² Walker, E. A., History of South Africa (first edition, Johannesburg, 1928), 210. See also footnote 69 below.Google Scholar

³ Although Zulu Aftermath was written in Nigeria and London, Omer-Cooper comes from the eastern Cape of South Africa, a region to which many of the foundations of mfecane theory can be traced back.Google Scholar

⁴ Edited by M. Wilson and L. Thompson. Chapters 8 and 9 by Thompson refer to the mfecane. For briefer overviews see Davenport, T. R. H., South Africa: A Modern History, 3rd ed. (Johannesburg, 1986), 12–21;Google Scholar and Maylam, P., A History of the African People of South Africa (Cape Town, 1986), ch. 4.Google Scholar

⁵ The first criticisms that I have traced are in Cornevin, M., Apartheid. Power and Historical Falsification (Paris, 1980), especially chs. 8–11 and 14.Google Scholar See also Satir, V. E., ‘The Difaqane: Fact vs. Fiction’, Educational Journal, LV, 2 (09 1983), 6–14;Google Scholar and Cobbing, J., ‘The Case Against the Mfecane’, African Studies Centre Seminar Paper, University of Cape Town, March 1983,Google Scholar which was developed in Cobbing, ‘The Myth of the Mfecane’, Seminar Paper presented to the History Department, University of Durban-Westville, June 1987.

⁶ Like the ‘mfecane’ the concept ‘Nguni’ is a twentieth-century invention of European academics. See the important essay by Wright, J. B., ‘Politics, ideology and the invention of the “Nguni”’, in Lodge, T. (ed.), Ideology and Resistance in Settler Societies (Johannesburg, 1986).Google Scholar

⁷ This internal revolution and suggestions of structural innovativeness were only given rudimentary definition by Omer-Cooper in the 1960s, and were welded onto earlier highly pejorative misinterpretations.Google Scholar

⁸ Omer-Cooper, Zulu Aftermath, 180. This is a theme of the Tomlinson Report of 1955–1956.Google Scholar

⁹ For inconclusive attempts to demonstrate rapid population growth after about 1790, see Hall, M., ‘Dendroclimatology, rainfall and human adaptation in the later Iron Age of Natal and Zululand’, Annals of the Natal Museum, XXII, 3 (11 1976), 693–703;Google Scholar and Guy, J., The Destruction of the Zulu Kingdom (Johannesburg, 1982), 8–12. Evidence noted below, 504–507, suggests population must have declined.Google Scholar

¹⁰ Hedges, D. W., ‘Trade and politics in Southern Mozambique in the eighteenth and nineteenth centuries’ (Ph.D thesis, University of London, 1978), especially chs. 6 and 7;Google ScholarSmith, A., ‘The trade of Delagoa Bay as a factor in Nguni politics, 1750–1835’, ch. 8, in Thompson, L. (ed.), African Societies in Southern Africa (London, 1969).Google Scholar Hedges attributes the stimulus to state expansion in black societies to competition to supply cattle to American whalers at Delagoa Bay after a downswing in the ivory trade in the early 1790s – an unconvincing hypothesis in a nevertheless important study.

¹¹ Smith, ‘Trade of Delogoa Bay’, 176–7, mentions slaving but only to belittle it as a cause of change in African societies: ‘it can be seen that it [the slave trade] has little relevance for consideration of the trade hypothesis and the development of the Zulu nation’.Google Scholar

¹² Harries, P., ‘Slavery, social incorporation and surplus extraction: the nature of free and unfree labour in South-East Africa’, J. Afr. Hist., XXII (1981), 309–30. Some of the conclusions of this crucial article are summarized below.CrossRefGoogle Scholar

¹³ Cobbing, J., ‘The Ndebele under the Khumalos, c. 1820–96’ (Ph.D thesis, Lancaster University, 1976), 12–13.Google ScholarRitter, E. A. in Shaka Zulu: The Rise of the Zulu Kingdom (London, 1955), 150, 228–9, has a fictional account of Mzilikazi joining Shaka, rising to be a general, and then rebelling.Google Scholar

¹⁴ Macmillan, W. M., Bantu, Boer and Briton (Oxford, 1929), 19–20.Google Scholar

¹⁵ Cobbing, ‘Ndebele under the Khumalos’, ch. I; Rasmussen, R. K., Migrant Kingdom: Mzilikazi's Ndebele in South Africa (Cape Town, 1978).Google Scholar

¹⁶ See Fig. I. The argument requires careful reference to the maps. In an attempt at clarity I have used modern territorial names, e.g. the Transvaal, Orange Free State, Transkei etc., anachronistically. It should be remembered that apart from the Cape Colony - which reached the Fish River in 1812 and the Keiskamma in 1820 - none of these boundaries existed in the 1820s.Google Scholar

¹⁷ See below, p. 515.Google Scholar

¹⁸ Outlined in Omer-Cooper, Zulu Aftermath, ch. 6.Google Scholar

¹⁹ A concept invented by Ellenberger, D. W. in History of the Basuto Ancient and Modern (London, 1912),Google Scholar especially Second Period, ‘The Lifaqane wars’, 117 onwards. Unlike Walker's mfecane, which referred to a sub-continental pattern of destruction, Ellenberger's difaqane defined an alleged middle period, i.e. a time-span of Sotho history with particular reference to the Caledon valley. This period of bloody destruction (caused by Nguni invaders) separates a pre-1820 era of peace from a post-1833 (arrival of the missionaries) era of recuperation and progress towards civilisation, Mfecane and difaqane are now in practice used interchangeably, although the Caledon stress of the latter often remains.

