Slave Records Sale Deeds : www. stamouers.co.za , AM van Rensburg et al from research of  AJ  Boeseken and Prof R Shell

Copied for this website by Lara Seaward

Footnotes by the authors of this database:

Several attempts have already been made to establish the geographic origins of the imported slaves with quite different results. The first problem  arises because in establishing the origins of Cape slaves, all researchers have included the sizeable "van der Kaap" [=creole] population in their tabulations of origins. Since the rate of creolization (the percentage of Cape slaves who were locally born) changed quite independently of the slave trade, the inclusion of any creole slaves in any tabulation of origins confuses the issue of the exact breakdown of origins of imported slaves. Creolization, simply put, is a different process with discrete demographic and cultural consequences within the household.

​

Pars pro toto [part for the whole]
The second major problem is that scholars have worked with tallies of one Cape slave population (there were three: burghers' slaves, the officials' slaves and the slave lodge) and made assumptions about the others. The lodge slave trade alone is well documented since it was sanctioned and sponsored by the Company Directorsi.
The slave lodge was a discreet, small, and wholly atypical slave community, whose numerical significance dwindled as the privately owned slave populations grew. By 1795 the privately owned slave force had grown to 60 times the size of the slave lodge; by 1826, the last recorded census of the lodge, the lodge population was only .32 of one percent of all Cape slaves. The lodge, in short, increasingly became a numerically insignificant part of the Cape slave population. Thanks to the work of James Armstrong and James Ravel we may calculate the total number of slaves imported to the Lodge, by tabulating  origins and tallying purchases on all voyagesii. Despite its growing numerical insignificance, the lodge slave trade has dominated most discussion of the Cape slave trade. Nobody has ever made an estimate of the number of slaves imported to the officials, who were the main slave traders of the colony. A "100 % surface chart" reveals the true proportions of the three populations.

​

Under-representation
Under-representation is the third problem and concerns the bulk of the slave imports, the slave trade serving the individuals in the employ of the Company, the free burgher and free black populations. Because of the multi-facetted, illegal and clandestine nature of most of this private Cape slave trade, these records are missing, often arcane and always difficult to interpret. Perhaps for this reason, no attempt, even in the roughest form, has ever been made to establish the numbers of slaves imported to these individuals at the Cape. The detailed, though sometimes deliberately evasive and certainly incomplete slave transfers (1658-1732), probate records, annual censuses from 1658 through 1834, shipping movements, records of slave voyages themselves and traveller's remarks do however, allow for a preliminary, albeit arithmetical, reconstruction of the bulk of the slave trade. The techniques used in this chapter are familiar to historical demographers, namely stable population theory, first used by demographers for populations with inadequate data, techniques quite suitable for historiansiii.

​

The reconstitution of the slaves belonging to the port patriciate
An immediate problem to surmount is that the official census returns do not correlate with known individual slave trading voyages, nor do annual census increases correlate with the estimated totals of recorded slave trade imports, controlling for natural reproduction, mortality, manumission, running away, or even in and out-migration of ownersiv. In short, increases in the censuses are largely unexplained, at least by currently documented slave trading to the burghers. This problem is resolved when one realizes that no Cape census ever accounted for Company officials's slaves and those belonging to a certain select group of wealthy retired burgher urban burgher councilors, who obtained the privilege of staying off the burgher census rolls, but who were among the main actors in Cape slave trading. This group of slave owners, whom have never been identified, I have termed the port's "patriciate."

