Cecil Rhodes: Racial Segregation in the Cape Colony and Violence in Zimbabwe

Cecil Rhodes did an extraordinary amount in his short life (1853-1902). Hugely ambitious and driven, he made an impact in many different spheres. Born in England, he arrived in South Africa in 1870, gained a little experience with his older brother on a farm in Natal, but soon migrated to the Kimberley diamond fields in 1871. There he flourished. By 1881 he had bought up sufficient claims to be one of the largest diamond producers. In that year he won election to the Cape Legislative Assembly and completed a degree at Oriel College, following irregular visits to Oxford over an eight-year period.

Rhodes became an increasingly influential member of the Cape parliament and served as Prime Minister from 1890 to 1896. He was the second longest-serving Prime Minister of the Colony and would have been in office longer if he had not staged the Jameson Raid (1895-6, see below). At this time, Rhodes, in his late 30s and early 40s, was at the peak of his power. He was simultaneously: chairman of De Beers diamond company, which monopolised production after 1888; joint managing director of Goldfields of South Africa; one of the richest men in South Africa; and managing director of the British South Africa Company, which colonised Zimbabwe and areas to the north from 1890. It is extraordinary that he was allowed to hold all these positions while Prime Minister but also some indication of the range of activities in which he was engaged, his influence, and his ability as politician, mining magnate, businessman and empire-builder.

This memorandum focusses on two key questions that have been central in debates following the Rhodes Must Fall protests in 2015-6: whether Rhodes contributed to racial segregation in the Cape Colony; and how to characterise the violence in the conquest of Zimbabwe in the 1890s.

With respect to the Cape, the evidence shows that Rhodes made a number of important decisions, or supported developments, that intensified racial segregation in the late nineteenth century. He had some power to influence an alternative political direction in the Colony but advocated a racially restrictive franchise, punitive racially-based Masters and Servants legislation, a labour (poll) tax for African people only, a segregated local government system and segregation in the South African cricket team. He was increasingly in favour of segregated urban ‘locations’ and rural districts. Rhodes was involved in the beginning of coercive compounds for black workers and other racially restrictive practices as an employer. To a limited degree a pragmatist in Cape politics, prepared to work with a range of people who would be useful to his interests, Rhodes was a deeply committed British imperialist, intent on white, specifically British, authority and committed to the idea that ‘the natives’ should be a ‘subject race’ (Samkange, 1982, 15; Vindex, 1900, 159).

In respect of Zimbabwe, 1890-97, Rhodes and his Company were responsible for extreme violence against African people. Wars were fought in 1893 and 1896-7: unbridled use was made of the Maxim gun; cattle were looted by the Company and its agents on a large scale; in the 1896-7 war, grain stores and crops were appropriated or destroyed over a sustained period as a deliberate strategy; many Ndebele soldiers were shot in flight; supposed rebels were sentenced and hung or shot without due process of law. Over a period of nine months in 1896-7, African men (including armed men), women and children sheltering in caves were blown up with dynamite, when it was clear that many were being killed. Rhodes was well aware of these practices, at times participating or present while they were taking place, and involved in strategic discussions.

Although these are difficult to quantify, it is likely that African deaths in the conquest of Zimbabwe in the 1890s were over 20,000 and perhaps closer to 25,000. A number in the small settler population and in the colonial forces were also killed. The casualties overall were in the proportion of roughly 1 white to 40 African in 1893 and 1 to 20-30 in 1896-7. This memorandum does not attempt to analyse in detail the causes of these wars, nor the rich historiography that has built up around the colonisation of Zimbabwe. The material below suggests strongly that conflict arose as a result of British South Africa Company policies.

In discussion of Rhodes, particularly with respect to these controversial questions of race and violence, it is often said that he was a man of his time and should not be judged by present day standards. Quite clearly, the student protests of 2015-6 and 2020, as well as the appointment of the Oriel Commission, are a result of twenty-first century concerns. However, there were a range of views in the late nineteenth century and this analysis aims also to include some assessment of Rhodes’s options at the time.

This memorandum is detailed in parts, although it does not begin to exhaust the sources available. The main findings are summarised above. Descriptive material from the time is valuable because it is so difficult to quantify key points. The text illustrates disturbing aspects of the past, in which Oriel and Oxford are entangled. It provides a resource for further research, which both the College and the University should sponsor and support. It is designed to provide background for rethinking and recalibrating the role of Oxford and Oriel in relation to the Rhodes statue and legacy.