²⁰ 1822–4, after which they returned to normal. See Sanders, P., Moshoeshoe. Chief of the Sotho (London, 1975), ch. 4.Google Scholar

²¹ Apparently first alleged by Livingstone; see Schapera, I. (ed.), Livingstone's Private Journals 1851–1853 (London, 1960), 18, and footnotes 58 and 152 below.Google Scholar

²² Since Marion How's ‘An alibi for Mantatis’, African Studies, XIII, (1954), 65–76.Google Scholar The most accessible newer version is Lye, W. F., ‘The Difaqane The Mfecane in the Southern Sotho area, 1822–24’, J. Afr. Hist., VIII (1967), 107–31.CrossRefGoogle Scholar See also Smith, E. W., ‘Sebetwane and the Makololo’, African Studies, xv (1956), 50, repeated in OmerCooper, Zulu Aftermath, ch. 8. See also below, n. 30.Google Scholar

²³ E.g. Cory, G. E., The Rise of South Africa 1820–34 (London, 1913), 363.Google Scholar

²⁴ Necessary to establish the truth of the rumours that the Zulu were operating in the Transkei in 1828.Google Scholar

²⁵ Bannister, S., Humane Policy or Justice to the Aborigines of New Settlements (London, 1968 reprint [originally London, 1830]), 228.Google Scholar

²⁶ For the ‘battle’ see Schapera, I. (ed.), Apprenticeship at Kuruman (London, 1951), 77–203.Google Scholar Moffat's account was first published in The Cape Gazette, 26 July 1823. See also below, nn. 28–30, for a discussion of the identification problem.Google Scholar

²⁷ See Thompson, George, Travels and Adventures in Southern Africa (Cape Town, 1967 reprint [originally London, 1827]), I, chs. xv and xvi.Google Scholar

²⁸ Schapera, Apprenticeship, 92, the coming on the old man and child, for example; and 95, where the Mantatee ‘poorer class’ are depicted seizing and eating meat in the middle of the fighting.Google Scholar

²⁹ Moffat does not explain why he returned to fetch the Griqua. It is probable that certain information is being omitted. See also footnote 142 below.Google Scholar

³⁰ Thompson, Travels and Adventures, 1, ch. xv; Schapera, Apprenticeship, 89. Note the surreal atmosphere of Thompson's alleged sighting of the Mantatees and the deserted village reminiscent of the Marie Celeste in Travels and Adventures, 1, 108–10.Google Scholar

³¹ Moffat (Schapera, Apprenticeship, 95) claimed ‘the slain of the enemy was between 400 and 500’, but it is probably advisable to allow for exaggeration.Google Scholar For the claim that Mantatees burnt their own villages see ibid., 93 and Thompson, Travels and Adventures, I, 146.Google Scholar

³² Ibid., I, 153.

³³ For the detail about Melvill see ibid., 1, ch. xvi, headed ‘Mr Melvill's Narrative of Transactions after the Battle, and of His Excursion to Rescue the Women and Children of the Invaders’; the quote about ‘cannibal ferocity’ is ibid., 1, 155; for Moffat in Cape Town see the South African Commercial Advertiser, 7 January 1824,Google Scholar reprinted in Theal, G. M., Records of the Cape Colony, 34 vols. (Cape Town, 1903–1906), XVI, 497–505;Google Scholar see also Philip, J., Researches in South Africa, 11 (London, 1828), 142–6,Google Scholar for the quarrel over the prisoners; for Melvill's despatch of prisoners to Graaff Reinet see Cope, R. L. (ed.), Journals of the Rev. T. L. Hodgson (Johannesburg, 1977), entry for 23 07 1823, p. 182;Google Scholar for the payment see Bird to Landdrost Reinet, Graaff, 27 08 1823, in Theal, Records, XVI, 223.Google Scholar

³⁴ For a rare intervention in this field see Newton-King, S., ‘The labour market of the Cape Colony, 1807–28’, ch. 7, in Marks, S. and Atmore, A. (eds.), Economy and Society in Pre-Industrial South Africa (London, 1980).Google Scholar

³⁵ For these events see Maclennan, B., A Proper Degree of Terror. John Graham and the Cape's Eastern Frontier (Cape Town, 1986);Google Scholar for the shootings see Theal, Records, XXIII, 212.Google Scholar

³⁶ For liberal mystifications of these serf codes see De Kiewiet, C. W., A History of South Africa. Social and Economic (London, 1941), 45–6;Google Scholar and Davenport, T. R. H., section: ‘The extension of freedom under the law’,Google Scholar in Wilson, and Thompson, , Oxford History, I, 293 and 303.Google Scholar For Davenport the Caledon Code was a point where ‘the removal of the disabilities of coloured persons reached an important stage’! For an admission that the Codes led to enserfment see Walker, History of South Africa (first edition, 1928), 155–6.Google Scholar

³⁷ Newton-King, ‘Labour market’, 182–91.Google Scholar

³⁸ Philip, Researches, II, chs. 14 and 15. This genocide needs a thorough study. The psychological dehumanization probably facilitated later massacres.Google Scholar

³⁹ See Newton-King, ‘Labour market’, 192–3 for the only source which correctly treats Mantatees as labour.Google Scholar

⁴⁰ Philip, , Researches, II, 85–6,Google Scholar reported Boers' claims that Sotho came voluntarily seeking food. Their account was they had been living peacefully ‘when a people (called Bergenaars) riding upon horses, and with fire-arms, came upon them and killed many of them, and took away all their cattle and many of their children’. Attempting to follow their children they were ‘detained by the boors’. See also ibid., 90–5. For the myth see Theal, G. M., History of South Africa 1795–1872, IV (4th edition, London, 1915), 446. Bergenaars were Griqua who broke away from the main group under Waterboer in 1822–23.Google Scholar

⁴¹ Somerset to Bathurst, 30 July 1825, in Theal, , Records, XXII, 419–22;Google Scholar Bird to Landdrost Reinet, Graaff, 27 August 1823,Google Scholaribid., XVI, 223; Secretary to Government to Landdrosts Graaff Reinet and Somerset, 21 July 1825, ibid., XXXII, 425, my emphasis.