​

There was, in short, another sizeable Cape slave population, which was never systematically recorded, nor mentioned in the secondary literature.
Victor de Kock, was probably most influential in obscuring the patrician slave population when he wrote: "Slaves at the Cape fell into two main classes: those belonging to the Dutch East India Company [the lodge slaves] and those owned by individual officials, burghers and farmers.v" A revised taxonomy of the Cape slave populations is the key to achieving a workable census of the Cape slave trades. The existence of the patriciate slave population is by no means speculative: we know of the slave population independently through the oceanic and domestic slave transfers up to 1732, through manumissions from 1658 to 1795, through baptismal records from 1658 to 1795vi. We also know of this slave population through Company official's wills, auction [vendu] lists and repatriations from 1658 through 1798, all of which indicate sizeable households of slaves. Their aggregated slave holdings have been reconstituted from several independent sources, from slave transfers from 1658 to 1732 and from specific registration events, for example, from the jump in slave enumerations in the 1706/7/8 censuses. This unexplained surge in burgher holdings between 1706 and 1708 was a result of the plantation patriciate being recalled to the Netherlands and being forced to divest themselves of their slaves to the burghers. A further such event was a one time slave corvée and registration of all company officials' slaves in 1762. The large jump in the Cape slave population between 1795 and 1798 is, in part, a result of resident Dutch East India Company officials, who for the first time, were obliged on oath, to give an account of their slave holdings to the new British administration. Adding the estimates of these patrician slaves to the slaves on the annual census allows one to estimate the total privately owned slave population much more accurately.

​

Patrician slaves could be sold directly from the patriciate to the burghers, or be bequeathed to burgher heirs (Company officials' offspring, became burghers): in such cases the slaves would suddenly appear on the next year's census. Furthermore, in economically favorable times individual company employees might leave the service (mostly to marry a burgher's daughter): they and their slave returns would appear on the following year's censusvii. Once the Dutch East India Company stopped supporting immigration of free people in 1717, resigning from the Company and applying for burgher status was the secondary means of immigration to the colony (the slave trade itself being the most important)viii. On the other hand, in hard times burghers sold their slaves to members of the patriciate, who always had substantially larger economic resources and salaries with which to weather rough periods: such slaves also "left" the census in the year following such a saleix. Similarly, in hard times a few burghers would apply to be taken back into Company service in which case, their slaves would also disappear from the subsequent burgher census.

​

The significant, but hitherto unexplained, annual fluctuation of slaves on the exclusively burgher and free black census is, in the main, a registration artifact. Moreover, this fluctuation was not a random process, but correlated to both Cape maritime commercial cycles and economic conditions within the colony. This movement of slaves off and on the censuses, i.e. slaves being traded and "moving" between the off census patriciate and the enumerated burghers, continued until 1795 when officials ceased to be "officials" and themselves became burghers. For the first time, the Dutch East India Company officials were not in charge of the record keeping of the colony, and they and their slaves were registered along with everybody else's. The year of 1795 thus provides the last observation point for the patrician slaves. After 1795 the ex-officials and their slaves are enumerated along with the burgher population. The incoming British officials were never required to account for their slaves and there is therefore something of a gap in their slave returns until 1834, which I have not reconstituted.

​

One way out of this dilemma posed by the non-registration of the patricians' slaves, is to add the new estimates of the patricians' slaves to the reliable burgher censuses and from that new series of base numbers of slaves, calculate the slave trade to both populations by arithmetical means, controlling for both vital rates (fertility and mortality), for manumission and running away, which have been independently collectedx. According to a visual observed matrix of this calculation showing formulae, approximately 1,770 calculations were solved simultaneously to retrodict the annual number of slaves imported which came to 62,964, a number including the lodge slaves, but excluding "prize negroes" and convicts.

​

Shortened time frames
A further problem is that no early researchers could have known how rapidly the slave trades to the Cape changed, both in magnitude and in geographical vectors. Their chosen periods, whether it was H. Blommart’s 1652-1662 period, A. Böeseken's 1652-1700 span, the present author's 1680-1731 dissertation period, H.P. Cruse's 1714-1795 period, J.L.M. Franken's 1672-1772 crime records, or even J. Armstrong's work on the entire Dutch period between 1652-1795, resulted in assumptions creeping in considering the total importation of all slaves from 1652 to 1808. In the early stages of research,  how could anyone know that the second half of the seventeenth century with its African majority was a precursor of the nineteenth century? How could anyone guess that after 1795 and before 1808, more slaves would be imported to the Cape than at any other comparable period, and moreover, from a  wholly different direction? The estimates made by earlier researchers working with data from their chosen periods were valid only for those periods and not  for the slave trade as a whole.