Robert Rotberg (1988, 455), Rhodes’s most thorough biographer, who worked hard to understand his complexities, wrote: ‘It is not wholly unfair to suggest that Rhodes’s legislative victories … proved essential precursors to apartheid’. Rhodes contributed to restricting the vote for black people in the Cape: this has been an important theme in the historiography of late nineteenth-century South Africa – a central element of the shift from Cape liberalism to segregationism. The Cape Colony, finally taken by Britain from the Dutch in 1806, was the largest and most significant of the four settler states in South Africa in the nineteenth century. It was granted representative self-government in 1853 with a non-racial qualified franchise; this was taken forward into responsible government (1872), after which Cape parliamentarians, then all white, could form their own executive.

Rhodes supported two major limitations on the black franchise. The first, in 1887, when Sprigg was Prime Minister, excluded land held in communal or customary tenure from the property qualifications for the franchise. Very few whites held their land in this way but most Africans in the Colony did so. The second Act in 1892, when Rhodes was in office, raised the property qualifications and introduced an educational qualification. This applied to all voters (men only) but had the effect of excluding a higher proportion of black people.

Much of the pressure for this legislation came from the Afrikaner Bond with whom Rhodes made an alliance in order to take office in 1890. Afrikaners (Boers) were the largely white descendants of earlier Dutch colonists and a majority of the white population. But Rhodes clearly shared the view that the black franchise should be curtailed and this was the explicit intention of both Acts. He said in his speech in 1887: ‘Either you have to receive them on an equal footing as citizens, or to call them a subject race. Well, I have made up my mind that there must be class legislation, that there must be Pass Laws and Peace Preservation Acts, and that we have got to treat natives, where they are in a state of barbarism, in a different way to ourselves. We are to be lords over them … Treat the natives as a subject people as long as they continue in a state of barbarism and communal tenure; be the lords over them, and let them be a subject race, and keep the liquor from them’ (Vindex, 1900, 159). In 1892, he indicated that he would have preferred the much more restrictive Natal franchise (Tamarkin, 1996, 175).

The precise effects of this legislation are difficult to quantify because race was not recorded in voter registration till 1903. The latest analysis by Nyika and Fourie (2020) argues that earlier estimates of the reduction were exaggerated. However, they find that the 1887 Act reduced black voters by about 40 per cent in the districts that they have carefully researched. The Act was called tung’ umlomo (sewing up the mouth) in isiXhosa. Although this restriction spurred voter registration, and the numbers were partly restored by 1891, the 1892 Act again reduced them by over 20 per cent. By 1895 they were down 30 per cent in comparison with 1886. By 1910 roughly 15 per cent of Cape voters were black, the majority of them ‘Coloured’ rather than African people (Trapido, 1970), although they comprised about 75 per cent of the population.

No black representatives were elected to parliament, but they did influence the election of white liberals, especially in some Eastern Cape and Cape Town constituencies. Black voters were not initially organised into a party but found a focus in John Tengo Jabavu’s newspaper, Imvo Zabantsundu.  In this way, the franchise provided some protection and influence in the Colony – for example in respect of defending the franchise, funding for education, opposing pass laws, protecting the right to own land in private property outside of the customary tenure districts and protecting African customary landholdings. The African common roll franchise was further diluted in 1910, when the Union of South Africa was formed; Cape white liberals insisted that the remnant black vote be retained but it was not extended to the rest of South Africa. Africans lost any franchise on a common voters’ roll in 1936, at the height of the segregationist era, and people classified as Coloured finally lost such a vote in 1956, during the early years of apartheid.

Rhodes was not opposed to a small measure of representation in the central colonial legislature for black people. He accepted to some degree that African people could become educated and share in the progress (then a central idea) of the Colony. But at a time when the number of black voters had started to increase significantly, he was in favour of strongly restricting such expansion. He excluded the great majority of Africans from the category of civilised. This included Christians with some education and many of those who participated in colonial economy as innovative, small-scale agriculturalists and ox-wagon drivers (Bundy, 1979). In the Transkeian Territories, for example, there had been mission stations since the 1830s and a growing number of Africans with basic literacy. But in the great majority of Transkei districts, land was held in customary tenure so that it was difficult to qualify for the vote.