⁴² E.g. Cory, Rise of South Africa, 1820–34, 236.Google Scholar

⁴³ By June 1825 there were nearly three hundred acknowledged Mantatees in Graaff Reinet alone, excluding those already apprenticed under Somerset's order of 27 Aug. 1823; see Stockenstrom to Secretary to Government, 1 June 1825, in Theal, , Records, XXII, 422.Google Scholar

⁴⁴ Ibid., XXXIV, Minutes of Council, 2 February 1827. For non-Mantatee labour of 3/- a day (plus food and wine) see Somerset to Bathurst, 31 March 1825 in ibid., xx, 400–1.

⁴⁵ Philip, , Researches, II, 91;Google Scholar letter from ‘Investigatus’ in Graham's Town Journal, 30 01 1834;Google ScholarLegassick, M., ‘The Griqua, the Sotho-Tswana, and the missionaries, 1780–1840: the politics of a frontier zone’ (Ph.D thesis, University of California, Los Angeles, 1970), 354–5.Google Scholar

⁴⁶ Ibid., 353; Philip, , Researches, II, 82–3.Google Scholar

⁴⁷ Mantatees were sold, alleged to have been paid, and legalized under ‘apprenticeship’, i.e. serf, codes. Whether they were resold is not known.Google Scholar

⁴⁸ By taking the western chain before the eastern chain (see below, pp. 503–507) I do not mean to imply that the events in the west had a chronological or any other priority over those in the east. The ‘chains’ intermeshed simultaneously.Google Scholar

⁴⁹ Legassick, M., ‘The northern frontier to 1820’, ch. 7, in Elphick, R. and Giliomee, H. (eds.), The Shaping of South African Society, 1652–1820 (Cape Town, 1979), 259, 262.Google Scholar

⁵⁰ Rasmussen, Migrant Kingdom, 46, 51; Cobbing, ‘Ndebele under the Khumalos’, 21;Google ScholarPhilip, , Researches, II, 85–6.Google ScholarThompson, L., Survival in Two Worlds. Moshoeshoe of Lesotho, 1786–1870 (Oxford, 1975), 57; dates the raids into the Caledon from ‘about 1825’, but this is too late.Google Scholar

⁵¹ Lye, W. F. (ed.), Andrew Smith's Journal of His Expedition into the Interior of South Africa, 1834–36 (Cape Town, 1975), 74, where Moshoeshoe's father, Mokhachane, ‘called upon us to observe the many marks of wounds from musket balls’.Google Scholar

⁵² See below, pp. 508.Google Scholar

⁵³ Lye, Andrew Smith's Journal, 48–50, for Dantzer's group;Google Scholar see also Holt, B., Greatheart of the Border. A Life of John Brownlee, Pioneer Missionary in South Africa (King William's Town, 1976), 15.Google Scholar

⁵⁴ Legassick, ‘Politics of a frontier zone’, 326, 355–60; Schapera, Apprenticeship, 65–6, 73;Google ScholarKirby, P. R. (ed.), The Diary of Dr Andrew Smith, 1834–36 (Cape Town, 1939–1940), 1, 358, entry for 17 April 1835.Google Scholar

⁵⁵ Initial, but not always very reliable, information on the Taung is in Theal, G. M., Basutoland Records, 1833–52, I (Cape Town, 1964 reprint), 517–18;Google ScholarMacGregor, J. C., Basuto Traditions (Cape Town, no date [c. 1906]), 58–67;Google ScholarEllenberger, History of the Basuto, 54–67, 165–9, 173, 306. See also Legassick, ‘Politics of a frontier zone’, 336–9. Legassick's pioneering study is marred by the symptomatic confusion of referring to the Taung as ‘difaqane marauders’, an error which is traceable to the Mantatee = MaNtatisi elision discussed below, pp. 514–515.Google Scholar

⁵⁶ E.g. Broadbent, S., A Narrative of the First Introduction of Christianity amongst the Barolong Tribe (London, 1865), 158.Google Scholar

⁵⁷ Ibid., 131–3, 173; Schapera, Apprenticeship, 145–55;Google ScholarCope, , Journal of the Rev. T. L. Hodgson, 9, 246, 322; Legassick, ‘Politics of a frontier zone’, 339.Google Scholar

⁵⁸ Livingstone (see Schapera, Livingstone's Private Journals, 18Google Scholar) claimed in the early 1850s that Sebetwane's people told him Sekonyela's Tlokwa drove the Kololo from Kurutlele (near modern Welkom) ‘in the first instance’, after which they ‘fled to [my emphasis] Sekonyela's present country’ (upper Caledon). Ellenberger, , History of the Basuto, 60, 306,Google Scholar has one part of the Patsa driven north by the Taung, another by the Tlokwa. It is possible that this split is to accommodate both Livingstone and material from traditions. My suspicion is that both Livingstone and Ellenberger were conflating ‘Mantatees’ - a word applied to both Taung and Patsa-Kololo - with ‘MaNtatisi’ (hence Tlokwa), and that no Tlokwa attack took place. Ellenberger's dating of the Tlokwa attack to June 1822 (p. 306) is purely imaginary: we have no information as to the chronology of the Patsa-Kololo before 1824 (when they were in Ngwaketse country). Since Sekonyela was only a boy in 1821–3, Livingstone is likely to have supplied the name from missionary preoccupation with Sekonyela in the period 1834–54, during which time the mistaken Mantatees = MaNtatisi (Tlokwa) elision was becoming habitual. For the chronology of the elision see below, pp. 514–515.

