Objects of Personal Adornment at the Cape

Carolyn White

IT’S ALL ABOUT PEOPLE For various reasons people from Western Europe as well as the Far East arrived at the early Cape of Good Hope. Many Westerners served in the local VOC structure, while the Easterners arrived mainly as slaves or exiles. While disseminating the material contained in certain series of VOC documents vested in the Cape Town Archives notes were made of certain individuals who made an impression on the inquiring mind. It was surprising to see how much information could be gathered from these series in order to conduct case studies, for our purpose Angela of Bengalen, the royal princes Loring Passir and Dipa Nagera of Java, Nicolaas Ondatje of Colombo and Manual Thuart of Tutucurin, complemented by a case study of the French refugee, Prieur du Plessis, a Protestant Christian. The fact is that these people who honoured different religious beliefs, enriched the contents of the various cases.

The purpose of the research was to study changes that occurred in the material culture of the Cape during the period when the British took over control of the colony from the Dutch. There were three phases for investigation: the colony under the Dutch East India Company in the 18th century, twenty transitional years of interim British and Netherlands governments between 1795 and 1815, and the Cape as a British colony after 1815. An historical archaeological approach was applied to material remains surviving from those years, such as excavated artefacts, documents and buildings, that assumed these sources of material culture reflected the larger cultural, or cognitive, contexts in which they were conceived, made and used. Particular emphasis was placed on examination of household inventory manuscripts (lists of fixed and moveable properties, goods and chattels). Selected information from the inventories of more than 800 households was recorded, and further detailed analysis made of seventy-nine documents. Room-by-room appraisals indicate the layout (house plan), room numbers (house size), room names and activities (functions of spaces) within the house. These probate records thus provided invaluable information about houses, their contents and the placement of objects within the household, and could be investigated from the level of individual rooms on the day of appraisal to a range of houses over a number of years. By constituting the documentary evidence in a form compatible with assemblages of excavated artefacts, as a series of space and time blocks, integrated information provided enhanced material cultural detail. Patterns were observed through time and across a range of regional and socio-economic situations. The first period covered a Dutch Cape under the control of the eastern arm of the Dutch East India Company, but households were organised in a way distinctive to the Cape. Then there was a short period of relative freedom from governmental control, as transition was made from Dutch to British colonial status and trade options broadened, resulting in the wealthier urban households reflecting fashion, and to the benefit of many farmers. Finally, the Cape was fully incorporated into the networks of the British Empire, undergoing widespread adaptations to colonial society and changes in the material culture of households.

What is it about a landscape that allows it to become a repository for memories and cultural values? How do you protect this for future generations? Indeed, how do you even know what it is before it is gone? These are the types of questions that dominate People and Places ...

Historical Archaeology in South Africa Material Culture of the Dutch East India Company at the Cape Carmel Schrire With contributions by Jeffrey J. Durst, Adam Robert Heinrich, Stacey Jordan, Jane Klose, and Carolyn L. White Walnut Creek, CA Left Coast Press, Inc. 1630 North Main Street, #400 Walnut Creek, CA 94596 www.LCoastPress.com Copyright ã 2014 by Left Coast Press, Inc. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior permission of the publisher. ISBN 978-1-59874-164-3 hardcover ISBN 978-1-61132-774-8 institutional eBook ISBN 978-1-61132-769-4 consumer eBook Library of Congress Cataloging-in-Publication Data: [to come] Printed in the United States of America The paper used in this publication meets the minimum requirements of American National Standard for Information Sciences—Permanence of Paper for Printed Library Materials, ANSI/NISO Z39.48–1992. CHAPTER 9 Objects of Personal Adornment at the Cape Carolyn L. White Introduction This chapter provides an analysis of objects of personal adornment found in two Dutch East India, or VOC, sites at the Cape. The first is the Van der Stel Moat at the Castle of Good Hope (M90), and the second is Oudepost I, an outpost of the main settlement lying about 100 km north on the shores of Saldanha Bay (see Chapter 2). Together they provide insight into the gender, rank, and physical appearance of residents living under VOC rule in the late 17th and early 18th century Cape. Artifacts may be divided into four categories: clothing fasteners, jewelry, hair accessories, and miscellaneous personal accessories (White 2005: 10). The Moat (M90) assemblage consists solely of buttons; the Oudepost assemblage comprises buckles, buttons, finger rings, and a segment of chain, likely from a watch chain or waist-hung appendage. The clothing fasteners—buttons and buckles—can be categorized by size, material, and decoration, and may be used to identify the sorts of clothing worn by the site inhabitants. Buckles may be linked to particular articles of clothing (shoes, breeches, straps/swords), and buttons reflect the level of elaboration of the clothing as well as its overall diversity. In both cases, material and level of decoration also allude to rank and status differences among the wearers. The documentary record is largely unaccommodating in its recording of the clothing and accessories worn by the people who lived at Oudepost I and at the Castle. The uniforms of the military are carefully described in Mentzel (1925 [1725]) and the supply lists record a wide variety of clothing materials shipped to the Cape, consisting mainly of textiles, but these describe only a few types of buttons and a few accessories (Leibbrandt 1902; Thom 1952). On the other hand, the archaeological record provides material remains that reify the clothing and garb of the Cape inhabitants. The only comparable work to date is that of Elizabeth Jordan (2006). Her study of late 18th-19th century washerwomen postdates the collections studied here, but incorporates a thorough examination of many artifacts of personal adornment, particularly buttons (ibid.: 2006). Her analysis draws on (and critiques) work on contemporaneous collections in North America (e.g., Edwards-Ingram 1999; Galle 2004; Heath 1999; Mullins 1999; Wilkie 2003). Analysis On the whole, the personal adornment assemblage, like much of the material culture of the VOC, was imported from Europe. There might have been limited local production of bone buttons, since supply lists do not mention these being imported, but there is an absence of button making waste recovered at the sites. The presence of a button mold (discussed later in this chapter) suggests that there were some early attempts to supplement the European metal imports. The following analysis categorizes personal adornments according to the divisions laid out in American Artifacts of Personal Adornment (White 2005) and draws on resources contained in that volume for identification. Given the very restricted information that is available for personal adornment for this period, each individual artifact type is described. The artifacts are grouped by clothing type and discussed as sub-assemblages. Appendix F contains a list of the personal adornment artifacts and the associated catalog numbers used in the text. Carolyn L. White, “Objects of Personal Adornment at the Cape“ in Historical Archaeology in South Africa: Material Culture of the Dutch East India Company at the Cape, pp. 205–212. © 2014 Left Coast Press, Inc. All rights reserved. 205 206 H I S TO R I C A L A RC HA E OLOGY I N SOU T H A FR I C A Clothing fasteners Jewelry The clothing fasteners are broken into the following categories: aglets, buckles, and buttons (the last of which includes a single button mold as well). Jewelry is represented at the Cape sites in a very small quantity. Aglets Sites Castle: Van der Stel Moat: Discussion Buckles The buckles are divided into the following categories: Shoe buckles Whole buckles Frames Whole Fragments Chapes Knee buckles Whole Strap buckles Whole The buckles are analyzed using standard descriptions established in White 2005. Although the personal adornment artifacts from the Cape are Dutch in origin, they parallel styles and fashions seen in England and throughout Europe from the period (e.g., Abbitt 1973; Egan and Pritchard 1991; Mould 1979; Swann 1981). Certainly, they exhibit stylistic individuality. The personal adornment collection from the Moat is restricted to buttons. It is possible that a more varied component was excavated and that it has been misplaced, in which case we look forward to its re-appearance at some future date (Carmel Schrire 2009, pers. comm.). The distribution of the buttons in the excavation shows that most are found in the richest level of the site, A2, but this distribution does not permit generalizations about Cape dress or status (Table 9.1). Table 9.1. Van der Stel Moat (M90) distribution of buttons by Level Level Nos. % A 5 5.2 A1 11 11.3 A2 70 72.2 B 5 5.2 NP 6 6.1 97 100.0 Total Buttons As in the case of the buckles, the buttons are analyzed using standard descriptions developed by White (2005). The buttons are divided into the following categories: Sleeve Waistcoat buttons Metal Bone Shanked Sew-through Coat buttons Metal Bone Shanked Sew-through Unidentified function Button mold Within each major category, buttons are divided according to diameter, and then further separated into decorated and undecorated examples. The waistcoat buttons from the Castle Moat are made exclusively of bone. Most of the buttons are undecorated, with drilled shanks. Those that are decorated are simply elaborated with incised rims and range in size from 13 to 17 mm. The coat buttons are mostly bone, with the exception of two metal ones (BT 8, 9). BT 9 (Figure 9.1) is a very large copper alloy button (28 mm in diameter). It is the largest button recovered on the sites and suggests a high level of elaboration for the garment on which it was worn. Figure 9.1. Large coat button (BT 9) from the Castle Moat (M90) excavation, made of copper alloy. Scale in cm. (Photo C. White) OBJECTS OF P ERSONAL ADORNMENT The bone buttons range in size from 19 mm to 25 mm, with most falling in the range of 14 to 16 mm. Many are decorated with incised rims, but this is the extent of the elaboration of these buttons. A single button is green, perhaps tinted by a copper or fabric cover. Most of the buttons fall into the category of ‘unidentified function’ and cannot be slotted into the categories in the previous section with certainty. Within this category are several buttons that are notable. There are several very small bone buttons (BT 27-34) that range from 7 to 9 mm (Figure 9.2). They may be shirt or sleeve buttons. Five of these are small, undecorated buttons of the same size (9 mm) that may have belonged to the same garment. A single sew-through button (BT 96) was identified. It is a button core that was likely covered with textile when worn. 207 The Oudepost I clothing fasteners comprise aglets, buckles, buttons, and a button mold. Two aglets were identified. They are made of sheet copper, curled into a tube with a straight seam along the edge in the fashion used since the medieval period (White 2005: 31). They have rounded tips, formed by rotating the end against a flat surface. The base of the aglet is pinched, to hold the lace in place. The Oudepost I sub assemblage includes complete buckles as well as frames and chapes. Most of these are shoe buckles, as indicated by the size and form of the frame (White 2005). The buckles are similar in form to those identified in more well-documented British excavations (Egan and Pritchard 1991), but also possess characteristics that suggest an origin other than England. The complete shoe buckles are small and adhere to fashions of the late 16th and early 17th centuries (White 2005: 40). Though the two buckles offer contrasting styles, they are consistent with the scalloping and knops and serrated decorations favoured in the period of 1680 to 1720 (White 2005: 40). These buckles also possess stud chapes, which were common in the period up to 1720 (White 2005: 40, 41) (Figures 9.3, 9.4). Figure 9.2. Small bone button (BT 28) from the Castle Moat (M90) excavation. Scale in cm. (Photo C. White) The remaining unclassified buttons are mostly undecorated bone buttons with drilled shanks. These are convex buttons that vary in size from 12 to 20 mm. The buttons are very similar in that they are undecorated. Variation in the form of the drilled shanks suggests the individual hand of the maker. There is therefore tremendous uniformity in these fasteners in the convex shape and lack of elaboration on the buttons. Even the decorated buttons have restrained elaboration. Most of the decorated buttons are incised on the rim, but this decoration is far from sophisticated and reinforces the vision of the site inhabitants as rather simply clothed. There are several buttons that are somewhat more elaborate; for example, BT 58, a possible waistcoat button, has a turned design. The controlled designs attest to the overall conservative nature of dress at this site. This lack of elaboration may be expected at a far-flung VOC settlement where the bulk of the clothing materials were imported as provisions for the garrison (Mentzel 1925 [1725]: 59-61). However, the simplicity of the buttons also suggests uniformity in physical appearance in terms of dress across the population. Oudepost I Unlike the Castle collection, the personal adornment from Oudepost I exhibits far more variation than the button collection from the Castle Moat. We are unable to assume, however, that buckles and jewelry were never present in the Castle Moat because some small components of the collections are currently missing. Figure 9.3. Complete shoe buckle (BK 1) from the Oudepost I excavation. Scale in cm. (Photo C. White) Figure 9.4. Complete shoe buckle (BK 2) from the Oudepost I excavation. Scale in cm. (Photo C. White) The rest of the shoe buckles are fragmentary buckle elements: frames and chapes. The decorative motifs on the frames range from simple bevelling to knops and scalloping. Three of the frame fragments (BK 9, 10, 11) are elaborate openwork frames with ribbon and bows design (Figure 9.5). These buckle fragments suggest the presence of a pair of shoes and the high level of decoration associated with this particular element of attire. The shoe buckle chapes are cooking pot shaped with a single tongue. This form was utilized beginning around 1720 in order to secure larger buckles to the shoes (White 2005: 42). 