There were other options at the time. Key white liberals, who were in Rhodes’s government in 1892, as well as African leaders such as Jabavu, defended the vote although most compromised over the 1892 Act because they saw the alternatives as being even more restrictive. (They left government over a corruption case in 1893.)  Cape politics were fluid and not yet tightly divided into parties. In 1898 the Afrikaner Bond went into alliance with white liberals, under Prime Minister W. P. Schreiner, because they shared a common antipathy to the pro-imperial views of the Progressive party, so closely associated with Rhodes, that seemed to favour war with the Boer-controlled Transvaal.

After his resignation as Prime Minister, Rhodes developed a slogan, in various versions, of equal rights for every civilised man south of the Zambezi. It seems that this was first coined in 1896-1897, but included the word white, and was promised to white working-class men in Cape Town – not all of whom were voters (Bickford-Smith, 1995, 166, 199). In the closely fought election of 1898, Rhodes used a similar slogan about civilised white men in speeches to Afrikaners, whose support he wanted to win back after the Jameson Raid (1895-6). Rotberg (1988, 610-11) affirms it was not initially intended to include African or Coloured people. But when pressed by Coloured voters in Kimberley in 1898, Rhodes responded by generalising the statement and specifying that equal rights should be based on conditions similar to the existing franchise requirements. The slogan was thus deployed after the restrictions on the franchise with no implication that they would be removed for black people.

Rhodes’s political pragmatism was also evident in his dealings with African chiefs and politicians. In 1896, he met Ndebele (but not Shona) chiefs in Zimbabwe in a series of indabas in order to resolve the war with them (see below). He briefly funded the Cape newspaper Izwi laBantu, aimed at an African readership. This arose initially out of a split within African politics, when Rhodes was approached by opponents of Jabavu to assist in financing a rival paper (Odendaal, 2012, 146-7). Rhodes did not support the views of the paper, launched in November1897, but sought black votes in the March 1898 election. He and the Progressive party had fallen out with Jabavu, who worked with Schreiner and the Cape liberals. The Progressives did win a majority of the votes, but lost the constituency-based election. Izwi laBantu was later edited by A. K. Soga, who was educated partly in Scotland and a radical in the spectrum of African opinion at the time. Jabavu opposed the British decision to go to war with the Boer republics while Soga, as well as some other African proto-nationalists, defended it.

Rhodes supported a Masters and Servants Amendment (Strop) Bill (1890), proposed by the Afrikaner Bond, that would allow flogging of black servants found guilty of breaking contracts. He was one of the few English-speakers who voted for it, but it was defeated because even some Afrikaner representatives voted against it (Tamarkin, 1996, 140). He also supported racially segregated locations for Africans in Cape Town. At least the pavements remained open in Cape Town, although Africans were barred in 1894 from using them in the new colonial town of Bulawayo, recently founded by Rhodes and Jameson (Ranger, 2010, 24).

The Glen Grey Act (1894) is often mentioned in discussions of the origins of segregation and apartheid. This complex piece of legislation cannot be analysed in full here, but one central feature was the priority to mobilise African workers, following widespread colonial concerns about labour shortages on the farms and in the mining industry. A primary mechanism was the labour tax on every adult African man – in effect a poll tax that would be levied in addition to the existing hut tax on all African household heads. In Rhodes’s words, he wished to teach Africans ‘the dignity of labour’ (Vindex, 1901, 374-5). The labour tax was little implemented, because Rhodes soon lost his position and officials, as well as his political successors, thought it too draconian and unnecessary. This is another example where Rhodes had a choice in connection with a racially based, discriminatory tax, which was opposed not only by white liberals but Cape officials. A South African-wide poll tax on African men was introduced by the Afrikaner nationalist Union government in 1925.

Rhodes also increasingly conceived of territorial segregation: ‘My idea is that the natives should be kept in these native reserves and not be mixed with the white men at all. Are you going to sanction the idea, with all the difficulties of the poor whites before us, that white children grow up in the middle of native locations?’ (Vindex, 1901, 386). However, Africans were not barred from purchasing private land in districts that were largely owned by whites. This was envisaged in the 1913 Natives Land Act and gradually implemented over succeeding decades.