⁵⁹ Lister, M. (ed.), Journals of Andrew Geddes Bain (Cape Town, 1949), 65–71;Google ScholarLegassick, ‘Politics of a frontier zone’, 352.Google Scholar

⁶⁰ Rasmussen, Migrant Kingdom, 62–7. For allegations about Zulu devastations in the Orange Free State see most Standard Eight school textbooks in South Africa.Google Scholar

⁶¹ Discussed below, pp. 503–507. The same caveat is necessary when considering the impact of the Ndebele in the southern and western Transvaal.Google Scholar

⁶² Sanders, Moshoeshoe, 51–2. Sanders has the difaqane end with Moshoeshoe's defeat of the Koranna in 1835–6. This epitomizes the confusion about the difaqane.Google Scholar

⁶³ Stockenstrom to Secretary to the Government, 1 June 1825, in Theal, , Records, XXII, 423, for example.Google Scholar

⁶⁴ E.g. Cope, Journals of the Rev. T. L. Hodgson, 184–5, where Hodgson describes a case which he associated with the consequences of the Griqua raid at Dithakong.Google Scholar

⁶⁵ Moffat and Melvill at Dithakong, for example, and Bain at Dithubaruba in 1826, who claimed he loaded his guns with blank ammunition (for his readers), but in the attack used live bullets of course; see Lister, , Bain Journals, 58, 69–70, and footnote 117 below.Google Scholar

⁶⁶ The initial estimate of the number of Mantatees at Dithakong was 30,000. Melvill calculated this by guessing the area of ground covered by the ‘enemy’, and allowing one square yard for each person. Within a few days 30,000 had become 50,000. To boost the alibi further, these engaged Mantatees were said to be half of the total Mantatee army, which produces 100,000, just enough to support claims of a danger to the Cape Colony. It is unlikely that the victims numbered more than about two thousand, and possibly less.Google Scholar

⁶⁷ Thompson, , Travels and Adventures, I, 179; II, 115.Google ScholarPringle, Thomas, in African Sketches (London, 1834), 359–60, has the labour driven partly by ‘Mantatees’ and partly by Griqua – the first stage of the alibi.Google Scholar

⁶⁸ See below, pp. 518–519.Google Scholar

⁶⁹ The word comes in many spellings: infanicama, imfetcanie, Il-Fitcanie, fickanees etc., but I have standardized on Fetcani. The word was used by Walker for his 1928 neologism ‘mfecane’. The linguistic connexions between the ‘Sotho’ difaqane and the Tembu-Xhosa Fetcani are worth exploring.Google Scholar

⁷⁰ Van Wyk to Mackay, in Theal, , Records, XXXI 51–2.Google Scholar

⁷¹ Pringle to Thompson, May 1825, in Holt, Great Heart, 58–9; Rogers to Forbes, 27 May 1825, and Pringle to Massey, 29 May 1825,Google Scholar both in Theal, , Records, XXII, 428–33;Google Scholaribid., XXXIV, 18–22, 453–69.

⁷² Shrewsbury to Dundas, 12 June 1828, in Cory, Rise of South Africa 1820–34, 354–5; Dundas to Lt.Col. Henry Somerset, 14 July 1828,Google Scholar in Van Warmelo, M. J. (ed.), History of Matiwane and the AmaNgwane Tribe (Pretoria, 1938), 238.Google Scholar

⁷³ Ngwane movements between about 1825 and 1828 are unclear. It is not known when they left the Caledon. Either they directed their raids on the Tembu in 1827 from the Mbashe, or they returned north of the Orange in 1827 and migrated later. It seems unlikely they were still engaging Moshoeshoe at the turn of 1827–8, as many accounts claim. Their route was probably across the Witteberg towards the Stormberg.Google Scholar

⁷⁴ Lord Somerset, Charles to Bathurst, 31 March 1825,Google Scholar in Theal, , Records, XX, 400–1;Google Scholar see also Van der Poel et al. to Bourke, 30 June 1826, ibid., XXVII, 90–8: ‘We inhabit a country of which the population is not and never has been equal to the extent of Territory nor adequate to the proper cultivation thereof’.

⁷⁵ Philip to Directors of London Missionary Society, Nov. 1826,Google Scholar in Theal, , Records, XXX, 150–1:Google Scholar ‘and when these [demands by the Graham's Town military] could not be complied with by the missionaries, when they could not send the number or the particular persons wanted, they received threatening letters or were summoned to appear at Graham's Town, as if they had been slaves themselves’.

⁷⁶ Bourke to Bathurst, 30 June 1827, ibid., XXXII, 53.