​

Sampling errors
A third problem with establishing origins has been inappropriate sampling. Franked, the Stellenbosch historian, working in the nineteen thirties, first tabulated slave origins using slave names mentioned in the voluminous crime records, assuming they were representative of all slaves. He estimated that  from 1672 to 1772, 63 percent of the imported slaves came from the Indonesian Archipelagoxi. Since slaves mentioned in crime records were not a representative sample of all Cape slaves (rural crimes were under-represented), little reliance should have been placed on Franken's work. Cruse, a Stellenbosch theologian, also worked on slave origins and published his work shortly after WWII. He used only origins mentioned in published manumission requests between 1713 and 1795xii. Just as Franked made assumptions about his crime records, so too did Cruse assume that his manumissions were a representative subset of all Cape slaves. This was also mistaken, but how could Cruse guess that Dutch-speaking slaves from the Eastern possessions managed to manumit themselves more easily than Africans, and especially Malagasy slaves? Moreover, since manumission was an urban process, the urban slaves, with their specific eastern geographic origins, were over-represented in his tabulation. If one believed that Cruse's sample was sound, one would have to believe that all slaves in the lodge were born there, since lodge manumissions, with a handful of exceptions, were of creole slaves. But over 5,000 slaves were imported into the lodge between 1652 and 1812. In short, manumitted slaves are not representative of any Cape slave populationxiii. As it turned out, Cruse's calculations meshed with and therefore reinforced Franken's figures. Later scholars, such as I.D. Du Plessis, therefore used Franken's and Cruse's figures in popular works, from where they entered the historical consciousness of South Africa. Consequently, most schoolchildren were traditionally taught that "a few Malay" slaves were brought to the Cape, although I.D. du Plessis simply added the "few."xiv Frank Bradlow and Margaret Cairns, two South African based historians, have made perhaps the most scientific and broad-ranging study. Both researchers were fully alive to the changing origins of the Cape slave force and for the first time, probate documents and the slave registry of the Slave Office (1816-1824) were used for the entire period of Cape slavery. Bradlow and Cairns also used a larger sample size, drawn from a wider range of documentsxv. There were however, built-in biases in their otherwise careful work. First, they considered "free blacks" synonymous with ex-slaves. The reason for this understandable, but mistaken, conflation is not obvious: all manumitted slaves did enter the "free black" community. However, not all free blacks had been slaves, since at the Cape, many "free blacks" also originated from the population of convicts and political exiles, who all came from Indonesia and other Eastern possessions, from Jafnapatnam in Sri Lanka to Mocha on the Red Sea coast. When these exiles and convicts had served their time, they also entered the census free black population, and were confidently tabulated by these researchers as ex-slaves. However, they were not slaves and many of these exiles simply left the colony after their sentences had expiredxvi. Since the free blacks were a small group and the convicts and exiles sizable, the bias was considerable. Assuming that free blacks were drawn from the slave population, therefore inflated their estimation of the Eastern component of all Cape slaves, since nearly all the exiles were Indonesian and nearly all convicts were Indo-Chinese, Indonesian, or Sri Lankanxvii.

​

Double counting
Second, Bradlow and Cairns also used manumission data, which are flawed as a source for slave origins on two counts. Since they incorporated slave transfers from the oceanic trade, their most important bias is double counting, i.e. all imported manumitted slaves, had already been counted as slave imports in their lists recording ship-to-shore transactions of the slave trade. Since some manumitted slaves took on a new name (sometimes, but not always, baptismal) at the time of manumission, I am convinced that such slaves were counted twice. For instance, "Ramtom van Matije [a village in Bali]" was  imported in 1698, in 1712 he was manumitted as "Rangton van Bali." The two records, housed in different archives, each with a different spelling for the same individual would be counted as two individuals from the Indonesian archipelagoxviii. Of all events in a Cape slave's life manumission was recorded with the most fidelity. Over fifteen hundred manumission requests are in published form: Bradlow and Cairns did not sample this class of documents, but counted   them all. Manumissions therefore represent a sizeable source of systemic error in the Cairns/Bradlow tabulations. The same objection must be levelled at  their inclusion of imported slaves mentioned in domestic sales, since such slaves were also previously counted in ship-to shore transactions. Also, there was no limit on the number of times a slave could be sold on the domestic market which could therefore result in multiple counting of the same slave. Slaves imported from the Dutch possessions in Indonesia were invariably resold in Cape Town by Dutch officials wishing to make a quick profit; African and  Malagasy slaves, on the other hand, were sold directly from ship to farm and were rarely resold. Indonesian slaves therefore appear more frequently in their tallies. In their system, a single slave could be imported, sold, sold again and then manumitted, but be counted as four slaves. The thrust of their collaboration has not surprisingly tended to magnify the Eastern component of the origins of Cape slaves and confirm both Franken's and Cruse's research of a generation before, whose work they duly incorporated in one table, albeit with reservationsxix. Having said that, the remainder of the Cairns/Bradlow collaboration is still more than useful -- they brought to light the sizeable number of slaves from the Indian subcontinent. Thanks to their explanatory notes, accurate tallying, and carefully stated assumptions behind each table, one can retrieve and rework much of their data on slave originsxx.