The Glen Grey Act also introduced councils, paid for by an additional tax on Africans, which created a segregated system of local government in districts where African people were in the great majority. Rhodes and others saw it at the time as increasing local government responsibility for African people. It did include a restricted element of election and foresaw the political emergence of an educated African elite, rather than placing authority, as in the later British system of indirect rule, in the hands of chiefs. In this latter sense it differed also from the apartheid Bantu Authorities Act (1951), which privileged traditional leaders. But councils were underpinned by an evolving Cape and later national policy to set up segregated African reserves. These areas in the Cape became the geographic base for two of the apartheid Bantustans.

There are also important examples where Rhodes intervened informally to support racial segregation. In the 1880s, some inter-racial cricket was possible (Odendaal et al, 2016) and in 1892, the English touring side played against a team described as Malay in Cape Town. Fast bowler H. ‘Krom’ Hendricks impressed the English captain, who was quoted in the Cape Times as saying ‘if you send a team [to England], send Hendricks; he will be a drawcard’. A South African-wide cricket association was formed in 1894 and organised a tour of England in that year; Hendricks was enthusiastically proposed by key white cricketers. But the chairman of selectors William Milton, Rhodes’s private secretary and future Administrator of Rhodesia, refused to include him. Rhodes later said ‘They wanted me to send a black fellow called Hendricks to England … but I would not have it’ (Winch, 2014; Bundy, ‘More than a Game’). This was a fateful decision, helping to confirm racial segregation in sport when other possibilities were still open and could have been facilitated by a decision to include Hendricks. Later in 1894, Milton intervened to stop the selection of Hendricks for a ‘Colonial Born’ side to play against ‘Mother Country’ in Cape Town.  Organised sport became a particularly important cultural expression, especially for whites, in South Africa.

Rhodes was deeply immersed in diamond mining at Kimberley for much of his adult life: his role in buying claims, developing technology, marketing gems as well as financing and amalgamating diamond companies has been extensively analysed. Together with others, his company De Beers introduced closed compounds for African migrant workers in the mid-1880s (Turrell, 1987; Worger, 1987). Compounds were initially in part a means of suppressing ‘Illicit Diamond Buying’, but increasingly they became a means of reducing costs, mobility, wages and African bargaining power. Black workers were rigorously searched before and after work and their movements restricted during their contracts. Compounds were not imposed on white workers who could live in town – as could the small African elite (Willan, 2018).

Rhodes clearly had an intense interest in new technology and an understanding of its potential in mining. Although there was already a budding bureaucracy in the shape of a government veterinary surgeon, an Agricultural Journal, and an official to deal with the scourge of phylloxera in vineyards, his administration established a Ministry of Agriculture in 1893. Following a lengthy commission,  a Scab Act was passed in 1894 against widespread opposition (Tamarkin, 1996): it introduced compulsory eradication of scab disease in sheep that diminished the yield of wool – alongside diamonds, the most valuable Cape export.

Yet his technical alertness, Maylam argues, went hand in hand with general lack of concern for his black employees; a Kimberley inspector noted his ‘reckless disregard for human life’. Rotberg (1988, 630) refers to an episode during the siege of Kimberley (1899-1900) by the Boers, in the early phases of the South African War (1899-1902), when the diamond mines had to close. Rhodes was present and initially deployed African workers on various public works. He tried to persuade them to leave town, so that they would not be a financial burden to De Beers, but the Boers drove many back. Kekewich, commanding the British forces in Kimberley, learned that there was an increasing rate of scurvy and death amongst black workers and it transpired that they were neither being paid nor given adequate food. When Kekewich raised the matter with Rhodes, the latter told him ‘not to meddle in his affairs’, adding that if the Africans ‘would not leave the town, they must be forced to, and giving them only bread and salt had this effect’. Yet during the siege, whites – in part as a result of Rhodes’s largesse –were ‘never short of provisions’. This was a couple of years before Rhodes’s bequest to Oriel that included support of high table.

In a less restrictive form, compounds were transposed to the Witwatersrand gold fields and became a central feature of the exploitative migrant labour system in South Africa (Wilson, 1972). Successive governments imposed rigorous controls on African freedom of movement by pass laws. Recent participants in debates about Rhodes have called these practices in Kimberley and Johannesburg a form of slavery. Africans were contracted workers, not slaves, and the distinction is important. But their rights as workers were increasingly curtailed and racially-based job colour bars formalised; compounds have been seen as coercive institutions at the heart of South African segregation (1910-1948) and apartheid (1948-1994).