⁷⁷ Dundas to Lt.-Col. Somerset, 1 Aug. 1828, in Van Warmelo, History of Matiwane, 239–40.Google Scholar Dundas and Somerset persisted in claiming they believed the Fetcani were ‘Chaca's people’. It is inconceivable that the combined Anglo-Tembu intelligence system did not know that their victims were the Ngwane. For an account from the era of Theal by one present see Bowker, B. E., ‘The Fetkanie commando of 1828’, Cape Illustrated Magazine, V (09 1894–08 1895), 347–54,Google Scholar where the effect of fire-arms on the Ngwane is brought out. Omer-Cooper, Zulu Aftermath, 92, calls this ‘a skirmish’.Google Scholar

⁷⁸ About 16,000 men, including C. 531 gun-armed whites (five times the number who dispersed the Ndebele at Mosega in January 1837), which included most of the 55th Regiment; see Cory, Rise of South Africa 1820–34, 360–1.Google Scholar

⁷⁹ Van Warmelo, History of Matiwane, 250–66, especially Somerset to Bourke, 29 Aug. 1828.Google Scholar The information about the burning forests was only obtained from several informants in early 1938. For Hottentot participation see Bannister, Humane Policy, cclxxxii. Somerset said between four hundred and one thousand Ngwane were killed: hopefully this is an exaggeration?Google Scholar

⁸⁰ Somerset to Bourke, 29 Aug. 1828, in Van Warmelo, History of Matiwane, 254.Google Scholar

⁸¹ Ibid., 255.

⁸² Bird, J., Annals of Natal (Pietermaritzburg, 1888), 1, 123.Google Scholar But n.b. the Mantatees, though technically they were not ‘Kafirs’ ( = Xhosa etc.). The Ngwane prisoners of 1828 were the first substantial contingent of labourers to be called Fingos. Previously the word had applied only to the odd individual who had gravitated to the mission stations. See also footnote 140, below.

⁸³ Hedges, ‘Trade and politics’, 193, my emphasis. Like Hedges, both Omer-Cooper, Zulu Aftermath, 29, n. 3,Google Scholar and Bonner, P., Kings, Commoners and Concessionaires: The Evolution and Dissolution of the Nineteenth-Century Swazi State (Johannesburg, 1983), 28,CrossRefGoogle Scholar relocate the inception of the mfecane geographically from Zulu territory on the Mfolosi north to the Pongola and chronologically to the period before Shaka, but also without realizing the implications.

⁸⁴ Harries, P., ‘Labour migration from Mozambique to South Africa, with special reference to the Delagoa Bay hinterland, c. 1862–1897’ (Ph.D thesis, University of London, 1983), 147–83;Google Scholar and Harries, , ‘History, ethnicity and the Ingwavuma land deal: a short account of the Trans-Mkuzi in the nineteenth century’, J. Natal and Zulu Hist., VI (1983), 1–27.Google Scholar

⁸⁵ Bannister, Humane Policy, 143;Google ScholarKay, S., Travels and Researches in Caffraria (London, 1833), 403, for example.Google Scholar

⁸⁶ Harries, ‘Free and unfree labour’. The material in the first half of the article has not received anywhere near the attention it needs.Google Scholar

⁸⁷ For general analyses see Williams, E., From Columbus to Castro: The History of the Caribbean, 1492–1969 (London, 1970), ch. 17;Google Scholar and Vail, L. and White, L., Capitalism and Colonialism in Mozambique. A History of the Quelimane District (London, 1980), ch. I.Google Scholar

⁸⁸ Harries, ‘Free and unfree labour’, 316.Google Scholar In eighteen months of 1827–8, 2,800 slaves were exported from Lourenço Marques and Inhambane to Réunion alone. In the six years 1825–30 Rio de Janeiro alone received over four thousand officially declared slaves from Lourenço Marques, and 3,400 from Inhambane (ibid., 315).

⁸⁹ Harries, ‘Labour migration’, 148, estimates a total population of 80,000 to 120,000 for a hinterland extending to the lower Mkuzi and upper Pongola in the south-west and to the Limpopo and Olifants in the north.Google Scholar The pivotal importance of the slave trade is likely to lead to a debate over preciser definitions and estimates. Bryant's estimate for ‘Zululand’, i.e. the area between the Pongola and Tugela, in about 1820 was 80,000 plus; see Bryant, A. T., Olden Times in Zululand and Natal (London, 1929), 81.Google Scholar

⁹⁰ The losses would presumably have been concentrated in certain areas.Google Scholar

⁹¹ The exception that proves the rule is Smith (see footnote 11, above).Google Scholar

⁹² Harries, ‘Labour migrations’, 160, 161–75 (163 for Albasini);Google ScholarDelius, P., The Land Belongs to Us: The Pedi Polity, the Boers and the British in the Nineteenth-Century Transvaal (Johannesburg, 1983), 15–19.Google Scholar Delius, 15, admits: ‘The full reasons for the apparent escalation of conflict in the region [c. 1780–1820] and the extension of political and military scale that emerged from it remain elusive’. The concept ‘war-lord’ is borrowed from Martin Hall (private correspondence) who, I gather, would apply the idea to a wide range of individuals, including Fynn, Shaka, Moritsane, Moshoeshoe, Barend Barends etc.

