla," after the Latin manus (hand) or from monilia, plural of monile (necklace).
    While copper bracelets dating prior to 1600 AD which likely had some exchange function continue to be excavated around Jenne-Jeno and related sites, we can only guess today at what prototypes may have inspired the distinctive flare-ended crescent shape. One theory is that Europeans copied a splayed-end raffia cloth bracelet worn by women, another that the well-known Yoruba Mondua with its bulbous ends inspired the manilla shape. Much closer in form to modern manillas, however, is this type, excavated at Igbo-Ukwu.. In "Die sog. Geldeifen aus Benin," Der Primitivgeldsammler #28/1, p.29-35, Rolf Denk summarizes what is known about heavy, faceted pieces with enlarged ends such as this one (25cm across and 4.5cm gauge) from the Museum für Volkerkunde in Vienna, which may predate European manillas.
    The first solid information we have on Manillas comes from records of a contract between the Portuguese government and Erasmus Schetz of Antwerp, who supplied the Portuguese factory at Mina with as many as 150,000 manillas per year. The standard in 1529 was for a piece about 240m long, about 13m gauge, weighing 600 gram. However, no examples of torque-shaped bracelets in this weight range are known today, suggesting that the standard was not honored, or that later negotiations provided for a reduction. Duchateau, Royal Art of Benin, p.15 shows a plaque with a European holding two pieces with barely flared ends whose apparent size could match these specifications.
    Recently recovered from the Guetaria bay off the Basque coast of Spain from a wreck dated to 1524, and described in detail in Der Primitivgeldsammloer #26/1 p.9-12 (M.A. Artica), are brass manillas averaging 306gms, 103 x 87mm size, and gauge increasing from 12mm at center to 22mm at ends. The shape is somewhat similar to the familiar French Popo manilla, but even less pronounced in the flare. This piece from Opitz's Ethnographic Study p.213 upper left is apparently of the Guetaria type. In 1547 the Portuguese switched suppliers to Cristoff Fugger, and pieces called tacoais with a standard in the 235-290 gram range were ordered. Known specimens of what the Africans call Mkporo manilla (939-941 below, Opitz p.211 upper left) range from 226 to 294 grams, and are likely Portuguese Fugger products. The existence of at least three "varieties" in metal, end-flare, and degree of crimp (#939-941 below) could be incidental or, more likely, a bow to regional preferences.
    Although Gold was the primary and abiding merchandise sought by the Portuguese, by the early 16th century they were participating in the slave trade for bearers to carry manillas to Africa's interior, and gradually Manillas became the principal money of this trade. By the end of the 1500s the Portuguese had been shouldered aside by the British, French, and Dutch, all of whom had labor-intensive plantations in the West Indies, and later by the Americans whose southern states were tied to a cotton economy . A typical voyage took manillas and utilitarian brass objects such as pans and basins to West Africa, then slaves to America, and cotton back to the mills of Europe.
    Early in the 18th century Bristol, and then Birmingham, became the most significant European brass manufacturing city. It is likely that most types of brass manillas were made there, including the "middle period" Nkobnkob-Onoudu whose weight apparently decreased over time, and the still lighter "late period" types such as Okpoho and those salvaged from the Duoro wreck of 1843. Among the late period types, specimen weights overlap type distinctions suggesting contemporary manufacture rather than a progression of types. The Popos, whose weight distribution places them at the transition point between Nkobnkob and Onoudu, were made in Nantes, France, and possibly Birmingham as well. They are wider than the Birmingham types and have a gradual, rather than sudden, flare to the ends.
     The Africans of each region had names for each variety of manilla, probably varying locally. They valued them differently, and were notoriously particular about the types they would accept. The price of a slave, expressed in manillas, varied considerably according to time, place, and the specific type of manilla offered. Internally, manillas were the first true general-purpose currency known in west Africa, being used for ordinary market purchases, bride price, payment of fines, compensation of diviners, and for the needs of the next world, as burial money. Cowrie shells, imported from Melanesia and valued at a small fraction of a manilla, were used for small purchases. In regions outside coastal west Africa and the Niger river a variety of other currencies, such as bracelets of more complex native design, iron units often derived from tools, copper rods, themselves often bent into bracelets, and the well-known Handa (Katanga cross) all served as special-purpose monies.
    As the slave trade wound down in the 19th century so did manilla production, which was already becoming unprofitable. By the 1890s their use in the export economy centered around the palm-oil trade. Although manillas were legal tender, they floated against British and French West African currencies and the palm-oil trading companies manipulated their value to advantage during the market season. Probably for this reason the British undertook a major recall dubbed "operation manilla" in 1948 to replace them with British West African currency at a rate of 3 Pence for the commonest type. The campaign was largely successful and over 32 million pieces were bought up and resold as scrap. The manilla, a lingering reminder of the slave trade, ceased to be legal tender in British West Africa on April 1, 1949.