208 H I S TO R I C A L A RC HA E OLOGY I N SOU T H A FR I C A Figure 9.7. Sleeve button (BT 2) set with stone, from the Oudepost I excavation. Scale in cm. (Photo C. White) Figure 9.5. Partial shoe buckle frame with ribbon and bow design (BK10) from the Oudepost I excavation. Scale in cm. (Photo C. White) Figure 9.8. Four-holed bone button (BT 33) with incised border from the Oudepost I excavation. Scale in cm. (Photo C. White) Figure 9.6. Iron knee buckle (BK 21) from the Oudepost I excavation. Scale in cm. (Photo C. White) There are two remaining buckles aside from shoe buckles. The first (BK 21) is a knee buckle made of iron (Figure 9.6). It is identifiable as such due to the form of the chape, which is anchor shaped. The small size of the buckle suggests that it was used to fasten breeches at the knee. This likely dates from the later portion of the occupation of the site, as knee buckles became popular in the 1730s (White 2005: 43). The second buckle (BK 22) is a sword or baldric buckle. Turning to buttons, they comprise the bulk of the Oudepost assemblage, and include sleeve, waistcoat, and coat buttons as well as a number that are not readily identifiable by garment. The Oudepost I buttons are made of metal (primarily copper alloy), bone, and shell. Eight metal sleeve buttons were identified in the Oudepost I excavations. Two sets are fragmentary and two are complete. The fragments are identifiable as sleeve buttons based on the flattened U-shaped shanks and the small size of the buttons (White 2005: 61; Noël Hume 1961: 383). One of these (BT 1) is undecorated, and the other (BT 2) is set with a white stone and has a beaded border (Figure 9.7). A connecting wire loop remains attached to the shank. The first intact set is decorated with tiny dots on the border; they remain attached with a wire loop. A second intact set is large in size, with a chained decorative border and a central floral design; the large size is unusual. The waistcoat buttons are made of metal and bone and exhibit variation in the decoration as well as in the form. The metal buttons have two primary forms: hollow cast two-piece buttons that are brazed together with pin shanks and solid cast buttons with cast or brazed shanks. All of the metal waistcoat buttons are undecorated, with convex or flat shiny surfaces. The bone waistcoat buttons are both decorated and undecorated. The undecorated buttons, like the metal ones, possess smooth convex or flat surfaces. The decorated buttons are either incised on the rims or possess a border and nippled decoration in the center. Most of the bone waistcoat buttons are shanked, save four sew-through buttons. Two of these are cores with a domed shape, while one is flat. These buttons would have been covered with textile when worn, which would have served to attach the button to the garment. A single fourholed sew-through button with an incised border (BT 33) was also recovered (Figure 9.8). This button is likely a waistcoat button, but may have been a breeches button. The coat buttons in the Oudepost I assemblage are made of bone and metal and are made in a manner similar to the waistcoat buttons in terms of production and decoration. They measure between 14 and 32 mm. The metal coat buttons exhibit a range of decoration. Two buttons (BT 34, 40) possess an octagonal shape as well as a nipple decoration (Figure 9.9). Another (BT 35) has a domed surface with crossed lines and a central nipple. Central nipple motifs are found on additional coat buttons. Organic designs are found on other buttons. One (BT 38) is a stamped copper alloy button cover with a stamped oval design with sunburst (Figure 9.10), and another (BT 36) has an engraved floral design on the surface (Figure 9.11). OBJECTS OF P ERSONAL ADORNMENT Figure 9.9. Octagonal metal button (BT 40) with nipple decoration from the Oudepost I excavation. Scale in cm. (Photo C. White) 209 Figure 9.13. Metal button (BT 92) with sunburst design from the Oudepost I excavation. Scale in cm. (Photo C. White) Figure 9.14. Small shell button (BT 102), probably expensive, from the Oudepost I excavation. Scale in cm. (Photo C. White) Figure 9.10. Copper alloy button cover (BT 38) with sunburst from the Oudepost I excavation. Scale in cm. (Photo C. White) Figure 9.15. Intact frog (BT 104) from the Oudepost I excavation. Scale in cm. (Photo C. White) Figure 9.11. Copper alloy button cover (BT 36) with floral design from the Oudepost I excavation. Scale in cm. (Photo C. White) Figure 9.12. Probable doublet button (BT 70) cast, with rosette design from the Oudepost I excavation. Scale in cm. (Photo C. White) The bone buttons are almost exclusively undecorated and all but two are shanked. The coat buttons are identified by size, and these large buttons mirror the form and decoration found in the waistcoat bone button sub assemblage. Two large sewthrough buttons are also present in this sub assemblage. Many of the Oudepost buttons are not readily classified within the categories described earlier in this chapter and fall into ‘Unidentified function.’ The metal buttons can be roughly divided into several groups according to diameter. Buttons that range from 9 to 12 mm may be sleeve buttons or shirt buttons. There are several (BT 69, 70) that are small cast buttons with a rosette design (Figure 9.12), and these are likely doublet buttons. The second major group of unidentified metal buttons range in size from 13 to 18 mm. Most of these are undecorated and may be waistcoat, doublet, or breeches buttons. These buttons are both cast buttons with drilled and brazed eyes and two-piece hollow cast brazed buttons. Most possess flat or convex, smooth surfaces. A sunburst design and octagonal shape (BT 92, 93) are shown on two of these buttons (Figure 9.13). The bone buttons that are unidentified in function are small and undecorated (B 97-99). These may be breeches buttons. A single bone core also may be a breeches button. These were likely plain and otherwise unremarkable elements of simple garments. Two small shell buttons (B 101, 102) were recovered at Oudepost (Figure 9.14). Measuring 10 and 11 mm respectively, these would have been among the most expensive buttons worn at Oudepost, as the shell would have been imported from the East (White 2005: 71). Several additional fasteners stand apart from the rest of the button sub assemblage. A small metal stud (BT 103), an intact frog (BT 104; Figure 9.15), and a frog fragment (BT 105) make up this group. The function of the stud is not known; the frog and frog fragment may have been used to attach a weapon to the body or to hold edges of a garment together. 210 HI S TO R I C A L A RC H A E OLOGY I N SOU TH A FR I C A A button mold was also identified at Oudepost (Figure 9.16). This is a portion of a three-piece button mold used to make frog-style button fasteners. The mold has tenon holes present on two sides and a groove is present to receive the molten metal. This mold is clear evidence of button manufacture occurring at Oudepost (Ivor Noël Hume 1988, pers comm.). Figure 9.16. Stone button mold from the Oudepost I excavation. Scale in cm. (Photo C. White) Finally, three links of chain (C 1) were also identified at Oudepost. These links are made of round wire. Chains were expensive items, as they were handmade and very labor intensive. It is difficult to ascribe a particular function to a small piece of chain, but the size of the links suggests its use as a watch chain or waist-hung appendage. The location of materials from Oudepost I shows that they are distributed all over the site, but that most come from the fort area in Level I (Table 9.2). This is similar to the distribution of gunflints but differs from the distribution of other artifacts, such as food remains and coarse earthenwares that concentrate in the Lodge. Since the buckles and buttons were dislodged from clothing, it may be that their distribution in the fort reflects the fact that more stress and activity took place here than in the Lodge. Table 9.2. Oudepost I (OPI) distribution of personal adornments by Unit and Level Object Unit Fort The jewelry in the Oudepost collection includes two finger rings. The first (R 1) is a copper alloy ring with a stone set in a cabochon setting (Figure 9.17). It may have been a signet ring, but the condition of the stone is poor and the surface has deteriorated. The second (R 2) is a copper alloy band decorated with etched patterns (Figure 9.18). Though this ring resembles a modern wedding band, it could have served a variety of symbolic functions in the 17th and early 18th century (White 2005: 94-96). It may have been a symbol of wealth, a mourning ring, a wedding ring, or a betrothal ring. In addition, although there is no hint of Indian or Asian beadmaking, two broken agate rings of South Indian manufacture are noted (Karklins and Schrire 1991: 67). Figure 9.17. Copper alloy ring (R1) from the Oudepost I excavation. Scale in cm. (Photo C. White) Figure 9.18. Copper alloy band (R2) from the Oudepost I excavation. Scale in cm. (Photo C. White) Lodge Level NP X Total I II Nos. % 1 1 2 .75 Aglets 2 Buckles 9 11 3 2 8 10 23 17.4 Buttons 57 39 7 14 53 29 103 78.0 Button set 2 2 2 1.5 Chain 1 1 1 0.8 Mold 1 1 0.8 1 Rings 2 70 50 12 2 17 64 39 .75 132 100.0 Summary and Conclusions The assemblages from the Castle Moat and from Oudepost I exhibit general similarity in form and appearance. The Castle Moat assemblage possesses a more limited range of materials than the Oudepost assemblage, but as in the case of other personal adornment assemblages (see White 2008, 2004) these differences may be a product of preservation and excavation strategies more than anything else. Mentzel’s description of the Cape’s inhabitants provides great detail for the military uniforms worn by the men (1925 [1785]: 59-61). They were worn only on special occasions, however, and the artifactual record contains little or no evidence of the expensive items associated with the officers, adjutants, sergeants, and others with distinctive uniforms. The personal adornments from the two VOC sites are dominated by clothing fasteners. Overall they reflect the appearance of the majority of the presumably European inhabitants of the site, who were males, both men and OBJECTS OF P ERSONAL ADORNMENT boys. The archaeological evidence suggests a relative lack of diversity in the clothing worn at these sites. Most of the buttons, and buttons comprise the bulk of the personal adornment, are very simple forms with smooth surfaces, suggesting that most of the VOC men were adorned in similar garb or in uniform. At the same time, within the assemblages is a smaller number of clothing fasteners and jewelry that suggest variation in garb across the population, likely across associated military ranks. Some buttons and buckles are decorated with rosettes, floral patterns, nippled decorations, and geometric designs, which would have set the wearers apart from the bulk of the group. The very small number of additional artifacts of personal adornment, finger rings and a watch chain, further underscores the presence of physical differentiation among members of the group. Mentzel’s description of the uniforms also emphasizes the associated class differentiation. The dominance in each assemblage of simple undecorated buttons likely correlates with the dominance of low ranking soldiers at the sites. As described elsewhere, the communities were largely divided into two classes (Jordan and Schrire 2002: 255), and the plain clothing that was associated with the simple buttons would have underscored that rank difference. Supply lists from the 17th century list a wide range of textiles and other clothing articles that were imported to the Cape. While the archaeological evidence does not preserve any of the textiles to compare with these lists, there are records of silvered and copper gilt buttons imported August 18, 1676 (Leibbrandt 1902: 277) and “hair buttons for coats” imported December 7, 1699 (ibid.: 236). The hair buttons likely degraded in the soil, but there is ample evidence of the metal buttons recovered on these sites. The personal adornment assemblages provide an image of the individual appearance of many of the elite inhabitants of the VOC sites. While the plain undecorated buttons suggest that many of the inhabitants dressed in ways that associated them with portions of the larger ‘unqualified’ class, the fancy shoe buckles, small buttons with cast rosette designs, and unusual articles of clothing associated with some of the artifacts found in small numbers reinforces the rank of those people in the ‘qualified’ class. There is a marked variation across the decorated items, and while there are some unifying themes in decoration that have to do with popular styles of the period, there is also a notable amount of differentiation within the decorated artifacts. The men of higher rank would have also differentiated themselves from one another through the dress they wore. There has been very little written about South African (or Dutch for that matter) clothing in the 17th and early 18th century. In this chapter I have focused on describing and illustrating in detail the kind of clothing worn on the VOC sites, rather than using the personal adornment to further 211 understand the sequencing of the site (see Chapter 2 for detailed stratigraphic sequences). As more of this material is uncovered and published, further refined sequences and illustrations of the variability and continuity of personal appearance through clothing and adornment will be possible. Acknowledgements Many thanks to Carmel Schrire for inviting me to study the personal adornment of the VOC. References Abbitt, M. W. (1973). The eighteenth-century shoe buckle. In A. Noël Hume, M. W. Abbitt, R. H. McNulty, I. Davies & E. Chappell (co-authors), Five Artifact Studies (pp. 25-53). Williamsburg, VA: Colonial Williamsburg. Edwards-Ingram, Y. (1999). The recent archaeology of enslaved Africans and African Americans. In G. Egan & R. L. Michael (Eds.), Old and New Worlds (pp. 155-164). Oxford: Oxbow Books. Egan, G. & Pritchard, F. (1991). Dress accessories c. 1150-1450: Medieval finds from excavations in London, Vol. 3. London: HMSO. Galle, J. E. (2004). Designing women: Measuring acquisition and access. In J. E. Galle & A. L. Young (Eds.), Engendering African American archaeology: A Southern perspective (pp. 3972). Knoxville: The University of Tennessee Press. Heath, B. J. (1999). Buttons, beads, and buckles: Contextualizing adornment within the bounds of slavery. In M. Franklin & G. Farrett Fesler (Eds.), Historical archaeology, identity formation, and the interpretation of ethnicity (pp. 47-69). Williamsburg, VA: Colonial Williamsburg Research Publications. Jordan, E. G. (2006). “From time immemorial:” Washerwomen, culture, and community in Cape Town, South Africa. (Unpublished Ph.D. dissertation). Department of Anthropology, Rutgers, The State University of New Jersey, New Brunswick. Jordan, S. C. & Schrire, C. (2002). Material culture and the roots of colonial society at the South African Cape of Good Hope. In C. L. Lyons & J. K. Papadopolous (Eds.), The archaeology of colonialism: Issues and debates (pp. 241272). Los Angeles: The Getty Research Institute. Karklins, K. & Schrire, C. (1991). The beads from Oudepost I, a Dutch East India Company outpost, Cape South Africa. BEADS 3, 61-72. Leibbrandt, H. C. V. (1902). Précis of the Archives of the Cape of Good Hope: Journal, 1671-1674. Cape Town: W. A. Richards & Sons. 212 HI S TO R I C A L A RC HA E OLOGY I N SOU TH A FR I C A Mentzel, O. F. (1925) [1785]. A complete and authentic geographical and topographical