⁹³ Harries in ‘Labour migration’, ch. 4, and in ‘Trans-Mkuzi in the nineteenth century’ makes a start.Google Scholar

⁹⁴ Burrows, E. H., Captain Owen of the African Survey, 1774–1857 (Rotterdam, 1979), 120; see also 155.Google Scholar

⁹⁵ Bonner, Evolution and Dissolution, chs. 2 and 3. The Dlamini were extruded from the south Bay region in approximately the 1780s, from where they migrated to the north bank the middle Pongola. Under Ndvungunye in the period c. 1790–1810 they contributed the regional violence. The subsequent move north-west under Sobhuza has neither been satisfactorily explained nor dated.Google Scholar

⁹⁶ A study of Ndwandwe is urgently needed. For a beginning see Hedges, ‘Trade and politics’, 155–65.Google Scholar

⁹⁷ Stuart, J. and Malcolm, D. (eds.), The Diary of Henry Francis Fynn (Pietermaritzburg, 1950), 7; see footnote 120 below.Google Scholar

⁹⁸ Harries, ‘Free and unfree labour’, 314.Google Scholar

⁹⁹ Macmillan, Bantu, Boer and Briton, 36.Google Scholar

¹⁰⁰ Kay, Travels and Researches, 403.Google Scholar

¹⁰¹ Stuart and Malcolm, Fynn Diary, 48, my emphasis. Dingane was doing the same in the 1830s; see Harries, ‘Free and unfree labour’, 314, footnote 21.Google Scholar

¹⁰² Bannister, Humane Policy, 137–9, cclxv.Google Scholar

¹⁰³ Thompson, Survival in Two Worlds, 18–20.Google Scholar

¹⁰⁴ Theal, Basutoland Records, 1, 1–2.Google Scholar

¹⁰⁵ Ellenberger's largely fictional chronologies and battles have too frequently been accepted uncritically by recent historians. Sanders, Moshoeshoe, 31, reveals Ellenberger's mistaken dating of the alleged Ngwane–Hlubi battle to 1825. In fact the conflict was several years earlier. The interpretational consequences of such errors have been farreaching. It is best not to rely on Ellenberger unless he is corroborated by completely independent sources.Google Scholar

¹⁰⁶ Thompson, Survival in Two Worlds, 35–40.Google Scholar

¹⁰⁷ Mackay to Secretary to the Government, 8 Aug. 1827, in Theal, Records, xxxiv, 464, who described the Ngwane as ‘two tribes formerly distinct “Masotu” and “Manguana”’. For the Ndebele see Cobbing, ‘Ndebele under the Khumalos’, 32.Google Scholar

¹⁰⁸ Thompson, Survival in Two Worlds, 44. Thompson describes Moshoeshoe as a vassal of Matiwane's. But the two growing states may have been more equally balanced in power. Note that after Mbolompo many Ngwane returned to join Moshoeshoe.Google Scholar

¹⁰⁹ Godlonton, R., A Narrative of the Irruption of the Kaffir Hordes into the Eastern Province of the Cape of Good Hope, 1834–35 (Struik reprint, 1965 [originally Graham's Town, 1835]), 51.Google Scholar

¹¹⁰ I do not regard the allegations either of Ellenberger or those in Van Warmelo as reliable. There is no evidence from the Zulu or Ndebele sides. ‘Zulu’ meant any Nguni. The Zulu stories were part of the myth of Shaka which took off after about 1826, and which needs analysing.Google Scholar

¹¹¹ Rogers to Forbes, 27 May 1825, in Theal, Records, XXII, 430;Google ScholarTooke, W. D. Hammond (ed.), The Journal of William Shaw (Cape Town, 1972), 160;Google ScholarBannister, Humane Policy, 157;Google ScholarStockenstrom, A., The Autobiography of the Late Andries Stockenstrom (Cape Town, 1887), I, 213.Google Scholar See also Van Warmelo, History of Matiwane, 263.Google Scholar

¹¹² The word ‘bounce’ was a favourite of Jameson's and is used here to signal the large number of parallels between what is now described and the machinations of Jameson and the British South Africa Company against Lobengula and the Ndebele in 1888–93.Google Scholar

¹¹³ Theal, G. M., Records of South-Eastern Africa, ix (London, 1903), 24–8.Google Scholar For the competing cases of the Portuguese and British Governments and the MacMahon Award, in which the Owen Treaty was exposed, see 63–268. The failure of historians to connect the attempt on Delagoa Bay in 1823 with the 1824 Natal ‘treaty’ is partly the result of the particularist cutting off of southern Mozambique mentioned on p. 504. A very different picture emerges of the regional expansion of the British when the two treaties are reconnected.

¹¹⁴ Bird, Annals of Natal, I, 71–3; Stuart and Malcolm, Fynn Diary, 56–7.Google Scholar

¹¹⁵ This treaty is usually listed without comment by Cape historians; see e.g. Davenport, T. R. H. and Hunt, K. S., The Right to the Land (Cape Town, 1974), 19.Google Scholar

¹¹⁶ King, James, ‘Some account of Mr Farewell's settlement at Port Natal, and of a visit to Chaka’, 07 1826, in Thompson, Travels and Adventures, 11, 243–52. Note the change of tone from factual description to anti-Shaka propaganda at the bottom of page 248.Google Scholar

¹¹⁷ Webb, C. and Wright, J. (eds.), The James Stuart Archive, vols. to date (Pietermaritzburg, 1976–1986), 11, 269Google ScholarStuart and Malcolm, Fynn Diary, 122–8; Hedges, ‘Trade and politics’, 202.Google Scholar Fynn received enough cattle in reward to establish his umusi of Insimbi. There is much in Farewell';s background to suggest involvement with the slave trade. The possibility that Ndwandwe prisoners ended up at the Bay should be kept in mind. Fynn in his reminiscences has Farewell fall ill just before the battle, and himself watch proceedings from a hill. Hedges writes overcautiously: ‘it is possible that both writers adjusted their testimony to minimise their role’.