    An unanswered question is whether "manillas" were made in Africa by native smiths during the period when European types were imported. It is hard to believe that no such attempts were made. Offered for sale below and linked to scans are obvious counterfeits, but such pieces made of lead, or underweight pieces in impure brass are rather uncommon and it is unclear where they were made. More interesting is the group of horseshoe-shaped pieces as well as gleanings from several years' worth of poring over Africa Traders' stocks, looking for pieces with flared ends within the known manilla weight range. Are they proto-manillas, early European manillas, or African-made pieces during the period of European importation which - deliberately or coincidentally - resemble the European type? All are copper, rough with verdigris indicating some age.
     Many other things are called manillas by authors and collectors. Commonly available are many distinctive regional bracelet forms of copper, brass, nickel, and iron made in Africa in the 19th-20th centuries with varying monetary functions and ranges of use. Legbands, collars, and coiled forms made of Calabar rod. Those with flared ends are often called manillas. Other flare-end forms of large size, called "King" and "Queen" manillas, are ceremonial rather than trade manillas, and from Zaire come large copper crescents in several distinctive shapes, sometimes including flared ends, which are likely forms of bullion storage (like the Katanga cross).
     For more depth on Manillas and other bracelet monies see InfoSheet 62.

MANILLA TYPOLOGY

     Manilla typology is a largely unexplored subject. While trader and traveler accounts are both plentiful and specific as to names and relative values, no drawings or detailed descriptions have survived which could link these accounts to specific manilla types found today. Historians and economists emphasize patterns of trade and show no interest in the specific appearance or variations among trade goods. Collectors rely on a confusing and incomplete chart by Johansson, Nigerian Currencies, unsourced, but reproduced endlessly, including in scholarly works. The chart shows nine named pieces, and he cites other types by area of use, which are not shown. Distinguishing factors are thickness and the diameter and degree of flare to the ends, size / weight, and shape. I believe that there really are few discreet types, mainly an evolutionary process: from large to small, heavy to light, crude to finished; end flare subtle to exaggerated, and footprint wide and rounded to small and elongated. The proliferation of African names is probably due more to regional customs than actual manufacturing practices. My own approach has been to examine and sort thousands of pieces obtained from different places over time, and try to apply the names in Johansson's chart, where possible, to varieties as found.
     Portuguese manillas, being significantly heavier than even the early British, tend not to turn up in mixed lots. Rolf Denk agrees in assigning the three varieties shown below (#939-941) to the Portuguese, and the distinctive Guetaria type is securely dated to the Portuguese era. The French Popo manilla, with three rather subtle variations, used to be the most available type from African suppliers. About 1990 these began to dry up and the British types became common. I have probably handled 10,000+ pieces, selling mostly to wholesale and promotional markets. Groups from Africa tend to include a limited range, sometimes predominantly heavy, early types, but more often the small, compact late types exemplified by the Duoro wreck pieces.
    In March 2007 I purchased a group of 72 pieces with similar patination and soil crusting, suggesting common burial. There were 7 brass Mkporo (941); 19 Nkobnkob-round foot (945) ranging 108 - 184gm; 9 Nkobnkob-oval foot (943) ranging 148 - 206gm; and 37 Popo-square foot (988). Since the lightest Nkobnkobs in the hoard are 108 and 114gm, while they are routinely found (called Onoudu) under 80gm, this strongly suggests that the group was buried at a certain point in the size devolution of the Manilla. The weight correspondence of the oval-foot Nkobnkob with the high end of the round-foot range suggests that it is either the earlier variety, or contemporary with the earliest round-foots. Finally, the exclusive presence of the square-foot variety of French Popo, normally scarce among circulation groups of Popos, confirms that this is the earliest variety, as their more irregular feet suggests. It also dates the earliest French manillas as likely contemporaries of the earliest British pieces, and virtually rules out the possibility that this hoard was a creaming of large pieces from a later burial, as square-foots are not heavier than later Popos, nor easily distinguishable. A May, 2009 purchase of 134 heavily patinated pieces which the supplier had culled for size out of a larger group, contained only 943s and 945a's in a rough 2/3 ratio, plus three 940s and a scattering of lighter 945b's. This suggests to me that, as per Gresham's law, the heavier Portuguese types left circulation, possibly hoarded, melted, or retained for ceremonial use, when the earliest Birmingham manillas were in use.