​

In strong contrast to Bradlow and Cairns, James Armstrong has tentatively targeted Madagascar as the main source for the privately owned Cape slaves up to 1795, his cut-off date. He has made only tentative estimates of "100-200 slaves per annum" for the bulk of the Cape slave trade: even in combination with the most generous rate of natural increase (which would result in a smaller slave trade), his estimates cannot account for the increase in the old or revised census returnsxxi. Moreover, Armstrong has not made any discrete estimates of the slaves privately imported by the port's patriciate, Cape Town's most important slave traders, nor has he made any estimates beyond 1795, the period of heaviest tradingxxii. James Armstrong has steadfastly held that the major source of Lodge slaves was Madagascar. His research, conducted over many years and in numerous archives in the Indian ocean basin, has concentrated on tabulating Company slave voyages, the only such work to use this painstaking approach. His austere research (almost every sentence is based on several archival sources) still points in the direction of Madagascarxxiii. This source of slaves remained relatively constant, although Company  imports in the first decade also came from the West Coast of Africa, and between 1721 and 1730 from Delagoa Bay. Later company-sponsored trade to the East Coast of Africa, resulted in most Lodge slaves originating from the African mainlandxxiv. A slim majority of all lodge slaves did come from Madagascar in the period with which he deals, i.e up to 1795, but the Lodge survived another generation before all the slaves were released in 1828. Many more slaves from Mozambique were imported into the lodge after 1795 and before 1826: for example, from "prize" ships (1795 to 1803) and even after 1808, for example, slaves from the 1812 voyage of the Restuardor ended up in the lodge. By 1826 the lodge also included some "prize negroes" who were all from Mozambiquexxv. Mozambique and the East Coast of Africa became the major source of lodge slaves after 1786.

​

Robert Ross summarizes the current lack of knowledge of the Cape slave trade:
[The Cape slaves] derived almost exclusively from five areas: the Indonesian Archipelago, Bengal, South India and Sri Lanka, Madagascar and the East African coast. They had been bought in the slave markets of Batavia, Chinsura, Cochin, Boina and Delegoa Bay or Mozambique island, brought there by an as-yet-little-understood network of traders including Bengali Banians, Buginese trader pirates, Chinese junk captains, Saklava kings Prazeros on the Zambezi, Portuguese officials in Delegoa Bay and south to Natal or [from] kidnappers in South India. From there they were sent to  the Cape, very often in small numbers and semi-legally as the cargo of sailors and officials on the VOC ships, otherwise in more regularized, probably large French and Portuguese slaving ships which sold off a number of their generally sickly slaves on their way to the New Worldxxvi.

​

How persistent the notion of Asian origins of the imported Cape slave population proved to be is exemplified in the most recent -- and otherwise excellent -- work on Cape slavery where the author refutes an accurate nineteenth century source on the African origins of Cape slaves in that period, with the quite erroneous assertion that "large numbers were, in fact, Asians." Not even a footnote was needed to refute the original source, evidence of how ingrained the early assumptions still are in the current literaturexxvii.