¹¹⁸ Isaacs, N., Travels and Adventures in Eastern Africa (reprinted Cape Town, 1935–1936 [originally 1836]), 1, 162–71.Google Scholar

¹¹⁹ Stuart and Malcolm, Fynn Diary, 130.Google Scholar

¹²⁰ Stuart and Malcolm's Fynn Diary is not a diary, but a welding together of a number of later propaganda essays written by Fynn mainly in the late 1850s and early 1860s. It contains interpolations by Stuart from the early twentieth century. The combined editing of Stuart and Malcolm is one of the major disasters of South African historical literature.Google Scholar

¹²¹ Dundas to [Somerset?], 15 Aug. 1828, in Van Warmelo, History of Matiwane, 243.Google Scholar

¹²² Stuart and Malcolm, Fynn Diary, 149; Isaacs, Travels and Adventures, 1, 227–8.Google Scholar

¹²³ Fynn, in Diary, 148, wrote of ‘reports being circulated in the Colony of an intended invasion by the Zulus, with me at their head…, supposed to be a division of Shaka's army. Reports such as these induced the Colonial Command of 1828 to advance against [Matiwane].’ This lets the cat out of the bag in more ways than one. For another version of the alibi see Isaacs, Travels and Adventures, I, 227–8. For a suggestion that Shaka's izinduna in the Cape and later informants of Stuart denied there was a Zulu army, see Webb and Wright, James Stuart Archive, 11, 267, 274. It is possible this evidence has been interfered with by Stuart and ‘straightened out’ by the editors in footnotes 21 and 22 on page 303.Google Scholar

¹²⁴ Stuart and Malcolm, Fynn Diary, 153. Fynn claimed that the ‘Transkei army’ was subsequently despatched north, but this was surely physically impossible.Google Scholar

¹²⁵ Ibid., 156.

¹²⁶ Isaacs, Travels and Adventures, 1, 238.Google Scholar

¹²⁷ Stuart and Malcolm, Fynn Diary, 157. There is a chance that the land concession to King in 1828 was genuine, a sort of Lippert Concession of the era. King's own death a week before Shaka's has always been proclaimed (by Fynn et al.) as natural; but its timing was very convenient for Farewell. The possibility that King was murdered should be kept open.Google Scholar

¹²⁸ Webb and Wright, James Stuart Archive, 11, 232, 295; III, 70.Google Scholar One suspects Stuart knew more about Fynn's activities than he lets appear in his notes. Questions about Fynn are infrequent in his interviews.

¹²⁹ Stuart and Malcolm, Fynn Diary, 161–2.Google Scholar

¹³⁰ Fynn, ibid., 171, refers coyly to the Qwabe knowledge ‘that Dingane would come by the whole stock Farewell was conveying to Natal’; and, ibid., 170, refers to ‘much property of considerable value’. Many of the guns fell into Qwabe hands.

¹³¹ Even Beinart, W., ‘Production and the material basis of chieftainship: Pondoland, c. 1830–80’, ch. in Marks and Atmore, Economy and Society, 121–6, falls for this. There are several references to the prosperity of the Mpondo at the turn of the 1830s, including their wealth in cattle.Google Scholar

¹³² Qwabe, Mbo, Cele and several other groups refused to recognize Dingane, the reasons for which have not yet been analysed.Google Scholar

¹³³ Kay, Travels and Researches, 331.Google Scholar

¹³⁴ Bannister, Humane Policy, 43, 238 (both italics in original); Stockenstrom, Autobiography, 1, 279. Bannister was an enthusiast for expansion into Natal, hence his animosity to the Zulu.Google Scholar

¹³⁵ Bannister, Humane Policy, 149–58; Kay, Travels and Researches, 330.Google Scholar Kay writes (ibid., 332): ‘their weapons were met with rockets; and for spears they received balls’.

¹³⁶ Godlonton, Irruption of the Kaffir Hordes, Introductory Remarks, 54.Google Scholar

¹³⁷ Graham’s Town Journal, 14 Feb. 1833. This was (allegedly) prefaced: ‘The speaker observed that that expedition had been the cause of the salvation of thousands and tens of thousand lives, - inasmuch as the people dispersed by the commando were wretches of the most atrocious character.’Google Scholar

¹³⁸ Godlonion, Irruption of the Kaffir Hordes, Introductory Remarks, 269–70; and Main Text, 229.Google Scholar

¹³⁹ Ayliff, J. and Whiteside, J., History of the AmaMbo Generally Known as Fingos (Butterworth, 1912), chs. 1 to 7.Google Scholar Ayliff alleged the prisoners came from the MzinyathiTugela region, survivors from an original 720,000 (sic) people. But no evidence has ever been cited to back this up. Most of the 17,000 were likely to have been Gcaleka and Rharhabe.

¹⁴⁰ For the most detailed study of the Fingos see Moyer, R. A., ‘A history of the Mfengu of the Eastern Cape, 1815–65’ (Ph.D thesis, University of London, 1976).Google Scholar Moyer, although trapped by the mfecane mythology in which he works, provides copious evidence that Fingo was a word given to any ‘Kafir’ who showed ‘willingness’ to work for or in other ways identify with the whites, their missionaries, or mode of life, They were collaborators whose lives had been thrown into disarray by British militarism. But the Fingos need separate treatment. See also footnote 82, above.

¹⁴¹ Schapera, Apprenticeship, 102. Moffat quotes the women as familiar with the destruction of Kaditshweni, the Hurutshe capital, which is reputed to have been attacked by Koranna.Google Scholar

¹⁴² In May on his outward journey and June on his return dash to Griqua Town Moffat was provided with food and hospitality by the inhabitants of ‘Old Lathakoo’ (ibid., 77, 87). He is, perhaps significantly, silent about who they were, which contrasts with his usual attempts at precise identification.