description of the African Cape of Good Hope, Part Two, (H. J. Mandelbrote, Trans.). Cape Town: Van Riebeeck Society 6. Mould, P. (1979). The English Shoe Buckle. Neston, England: Leemans Seel House. Mullins, P. R. (1999). Race and affluence: An archaeology of African America and consumer choice. New York: Kluwer Academic/Plenum Publishers. Noël Hume, I. (1961). Sleeve buttons: Diminutive relics of the seventeenth and eighteenth centuries. Antiques 79(4), 380-83. Swann, J. (1981). Shoe buckles: Catalogue of shoe and other buckles in Northampton Museum. Northampton, Canada: Northampton Borough Council Museums and Art Gallery. Thom, H. B. (1952). Journal of Jan van Riebeek, Vol. 1 (16511655). Cape Town: A. A. Balkema. White, C. L. (2004). What the Warners wore: An archaeological investigation of visual appearance. Northeast Historical Archaeology 33, 39-66. White, C. L. (2005). American artifacts of personal adornment, 1680-1820: A guide to identification and interpretation. Lanham, MD: AltaMira Press. White, C. L. (2008). Personal adornment and interlaced identities at the Sherburne site, Portsmouth, New Hampshire, Historical Archaeology 28(2), 17-37. Wilkie, L. A. (2003). The archaeology of mothering: An AfricanAmerican midwife’s tale. New York: Routledge. C H A P T E R 10 Metals and Associated Artifacts from Oudepost I, Cape Carmel Schrire Introduction Metal finds from Oudepost I include iron, copper, brass, bronze, pewter, and lead (see catalog). This chapter does not deal with the large iron collection, which has yet to be analyzed and conserved. The metal artifacts used for personal adornment are discussed in Chapter 9. For ease of analysis I have separated the copper, brass, bronze, tin, and pewter from the lead, which is described later in the chapter, together with the molds that were mainly used to produce lead shot. Copper, Brass, Bronze, Tin, and Pewter Although copper is often subsumed with scrap alloys or brass in both the early Cape accounts (Raven-Hart 1967) and also in colonial metal analyses (Hudgins 2005: 2, 21 footnote 5), it is discussed as a separate category here (Tables 10.1, 10.2). The Oudepost I copper utensils include a skimmer or strainer (C 1; Figure 10.1), a dish (C 2; Figure 10.2), and a barrel strap (C 3), as well as rivets and washers (C 214-226, 229, 329), nails and tacks (C 302-313), a fish hook (C 332; Figure 10.3), possible bandolier caps (C 334, 335, 342), hinges and hardware from books (C 338, 339, 343), a fork (C 345), and a hand cut spur (C 349; Figure 10.4). Twenty two copper beads from Oudepost are described elsewhere (Karklins and Schrire 1991: 67), ten of which can be broadly dated because they resemble those from the wreck of the De Liefde (1711) (Bax and Martin 1974; Karklins 1988). The small brass series includes clasps or hinges from books (Br 4, 9, 12; Figure 10.5) and two key escutcheons (Br 7; Figure 10.6). The bronze artifacts include a fragment of a small scissors (Bz 2), two gun sights (Bz 6, 7), a possible spigot (Bz 12), and the handle of an eating utensil initialed ‘SVDM’ (Bz 16; Figure 10.7). Finally there is a scrap of tin and two pewter lids (P 1, 2) off measures (Welz, 1997: 30-31, Pl. 136). The distribution of identifiable objects is shown in Tables 10.1, 10.2, and 10.3. They are very few in number, but their presence carries important connotations about the garrison posted there. For example, book clasps and hinges probably came from Bibles or religious tracts, which, together with the archival correspondence and the presence of slate pencils, points to literate members of the garrison posted there (Schrire 1990: 13-15; Sleigh 1993: 411-468). The escutcheons might denote the presence of private property. Finally, if we exclude the burial found at Oudepost (Schrire et al. 1990), the initialled handle of a piece of flatware (Bz 16) constitutes one of only two personalised objects found here—the other being a lead cutlery handle described later in this chapter (L 19). Unfortunately, VOC records listing the occupants of the post at various times failed to turn up anyone like a Van der Merwe, let alone one with these initials (Remco Raben 1998: pers. comm.). Carmel Shrire, “Metals and Associated Artifacts from Oudepost I, Cape“ in Historical Archaeology in South Africa: Material Culture of the Dutch East India Company at the Cape, pp. 213–220. © 2014 Left Coast Press, Inc. All rights reserved. 213 214 HI S TO R I C A L A RC HA E OLOGY I N SOU TH A FR I C A Table 10.1. Oudepost I distribution of metal artifacts by Area and Level (numbers) Metal Area (nos.) Fort Lodge DP 31 20 - 199 50 5 Copper artifacts Copper fragments Brass Bronze Level (nos.) NP Total (nos.) % X I II 10 12 22 17 61 3.6 - 38 106 109 34 287 17.0 6 - - 2 3 6 11 0.7 14 2 - - 4 11 1 16 0.9 Tin 1 - - - 1 - - 1 0.1 Pewter - 1 - 1 - 1 - 2 0.1 Lead artifacts 117 119 2 21 35 129 72 259 15.4 Lead shot 984 59 - 7 14 928 101 1050 62.2 1351 257 2 77 174 1203 231 1687 99.7 Total Table 10.2. Oudepost 1 distribution of metal artifacts by Area and Level (weight) Metal Copper artifacts Copper fragments Area (gms.) Fort Lodge 320.1 238 - 199.7 1249.4 354.1 - 218 Brass Bronze Total NP X % I II 65 270.4 222.7 757.8 6.5 542.5 820.1 240.9 1821.5 15.7 19.4 13.3 - - 4.9 14.5 13.3 32.7 0.3 102.8 - 6.9 21.3 165.7 100.3 294.2 2.5 0.6 - - - - 0.6 - 0.6 - 4.8 - 39.5 - 4.8 - 44.4 0.4 3608.1 2962.9 120.7 638.5 358.9 4838.5 1373.6 7330.2 63.2 856.6 312.9 - 141.8 70.4 891.4 207.7 1311.3 11.3 6238.7 3988.8 120.7 1244.5 1063 7006 2158.5 11592.7 99.9 Pewter Lead shot DP Total (gms.) 184.5 Tin Lead artifacts Level (gms.) x Table 10.3. Oudepost I distribution of copper artifacts by Area and Level (numbers and weight) Metal Area Level Total (nos./ gms.) % Fort Lodge DP NP X I II 31 20 - 10 12 22 17 61 17.5 Copper fragments nos.) 199 50 - 38 106 109 34 287 82.5 Total 230 70 - 48 118 131 51 348 100.00 320.1 238 - 199.7 65 270.4 222.7 757.8 29.4 Copper fragments (gms.) 1249.4 354.1 - 218 542.5 820.1 240.9 1821.5 70.6 Total 1569.5 592.1 - 417.7 607.5 1090.5 463.6 2579.3 100.0 Copper artifacts (nos.) Copper artifacts (gms.) METAL S AND ASSOCIATED ARTIFACTS Figure 10.1. Copper skimmer or strainer from the Oudepost I (OPI) excavation (C 1). Scale in cm. (Photo C. Schrire) Figure 10.2. Copper dish from the Oudepost I (OPI) excavation (C 2). Scale in cm. (Photo C. Schrire) Figure 10.3. Copper/?alloy fish hook (C 332) from the Oudepost I (OPI) excavation compared with similar one (C 333) from Castle Moat (M90) excavation. Scale in cm. (Photo C. Schrire) 215 Figure 10.5. Brass clasps or hinges from books from the Oudepost I (OPI) excavation (Br 4, 9, 12). Scale in cm. (Photo C. Schrire) Figure 10.6. Brass octagonal plate (Br 6) and key escutcheon (Br 7) from the Oudepost I (OPI) excavation. Scale in cm. (Photo C. Schrire) Figure 10.7. Bronze utensil handle inscribed ‘SVDM’ (Bz 16) from the Oudepost I (OPI) excavation. Scale in cm. (Photo C. Schrire) Figure 10.4. Copper spur from the Oudepost I (OPI) excavation (C 349). Scale in cm. (Photo C. Schrire) In contrast to the recognizable artifacts, the vast majority of the copper collection is classified as ‘Fragments’ (Tables 10.1, 10.2). They constitute 82.5% by number and 70.6% by weight of the total copper finds (Table 10.3). Most of these are pieces of sheet copper, some of which was pierced and folded and then cut with shears and chisels into various shapes (Figure10.8). Some are patent scraps or cutouts echoing the extraction of squares and circles (Kelso 2006: 179). In addition, there are irregular and curved pieces that were apparently parts of kettles and pans, with one (C 195) being a leg fragment from a tripod pot. Figure 10.8. Fragments of chisel cut sheet copper possibly used in trade, from the Oudepost I (OPI) excavation (C 258, 260, 262, 272, 285). Scale in cm. (Photo C. Schrire) Where the distribution of all fragments at Oudepost is concerned, there are over three times as many by number and weight in the fort than the lodge (Tables 10.2, 10.3), suggesting that this may have been where metal sheets and broken metal objects were reduced to small pieces. The purpose of this might have been to patch and repair other broken objects, or to create tokens for exchange in the VOCKhoekhoen copper trade. The Oudepost fragments resemble those found in other colonial sites worldwide, including the French Fort Pentagoet in Maine and the fort at the APVA excavation at Jamestown, Virginia (Faulkner and Faulkner 1987; Kelso 2006: 179). A recent study of trace elements using inductively coupled-plasma atomic emission spectrometry (ICP-AES ) shows that most of the unalloyed copper at Jamestown came from England and Sweden, with the sheet brass emanating from Continental Europe (Hudgins 2005: 43). Furthermore, although some of the scraps were undoubtedly traded, most of the material appears to have been industrial waste that was generated in English factories during the production of kettles, bowls, and lanterns, and 216 HI S TO R I C A L A RC HA E OLOGY I N SOU TH A FR I C A shipped out to Virginia for processing with Virginian zinc to produce brass (Hudgins 2005: 51-52; Kelso 2006: 178-8). In any case, Virginia did not produce zinc and the venture was abandoned together with the large heaps of scrap metal that now lie buried alongside the failed crucibles (Hudgins 2005: 60-61; Kelso 2006: 181). There is no evidence at present that the VOC shipped industrial waste to the Cape, but at the same time, no Cape collection has yet been systematically sourced. The early trade copper carried to the Cape probably emanated from European mines in Sweden, as well as from Germany and Slovakia (Hudgins 2005: 38). The Fugger family played a powerful role in the distribution of Central European copper not only throughout Europe but also in Asia. Their wealth is strikingly underscored by the discovery of an immense cargo of hemispherical copper ingots stamped with the tripod seal of the Fuggers in the cargo of a 16th century Portuguese ship that was wrecked off the Namibian coast en route to purchase pepper in Asia (Jensen 2008; Pringle 2009: 25; Werz 2011). Whether shipments such as these ever came to the VOC Cape is unknown, but Japanese copper is specified in the 17th century VOC storehouses there (Leibbrandt 1902: 275). The VOC was deeply involved as a middle-man in the Japanese copper trade in the 17th-18th centuries. It held a monopoly in South Asia, China, the Middle East, Indonesia, and India, where it shipped raw copper from Japan to factories for rendering into coinage. In fact, the importance of Asian copper to the VOC was such that the Van Imhoff, the Governor General in Batavia, referred to it as the “bride” for whom the VOC danced (Shimada 2006, 2007). What is known for the Cape is that a booming trade in copper and brass began almost as soon as Europeans first rounded the Cape in the late 15th century. Archival references document the presence of three kinds of copper (sometimes called ‘brass’)—native, European, and Asian. Native copper was seen by Da Gama in 1497 as small copper beads worn in the ears of the people of St Helena Bay (Axelson 1998: 24). There is mention in 1608 of a bracelet that the Khoekhoen might have made themselves (Raven-Hart 1967: 39), and red copper bracelets, listed as being on the south coast in 1595, might also have been locally traded (RavenHart 1967: 18). European metal might be denoted in a cattle exchange for “…a little Brasse cut out of two or three old Kettles” at the Cape (Raven-Hart 1967: 56, 59). An early source specifies that thin pieces, two inches square cut out of a kettle, were fashioned into arm rings (Raven-Hart 1967: 48). As the demand for a hard and shiny material grew, copper and iron became less valuable than brass (Raven-Hart 1967: 73). Brightness was paramount and scraps no longer sufficed; natives demanded pieces as big as their cloaks, as well as housewares like pans, basins, and skimmers (RavenHart 1967: 64-67, 70; Elphick 1985: 78). By 1615, the size of sheet copper in an exchange had escalated to great pieces “1 ½ feete square, or one foote at the least” (Raven-Hart 1967: 73). For what it’s worth, an early reference to a copper bracelet traded at Table Bay in 1608 might actually denote an Asian ring in that it came from a Dutch ship homebound after a voyage to the East Indies (Raven-Hart 1967: 37-38), and later sources specify Japanese copper in the VOC stores (Leibbrandt 1902: 275). Copper and brass functioned alongside alcohol and tobacco as mainstays in a cattle trade that shaped the future course of indigenous-colonial relations until today (see Elphick 1985; Elphick and Giliomee 1989; Heinrich and Schrire 2011; Schrire 1995, 2009; see also Chapter 3). It began in 1497 with the purchase of meat for metal from the indigenous Khoekhoen (Axelson 1998: 23-25). Iron was initially preferred by the Khoekhoen because of its resilience when used as knives and weapons, but once the VOC settled at the Cape, indigenous people preferred to trade for copper, which was soft and malleable and could be fashioned into highly visible jewelry (Elphick 1985: 77, 166-67). The early diaries document a wide variety of forms including wire, staves, plates, bars, and beads (Leibbrandt 1901: 7, 38, 81,123, 193) as well as red plate and yellow bar copper (ibid.: 123). So intimately was the cattle trade linked to copper that strings of copper beads, like glass beads and twists of tobacco, were meted out as ‘quispelgreijn,’ an amalgam of ‘quispel’ (the tuft at the end of a cow’s tail), and ‘greijn’ (a bead), denoting that the payment, whether in beads or tobacco, was measured as the length from the tip of the tail, over the back, to the front of the bartered animal’s horns (Leibbrandt 1901: 99; Thom 1958: 432-433; see also Karklins and Schrire 1991). Our most specific listing of the metal trade comes from an 1676 inventory of the Castle warehouses, made several years after Oudepost I was manned, which includes vessels like kettles, cooks’ pans and spoons, watering can (Leibbrandt 1902: 274), as well as “6264 lbs. copper, viz.: 787 lbs plate copper, consisting of 651 lbs. small yellow copper plates; 136 red do.; 2620 lbs. wire copper, e.g., 1891 lbs. yellow and 720 lbs. red; 262 lbs. old copper; 530 lbs. chips; and 2065 Japan rings” (ibid.: 275). As though this were not enough, the record also specifies two “large iron scissors for cutting copper” (ibid.: 274). These sources never document industrial waste copper such as was shipped to Jamestown. The Oudepost copper was always intended for trade, as can be seen in the first list of provisions that were dispatched to the outpost and that itemizes 140 lbs of copper (50 lbs sheet, 90 lbs wire) and seven copper kettles (VOC 4004; see Schrire 1995: 94-95). In 1672, 100 chains of copper beads were supplied to the post (VOC 4008 f. 453), and a further 1000 copper beads went off in October 1700 (VOC 4047 f. 539v). An inventory made at the post in 1729, three years before its relocation, listed a single copper kettle, suggesting that earlier vessels had probably disappeared into the bottomless pit of the exchange system (Western Cape Archives and Records Service LM 20: 121). METAL S AND ASSOCIATED ARTIFACTS Figure 10.9. Rolls of sheet lead from the Oudepost I (OPI) excavation (L 3a-c). Scale in cm. (Photo C. Schrire) Figure 10.10. Lead toys from the Oudepost I (OPI) excavation (L 2a,b). Scale in cm. (Photo C. Schrire) 217 Figure 10.12. Lead cutlery or flatware from the Oudepost I (OPI) excavation (L 18, 17, 19, 22). Scale in cm. (Photo C. Schrire) Figure 10.13: Close-up of lead handle from the Oudepost I (OPI) excavation (L 19) inscribed ‘M C X.’ Scale in cm. (Photo C. Schrire) Figure 10.11. Lead discs from the Oudepost I (OPI) excavation (L 6ad). Scale in cm. (Photo C. Schrire) Lead The lead remains from Oudepost include finished artifacts and manufacturing debris (see catalog; see also Tables 10.1, 10.2). Raw material includes rolled and unrolled sheets (L3 (Figure 10.9), L4), as well as 133 lumps of lead slag (L16) that vary in size from tiny splatters weighing less than a gram to immense lumps, two of which weigh 604.2 gm (L 248) and 1761.9 gm (L 249). Smelting is evident in 49 fragments of bullet sprues (L 9) that emanate from the bullet molds described later in this chapter. In addition there are two lids (L1, L14) that possibly come from measures and that are comparable to the pewter examples (P1, P2) listed earlier (see Welz 1997: 30-31 Pl. 136). Two possible whirligigs, a toy called ‘woer’ in Afrikaans and ‘snorrebot’ in Dutch (L 2a, b), are present ( Figure 10.10), as are a pierced disc (L7) and four cut sections of discs (L 6) that might have served as counters or weights (Figure 10.11). Rectangular fragments (L10) are also present. There are two musket balls with drilled holes (13a, b) that might have been fishing weights or decorative objects, and six folded oval pieces (L 15a-f) that might be snaplocks (Geoffrey Egan 2007, pers. comm.) In addition, there are six pieces of cutlery, including three spoons (L17, 18, 22; Figure10.12). An archival source notes the presence of spoons at the post (LM20, 121) and the excavated objects look as though they might have been locally made, but in the absence of spoon molds, this cannot be verified. One handle (L 19) that carries the initials ‘M C’ over an ‘X’ (Figure 10.13) may represent the second personalised object from the site, the first being another handle, Bz 16, which was described earlier in this chapter. A single lead bead made from a narrow strip of sheet lead rolled into a cylinder is described elsewhere (Karlins and Schrire 1991: 67). The largest numerical category of lead artifacts includes 1050 fragments of lead shot. Four types are present, including cast shot, made in a mold (892 fragments: 85%); cut shot, made from strips or rolls for later rounding (42 fragments: 4%); Prince Rupert shot, made using arsenic (29: 2.8%); and dribble, made in a ladle containing dust (27: 2.6%). The cut strips in the artifact listing (L 8, 12) may have been on their way to becoming cut shot. We can make a guarded estimate of the size or caliber of the firearms used here from the size of the shot. It is hard to estimate precisely the size of the gun that fired each ball, because shot expands after firing, as seen in a few patently misshapen balls (L 13a, b). Bearing in mind that small shot need not denote a tiny gun in that the bearer might cram small shot into a large barrel, it would nevertheless seem that there were at least four sizes of guns being used here, namely, pistols (0.19-0.52”), carbines (0.53-0.65”), muskets (0.69-0.75”), and ramparts (0.83-0.92”). The vast majority of the shot is of the smallest, pistol size (0.19”) BBB shot (E. H. H. 1961). This is consistent with the analysis of the Oudepost gunflints, where 62.2% might have been have been used in such pistols (Chapter 11; Table 11.5). It also conforms with the associated bullet molds (discussed in the next section), all of which have very small shot holes (0.19-0.25”). 218 HI S TO R I C A L A RC HA E OLOGY I N SOU TH A FR I C A Table 10.4. Oudepost I distribution of lead artifacts by Area and Level (numbers and weight) Metal Area Level Total (nos./ gms.) % Fort Lodge DP NP X I II Lead artifacts (nos.) 117 119 2 21 35 129 72 259 19.8 Lead shot (nos.) 984 59 - 7 14 928 101 1050 80.2 Total 1101 178 2 28 48 1057 173 1309 100.0 3608.1 2962.9 120.7 638.5 358.9 4838.5 1373.6 7330.2 84.8 856.6 312.9 - 141,8 70.4 891.4 207.7 1311.3 15.2 4464.7 3275.8 120.7 780.3 429.3 5729.9 1581.3 8641.5 100.0 Lead artifacts (gms.) Lead shot (gms.) Total The distribution of lead in the site mirrors that of copper to some extent (Table 10.4). It predominates at the fort by numbers, mainly due to a marked concentration of small shot there, which constitutes 79% of the shot in the entire site. The presence of several large melted pools in the lodge, especially in Level I, tips the distribution by weight and points to smelting in both the fort and lodge. Further insight into the lead assemblage at Oudepost may be found in the archival records. Lead was used for paint, repairs, bullets, window glazing, bullets, and repairs. It served as a base for ceramic glazes, and a monthly supply list of the Company dispatches 12 pounds of “lead… to the potters” (ARA1684: 676; see Chapter 5). Archival records list the stocks of lead held at the Castle, as well as those consigned to the post. The Castle warehouse lists of 1676 specify “22,077 lbs lead e.g. 10,967 lbs. ‘Schuyt’ lead, 11,110 lbs. flat do” and specify 3078 lbs flat lead and 400 lbs white lead (Leibbrandt 1902: 275). These technical names probably refer to lead used for mending ships (schuyt is a small flat bottomed ship) and for paint derived from corroding lead into white lead. The requisitions for the Saldanha Bay outpost, Oudepost I, between 1669 and 1729 include shipments weighing between 15 lbs and 100 lbs and specify 100 lbs of lead ‘pigs,’ or ingots, in 1701 (VOC 4047 f. 586), 30lbs of flattened lead in 1701 (VOC 4047 F, 586), and 100 lbs to mold bullets in 1669 (VOC 4004 f. 591v-592v). Stone Molds The collection of stone molds was originally examined by Professor Donald Baird of Princeton University in 1988, whose technical comments are incorporated here. There are 18 fragments labelled ‘BM’ for ‘Bullet molds’ in the stone molds catalog. Some may not have been parts of molds but are included here for want of a better category. They are all made of fine grained shale or slate, ranging in color from dark grey to grey/green, as well as one reddish example. All are broken, some more than others, but all have distinguishing holes and grooves. Their source is not certain; they may have been imported, but since similar rocks are present at the Cape and many of the molds are roughly made, they might have been produced at the Cape, if not on-site. The most complete examples (B 1, 3) have holes cut with a professional cherry or bulbar file, with the sprue and ball apparently cut at the same time (Figures 10.14, 10.15). Many have irregular spacing, and incomplete holes and wells, and in one case (BM 5) the wells are way out of round, suggesting that the cherry wobbled in inexpert hands (Figure 10.16). Figure 10.14. Bullet mold from the Oudepost I (OPI) excavation (BM 1), shown with associated shot and pulls. Scale in cm. (Photo C. Schrire) Figure 10.15. Bullet mold from the Oudepost I (OPI) excavation (BM 3). Note that one hole is spiked with lead. Scale in cm. (Photo C. Schrire) Figure 10.16. Bullet mold from the Oudepost I (OPI) excavation (BM 5). Peg hole visible. Scale in cm. (Photo C. Schrire) METAL S AND ASSOCIATED ARTIFACTS The size of the shot produced in the unweathered holes ranges from about 0.125» to 0.25», which is consistent with the small shot and lead pulls (L 9) found on site. As noted earlier, the size of shot that would have been produced in these molds is way below the normal caliber (0.52») of pistol shot and was probably crammed into guns of various sizes. The distribution pattern is not very clear, and seven have no definite provenance. The remaining 11 are distributed mainly in the fort. The presence of bullet molds is consistent with the abundant evidence of lead smelting in this site, as well as with many associated artifacts including rolls and sheets of lead, bullet pulls, sprue, gun sights (B7, 8), snaplocks (L 15), and numerous gunflints. The analysis of fauna points to a plethora of wild animals that must have been taken with guns (Cruz Uribe and Schrire 1991). Archival records attest to the provision of lead, and the original supply list specifies six bullet molds and 100 lbs of lead to mold bullets (VOC 4004). Exactly what these molds looked like is not known, though a variety is listed in the stores of the Castle, including “30 moulds, e.g. 8 metal bullet moulds, two for shot, and 22 single bullet moulds” (Leibbrandt 1902: 272), as well as “two copper cartridge moulds” (ibid.: 273). Conclusion The metal remains at Oudepost include a number of raw materials. A cursory examination of the unconserved iron from the site showed the presence of a wide variety of materiel including pulleys, pintles, locks, cannon balls, trowels, a mattock, and a shovel (Schrire 1995: 94). A future scholar might well expand our grasp of the post with a detailed analysis of those remains. The non-ferrous artifacts described in this chapter include objects made elsewhere for use on the post, as well as copper, lead, and stone molds that were used to manufacture artifacts on site. The collections attest to artifact repairs and maintenance and to shooting, hunting, fishing, and trade, as well as to certain aspects of adornment like belts and laces (see Chapter 9). The key escutcheons and initialled items of flatware speak of personal and possibly private property. Reading and literacy is implied by the presence of book (possibly Bible) hardware. Books were more common in early Cape households than in later 18th centuries ones (Worden et al. 1998: 74), but it is nevertheless surprising to find such remains on a distant outpost of what was already a distant shore. Their presence confirms the well-known aphorism of VOC employees as ‘soldiers of the pen,’ and is in keeping with the fact that the post sent regular bulletins back to the Castle (Schrire 1990: 13-15; Sleigh 1993: 411-468). The small shot is particularly important where the relative abundance of bird species in the site is concerned (Schrire 1995: 107). Forty two species were identified according to their bones, with ostrich eggshell fragments raising the count to 43. The predominant species are Jackass Penguin, Cape Gannet, Cape Cormorant, Greater and Lesser 219 Flamingo, Egyptian Goose, South African Shelduck, Cape Teal, Cape Francolin, and Kelp Gull (Graham Avery 1989: pers. comm.). The list contrasts that found on a precolonial midden at Stofbergsfontein, about two kilometres away, where only 13 species of birds appear (Robertshaw 1978: 144-146). A radiocarbon date of main horizon of the midden places it around 1500 years ago when a similar environment prevailed in the region (Robertshaw 1978: 143). This suggests that although the occupants of both sites had access to similar resources, the use of guns and small shot at Oudepost quadrupled the hunting success as seen in the relative diversity of prey taken there. The supply of materiél to the outpost raises two final issues. The copper trade, unlike the exchanges of consumable like liquor and tobacco, involved a commodity that might be hoarded and traded as capital (Elphick 1985: 166-67). It is possible that it tipped the scales of wealth in indigenous groups, allowing men to buy more cattle or get more wives. The actual amount traded contrasts strongly with the number of metal artifacts recovered in our excavations, begging the question of where it all ended up. Some may have been transferred to the new post which was established in 1732, but the rest was probably broadcast widely across the land, scattered across the domain of the former pastoral Khoekhoen, and is waiting to be discovered in the archaeological traces of the villages, huts, and caves that were swamped in the bow wave of European settlement. Acknowledgements The identification of the shot and bullet molds discussed in this chapter was made by the late Professor Donald Baird, Geology Department, Princeton University. I owe him a debt of gratitude for his identification of this collection as well as its associated gunflints. I thank also Adam Heinrich for help with the identification of the artifacts and the late Geoffrey Egan for helping to identify the snaplocks. References Axelson, E. (1998). Vasco da Gama: The diary of his travels through African waters 1497-1499. Somerset West, Cape: Stephan Phillips (Pty) Ltd. Bax, A. & Martin, C. J. M. (1974). De Liefde: A Dutch East Indiaman lost on the Out Skerries, Shetland in 1711. International Journal of Nautical Archaeology and Underwater Exploration 3(1), 81-90. Cruz-Uribe, K. & Schrire, C. (1991). Analysis of faunal remains from Oudepost I, an early outpost of the Dutch East India Company, Cape Province. South African Archaeological Bulletin 46, 92-106. E.H.H (1961). Choosing shot size. The American Rifleman. February. 220 H I S TO R I C A L A RC HA E OLOGY I N SOU T H A FR I C A Elphick, R. (1985). Kraal and Castle: Khoikhoi and the founding of White South Africa. New Haven and London: Yale University Press. Elphick, R. & Giliomee, H. B. (Eds.). (1989). The Shaping of South African Society, 1652-1840. Middletown, CT: Wesleyan University Press. Faulkner, A. & Faulkner, G. (1987). The French at Pentagoet 1635-1674: An archaeological portrait of the Acadian frontier. Augusta, MN: Occasional Publications in Maine Archaeology Number Five and Special Publications of the New Brunswick Museum. Heinrich, A. R. & Schrire, C. (2011). Colonial fauna at the Cape of Good Hope: A proxy for the colonial impact on indigenous people. In M. Leone & J. M. Schlablitsky (Eds.), Historical archaeology and the importance of material things, Vol. II. ( pp. 119-139). Washington, DC: Society for American Archaeology. Hudgins, C. C. (2005). Articles of exchange or ingredients of New World metallurgy? An examination of the industrial origins and metallurgical functions of scrap copper at early Jamestown (c. 1607-17). Early American Studies: An interdisciplinary Journal 3(1), 32-64. Jensen, F. (2008). Uncovering Namibia’s sunken treasure. BBC News, Oranjemund, Namibia. 