​

New methodology used
In the light of both conflicting estimates of the proportions of origins of Cape slaves, how does one establish the origins of the Cape slave population? The approach used here is to combine all the newly collected ship-to-shore transactions from 1656 to 1732xxviii with later (1732-1799), random samples of probated slave owners' inventories, sources which all included the individual slaves' names and origins. In addition, I have tabulated slaves' origins mentioned in surviving Cape Town domestic sales (eliminating any double countingxxix ) and blended them with the origins of individual rural slaves' origins collected by Nigel Worden and Margaret Cairns and other Cape historians. I have also tallied slaves' origins mentioned in Cape Vendus and repatriations between 1658 through 1806, to obtain annual proportional estimates for the entire periodxxx. All these quite independently collected data (1656-1824) then, have been weighted, laggedxxxi and multiplied by the independently calculated census estimates of the annual slave trade to make the following graphs of the changing origins of the imported slave population. To these retrodicted annual totals, I have added the 5,100 lodge slaves, data which J.Armstrong has provided or overseen.

​

Probate Records
The probate data is mainly taken my own (1652 through 1799) and from Nigel Worden's, Margaret Cairns' and Frank Bradlow's tabulation of rural inventories. Their data have been adjusted to exclude "creole" and "unknown" components and also weighted from the relative regional proportions in the censuses to eliminate their rural bias. For the early nineteenth century I have exclusively relied on Margaret Cairns' tabulation of the slave registry returns from the second British occupation, both urban and rural, only pruning them of their "creole" and "unknown" components: both proportions changed independently of the operations of the slave trade.

​

Inventories Disadvantages
Inventories are troublesome sources for the origins of slaves since the probated slave was always passed on to another owner, and could therefore conceivably be counted again in another inventory of an heir. Since half of the slaves in a typical will went to the younger generation and since most slaves at the Cape suffered higher mortality than their owners it is unlikely that many imported slaves survived through two generations of owners to be counted twice   in consecutive generations of family wills. However, in the Cape's partible inheritance scheme, half of the estate went to the spouse, who was often a younger person, whether male or female. It is therefore conceivable, but not very likely, that some probated spouses' slaves could be counted twice when the spouse or, for that matter, any laterally related heir, died. Since imported slaves from different regions had differing mortality rates, probate sampling could therefore also result in some, but I believe, minor, double-counting and under-countingxxxii. Thus, (say) a long-lived Indonesian slave might survive through two owners, while a short-lived Malagasy slave might not live long enough to be probated.

​

The second drawback in using probated slaves' origins is that one can make few assumptions about exactly when an individual probated slave was imported. Since direct record linkage between a slave mentioned in a ship-to-shore transaction and the same slave recorded in a will, or vendu list is virtually impossible, because of changes of name (either baptismal changes or owner's preference), one is obliged to make an average estimate. The average period between importation and the probate date was obtained by subtracting the average age of all slaves in ship-to-shore transactions from the average age of all probated non-creole slaves. I have somewhat glibly assumed that both averaged ages remained relatively constant throughout the period 1658-1808 on the basis that the age composition of imported slaves stayed relatively stable while there were no drastic changes in the life tables of Cape slave ownersxxxiii.
Thus, we can say with some confidence that probated slaves were imported on average 17.5 years before they were probated. This is a very useful figure since it can be used to "lag" all probated slaves' origins to obtain a picture of the probable periods of importation for all imported slaves who were probated. Thus, to make an example of my assumption: I assume that Jephta of Bali, who was probated in the will of Jan Cornelissen in 1722, was probably imported in 1704. I rely here entirely on the law of large numbers.

​

Advantages
The advantages of using probate material far outweigh these considerable drawbacks and forced assumptions. First, wills, vendu lists and inventories concern transferable property and the compilers had a special interest in both their authenticity and accuracy. Second, inventories are the only source which reveals the origins of slaves belonging to Company officials after 1731, after which oceanic transfers recording their purchases were no longer maintained. A third attraction is that Cape probate material invariably provides information about both the locality and social status of the owner. Their final and compelling advantage lies in that they are also the only sources which systematically reveal the origins of all imported Cape slaves throughout the period, 1652 to 1833.
They will assuredly provide the sources from which refinements of these estimates will emerge.