With respect to land and conquest, South Africa should be distinguished from Zimbabwe. Rhodes had barely arrived from England when the Cape Colony forcibly annexed the diamond fields in 1871. He was certainly an expansionist, but the great majority of what became South Africa was already annexed by the time he had significant political power. Rhodes did, however, have some role in the colonisation of Tswana-speaking people, not discussed here, and oversaw the annexation of Mpondoland in 1894 – the last independent African kingdom that came under the Cape.

By this time, Rhodes was directly involved in violent conquest in Zimbabwe, where his British South Africa Company deployed Maxim machine guns to devastating effect (see below). A number of texts report that when he visited Mpondoland in 1894, Rhodes ordered a field of maize to be flattened by machine gun fire in order to demonstrate what would happen to the Mpondo if they tried to fight annexation. These reports, probably first recorded in McDonald’s sympathetic biography Rhodes: a Life (1927), may be apocryphal. But Rose Innes (1949, 106), liberal Cape parliamentarian, and later Chief Justice of South Africa, saw this period as one in which Rhodes turned to violence and the machine gun, which ‘woke in the minds of the possessors of the new weapon – an overweening sense of its importance’. Rhodes ‘publicly displayed that dictatorial and impatient vein with which, in the near future, we were to become familiar’ and he moved ‘from the constitutionalism of the statesman to the lawlessness of the revolutionary’. Rhodes and Jameson had already become violent adventurers in Zimbabwe and observers increasingly noted his intemperance. Rose Innes, looking back, was also referring to the ‘lawlessness’ of the Jameson Raid.

At the end of 1895, Rhodes and Jameson tried to orchestrate a simultaneous invasion and internal rebellion by ‘Uitlanders’ (largely British) in the Transvaal. It failed but it was illegal, highly aggressive and careless. Jameson took Maxim guns but could not use them.  Most of the deaths – probably less than 100 – were among the small invading force, which was based on the British South Africa Company police and volunteers. Chamberlain, Secretary of State for the Colonies in Salisbury’s administration, was party to the raid and encouraged it. It could have precipitated a civil war, when alternatives were again possible. Rhodes had to resign as Cape Prime Minister because he lost the support of the Afrikaner Bond. Having spent much of his political career working with Afrikaners, he helped to polarise white society, intensified Afrikaner suspicion of British imperialism and probably made the South African War (1899-1902) more likely. He did not, however, stay out of politics, engaged energetically in the 1898 Cape election, and supported the British move to war against the Transvaal. Representatives of Rhodes Must Fall associated Rhodes with the concentration camps used in that conflict. This connection is not justified; they were the responsibility of the British army.

Rhodes continued to attract widespread support amongst English-speakers in the Cape. In the 1891 Cape census, about 25 per cent of population was recorded as white, of which British colonists, estimated at 130,000, made up about 35 per cent were a minority. They were probably less than 10 per cent of the whole population, yet they were socially and culturally dominant in the cities, in politics, in the professions, in the military and especially in business. Despite this, Rhodes was widely criticised in his time by English-speakers, as well as by Africans – and by Afrikaners especially after the Jameson raid.

Among liberals, John X. Merriman, who was once a friend, and served in the Rhodes cabinet 1890-93, wrote in 1897: ‘Rhodes is a curious product of his time. People who compare him with Clive or Warren Hastings are those who take their history from the Daily Telegraph or Tit Bits. He is a pure product of the age, a capitalist politician … and has neither moral courage nor convictions, but he has the sort of curious power that Napoleon had of intrigue and of using men – the worse they are the better for his purpose which is self-aggrandisement under one high-sounding name or another’ (Lewsen, 1963, 254-5). Contrary to Merriman, the comparison with Clive is appropriate: the latter was also committed to conquest for a colonising company and intent on enriching himself while also expanding British interests.

The author Olive Schreiner was initially attracted to Rhodes’s modernising ambitions, but turned against his policies. Her novel, Trooper Peter Halket (1897), focussing on the conquest of Zimbabwe, was a sustained attack. She wrote to Merriman in that year: ‘we fight Rhodes because he means so much of oppression, injustice, and moral degradation in South Africa’ (Lewsen, 1963, 265). John Charles Molteno, son of the first Prime Minister of the Cape Colony, who represented the Thembuland constituency, which had a significant African vote, was an outspoken critic of Rhodes both in relation to the franchise and his imperial ambitions.