26 September 2008. Karklins, K. (1988). Beads from the wreck of the Dutch east Indiaman De Liefde (1711). Bead Forum 12, 11-17. Karklins, K. & Schrire, C. (1991). The beads from Oudepost I, a Dutch East India Company outpost, Cape, South Africa. Beads 3, 61-72. Kelso, W. (2006). Jamestown: The buried truth. Charlottesville and London: University of Virginia Press. Leibbrandt, H. C. V. (1901). Précis of the archives of the Cape of Good Hope. Journal, 1662-1670. Cape Town: W. A. Richards & Sons. Leibbrandt, H. C. V. (1902). Précis of the archives of the Cape of Good Hope. Journal, 1671-1674 & 1676. Cape Town: W. A. Richards & Sons. Pringle, H. (2009). Portuguese Indiaman/Namibia. Archaeology 61, (1): 25. Raven-Hart, R. (1967). Before Van Riebeeck: Callers at South Africa from 1488 to 1652. Cape Town: C. Struik (Pty.) Ltd. Robertshaw, P. T. (1978). Archaeological investigations at Langebaan Lagoon, Cape province. In E. M. Van Zinderen Bakker & J. A. Coetzee (Eds.), Palaeoecology of Africa (pp. 139-148). Rotterdam: A. A. Balkema. Schrire, C. (1990). Excavating archives at Oudepost I, Cape. Social Dynamics 16(1), 11-21. Schrire, C., Deetz, J., Lubinsky, D. & Poggenpoel, C. (1990). The chronology of Oudepost I, Cape, as inferred from an analysis of clay pipes. Journal of Archaeological Science 17, 269-300. Schrire, C. (1995). Digging through darkness: Chronicles of an archaeologist. Charlottesville: University Press of Virginia. Schrire, C. (2009). The material world of the English at Jamestown VA and the Dutch at the Cape of Good Hope. In W. Kelso (Ed.), Archaeology of early European colonial settlement in the emerging Atlantic world (pp. 75-86). Rockville MD: Society for Historical Archaeology, Special Publication Number 8. Shimada, R. (2006). The Intra-Asian trade in Japanese copper by the Dutch East India Company during the eighteenth century. Leiden: Koninklijke Brill NV. Shimada, R. (2007). Fight for the bride: The Inter-Asian competition for Japanese copper, 1690-1760. TANAP workshop, 2003. Xiamen. Ms. 10 pp. Sleigh, D. (1993). Die Buiteposte. VOC-buiteposte onder Kaapse bestuur 1652-1795. Pretoria: HAUM Uitgewers. Thom, H. B. (1958). Journal of Jan van Riebeeck: Vol. III, 16591662. Cape Town and Amsetrdam: A. A. Balkema. Welz, S. & Co. (in association with Sotheby’s) (1997). Decorating and fine arts including artifacts recovered from the wreck of the Oosterland. Sale catalogue May 27th-28th. Cape Town. Werz, B. (2011). Treasure ship, a world legacy: The Oranjemund shipwreck in Namibia. The digging stick 28(3), 20-21. Worden, N., van Heyningen, E. & Bickford-Smith, V. (Eds.). (1998). Cape Town: The making of a city. An illustrated social history. Claremont, Cape Town: David Philip Publishers. Archival Document Algemeen Rijksarchief, The Hague ARA1684: 676. Monthly Supply List, February 1684. VOC 4004. Overgekomen brieven en papieren van Kaap de Goede Hoop 1669. Folio 558-652. Journael en grootboeck van de winckel. VOC 4008. Overgekomen brieven en papieren van Kaap de Goede Hoop 1672. Folio 453. Journael en grootboeck van de winckel. VOC 4047. Specification book of expenses 1700 Sept. 11701August 31. Western Cape Archives and Records Service Western Cape Archives and Records Service LM 20. Precis and translation of Letters received 1726-1732. H. C. V. Leibbrandt c (1900). C H A P T E R 11 Analysis of Gunflints from VOC Sites at the Cape Jeffrey J. Durst Introduction In a collection of 520 lithic samples from the combined excavations at the Castle of Good Hope and Oudepost I, 196 have been identified as being complete enough for cataloging as intact or fragmentary gunflints. The collection offers a unique opportunity to examine gunflints brought to South Africa primarily through the Dutch East India Company or VOC trade beginning in the mid-17th and continuing into the late 18th century. An examination of the collection identified samples that can be grouped into several categories based primarily on distinguishing characteristics indicative of their differing sources of manufacture. Research conducted on the gunflints from these sites focused on three primary objectives. The first was to sort the collection into various categories based on a detailed attribute analysis. Previous researchers (Hamilton and Emery 1988) had established acceptable methods for categorizing gunflints based on size and morphological characteristics, and where applicable, gunflints from the Cape sites were categorized in a like manner to facilitate cross comparisons (see catalog). The second aim was to interpret what the attribute analysis was indicating about the process of production, and what if anything, this might tell us about where the gunflints were produced. Three previously analyzed collections were selected as a comparative data set: the first from the shipwreck of the French vessel La Belle (1687) recovered off the coast of Texas (Durst 2009: 20), the second from the French site of Fort St. Louis established in 1685 along the Texas coast (ibid.: 20), and the third from the English site of Fort Frederica (1736-42) located on St Simon’s Island off the coast of Georgia (Hamilton and Emery 1988: 28). All three collections are roughly contemporaneous with the Cape sites, with comparable ratios of spall- to blade-type gunflints. The third objective was to try and establish the country of manufacture using methods that would be scientifically verifiable, as opposed to the intuitive methodologies used in several previous efforts to source gunflints (Witthoft 1966; Hamilton and Emery 1988). Historical Background The earliest production of gunflints is poorly documented and little is known of their production centers. The need for gunflints arose around the 1530s with the invention of the German snaphaunce (Brown 1980: 68), followed by the Spanish snaplock, which was essentially a simplified version of the original snaphaunce (Lavin 1965: 148). The basic flintlock mechanism used at the Dutch sites of the mid-17th to late 18th century appears to have been developed in France around 1630 (Held 1970: 83). By the mid1600s, its production was commonplace throughout much of Europe. The newly refined and somewhat standardized mechanism demanded a more uniformly shaped flint for maximum performance (Lavin 1965: 159). While exact information regarding the early production of gunflints is very sparse, it appears that a burgeoning cottage industry soon emerged to fill the increasing demand for the uniformly shaped gunflints. By 1661 the British Board of Ordnance was placing thousands of orders, and by 1685 it was specifying particular sizes to be used in pistols, carbines, and muskets (Forrest 1983: 481). Jeffrey J. Durst, “Analysis of Gunflints from VOC Sites at the Cape“ in Historical Archaeology in South Africa: Material Culture of the Dutch East India Company at the Cape, pp. 221–229. © 2014 Left Coast Press, Inc. All rights reserved. 221 222 H I S TO R I C A L A RC H A E OLOGY I N SOU T H A FR I C A France and England were the two European countries with major production during the late 17th century and through the 18th century. While Brandon was the center of English, blade-type production starting around the time of the Napoleonic War (ca. 1790), earlier production of spalltype gunflints is believed to have been located elsewhere in England, where “…some twenty wedge [i.e., spall-type] manufacturing sites have been located in the south of England” (de Lotbiniere 1980: 155). If the predominance of spall-type over blade-type gunflints on most 17th century colonial sites is any indicator of market forces, it might be assumed that the nation that produced this type of gunflint was the controlling force at that time. After this time, and through most of the 18th century, the French apparently dominated much of the world market with their production of blade-type gunflints (White 1976: 27). In addition, the discovery of large nodules of very pure flint buried in veins 20-40 ft below the surface, just outside the village of Brandon, Sussex, allowed the English to produce blade-type gunflints and regain control of the market. Previous Research Witthoft published an early history of gunflint manufacture and distribution in which he proposed a four-stage chronology of European gunflint production (1966). The earliest stage gunflints, termed ‘Nordic,’ were produced from Danish flint, extensively worked on both sides, and were generally square or rectangular in shape (Witthoft 1966: 23-24). The second stage was attributed to Dutch makers primarily on the basis of the range of color in the source materials, and included ‘wedge shaped’ gunflints. The third stage was attributed to French makers and was characterized by ‘D-shaped’ or ‘horseshoe-shaped’ gunflints, manufactured using blade technology. The final stage was assigned to the English, who used blade technology, but who, unlike the French, did not round the heels of the gunflints. Subsequent research has greatly modified several of Witthoft’s conclusions. First, the ‘Nordic,’ bifacially worked gunflints that he believed to have been made from Danish source materials have since been recovered from numerous sites in the Americas and are now believed to have been made from locally procured sources, utilizing the same technology evident on lithic tools made by Native Americans (Kent 1983: 28-29). Second, there is no evidence to support Witthoft’s belief that the Dutch were actually producing gunflints themselves; instead, historical accounts suggest that the Dutch were purchasing gunflints from any available source and redistributing them for profit (Clarke 1935: 40; Forrest 1983: 51). Third, the ‘D-’, or ‘horseshoeshaped’ gunflints that were manufactured using blade tech- nology are probably French, but the French also produced blade-type gunflints that were not rounded at the heel. Finally, Witthoft’s assertion that the English blade-type gunflints were not rounded at the heel appears at this time to be correct. Analysis of Gunflints from VOC Sites at the Cape The Cape collections include gunflints from three sites. Two come from locales in the Castle of Good Hope, namely the Van der Stel Moat (M90) and the Old Granary (F2), and the third comes from the VOC outpost at Oudepost I (see Chapter 2). These sites overlap in time with the Moat (M90) dating from about 1690-1740, the Old Granary (F2) series from the late 17th-early 18th century, and Oudepost I from 1669-1732 (see Chapter 2 for details). Fragments recognizable as gunflints or byproducts of gunflint manufacture, include 151 from the M90, seven from F2, and 362 from Oudepost I. Of these, 196 are listed in our catalog. They display two primary manufacturing techniques—spall-type and blade-type—with most being attributed to European workshops where gunflints were being mass-produced. The majority of the raw materials used to produce the gunflints from both the Moat and Oudepost I sites appear to be identical, and the overall similarity of production types suggests that both sets come from shipments imported to the Cape by the VOC. The exceptions include seven gunflints from Oudepost I made of local silcrete. They appear to utilize a more expedient technology, and are generally seen to lack the characteristics associated with mass production. The presence of microscopic traces of iron on four of these confirms their use as gunflints or strikea-lights (see Schrire and Deacon 1989: 108-109, Fig. 4). Distribution The majority of the gunflints recovered from the Cape sites fit into Witthoft’s ‘Dutch, wedge shaped’ category, now commonly referred to as ‘spall-type’ (Table 11.1). The Van der Stel Moat collection includes 42 (65.6%) spall-type and 22 (34.4%) blade-type gunflints (Figures 11.1, 11.2). There are only two gunflints in the Old Granary collection. The Oudepost I collection includes 72 (55.4%) spall-type and 58 (44.6%) specimens, produced using the blade technique (Figures 11.3, 11.4). The distribution of different types of gunflints within the sites is shown in Tables 11.2, 11.3, and 11.4. Gunflints and fragments are scattered throughout the Van der Stel Moat (M90) deposit and are concentrated in the deepest and richest level A2, with no apparent sequence of spalls and blades (Table 11.2). 223 ANALYSIS OF GUNFL INTS Table 11.1. Incidence of spall and blade-type gunflint types at Cape sites Site Gunflint Type Spall Total Blade Nos. % Nos. % 42 65.6 22 34.4 64 100.0 Old Granary (F2) 1 50.0 1 50.0 2 100.0 Oudepost I (OPI) 72 55.4 58 44.6 130 100.0 Van der Stel Moat (M90) Total 115 Nos. 81 % 196 Figure 11.3. French, blade-type gunflints from the Oudepost I (OPI) excavation. Scale in cm. (Photo C. Schrire) Figure 11.1. French, blade-type gunflints from the Castle Moat (M90) excavation. Scale in cm. (Photo C. Schrire) Figure 11.2. English, spall-type gunflints from the Castle Moat (M90) excavation. Scale in cm. (Photo C. Schrire) Figure 11.4. English, spall-type gunflints from the Oudepost I (OPI) excavation. Scale in cm. (Photo C. Schrire) Table 11.2. Van der Stel Moat (M90) distribution of gunflint types by Level Unit GunflintType Spall Nos. Blade % Nos. Total Unclear % Nos. % Nos. % Fort 48 66.7 31 53.4 154 66.4 233 64.4 Lodge 18 25.0 20 34.5 59 25.4 97 26.8 DP (Dump) - - 1 1.7 3 1.3 4 1.1 NP (Non-Prov) 6 8.3 6 10.3 16 6.9 28 7.7 72 100.0 58 99.9 232 100.0 362 Total % 19.9 16.0 64.1 100.0 224 H I S TO R I C A L A RC H A E OLOGY I N SOU T H A FR I C A Table 11.3. Oudepost I (OPI) distribution of gunflint types by Unit Unit GunflintType Spall Nos. Blade % Nos. Total Unclear % Nos. % Nos. % Fort 48 66.7 31 53.4 154 66.4 233 64.4 Lodge 18 25.0 20 34.5 59 25.4 97 26.8 DP (Dump) - - 1 1.7 3 1.3 4 1.1 NP (Non-Prov) 6 8.3 6 10.3 16 6.9 28 7.7 72 100.0 58 99.9 232 100.0 362 Total % 19.9 16.0 64.1 100.0 Table 11.4. Oudepost I (OPI) distribution of gunflint types by Level Level Gunflint Type Spall Nos. % X 11 16.7 I 35 II 20 Total 66 % 20.0 Blade Nos. Total Unclear % Nos. 3 5.9 24 53.0 32 62.7 30.3 16 31.4 100.0 51 100.0 15.5 % Nos. % 11.3 38 134 62.9 201 60.9 55 25.8 91 27.6 213 100.0 330 100.0 64.5 Gunflints found at Oudepost I are markedly concentrated in the Fort (Table 11.3) and in Level I (Table 11.4), but the distribution of spall-types and blade-types is not correlated with place (Table 11.3) or time (Table 11.4) These conclusions are consistent with wider findings. The spall-type gunflint was originally thought to pre-date the blade-type gunflint (Witthoft 1966; Hamilton 1968; White 1975). Contrary to this view, recent findings suggest that both types occur in association at several sites, dating earlier1 than 1685, so that the time of transition from spalltype production to blade-type production may be difficult to pinpoint (Honerkamp and Harris 2005: 105). A number of sites have produced evidence for the early production of blade-type flints, including Chicoutimi, Québec, Canada, where they were found in association with spall-type gunflints in a sealed context with a terminus post quem of 1663 (Blanchette 1975: 43). Likewise, at Fort Pentagoet, Maine (1635-1674), 65% of the gunflints recovered were of the blade-type (Faulkner 1986: 83), and early blade-type gun- 11.5 100.0 flints were also recovered from the sealed context of La Belle shipwreck located off the Texas Gulf Coast, dating to 1687. The occupation dates for Oudepost I (1669-1732) postdate the presumed date for the initial production of bladetype gunflints, so it is of no surprise to find that both spalltype and blade-type gunflints appear together here. In the Castle Moat collection, the ratio of spalls to blades is 2:1 and blades constitute 34% of the recognizable flints there (Table 11.2). At Oudepost, the ratio of spalls to blades is 1.2:1 and blades make up 45% of the recognizable gunflints there (Table 11.3). These figures may be compared to those from other contemporaneous colonial sites in the New World. At the lower end of the spectrum the La Belle (1687) shipwreck collection contained only 7% blade-type gunflints, but the associated terrestrial site of Fort St. Louis contained 32% blade-type gunflints (Durst 2009). A recent examination by the author of 482 specimens from the site of Fort Frederica showed that blade-type flints constitute 33.4% of the collection, which corresponds more closely with the Cape sites. ANALYSIS OF GUNFL INTS Hamilton and Emery relate these size categories to their use in different types of guns, saying: Raw Material The blade- and spall-type gunflints recovered from the two South Africa sites were produced from different source materials. The spall-type gunflints have a broader spectrum of colors, ranging from light grey to an almost black color and from light tan to dark brown. Color has generally been found to be of little assistance in sourcing gunflints to their origins of production due to broad variations being noted among samples from a common source (Luedtke 1992: 118). Conversely, chert samples of visually identical material have been discovered from different sources across Western Europe and Great Britain. The blade-type gunflints from the Cape collections resemble those from La Belle, Fort St. Louis, and Fort Frederica, all of which were made from a nearly translucent, honey-colored chert. The exception to this rule came from several dark grey flints from La Belle; however, testing them in a solution of hydrogen peroxide showed that they were the same honey-colored chert as the examples from Fort St. Louis and, thus, that they had acquired their present color due to exposure to certain underwater contaminants for a period of 300 years (Durst 2009: 22). Musket flints are more than 34 mm from side to side. Fowler or carbine flints are between 34 and 28 mm from side to side. Trade gunflints are between 28 and 20 mm from side to side. Flints that are less than 20 mm can be from either trade guns or pistols (1988: 21). Overriding this seemingly rigid categorization is a certain amount of overlap in the sizes of gunflints that would make a particular type of flintlock mechanism fire adequately. In the Cape collections, only one gunflint (OPI-F2) was larger than 34 mm and falls into Category 1. According to Hamilton and Emery (1988: 20), “[f]lints larger than 34 mm are definitely military, but the probabilities are that many flints in the upper ranges of categories 1 and 2 could have served equally well in the French military locks of the period.