    In the listings below I designate "top" as the middle of the crescent opposite the "feet" and "gauge" as measured here.

   ⇓  CLICK ON STOCK NUMBER FOR PHOTO
PORTUGUESE MANILLAS (TACOIS)
939Torque Shape [Portuguese Fugger?]   Brass, Torque profile with close ends and gradual flare. Feet irregularly round to oval. Aside from the different metal, these appear to be bent versions of the copper Mkporo, but the outside length is significantly greater. The weight range and evenness of the end flare appears transitional between the pre-Fugger Guetaria manillas and the as-published Mkporo types below. For 2 specimens seen: 84, 85m across by 82, 90m high x 251, 255m outside length, 11-13m gauge, 228, 294g NA
940Copper Mkporo [likely Portuguese Fugger]   Copper, Rounded crescent profile with ends slightly less flared than the brass versions. Feet irregularly round to oval. Broader back and asymmetrical shape with thicker "ankle" on the short side also distinguish these from the brass pieces. For 3 specimens seen: 99-100m across by 82m high x 236-238m outside length, 11-13m gauge, 226, 244, 264g Similar to Johansson plate profile. Transitional in shape between barely-flared Guetaria sea-salvage type, and distnctly flared British Nkobnkob / Onoudu types. All specimens are heavier than the heaviest Nkobnkob., and their weight matches the specifications (235 - 240 grams) given by the Portuguese to their suppliers, the Fuggers of Augsburg in 1547. In later times, the Mkporo became scarce in Africa and was "promoted" from everyday use to a standard of wealth and burial money. SOLD
941Brass Mkporo [likely Portuguese Fugger]   Brass, Rounded crescent profile with ends gradually flaring. Feet irregularly round. For 7 specimens seen: 88-98m across by 90-95m high, 10-13m gauge, 228-250g. Just as Johansson plate profile. Transitional in shape between barely-flared Guetaria sea-salvage type, and distnctly flared British Nkobnkob / Onoudu types. All specimens are heavier than the heaviest Nkobnkob., and their weight matches the specifications (235 - 240 grams) given by the Portuguese to their suppliers, the Fuggers of Augsburg in 1547. In later times, the Mkporo became scarce in Africa and was "promoted" from everyday use to a standard of wealth and burial money.SOLD
BRITISH MANILLAS   Birmingham
943Nkobnkob-oval feet [Early Period]     Ends sharply flaring creating large feet with oval footprint. 81-85mm across by x 70 -73mm, 9-12mm gauge, 148-206gm. Essentially a Nkobnkob profile as shown in Johansson, but with more distinctive heels. Earlier than, or contemporary with earliest of round-foot Nkobnkob. Likely the earliest Birmingham Manilla.40.00
945ANkobnkob     Ends sharply flaring producing roughly rounded but asymmetrical footprint. 75-76m across by 77-85m, 11m gauge, 160-225 grams. Nkobnkob are often found roughly cast with flashing at the seams, as these examples. Likely the earliest Birmingham Manilla.40.00
Flashing SetNkobnkob     Three pieces as shown in this scan (943 & 945) with casting flash or a stongly offset seam on the foot. Casting flashing is not unusual, but these are extreme examples. Many specimens are found filed smooth.90.00
945B    —  [Middle Period]    Smaller specimens, 9-11m gauge, 68-71m across by 73-78m high, 130-160gm 25.00
946Onoudu    Similar to Nkobnkob, but smaller, 65-71mm across by 72-75mm, under 80-120gm, 8-9m gauge, often showing casting seam. Large, broad footprint ranges 23-26mm across. Specimens offered are 100gm and up.13.50
946F    —    Distinctive variant. with a broader back, rather than crescent shape. Same size & weight range.13.50
948Okpoho [Late Period]    Smaller than earlier types, foot significantly smaller and more symmetrical teardrop shape, no seam. 57-68mm across by 59-64mm, 8-9m gauge, 65-85gm (most 70-80); footprint 19-20mm across. A later time period than the Nkobnkob & Onoudu types. Commonest Manilla variety. 12.50
950Ejema?    Footprint more rounded. 59-65mm across by 59-65mm, 8m gauge, 60-90gm (most 65-75); footprint 18-19mm across.13.50
969"Duoro type"    Late type with more compact form: thick back and feet more pulled together Dateable to 1843. Foot is a rounded oval, almost circular. Side view of foot is thick. Two specimens from the wreck are 8-9m gauge, 57-58m across by 54-55m across, 71.29 & 71.35gm with corrosion and patination different from circulation manillas. Specimen SIMILAR to Duoro, but not from the wreck:N/A