​

i Nevertheless there was much deviousness in the local record keeping. The accounting procedures are deeply flawed on every voyage: totals do not agree and deliberate confusion is suspected. Some voyages are not mentioned, but did take place. The May 1679 voyage of the 1,000 ton Oliphant, oneof the very largest ships in the Dutch fleet is a case in point, according to the authors of Dutch Asiatic Shipping, the ship was "laid up in Batavia in 1665," but there was a voyage to Madagascar recordedin the Brieven Overcomen, however the journal in the inventory is marked uijtgenomen, (i.e "takenout"). Cf. VOC 4014 (May 1679) folios 546, 719. Nor is the voyage mentioned in the heavily indexed Resolutions (vol. 2). The 1698 voyage of the Swaag is not mentioned in Dutch Asiatic Shipping, but it did bring slaves to the Cape, cf. Transporten T1697/8 (27 January, 1698), no pagination. Clear deceit is evident in the daily journal of February 10, 1711: "The Council decides to retain a vessel here until the commissioner's arrival...none have been imported since the time of Simon van der Stel,"LJ, page 245. No trip is recorded and yet we know from the transporten that hundreds of slaves were imported from the time of Simon van der Stel. A "list of articles required for a slave trip to that Island[Madagascar]" in 1719 is mentioned in Leibbrandt's Requesten, 5 index, page 7371 [incorrect], but see LR 2 (no 59 18th July, 1719): 737, but there was no voyage mentioned in the carefully indexed and annotated Resolutions! We can be therefore be fairly certain that the documents concerning the lodge slave trade are not as straight forward as they appear to be.
ii  Some of the Company slave voyages have been published: viz. Snuffelaar [and de Zeepost] in E.C.Godée-Molsbergen (comp.), Reizen in Zuid-Afrika ('S Gravenhage: Martinus Nijhoff, 1976 edition,)3:220-257; De Brak captained by Otto Lüder Hemmy, in A. Grandidier et al., Collection des Ouvragesanciens concernant Madagascar 6 (1913) virtually impossible to obtain; De Brak in Maurice Boucher,"The Voyage of a Cape Slaver in 1742" in Historia 24, 1 (May 1979): 50-58; see also his fragment of a Danish Company voyage and the journal of a supercargo, Ibid., "An Unexpected Visitor: Charles Barrington at the Cape," South African Historical Journal 13 (1981): 20-35; Robert Ross ed., "The Dutch on the Swahili Coast, 1776-78," part one in International Journal of African Historical Studies19 2 (1986): 479-506; and part two in Ibid., 19 3 (1986), 305-358. The director of the South African Library, Dr Westra, is busy with the voyage of the Leidsman in 1715 and it will appear shortly in the Quarterly Bulletin of the South African Library. Non-company sources, such as travellers' accounts, not being subject to the purview of the Company directors provide invaluable commentary on the lodge's archival sources.
iii  The main techniques are drawn from a handbook published by the United Nations, Department of Economic and Social Affairs, The Concept of a Stable Population: Application to the study of populations of countries with incomplete demographic statistics. (New York: United Nations, 1968). Thischapter draws its evidence from the micro level of individual slave sales (1656-1732), from slavevoyages (1656-1822), from aggregated imports derived from the censuses (1656-1833), the "compensation lists" 1823-1829, CAD SO 17/1 and from a detailed estimate of the population madeby British officials in August of 1833, PRO CO 53/57. The first class of records are the transfers, detailed records of individual slaves from 1658 through 1717. These slave transfers were until recently inter filed with the cadastral records of the Cape Town municipal Deeds Office (DO): "Transporten enSchepenkennis" (1680 to 1715), and the over lapping and continuing series 1717 to 1732 in CAD:"Obligatien, Schultbrieven" and "Misc. notarial documents," CJ 2914 to CJ 3081. Since some statements and all graphs in this book are each based on several thousand observations, I could scarcely footnote all, instead, an appendix reference in the present author's dissertation is alluded to in the footnote. See specifically the "sales appendix" in the author's dissertation for a full description of the compilation and coding of data. After 1717 -- the year of the decision to base the colony on slave labor -- the individual transfer records start deteriorating until they disappear after 1732, in which year a new system of registering slaves came into being. Thereafter, slave traders and buyers only had acopy of the deed of sale. Very few of these private deeds of sale have survived in miscellaneous collections. There was no central registry of slaves from 1732 until 1817, the year of the establishment of the Slave Office. After 1732, therefore, the researcher is obliged to turn to the censuses, which begin in 1658 and run to 1793, start again in 1806 and run through to 1826.