Rotberg is justified in pointing to Rhodes’s increasing pursuit of segregationist measures at a critical moment when policies in the Cape, at least, could have taken a different route. He advocated restrictions on the black franchise, a punitive tax for African people only, locations for African people in Cape Town, a segregated local government system, segregation in cricket, coercive compounds for black workers only and other racially restrictive practices as an employer.

The arguments in Oxford characterising Rhodes as violent, criminal and even responsible for genocide focused particularly on the colonisation of Zimbabwe by his privately owned British South Africa Company in 1890-97. Rhodes’s actions in Zimbabwe certainly involved force and violence when he was at the height of his political power; his sense of urgency, and perhaps his hubris, increased. This memorandum avoids discussion of Rhodes’s specific motivations, an issue extensively covered in many biographies.

The term ‘extreme violence’ is used here to include ‘looting, a disregard for international standards of warfare, the use of collective reprisals on civilians …, scorched earth policies, starvation tactics on the enemy, as well as the wider population’ (Gordon, 2020). Ndlovu-Gatsheni (2009,146) calls it ‘savage’ warfare, inverting colonial language. F.C. Selous (1897, 30-1), who was a leader of the pioneer column in 1890 and fought in both wars, talked of ‘merciless ferocity’ by settler militias in 1896. Such terms accurately describe the conduct of war by the Company and British imperial forces in Zimbabwe. They remove the necessity of proving genocide as ‘the ultimate yardstick of depravity’, for which the bar in relation to evidence is high. The term extreme violence acknowledges atrocities, systematic violence against civilians and excessive force beyond that strategically required; in this case it also includes racially motivated violence.

Lobengula, the Ndebele king, signed the Rudd concession voluntarily in 1888 although his indunas and chiefs were split and he soon tried to retract. He sent a delegation to England, but the British government decided to enforce the concession. It was an important step in winning support for the British South Africa Company charter in 1889. The Rudd concession covered minerals and not land rights over areas that Lobengula controlled. The pioneer column of 1890, guided by hunter and author Selous, was an armed invasion and the British South Africa Company went further than this concession in laying claim to land (Ranger, 1967, 31).  Jameson, administrator from 1891, based in the new colonial settlement of Salisbury, was particularly generous in handing out farms, first in Mashonaland. At a time that the Cape government was peacefully annexing Mpondoland and reserving its land for Africans, Rhodes and Jameson were responsible for an aggressive settler colonialism in Zimbabwe that precipitated rebellions.

I will focus largely on two issues. Firstly, the pattern of violence by the British South Africa Company is examined. In view of the debates in Oxford and elsewhere about the character of colonial conquest in Zimbabwe, this must be a central issue. Secondly, evidence of systematic appropriation or destruction of livestock and grain is illustrated; both were of fundamental importance to agrarian societies that had few other means of securing food.

Colonial commentators sometimes gave rough figures for the deaths inflicted on African people in armed clashes. Estimates are generally fuller for the more dramatic incidents and for battles with the Ndebele, rather than for deaths in Mashonaland where conflict was more diffuse. There is a range of revealing books by participants, such as Selous, Baden-Powell, Plumer and Alderson, as well as an extensive historical literature, some of which draws on archival records. In the short time available, it has not been possible to consult all sources nor to visit archives.

In writing about Zimbabwean history, caution should be shown in using the generic descriptions Ndebele and Shona for diverse pre-colonial African people. I use the terms Matabeleland and Mashonaland, introduced by the new colonial state in the 1890s, and still used for provinces. Both of these areas included diverse political and ethnic communities. Spellings and names of places have changed and are not always uniform in the sources. I have generally used the version in the original source, followed by current orthography when available.

Assessment of mortality in these conflicts must also take into account losses from conflict-related famine. After the war of 1893, the British South Africa Company organised and facilitated the looting of cattle, central to African diets, on a large scale. The term loot was widely used by the Company; it constituted and staffed a Loot Committee and set up loot kraals in Bulawayo and elsewhere. (The term kraal is widely used in sources at the time. It could mean any enclosure for livestock but often referred to an African homestead, base of an extended, multi-generational family, with multiple huts, or even a small village. A description of kraals in Mashonaland noted ‘from 25 to 100 huts each’ (Ranger, 1967, 281).)

In 1896-7 Company forces, and the imperial detachments that joined them, pursued a scorched earth strategy. There are many records of the appropriation of grain and livestock, burning of huts, villages and grain stores, as well as destruction of crops and gardens. Colonial contingents blew up caves and rock shelters where Shona communities had retreated; some of these were also grain stores.