970Atoni?    "Elephant foot:" foot is rounded oval, slightly pointed at bottom; side view of foot is thick, especially at bottom. 53.5-59.5m across by 56-60mm, 77-88gm, some heavier, 7-10m gauge . 9.50
971Unknown name    Foot is narrower with only a slight bulge at top. Foot profile is thick but not flared at top. Similar dimensions, 70-81gm. 10.00
972Okombo?    Foot is narrower still, almost a symmetrical oval. Foot profile is thick with a large bulge at top. Similar dimensions. 59-71gm. 10.00
975Late Manilla    Thin-profile foot with sharply pointed teardrop footprint. Sharp ridge behind foot. 7-8m gauge, 52-56 across by 58-60mm, 54-66gm (most 54-60).16.50
965LLead Manilla, general type of 970-972 late Manillas with good shape but prominent seam on foot. Counterfeits?? Scarce.17.50
965xCounterfeit Manilla?   Low weight (41-54gm) with deteriorated metal suggesting high lead content. Not too common. (3 pcs. $15)6.50
965dDamaged Manillas   Mixed types, but mostly early and middle period British with heavy corrosion, usualy feet are chipped or splitting. Per 10 Manillas:29.50
965bBroken Manillas   Mixed types, but mostly early and middle period British. These have broken roughly in half. Per 20 half-pieces.15.00
FRENCH (POPO) MANILLA   Nantes
988Popo-Square end    Squarish, often convex ends; nearly seamless body. Much evidence of post-cast filing on both ends and body. 102-160gm, 134gm avg. Slightly cruder than round-end type; presence in hoard of early British types suggests this is the earliest French variety.17.50
992Popo-Round end    Smaller, irregularly-roundish or heart-shaped ends. Often with prominent seam on end and body. 79-85x70-72mm, 9.5 -11mm gauge, 95-159gm, 131.4gm avg. 13.50
994Popo - Rim end    Similar but less evidence of seams or filing with a slight raised rim around the foot. Possibly a late variety using more advanced casting technique.13.50
MANILLA SETS
996SBasic Set:    Early and Late British, and French (946, 970/71/72, 992/94) 3 pieces    29.95
977SBasic British Set:    Early, middle, and late period examples (945b, 946, 970/71/72) 3 pieces    39.95
977TExpanded British Set:    Most of the types above, covering all periods: (943 or 945a, 946, 946F, 948, 950, 970, 971, 972, 975 plus miscast piece) 10 pieces    115.00
995SFrench Set   : All three varieties, probably representing different periods of manufacture: (988, 992, 994) 3 pieces    37.50
996TComprehensive Set:    All of the types above (except lead & Duoro) plus a miscast piece and broken fragment to show cross section. (British: 943, 945a, 946, 946F, 948, 950, 970, 971, 972, 975 plus miscast & broken pieces, Fench: 988, 992, 994) 14+ pieces 185.00
996vGrand Set:    Comprehensive Set as above, plus a large copper bullion Manilla which I call the "working Queen" 250.00
Bulk Quantities    I can supply unsorted quantities of British manillas per 10, 50, 100, etc. Please Inquire.
AFRICAN-MADE TRADE MANILLAS (?)
910aOverlap, slight flare to ends   310gm, 75mm acrossSOLD
910bOverlap, medium flare   about 640gm, 88mm across, one end chiselledSOLD
910cNarrow gap, medium flare   440gm, 105mm across, 15mm gauge at backSOLD
910dWide gap, slight flare   308gms, 124mm across, just as "a" without the overlap65.00
910eClosed gap, slight flare   180gm, 84mm across40.00
910fRounded shape with closed gap, sharp flares   60 oz, 72 oz, 115, 130mm across (for 2 specimens)SOLD
930a"Horseshoe" Manilla   NIGERIA, unpublished?, My made-up name for 9 intriguing pieces found in the stock of an African trader specializing in metal objects, along with many bracelet pieces of usual types. They are heavily corroded brass, suggesting significant age. The flared ends resemble ordinary trade manillas and the weight range places them with the early- mid period British Nkobnkob manilla type. My guess is an African-made "imitation" of European manillas, or could they be the "Dutch" manilla?. Width 85-95mm, Weights 118 (missing end), 130, 157, 160gm30.00
930b"Horseshoe" Manilla   Weights 173, 187, 195, 200gm35.00
AFRICAN-MADE CEREMONIAL MANILLAS
Used throughout west and central Africa, the large King and Queen Manillas are offered HERE


WEBSITE DIRECTORY    | | |    INDEX OF ARTICLES    | | |    Ethnographic Money For Sale    | | |    Ceremonial Manillas