iv  See appendix to this chapter.
v  De Kock, Those in Bondage, page 36.
vi  The latter also provide a guide to the number of creole slaves born in the households of the patriciate.
vii  See the five volume Requesten, which contain over 3,000 such applications.
viii  R. Shell, "Immigration: the forgotten factor in South African frontier expansion," SARP work shop paper, 1980.
ix See chapter on domestic trade.
x  See appendix to this chapter.
xi  J.L.M. Franken, "Vertolking aan die Kaap in Maleis en Portuguees," Huisgenoot (18 June 1930), page 41, et seq.
xii  Cruse, Opheffing, page 267-8.
xiii  See chapter on manumission.
xiv  I.D. du Plessis, The Cape Malays, page 3.
xv F.R. Bradlow, & Margaret Cairns, The Early Cape Muslims... (Cape Town: Balkema, 1978), pp.80-106.
xvi  Shell, "The Establishment and Spread of Islam," page 21.
xvii  I am aware that Hans Heese and Leon Hattingh have attacked this definition of freeblacks. Their criticism has arisen at least in part, since there was no real legal definition of free blacks in the Dutch East India Company period. Under the British administration this changed. A document detailing thelegal status of Free blacks before Ordinance 50 clearly includes the political, exiles and convicts as"Free Blacks." CAD M142 (6) "Laws relating to Freeblacks in the Colony of the Cape of Good Hope "Article one states clearly: "The class of people commonly called Freeblacks, has originated in this Colony partly from the manumissions of Slaves partly from the transportation of native Inhabitants ofsome of the East India Countries, removed to this Colony either as State- prisoners or as Convicts (vide:Interdiction of the 20 April 1779) and finally also from the arrival of Indians and Chinese, who have been allowed for a time to reside in this Colony subject to removal
at the pleasure of Government. The Freeblacks are thus originally foreigners, and as such distinguishable from the Hottentots and other Natives of this Colony." This definition of Free blacks is contested in Hans Heese, "Challenging certain aspects of Intergroup Relations in the Shaping of South African Society, 1652-1840" A Review Article," Kronos, 17, pp. 71-76, but see the authors' "A reply to Hans Heese" and also Heese's riposte, Kronos, 18 (forthcoming).
xviii  R.Shell, "The Short Life and Personal Belongings of One Slave, Rangton of Bali," Kronos, 18(1991) forthcoming.
xix Bradlow and Cairns, The Early Cape Muslims, Table 7, page 99.
xx  However, all the biased data must be ruthlessly excised; all creole slaves must be cut from each remaining table; all "unknowns" must be dropped, a value which changes quite randomly in each classof documents. Leon Hattingh, the former head of the Historical Research Institute at the University of the Western Cape, has also counted the transfers, carefully rechecking and publishing verbatim copies of the primary material which Anna Böeseken and Margaret Cairns had earlier collected and published in 1975. His meticulously published material, up to 1679, in so far as it corrects the Böeseken data, has been incorporated. J.L. Hattingh, "Kaapse Notariële Stukke Waarin Slawe vanVryburgers en Amptenare Vermeld Word" Kronos 14, (1988-9): 14-37; continued in subsequent issue, pages are repeated in my copy, no doubt "a printer's error." Hattingh, however, does not deal with the lodge slaves imported by the Company.