¹⁴³ Thompson, , Travels and Adventures, I, 178.Google Scholar

¹⁴⁴ Ibid., I, ch. xviii, ‘History of the Mantatees’; Godlonton, , Irruption of the Kaffir Hordes, Introductory Remarks, 51, 69.Google Scholar

¹⁴⁵ Ibid., Introductory Remarks, 67–70; Graham's Town Journal, 30 January 1834.

¹⁴⁶ Lye, Andrew Smith's Journal, 92, and Smith's footnote.Google Scholar

¹⁴⁷ Arbousset, T., Narrative of an Exploratory Tour to the North-East of the Colony of the Cape of Good Hope (Cape Town, 1846), speaks of ‘some labourers sent to the Basutos with the Paris Society for Evangelical Missions’, who visited Sekonyela in 1833. There is a chance that they are the first source of the elision.Google Scholar

¹⁴⁸ For an uncritical acceptance of the anti-Tlokwa propaganda see Sanders, P. B., ‘Sekonyela and Moshweshwe: Failure and Success in the Aftermath of the Difaqane’, J. Afr. Hist., X, 3 (1969), 439–55.CrossRefGoogle Scholar

¹⁴⁹ Theal, , History of South Africa 1795–1834 (Cape Town, 1891), 301–4.Google Scholar Or, to be cautious, somewhere between Casalis in the 1860s and Theal the connective extension took place. An examination of the influences on Theal might prove rewarding.

¹⁵⁰ Ellenberger's chapter on ‘Cannibalism’ (History of the Basuto, 217–26) is necessary reading. He claimed 300,000 people were eaten by cannibals in the Caledon region between 1822 and 1830, a figure arrived at by guessing there were initially four thousand cannibals, who each ate one person a month over six plus years (p. 225). Cf. his opinions on the origins of the San (Bushmen) pp. 4–5. And he is still cited as a respectable source!Google Scholar

¹⁵¹ Theal, G. M., History of South Africa 1794–1828 (London, 1903), 389.Google Scholar Theals motive was to put white destruction in its ‘proper perspective’: see Theal, History of South Africa 1795–1834 (1891), 305.Google Scholar

¹⁵² Ellenberger, History of the Basuto, 137–9. Ellenberger may have followed Livingstone in having the Kololo as part of the Mantatee horde at Dithakong;Google Scholar see Livingstone, D., Missionary Travels and Researches in South Africa (London, 1857), 84.Google Scholar Livingstone claims to have obtained this from Sebetwane's entourage in 1851. This is not conclusive. There may have been misunderstanding between Livingstone and his Kololo informant as to which conflict with the whites was being referred to: it could have referred to the known attack on the Kololo by whites at Dithubaruba in 1826 or to some other event. Besides, there is no relevant contemporary evidence whatever from 1823. Livingstone belonged to Moffat's group, amongst whom ‘the battle of Dithakong’ had become a well-known conversation piece. There was every possibility of misunderstanding, leading questions, or subsequent insertion. We are very short of precise information as to the migration path of the Kololo, particularly in the earlier stages.

¹⁵³ How, ‘An alibi for Mantatis’, 67.Google Scholar

¹⁵⁴ Ibid., 69; Smith, ‘Sebetwane and the Makololo’, 52–3.Google Scholar The attempts to link the shadowy Caledon groups, the Phuthing and Hlakwana, with Moffat's reference (Schapera, Apprenticeship, 102) to groups called ‘Maputee’ and ‘Batclaquan’ under, respectively, chiefs Chaane and Carrahanye, who were in the vicinity of Dithakong some days after 26 June 1823, are extremely dubious. Between them Ellenberger, How, Smith and Schapera (in footnotes) jump to as many unwarranted conclusions as Ellenberger did with the Tlokwa in the first place.

¹⁵⁵ How, ‘Alibi for Mantatis’, 69.Google Scholar

¹⁵⁶ Lye, ‘Mfecane in the Southern Sotho area’, 118–21.Google Scholar

¹⁵⁷ These include Tlokwa incursions into the central Orange Free State, a migration down the Caledon to the Orange confluence, and another migration up and down the Caledon, all within three years. The staticness of the Tlokwa after c. 1823 is perhaps more suggestive.Google Scholar

¹⁵⁸ For example, Forbes's annotation of Thompson's chapter on the Mantatees in Travels and Adventures, i, e.g. footnote 9, p. 81, footnote, p. 118, and footnote 4, p. 176.Google Scholar

¹⁵⁹ For the metaphor see Hole, H. M., The Passing of the Black Kings (London, 1932), 35–6.Google Scholar

¹⁶⁰ The Swazi built their power on supplying the eastern Transvaal Boers with ‘apprentices’ raided in the low veld in the 1850s and 1860s. See Bonner, Evolution and Dissolution, 80–84. Gaza slaving into the 1860s is outlined in Harries, ‘Labour Migration’, ch. 4.Google Scholar

¹⁶¹ Davenport, Modern History (3rd edition), 14, 36–53.Google Scholar

¹⁶² See e.g. the extreme right wing Action White Natal pamphlet, Durban, 19 May 1986, in which, after alluding to the ‘mfecane extermination wars’, they go on: ‘The whites of Natal thus have the unequivocal right historically and lawfully to own certain areas and to govern such.’Google Scholar