…” Category 2 contains nine gunflints from 28 to 33.9 mm in width, which would fit Hamilton and Emery’s description of gunflints suitable for use in French military locks and would overlap with use in fowlers as well. Forty two gunflints fall into Category 3, which ranges from 25 to 27.9 mm in width. Hamilton and Emery suggest that flints of this size would be best suited for use in trade guns. The fourth category of gunflints, ranging from 18 to 24.9 mm in width, contains 88 specimens. Those smaller than 20 mm are possibly for use in pistols but might also have been used in trade guns (20 to 28 mm), according to Hamilton and Emery (1988: 21). The size range for gunflints recovered from the Cape sites may be compared with those from the La Belle (1687) shipwreck and from Fort Frederica (Table 11.6). Whereas La Belle has smaller flints similar to the pattern from the Cape sites, Fort Frederica has a much larger percentage of larger gunflints falling into Category 1 and Category 2, possibly indicating that the English soldiers there were outfitted with a higher percentage of ‘military’ arms. Size The gunflints from the Castle Moat and Oudepost I are sorted in the catalogs into four size-related categories based on parameters that delineate size according to width as measured from side to side (Hamilton and Emery 1988: 20). Only those samples which retain enough of their original conformity were included in these calculations. Of the 196 samples from the two South Africa sites, only 140 were deemed sufficiently intact to be measured (Table 11.5). While the size of the gunflints in these collections offers little evidence toward determining their source of production, it offers a certain degree of insight regarding the types of arms that were most likely to correspond to these sizes. Table 11.5. Distribution of size range for gunflints at Cape sites Site Gunflint Size (Category: Range in mm) Cat. 1: 34 mm + Nos. Van der Stel Moat (M90) % Cat.2: 28-33.9 mm Nos. Cat. 3: 25-27.9 mm Total Cat. 4: 18-24.9 mm % Nos. % Nos. % Nos. % 30.0 26 65.0 40 100.0 - - 2 5.0 12 1 50.0 1 50.0 2 100.0 Oudepost I (OPI) 1 1.0 7 7.2 29 29.6 61 62.2 98 100.0 Total 1 0.7 9 6.4 42 30.0 88 62.9 140 100.0 Old Granary (F2) 225 226 H I S TO R I C A L A RC H A E OLOGY I N SOU T H A FR I C A Table 11.6. Distribution of size range of gunflints from La Belle (1687) and Fort Frederica Site Gunflint Size (Category: Range in mm) Cat. 1: 34 mm + Nos. La Belle (1687) Fort Frederica 128 Cat.2: 28-33.9 mm Cat. 3: 25-27.9 mm Total Cat. 4: 18-24.9 mm % Nos. % Nos. % Nos. % Nos. % - 95 29.9 132 41.5 91 28.6 318 100.0 29.5 205 47.1 61 14.0 41 9.4 435 100.0 Sourcing A major problem in determining the various early sources for the production of gunflints from Europe is the homogeneity of the source material across much of Great Britain and Western Europe. The primary source for flint in this region is the Upper Cretaceous chalk that covers roughly two-thirds of southern England, extending northward across much of northern East Anglia, Lincolnshire, and southern Yorkshire, and even beyond, covering a large portion of central northern France as well as portions of Italy and other western European countries (Rockman et al. 2003). Until very recently, the only means for assessing the source of spall-type gunflints found in the New World was to use criteria established by T. M. Hamilton and his various coauthors. Hamilton’s criteria for gunflints from Fort Michilimackinac, an early 18th century French fort constructed along the southern shore of the Straits of Mackinac in the Great Lakes region of North America, and Fort Frederica were based primarily on color, luster, quality of the rock, and the presence or absence of pressure flaking along the sides and heel of the gunflint (Hamilton and Emery 1988: 28). Thus, a French spall-type gunflint would be of good, uniform quality, mined flint, mostly brownish but ranging in color from “…a gray through gray-brown to a light brown, then through a darker brown and eventually merging into black” (Hamilton and Emery 1988: 30). In contrast, an English spall-type gunflint would be made from an excellent black flint with no gloss, a minimum amount of retouch around the edges, and large flake scars on the heel. The main problem with the classification is that the specific chert sources have not been verified. This may now be addressed using Inductively-Coupled Plasma Mass Spectrometry (ICP-MS). ICP-MS is a type of mass spectrometry in which a sample material is atomized and ionized in a charged argon gas torch in order to detect almost the full suite of elements in the periodic table at concentrations as low as a few parts per million (Rockman et al. 2003: 3). ICP-MS is not a new technology, but improvements to the process have only recently increased its applicability to the problem of sourcing lithic materials. The two methods of preparing samples for testing are laser ablation and acid digestion. Laser ablation (LA-ICP-MS) is virtually nondestructive and is gaining in popularity over acid digestion, which results in the obliteration of the sample. Probably the most significant recent advance in LA-ICP-MS technology is the reduction of changes in the operating parameters, which previously fluctuated from one run to the next. Additionally, LA-ICP-MS now has “the ability to reproduce data generated by other bulk analytical techniques” (Speakman and Neff 2005: 4). Researchers have gathered core materials from a broad spectrum of potential source locations of gunflint production in France, England, and Denmark to serve as a data base in identifying the origins of the gunflints from colonial sites in the Americas. These core samples, along with 150 gunflints from the sites of La Belle shipwreck, Fort St. Louis, and Fort Frederica, were sent to the Institute for Integrated Research in Materials, Environments and Societies (IIRMES) at California State University, Long Beach, for LA-ICP-MS testing in order to determine if trace-element levels found in the sample set would be high enough to identify potential European flint sources. Because the European source samples submitted for testing were relatively few in number, the IIRMES lab included results from previously tested European flint sources. French, British, and Danish source areas were represented in substantial numbers, permitting a reliable assessment of the range of chemical variation within each area. Historical documentation strongly suggests that France and England are the two most likely sources for gunflints in the VOC sites at the Cape of Good Hope. Twenty-four gunflints from Oudepost I were tested against this preexisting data set to determine if there was any congruity in production source locations based predominantly on concentrations of uranium, yttrium, and cerium. The results are shown in Figures 11.5 and 11.6 and in the catalog under ‘Source.’ The first group (Texas Group 1) includes only samples from the site of Fort St. Louis and consists of gunflints believed to have been produced at the site by the resident Spanish soldiers or by nonmilitary residents, conceivably from core material gathered locally. This group showed significantly higher levels of uranium and, as suspected, displayed no overlap with submitted European source materials or with ANALYSIS OF GUNFL INTS 227 Figure 11.5: Bivariate plot comparing trace element levels of uranium and yttrium in the samples. (graph after Neff 2007) Figure 11.6: Bivariate plot comparing trace element levels of uranium and cerium in the samples. (graph after Neff 2007). 228 H I S TO R I C A L A RC H A E OLOGY I N SOU T H A FR I C A the tested Oudepost I samples. In contrast to this, the South African Group 1 contains five blade-type gunflints, all blonde in color (OPI-F11, F18, F29, F33, F58) and six spalltype gunflints (OPI-F61, F62, F68, F73, F74, F95). The five blonde specimens cluster within the ellipse that indicates a 90% statistical likelihood of association with the French quarry samples, and the trace element levels for these five samples define a virtually identical range that may be attributed with a high level of confidence to France and perhaps even to the specific quarry site in Meusnes (Neff 2005). The six spall-type gunflints in Group 1 fall outside the ellipse indicating association with the French quarry site. The two blade-type specimens found in Group 2 (OPI-F26, F99) did not group with the French quarry samples and were not visual matches with the blonde samples, and as they grouped outside the French ICP ellipse, it is likely that they derive from a different source. The spall-type gunflints found in South Africa Group 2 from Oudepost I (OPI-F41, F46, F59, F63, F65, F72, F82, F92, F93, F96, F101) cluster within a bivariate ellipse indicating a 90% likelihood that they derive from a common source. As this group has higher yttrium and cerium concentrations than most European quarry samples tested, it is suggested that they derive from a previously untested source. Given their overlap with British quarries, this is currently the closest match. Conclusions The gunflints recovered from the Castle and Oudepost I provide a unique opportunity to examine the sources for gunflints used by the VOC to provision their outposts in South Africa. They also offer the opportunity to explore the provisioning strategies of the Dutch in South Africa, compared to strategies used by other nations at colonial sites in the New World. Where origins of the gunflints themselves are concerned, Cape gunflints appear similar to those recovered from sites in the New World. Given the level of mass production seen in France and England, we suggest that either one, or possibly both, of these countries are the source of the gunflints recovered at the South African sites. Sufficient numbers of spall-type and blade-type gunflints are present to assert that the Dutch were probably trading, either directly or indirectly, with both French and English gunflint manufacturers. The fact that the LA-ICPMS results from several of the blade-type gunflints from the two VOC sites matched identically the core material from Meusnes, France, strongly suggests that the Dutch were obtaining and distributing gunflints from the primary French supply source. Although all of the honey-colored gunflints subjected to the LA-ICP-MS testing do not display an identical match to the source material in Meusnes, it is a reasonable assumption that most of the blade-type gunflints recovered from the two VOC sites were manufactured in France. While many of the spall-type gunflints recovered from the South Africa sites show a close match to core materials recovered from Britain, none shows an exact match to this material, making it impossible to say for certain where the spall-type gunflints recovered from the South Africa sites were manufactured. The fact that no spall-type gunflints from either the South Africa sites or the sites used in this study for comparison were manufactured from honey-colored chert matching that from Meusnes suggests that, although the French may have made spall-type flints elsewhere, the spall technique was not used at Meusnes. Given that multiple spall-type production sites are known to have existed in southern England, this may be the source of the Cape series. Archival lists were examined for evidence of the source of gunflints stored in the Cape armory. Documentary evidence is very sparse, but a 1676 listing of material from the old armory at the Cape does mention two 100 pound kegs, one about three quarters full and the other about two thirds full, of “Fatherland flints,” as well as a case, about half full, of “Ceylon flints” (Leibbrandt 1902: 272). It is impossible to say whether this attribution refers to place of production or lading; however, the numbers can be calculated for the Cape armory of 1676. Gunflints were usually dispensed in small casks or “half barrels,” the smallest of which contained two thousand gunflints, the largest containing four thousand (Wyatt 1870: 587). Historic documents also tell us that casks of “sorts,” or mixed sizes, were also available (Forrest 1983). The 100 pound kegs referred to in the Cape of Good Hope inventory were the larger size barrels containing 4,000 gunflints. Thus we may calculate that there were approximately 3,000 gunflints in the keg recorded as three quarters full and 2,640 in the keg listed as two-thirds full. The size of the case of “Ceylon” flints is not known. The analysis of Cape gunflints links the VOC station into a web of trade that enmeshed the Old World, the New World, and the VOC realms in Asia. It also offers insight into the provisioning strategies for VOC outposts situated in remote locations. Acknowledgements I would like to acknowledge Jay Blaine, an extremely wise, avocational archeologist who made me first question the production locations for ‘French’ gunflints. Bibliography Blanchette, J. F. (1975). Gunflints from Chicoutimi Indian Site (Quebec). Historical Archaeology 9, 41–54. Brown, M. L. (1980). Firearms in Colonial America 1492-1792. Washington, DC: Smithsonian Institution Press. ANALYSIS OF GUNFL INTS 229 Clarke, R. (1935). The flint-knapping industry at Brandon. Antiquity 9, 38–56. Luedtke, B. (1997). Gunflints in the Northeast. Northeast Anthropology 57, 27-43. De Lotbiniere, S. (1980). English gunflint making in the seventeenth and eighteenth centuries. In T. M. Hamilton (Ed.), Colonial Frontier Guns (pp. 154-160). Chadron, NE: Fur Press. Neff, H. (2005). LA-ICP-MS analysis of flint samples from Texas, France and England. Manuscript. Austin, TX: Texas Historical Commission. Durst, J. J. (2009). Sourcing European-produced gunflints to their country of manufacture. Historical Archaeology 43(2), 19-30. Faulkner, A. (1986). Maintenance and fabrication at Fort Pentagoet 1635–1654: Products of an Acadian Armorer’s