xxi  See discussion on the revised opgaaf returns in appendix one. James Armstrong and Nigel Worden, "The slaves, 1652-1834" in Shaping, page 120; Nigel Worden, another historian of Cape slavery, has concentrated on rural Cape slavery and has also made some tentative estimates of slave origins in the rural area of Stellenbosch (1722-1799) in one table, based on probate material, which in broad measure supports Franken and Cruse. However, the oceanic slave trade was conducted in theport city and his conclusions about that trade are therefore heavy inferential; also, since he exclusively worked on burgher inventories, all officials's slave origins are missing. Also, according to the census(opgaaf) returns, rural rates of creolization were quite different from urban, another source of bias in his estimation of origins of imported slaves. There were, for instance more slave children per womenin the Stellenbosch district. This is based on slave child/women ratios by region in archival opgaafs,1658 through 1834, but a handy published summary (every five years) is found in Coenraad Beyers, Die Kaapse Patriotte (1779-1791) (Cape Town: Juta, 1929) Bylaag (H): Opgaaf rolle. . . Burghers,1701-1793, pages 240 to 249 and for annual totals see Ross et al., The Economy of the Cape Colony in the Eighteenth Century (Leiden: Intercontinenta, No 7, 1987), Appendices. Nigel Worden has however, tabulated many thousands of cases from Stellenbosch inventories, valuable data which he has graciously shared and explained and I have reworked into these graphs.
xxii  His methodology is based on identifying and counting slave voyages. Since over 10,000 ships visited the Cape in the Company period alone, it will be a long time before we know the result of this research, since there is no efficient way of identifying ships carrying slaves. Since the vast bulk of the Cape slave trade was conducted in a clandestine manner by Company officials and wealthy urban burghers building their fortunes, and since the same people oversaw the archival sources detailing the oceanic trade in their primary state, records of the illegal trade are, not surprisingly, both sparse and extremely misleading. In the light of deliberate distortion by officials concerning the importation of privately owned slaves, the historian must make estimates which go beyond the preserved records, whose primary purpose, I argue, was to mislead the Company directors.
xxiii  James Armstrong and Nigel Worden, "The Slaves," pages 122-125.
xxiv  There were attempts by the company to slave at Kilwa, Sofala, Zanzibar and the Comores islands later in the eighteenth century: Robert Ross, ed., "The Dutch on the Swahili Coast, 1776-78, "International Journal of African Historical Studies 19 2 (1986): 479-506; and Part Two in Ibid., 193 (1986), 305-358.
xxv  RCC., "Return of Government Slaves," 19:264-268.
xxvi  Robert Ross, Cape of Torments, page 13.
xxvii Richard Watson, The Slave Question (Middletown: Wesleyan Press, 1999), page 131.
xxviii  Because of the purposeful degradation of the record-keeping of oceanic slave transfers after the inauguration of Jan de la Fontaine, a slave-trading Governor, one cannot consult oceanic slave transfers after 1732. Cf. Guelke, Shell and Whyte (comp.), The de la Fontaine Report (New Haven: Opgaaf Project, 1990), introduction.
xxix  The latter two sets of 88 slaves provided by James Armstrong.
xxx  Repatriation from South Africa was a random event and therefore slave transfers resulting from repatriations provide a reliable guide to the origins of slaves. Since most repatriating owners were urban, these latter data, suitably weighted to their true proportion of the census, offset the considerable rural emphasis in the work of Cairns, Bradlow and Worden. An early list of repatriating owners maybe found in Guelke, Shell and Whyte (comp.), The de la Fontaine Report, New Haven, 1990. A more comprehensive list of repatriating owners is in Leibbrandt, Requesten, 5 Volumes, passim.
xxxi  All probated material was lagged by 17 years, which was the difference between the mean age of imported slaves and the mean age of probated slaves.
xxxii  See later section on mortality, page 000.
xxxiii  The average age of 434 imported slaves in all probated sales, i.e. sales in which either the Office of Sequestration, Executors, Orphan Chamber, or Vendu Master were mentioned, was 37 years; whilein 1,170 ship-to-shore transactions, which recorded the exact age of the slave, the averaged age was 19.5 years. Many of the "unknowns" in the latter data were simply listed as "son" or "daughter," i.e.children under 14, which would suggest that if we knew the exact age of those children, a more accurate average would result in a longer period between importation and probation.

​