Workshop. Historical Archaeology 20(1), 63–94. Forrest, A. J. (1983). Masters of Flint. Lavenham, Suffolk, UK: Terence Dalton. Hamilton, T. M. (1968). Review of “A History of Gunflints” by John Witthoft. Historical Archaeology 2, 116–117. Rockman, M., Glascock, M. & Baker, M. (2003). Learning the lithic landscape: Trace element characterization of flint using ICP-MS and the recolonization of Great Britain at the end of the last Ice Age. Paper Presented at the Society for American Archaeology 68th Annual Meeting, Milwaukee, WI. Schrire, C. & Deacon, J. (1989). The indigenous artifacts from Oudepost I, a colonial outpost of the VOC at Saldanha Bay, Cape. South African Archaeological Bulletin 44, 105-13. Speakman, R. J. & H. Neff. (Eds). (2005). Laser ablation-ICP-MS in archaeological research. Albuquerque: University of New Mexico Press. Hamilton, T. M. & Emery, K. O. (1988). Eighteenth-century gunflints from Fort Michilimackinac and other colonial sites. Archaeological Completion Report Series, No. 13. Mackinac Island, MI: Mackinac Island State Park Commission. White, S. W. (1975). On the origins of gunspalls. Historical Archaeology 9, 65–73. Held, R. (1970). The age of firearms: A pictorial history from the invention of gunpowder to the advent of the modern breechloader. Northfield, IL: Gun Digest Co. White, S. W. (1976). The French gunflint industries. (Unpublished M.A. dissertation). Department of Archaeology, University of Calgary, Alberta, Canada. Honerkamp, N. & Harris, N. (2005). Unfired Brandon gunflints from the Presidio Santa María de Galve, Pensacola, Florida. Historical Archaeology 39(4), 95-111. Witthoft, J. (1966). A history of gunflints. Pennsylvania Archaeologist 36(1&2), 12-49. Kent, B. C. (1983). More on gunflints. Historical Archaeology 17(2), 27–40. Lavin, J. D. (1965). History of Spanish Firearms. London: Herbert Jenkins. Leibbrandt, H. C. V (1902). Précis of the archives of the Cape of Good Hope. Journal, 1671-1674, 1676. Cape Town: W. A. Richards & Sons. Wyatt, J. (1870). On the manufacture of gun flints. In E. T. Stevens (Ed.), Flint chips: A guide to pre-historic archaeology as illustrated by the collection in the Blackmore museum, Salisbury (pp. 578-590). London: Bell and Daldy. APPENDIX F Objects Of Personal Adornment Castle Coat buttons Metal There are 97 buttons in the Moat collection, as follows: Waistcoat buttons Bone 13 mm BT 1 Complete bone button with drilled shank. Convex face with incised rim. Intact shank. D = 13 mm. 15 mm Undecorated BT 2 Complete bone button with drilled shank. Convex undecorated face. Intact shank. Face is spalled. D = 15 mm. 18 mm Undecorated BT 8 Cast copper alloy button with flat surface. Drilled shank cast with button. D = 18 mm. 32 mm Undecorated BT 9 Two-part button with convex surface. Two holes in button back to release gas during brazing process. Dented and punctured. Shank missing. D = 32 mm. Bone Decorated BT 3, 97 Complete bone buttons with drilled shank. Convex faces with incised rims. Intact shanks. D = 15 mm. 19 mm Undecorated BT 10 Complete bone button with drilled shank. Convex undecorated face. Edge worn. Intact shank. D = 19 mm. 16 mm Undecorated BT 4 Complete bone button. Convex undecorated face. Broken shank. D = 16 mm. Decorated BT 11 Complete bone button with drilled shank. Slightly convex face with incised rim. Intact shank. D = 19 mm. Decorated BT 5, 98 Complete bone buttons with drilled shanks. Convex faces with incised rims. Intact shanks. D = 16 mm. 17 mm Undecorated BT 6 Complete bone button with drilled shank. Convex undecorated face. Intact shank. D = 17 mm. 20 mm Undecorated BT 12, 100 Complete bone button with drilled shank. Convex undecorated face. Broken shank. D = 20 mm. BT 13 Bone button with drilled shank. Convex undecorated face. Edge broken on one side. Intact shank. D = 20 mm. Decorated BT 7, 99 Complete bone button with drilled shank. Convex face with incised rim. Intact shank. D = 17 mm. 269 Decorated BT 14 Complete bone button with drilled shank. Convex face with green coloring. Broken shank; threadwear present. D = 20 mm. 270 H I S TO R I C A L A RC HA E OLOGY I N SOU T H A FR I C A BT 15 Complete bone button with drilled shank. Convex face with incised rim. Intact shank. D = 20 mm. 21 mm Undecorated BT 16 Complete bone button with drilled shank. Convex undecorated face. Intact shank. D = 21 mm. BT 17 Bone button with drilled shank. Convex undecorated face. Intact shank; threadwear present. D = 21 mm. Decorated BT 18 Complete bone button with drilled shank. Convex face with incised rim. Intact shank. D = 21 mm. 22 mm Undecorated BT 19 Complete bone button with drilled shank. Convex undecorated face. Intact shank. D = 22 mm. BT 20 Complete bone button with drilled shank. Flat face with tapered rim; undecorated face. Intact shank. D = 22 mm. BT 21 Bone button with drilled shank. Convex undecorated face. Intact shank. Edge broken. D = 22 mm. Decorated BT 22 Complete bone button with drilled shank. Convex face with incised rim. Intact shank. D = 22 mm. 23 mm Undecorated BT 23 Complete bone button with drilled shank. Convex undecorated face. Intact shank. D = 23 mm. Decorated BT 24 Complete bone button with drilled shank. Convex face with incised rim. Broken shank; threadwear present. D = 23 mm. 24 mm Decorated BT 25 Complete bone button with drilled shank. Convex face; incised rim. Broken shank. D = 24 mm. 25 mm Undecorated BT 26 Complete bone button with drilled shank. Convex undecorated face. Intact shank. D = 25 mm. Unidentified function Bone 7 mm Undecorated BT 27 Complete bone button with drilled shank. Steeply pitched convex design. Broken shank. Possible sleeve or shirt button. D = 7 mm. 8 mm Decorated BT 28 Complete bone button with drilled shank. Convex face with incised rim. Intact shank. D = 8 mm. 9 mm Undecorated BT 29, 30, 31, 32, 33, 34 Complete bone button with drilled shank. Convex undecorated face. Intact shank. D = 9 mm. 12 mm Undecorated BT 35 Complete bone button with drilled shank. Convex undecorated face. Intact shank. D = 12 mm. BT 36 Complete bone stud. Convex undecorated face. Edge worn. Broken shank. D = 12 mm. 13 mm Undecorated BT 37 Complete bone button with drilled shank. Convex undecorated face. Intact shank. D = 13 mm. BT 38 Bone button fragment. Convex undecorated face. Intact shank. D = 13 mm. BT 39, 101 Complete bone button with drilled shank. Convex undecorated face. Intact shank. D = 13 mm. 14 mm Undecorated BT 40 Complete bone button with drilled shank. Convex face. Intact shank. D = 14 mm. BT 41 Complete bone button with drilled shank. Convex face. Intact shank. D = 14 mm. BT 42-53 Complete bone button with drilled shank. Convex undecorated face. Intact shank. D = 14 mm. BT 54, 55 Bone button fragment. Convex undecorated face. Broken shank. D = 14 mm. BT 56 Complete bone button with drilled shank. Flat undecorated face. Shank broken. D = 14 mm. BT 57 Complete bone button with drilled shank. Convex undecorated face. Intact shank. D = 14 mm. Decorated BT 58 Complete bone button with drilled shank. Turned design on slightly convex face. Intact shank. Possible waistcoat button. D = 14 mm. 15 mm Undecorated BT 59-67 Complete bone button with drilled shank. Convex undecorated face. Intact shank. D = 15 mm. APP ENDIX F BT 68-70 Bone button fragment. Convex undecorated face. Intact shank. D = 15 mm. BT 71 Complete bone button with drilled shank. Convex undecorated face. Broken shank. D = 15 mm. BT 72 Complete bone button with drilled shank. Convex undecorated face. Intact shank. D = 15 mm. 16 mm Undecorated BT 73 Complete bone button with drilled shank. Convex, steeply pitched undecorated face. Intact shank. Possible waistcoat button. D = 16 mm. BT 74 Complete bone button with drilled shank. Convex undecorated face. Intact shank. Possible waistcoat button. D = 16 mm. BT 75-77 Complete bone button with drilled shank. Convex undecorated face. Intact shank. D = 16 mm. BT 78 Complete bone button with drilled shank. Convex undecorated face. Broken shank; threadwear present. D = 16 mm. BT 79-83 Complete bone button with drilled shank. Convex undecorated face. Intact shank. D = 16 mm. BT 84 Bone button fragment. Convex undecorated face. Intact shank. D = 16 mm. Decorated BT 85 Complete bone button with drilled shank. Convex face with incised rim. Intact shank. D = 16 mm. 17 mm Undecorated BT 86 Complete bone button with drilled shank. Convex undecorated face. Intact shank. Possible waistcoat button. D = 17 mm. BT 87 Bone button fragment. Convex undecorated face. Intact shank. D = 17 mm. BT 88 Complete bone button with drilled shank. Convex undecorated face. Intact shank. Edge broken. D = 17 mm. 18 mm Undecorated BT 89 Complete bone button with drilled shank. Convex undecorated face. Broken shank. D = 18 mm. BT 90-91 Complete bone button with drilled shank. Convex undecorated face. Intact shank. D = 18 mm. BT 92 Bone button fragment. Convex undecorated face. Intact shank. D = 18 mm. BT 93 Bone button fragment. Convex undecorated face. Broken shank. D = 18 mm. 271 19 mm Undecorated BT 94 Complete bone button with drilled shank. Convex undecorated face. Intact shank. D = 19 mm. Sew-through BT 96 Bone button core. Central hole. Convex on one side, flat on the other. D = 9 mm. Van der Stel Moat. Oudepost 1 (Opi) Aglets A 1 Aglet made of copper alloy sheet metal with straight seam and rounded tip; crimped dimple at base to hold lace fast. L = 23; W = 3.5 mm. A 2 Aglet made of copper alloy sheet metal with straight seam and rounded tip; crimped dimple at base to hold lace fast. L = 20; W = 3.5 mm. Buckles Shoe Buckles Whole Buckles BK 1 Double framed annular shoe buckle made of copper alloy with an inner band of nailheads and outer ring of ovate beading. Decoration is imparted by the mold. Pin is cast with the frame. Stud chape; no tongue. L = 39 mm; W = 26 mm. Tongue = 31 mm x 11 mm. Stud = 7 mm. BK 2 Double framed trapezoidal shoe buckle with beveled interior edge of copper alloy. Lobed on the outer edge. Flared molded band applied to top of pin terminal. Pin is cast with the frame. Stud chape, no tongue. L = 42 mm; W = 23 mm. Tongue = 27 x 9 mm. Stud = 8 mm. Frames Whole BK 3 Whole rectangular framed copper alloy shoe buckle with rounded corners. Flared at buckle ends with a notched design over the pin terminals. No chape. L = 44 mm; W = 31 mm. BK 4 Whole rectangular framed copper alloy shoe buckle. Flared at ends on the interior with transverse lines over the pin terminals. No chape. L = 30 mm; W = 23. BK 5 Whole rectangular framed copper alloy shoe buckle. Sharply pitched beveled interior. No chape. L = 45 mm; W = 34 mm. 272 H I S TO R I C A L A RC H A E OLOGY I N SOU T H A FR I C A Fragments BK 6 Very corroded iron shoe buckle frame fragment. Rectangular with rounded corners. L = 40; W = 27 mm. BK 7 Very corroded iron shoe buckle frame fragment. Rectangular with rounded corners. L = 29 mm; W = 24 mm. BK 8 Corroded iron shoe buckle frame. Rectangular with rounded corners. L = 39 mm; W = 27 mm. BK 9 Openwork copper-alloy shoe buckle with interwoven ribbon design fragment. Sub rectangular in shape. Bow at center of frame end. Beading along the inner and outer edges and nailhead design along outer register. No chape. Matches BK 10 and 11. L = 46 mm; W = 45 mm. BK 10 Openwork copper-alloy shoe buckle with interwoven ribbon design fragment. Sub rectangular in shape. Bow at center of frame end. Beading along the inner and outer edges and nailhead design along outer register. No pin terminal present. Matches BK 9 and 11. No chape. L = 51 mm; W = 30 mm. BK 11 Openwork copper-alloy shoe buckle with interwoven ribbon design fragment. Sub rectangular in shape. Bow at center of frame end. Beading along the inner and outer edges and nailhead design along outer register. Frame edge only. No pin terminal present. Matches BK 9 and 10. No chape. L = 51 mm; W = 16 mm. BK 12 Rectangular copper alloy shoe buckle frame fragment. Simple design with flare and beveling in interior edge. Both pin terminals present. No chape. L = 33 mm; W = 32 mm. BK 13 Rectangular copper alloy shoe buckle frame fragment. Sharply pitched beveled interior. Worn surface. One pin terminal present. No chape. L = 33 mm; W = 32 mm. BK 14 Double framed sub annular shoe buckle frame fragment. One side and pin present. Undecorated and worn surface. No chape. L = 34 mm; W = 27 mm. BK 15 Sub rectangular copper alloy shoe buckle frame with molded contour design with chevrons at center of frame end. One complete side. Broken at pin terminal. Matches BK 16. No chape. L = 24 mm; W = 31 mm. BK 16 Sub rectangular copper alloy shoe buckle frame with molded contour design with chevrons at center of frame end. Broken at pin terminal. No chape. Matches BK 15. L = 24 mm; W = 31 mm. Chapes BK 17 Very corroded iron shoe buckle chape roll. Cooking-pot shape with single tongue. L = 38, W = 35. BK 18 Copper alloy shoe buckle chape roll. Cooking pot shape with nodule base. Single tongue. Dutch style. L = 31, W = 33. BK 19 Copper alloy shoe buckle chape roll. Cooking pot shape with single tongue. L = 22 mm; W = 27 mm. BK 20 Copper alloy shoe buckle chape roll. Cooking pot shape with single tongue. L = 28 mm; W = 29 mm. BK 23 Copper alloy shoe buckle chape roll. Cooking pot shape with single tongue. L = 32 mm; W = 26 mm. Knee buckle Whole BK 21 Complete iron knee buckle. Anchor shaped chape. Corrosion obscures any surface decoration, but likely undecorated flat frame. L = 42 mm; W = 34 mm. Strap buckle Whole BK 22 Complete strap buckle. Double framed sub annular strap buckle. Decorative flare at sides of pin. Tongue attached. Pin sleeve made of sheet metal. L = 41 mm; W = 30 mm. Buttons Sleeve buttons Metal Undecorated BT 1 Copper alloy sleeve button. Undecorated. Convex face. Broken flattened U-shaped shank. D = 10.5 mm. Decorated BT 2 Copper alloy sleeve button set with white stone. Beaded border. Flattened U-shaped shank. Connecting wire loop present. D = 12 mm. BT 106 Copper alloy sleeve button set. Chained circle decoration border. Central nipple with cast floral design. Flattened u-shaped shank. Connecting wire loop present. D = 15 mm.. BT 107 Copper alloy sleeve button set. Convex decorated face with tiny dots on the outer rim. Flattened ushaped shank. Connecting wire loop present. D = 8 mm. Waistcoat buttons Metal 12 mm Undecorated BT 3 Cast copper alloy button with convex surface and no decoration. Cast shank. D = 12 mm. 13 mm Undecorated BT 4 Cast copper alloy button with convex surface and no decoration. Drilled eye. D = 13.5 mm. AP PENDIX F 273 BT 5 Cast copper alloy button with convex surface and no decoration. Shank missing. D = 13.5 mm. BT 6 Cast copper alloy button with convex surface and no decoration. Shank cast with button. D = 13.5 mm. Decorated BT 21 Complete bone button with drilled shank. Flat face with bevelled edged. D = 14 mm. 15 mm 14 mm Undecorated BT 7 Fragment of cast copper alloy button with convex surface and no decoration. Shank broken. D = 14 mm. BT 8 Cast copper alloy button with convex surface and no decoration. Shank cast with button. D = 14 mm. BT 9 Cast copper alloy button with convex surface and no decoration. Shank missing. D = 14 mm. BT 10 Cast copper alloy button with flat surface and no decoration. Brazed shank with broken eye. D = 14 mm. BT 11 Cast copper alloy button with convex surface and no decoration. Shank missing. D = 14.5 mm. 15 mm Undecorated BT 22 Complete bone button with drilled shank. Convex undecorated face. Shank intact. Very worn. D = 15 mm. BT 23 Complete bone button with drilled shank. Convex undecorated face. Shank intact. D = 15 mm. BT 24 Complete bone button with drilled shank. Convex undecorated face. Shank intact. Very worn. D = 15 mm. BT 25 Complete bone button with drilled shank. Convex undecorated face with broken edge. Shank intact. D = 15 mm. 15 mm Undecorated BT 12 Hollow cast copper alloy button fragment. Face of two-part button. No decoration. D = 15 mm. BT 13 Cast copper alloy button face of two-part brazed button.. No decoration. D = 15 mm. 17 mm Undecorated BT 14 Cast copper alloy button with flat surface and no decoration. Applied omega shank. D = 17 mm. BT 15 Cast copper alloy button with convex surface and no decoration. Drilled shank. D = 17.5 mm. Bone 13 mm Decorated BT 16 Complete bone button with drilled shank. Concave face with border and nipple at center. Shank broken. D = 13.5 mm. 14 mm Undecorated BT 17 Complete bone button with drilled shank. Convex undecorated face. Shank intact. D = 14 mm. BT 18 Complete bone button with drilled shank. Convex undecorated face. Shank detached. D = 14 mm. BT 19 Complete bone button with drilled shank. Convex undecorated face. Shank detached. D = 14 mm. BT 20 Complete bone button with drilled shank. Convex undecorated face. Shank intact. D = 14.5 mm. Decorated BT 26 Complete bone button with drilled shank. Concave face with incised border and nipple at center. Shank intact. D = 15 mm. 16 mm Undecorated BT 27 Complete bone button with drilled shank. Convex undecorated face with broken edge. Shank intact. D = 16 mm. 18 mm Decorated BT 28 Complete bone button with drilled shank. Flat face with bevelled edged. Broken shank. D = 18 mm. BT 29 Complete bone button with drilled shank. Convex face with incised edge. Intact shank. D = 18 mm. Sew-through 15 mm BT 30 Complete bone button core with center hole. Domed shape. Irregular hole on surface. D = 15 mm. BT 31 Bone button core with center hole. Broken edge. Domed shape. D = 15 mm. 16 mm BT 32 Bone button core with center hole. Broken edge. Flat. D = 16 mm. 17 mm BT 33 Bone sew-through button with four holes. Incised border on front surface to create softly rounded border. D = 17.5 mm. 274 H I S TO R I C A L A RC H A E OLOGY I N SOU T H A FR I C A Coat buttons Metal 14 mm Decorated BT 34 Hollow cast copper alloy button. Octagonal shape and nipple decoration on domed button. Hole in button back to release gas during casting process. Shank missing. D = 14 mm. 15 mm Decorated BT 35 Cast copper alloy button with domed surface. Crossed lines and nipple decoration. Shank broken D = 15 mm. 17 mm Decorated BT 36 Cast copper alloy button with flat surface. Engraved flower design. Brazed alpha shank. D = 17.5 mm. Undecorated BT 37 Cast copper alloy button with flat surface.. Brazed alpha shank. D = 17.5 mm. 18 mm Decorated BT 38 Stamped copper alloy button cover for two-part stamped metal-covered button. Stamped oval design with sunburst. Stamped textured surface. D = 18 mm. BT 39 Hollow cast copper alloy button with domed surface. Nipple decoration. Brazed shank. Hole in button back to release gas during casting process. D = 18.5 mm. BT 40 Hollow cast copper alloy button with domed octagonal surface. Faceted with nipple decoration. Brazed shank. Hole in button back to release gas during casting process. D = 18.5 mm. 19 mm Undecorated BT 41 Cast copper alloy button with convex surface. Undecorated. Brazed alpha shank. D = 19 mm. BT 42 Cast copper alloy button with convex surface. Very poor condition. Brazed alpha shank. D = 19 mm. 23 mm Undecorated BT 43 Cast copper alloy button with flat surface. Undecorated. Cast shank with drilled eye. D = 23 mm. 28 mm Undecorated BT 44 . Cast copper alloy button back of two-piece brazed button. Loop shank inserted into button back. Crushed. D = 28 mm. BT 45 Cast copper alloy button back of two-piece brazed button. Loop shank inserted into button back and brazed. Crushed. D = 28 mm. BT 46 Cast copper alloy button front of two-piece brazed button. Undecorated. Crushed. D = 28 mm. BT 47 Cast copper alloy button with domed surface. Undecorated. Loop shank inserted into button back and brazed. Hole in button back to release gas during brazing process. D = 28.5 mm. BT 48 Hollow cast copper alloy button with domed surface. Undecorated. Brazed alpha shank. Hole in button back to release gas during casting process. D = 28.5 mm. 29 mm Undecorated BT 50 Hollow cast copper alloy button with domed surface. Undecorated. Brazed alpha shank. Hole in button back to release gas during casting process. D = 29 mm. 30 mm Undecorated BT 51 Cast copper alloy button back of two-piece brazed button. Brazed alpha shank. D = 30 mm. BT 52 Cast copper alloy button back of two-piece brazed button. Folded; edges bent. Inserted and brazed loop shank. D = 30 mm. 32 mm Undecorated BT 54 Two-part brazed button with convex surface. Loop shank inserted into button back and brazed. Hole in button back to release gas during brazing process. D = 32 mm. Bone 19 mm Undecorated BT 55 Bone button fragment of shanked button. Onethird of domed button face. No shank. D = 19 mm. BT56 Complete bone button with drilled shank. Convex undecorated face. Intact shank. D = 19 mm. BT 57 Fragment of shanked bone button. Domed face. No shank. D = 19 mm. 20 mm Undecorated BT 58 Complete bone button with drilled shank. Convex face. Intact shank. D = 20 mm. BT 59 Complete bone button with drilled shank. Convex face. Intact shank. D = 20.5 mm. AP PENDIX F 22 mm Decorated BT 60 Complete bone button with drilled shank. Flat face with bevelled edged. D = 22 mm. Sew-through BT 61 Bone sew-through button with four holes. Assymetrically drilled. Softly rounded edge. D = 19.5 mm. BT 62 Bone button core with center hole. Flat. D = 20 mm. Unidentified function Metal 9 mm Undecorated BT 63 Copper alloy button. Shank obscured by corrosion. Possible sleeve button. D = 9.5. 10 mm Decorated BT 64 Cast copper alloy button with nippled decoration. Surface otherwise smooth. Attached loop shank. Possible waistcoat or doublet button. D = 10.5 mm. 11 mm Undecorated BT 65 Cast copper alloy button with smooth slightly convex surface. Worn on opposite surfaces from use. Attached loop shank. Possible waistcoat button. D = 11 mm. BT 66 Cast copper alloy button with convex undecorated surface. Alpha shank cast with button. Possible waistcoat or doublet button. D = 11 mm. BT 67 . Cast copper alloy button with convex undecorated surface. Attached loop shank cast with button. Possible waistcoat button. D = 11 mm. BT 68 Cast copper alloy button with convex undecorated surface. Alpha shank cast with button. Possible waistcoat button. D = 11 mm. Decorated BT 69 Cast copper alloy button with convex surface. Cast rosette decoration. Attached loop shank cast with button. Possible waistcoat or doublet button. D = 11 mm. BT 70 Cast copper alloy button with convex surface. Cast rosette decoration. Attached loop shank cast with button. Possible waistcoat or doublet button. D = 11 mm. BT 71 Cast copper alloy button with flat surface with bevelled edge. Shank missing. Possible waistcoat button. D = 11.5 mm. 275 12 mm Undecorated BT 72 Cast copper alloy button with convex surface and no decoration. Shank broken. Possible waistcoat button. D = 12 mm. BT 73 Cast copper alloy button with convex surface and no decoration. Loop shank cast with button. Possible waistcoat button. D = 12 mm. BT 74 Cast copper alloy button with convex surface and no decoration. Shank missing. Possible waistcoat button. D = 12 mm. BT 75 Hollow cast copper alloy button with convex surface and no decoration. Crushed. Two holes on reverse product of casting. Shank missing. Possible waistcoat button. D = 12 mm. BT 76 Cast copper alloy button with convex surface. Very poor condition. No shank. Possible waistcoat button. D = 12 mm. 13 mm Undecorated BT 77 Hollow cast copper alloy button with steeply convex surface. Top portion of two-part button. Possible waistcoat or doublet button. D = 13 mm. BT 78 Cast copper alloy button with convex surface and no decoration. Cast seam on back of button. Shank cast with button; drilled eye. Possible waistcoat button. D = 13 mm. BT 79 Cast copper alloy button with convex surface and no decoration. Shank cast with button; drilled eye. Possible waistcoat button. D = 13 mm. BT 80 Cast copper alloy button with convex surface and no decoration. Shank cast with button; drilled eye. Possible waistcoat button. D = 13 mm. BT 81 Cast copper alloy button with convex surface and no decoration. Shank cast with button; drilled eye. Cast seam on back of button. Possible waistcoat button. D = 13 mm. BT 82 Cast copper alloy button with convex surface and no decoration. Button back corroded. Possible waistcoat button. D = 13 mm. BT 83 Cast copper alloy button with convex surface and no decoration. Poor condition; edges broken. Possible waistcoat button. D = 13 mm. BT 84 Cast copper alloy button. Very poor condition. Brazed omega shank. Possible waistcoat button. D = 13 mm. BT 85 Hollow cast copper alloy button with convex surface. Top portion of two-part button. Slightly dented and surface scratched (from use?). Possible waistcoat button.. D = 13 mm. 276 H I S TO R I C A L A RC H A E OLOGY I N SOU T H A FR I C A BT 86 Hollow cast copper alloy button. Back of twopart button. Shank missing. Possible waistcoat button. D = 13 mm. BT 87 Hollow cast copper alloy button. Front of convex two-part button. Possible waistcoat button. D = 13 mm. 15 mm Undecorated BT 88 Fragment of face of two-part brazed button. Crushed. D = 15 mm. 16 mm Undecorated BT 89 Fragment of cast copper alloy button. Broken. D = 16 mm. BT 90 Cast copper alloy button. Very poor condition. Crushed. D = 16 mm. BT 91 Cast copper alloy button face of two-part button. No decoration; three holes puncture surface. D = 16 mm. Decorated BT 92 Cast copper alloy button face of two-piece brazed button. Stamped sunburst decoration in center. D = 16 mm. 17 mm Decorated BT 93 Cast copper alloy or iron button. Octagonal shape. Drilled shank. Very poor condition. D = 17. Eyes BT 94 Button eye. L = 7 mm; H = 6 mm. BT 95 Wire button eye with portion of button attached. L = 6; W= 8. BT 96 Wire button eye with portion of button attached. L = 4; W= 4. 4.3 Shell BT 101 Shell button with shank. Shank very worn. Undecorated. D = 10. BT 102 Sew-through shell button with two holes. Double band engraved at edge. D = 11. Miscellaneous Metal BT 103 Cast metal stud. Undecorated flat face. D = 13 mm. BT 104 Cast copper alloy frog. Complete. Two buttons connected by solid bar. L = 31; bar width = 4 mm; button diameter = 13. BT 105 Cast copper alloy frog fragment. Button with fragment of metal bar. Button diameter = 13. Button mold M1 One piece of three piece button mold. Mold has two dimples to fit three pieces of mold together. Shank for frog type of button carved into mold present on anterior surface. Hole for pouring molten metal present on dorsal surface. Ventral surface exhibits smooth surface from use. L = 66 mm; W = 17 mm. Jewelry Rings R 1 Copper alloy ring with stone set in cabochon setting. Stone in poor condition; possible amythest. Hoop tapers and is slightly convex. Bezel width = 7 mm; band diameter = 27 mm. R2 Copper alloy band. Flat surface with etched design on either side. Possible wedding band. D = 27 mm; W = 4 mm. Bone 13 mm Undecorated BT 97-98 Complete bone buttons with drilled shanks. Convex faces. D = 13 mm. BT 99 Complete bone buttons with drilled shank. Convex face. D = 13 mm. Sew-through BT 100 Complete bone core with single hole. D = 13 mm. Miscellaneous accessory Chain C1 Three links of chain made of round wire. Possible watch or chatelaine chain. L = 26 mm. ABOUT THE AUTHORS Jeff Durst is the Project Director of the Ft. St. Louis Archaeological Project, the South Texas Regional Archaeologist, and Project Reviewer for the Texas Historical Commission. In recent years he has conducted the analysis of over 1,200 gunflints related directly and indirectly to both the 17th century shipwreck La Belle (1687) off the coast of Texas and the related terrestrial site of Fort St. Louis located in Victoria County, Texas. Adam Heinrich is a 2010 Ph.D. from Rutgers University who works in historical archaeology and zooarchaeology. His research into material culture includes examining gravemarkers as reflections of consumerism across the colonial American landscape. His zooarchaeological research includes advocating for more rigorous taphonomic perspectives to historical faunal analyses as well as investigating animal consumption in colonial and early post-colonial contexts. Stacey Jordan received her doctorate from Rutgers University, conducting research at the University of Cape Town on the local production and use of coarse earthenware. Her publications on coarse earthenware have appeared in The Getty Research Institute’s The Archaeology of Colonialism, International Journal of Historical Archaeology, and Journal of Archaeological Science. She has been involved in historical archaeological projects in South Africa, the United States, and Puerto Rico, spanning from early colonial contact to the recent past. She is currently a Cultural Resources Practice Leader at AECOM. Jane Klose has been analyzing ceramic assemblages excavated from historical archaeological sites throughout the Western Cape since 1986. Her pioneering work on Asian ceramics found in South Africa has led to international recognition. Klose is a member of the Oriental Ceramic Soci- ety in London and has collaborated widely, especially with colleagues with museums in the Netherlands and Oxford. She is an Honorary Research Assistant in the Department of Archaeology at the University of Cape Town, where she has taught specialist courses on the identification and analysis of excavated ceramics. Carmel Schrire is a Distinguished Professor of Anthropology at Rutgers, The State University of New Jersey in New Brunswick, NJ. She was educated at the University of Cape Town and Cambridge University and received her doctorate at The Australian National University. She has done archaeological research in Australia and South Africa and has specialized in the application of hunter-gatherer history and ethnography the interpretation of the past. She is the author of The Alligator Rivers: Prehistory and Ecology in Western Arnhem Land (1982) and the award winning Digging through Darkness: Chronicles of an Archaeologist (1995). Patricia Schwindinger is a graduate student of Nautical Archaeology at Texas A&M University. She completed her B.A. at Rutgers, The State University of New Jersey in 2012. She was awarded the Henry Rutgers Scholars Award for her Honors thesis reconstructing Oudepost I. She has also worked with Dr. Craig Feibel to create a series of images mapping the rise and fall of Lake Turkana, Kenya. She is interested in graphics and their applications in archaeology. Carolyn L. White holds the Mamie Kleberg Chair in Historic Preservation and is an associate professor in the Department of Anthropology at the University of Nevada, Reno. Her work spans four centuries and currently focuses on daily life in 1860s Aurora, Nevada, and on the built environment of Black Rock City, the temporary city built for the Burning Man festival. 287

RELATED PAPERS

RELATED TOPICS

Log In

Objects of Personal Adornment at the Cape

Objects of Personal Adornment at the Cape

Related Papers

RELATED PAPERS

RELATED TOPICS