Nineteenth-century wars often resulted in a higher number of deaths from disease and famine than from military casualties. This was the case on both sides in the South Africa War of 1899-1902, even though food supplies and medical care were more widely available than in Zimbabwe in the 1890s (Beinart and Dubow, 2021). The Boers lost about 27-28,000 from disease, largely women and children in the concentration camps, and perhaps 6-7,000 in conflict. Nearly two thirds of British deaths, roughly 14,000 out of 22,000, were also from disease, particularly typhoid. It is estimated that 20,000 Africans died, mostly in (segregated) camps. Deliberate scorched earth tactics by the British destroyed food supplies of both Boer and African people. Although the figures are elusive, it is also probable that more people died from war-related famine and disease in Zimbabwe than from military conflict. Numbers are significant in assessing the character and scale of colonial violence.

Ndebele people, whose ancestors originated from the peripheries of what became the Zulu kingdom, over 1,000 km to the south, migrated to Zimbabwe in the late 1830s. They established a militarised state, initially by conquest, with a relatively small heartland that was mostly within 80 km radius around Bulawayo – in the western part of what became Zimbabwe. Lobengula succeeded as King in 1868. The Ndebele elite (abezansi –people above) to some degree absorbed subject communities (amahole) and they collected tribute from, and raided, smaller chieftaincies on their peripheries. Historians of Matabeleland such as Cobbing, Ndlovu-Gatsheni and Ranger argue that Ndebele political authority, though it continued to involve demands for tribute, became more incorporative and stable in the later decades of the nineteenth century. Beach (1986, 33) notes a changing balance of power in late nineteenth century, as Shona chiefdoms were able to acquire firearms and fortify hilltop strongholds: ‘in the 1880s the Shona were indeed beginning to turn the tide of Ndebele power’.

For 1893, Beach suggests 2,000 Ndebele deaths in conflict and Selous 3,000. Looking at the figures from the three main encounters, and taking account of smaller incidents, 3-4,000 seems likely. When they could, Ndebele forces took their injured away with them from the battlefield, and these men, some of whom may have died later, were probably not included in colonial estimates. Additional deaths from famine were reported.

The original colonial intrusion was largely restricted to Mashonaland. In July 1893, a sizeable Ndebele army moved eastward in order to assert authority in a dispute with a chief in an area Jameson considered under Company control. The Ndebele did not threaten white settlers, but they did disrupt labour supplies and captured a limited amount of cattle. Jameson warned them to leave, which they did. On foot, they were construed to be moving too slowly and a mounted Company patrol fired on them, killing about 10.

Jameson thought this incident a legitimate reason for attacking Lobengula, because he believed that the remaining authority of the king potentially undermined Company power. A few months of planning and mobilisation were required; the precise motives of the Company and Rhodes’s role have been extensively discussed (Keppel-Jones, 1983). Whatever his initial view, Rhodes backed the decision, agreeing that ‘Lobengula has forced this question on us’ (Ranger, 1967, 94). He gave Jameson £50,000 for war expenses and one rationale offered publicly was to save the Shona from the Ndebele. Lobengula did not want war, nor was he prepared for it. Rhodes, who travelled from Cape Town to Zimbabwe in September 1893 (Rotberg, 1988, 440), could at this stage have negotiated, as Lobengula was requesting.

Instead sizeable contingents of white mounted men were mobilised in Salisbury, Victoria, and Tati in the Bechuanaland protectorate. Recruited from the settlers and from South Africa, they were promised land and loot. Colonial forces were supported by African auxiliaries including armed Shona men. In October the Company sent a formidable force, equipped with Maxim guns, from Salisbury towards Bulawayo – Lobengula’s home and main military encampment. The Ndebele did have a substantial number of rifles, but – as in the case of the Zulu in 1878-9 – they had not succeeded in adapting their military strategies. ‘Instead of breaking up their army into small, mobile units, trained to make full use of cover, relying on surprise and speed, and ready to wear down the white man in the bush, they continued to use the old “chest-and-horn” …formation’ (Gann, 1965, 116).

On 25 October 1893 the advancing Company army, moving into the Ndebele heartland, set up camp at the Shangani River, about 80 km from Bulawayo, where they were attacked by Ndebele forces generally estimated at 5-6,000 men. In a battle that lasted about four hours, the Ndebele made very little impact and were mown down by the Maxim guns. Most sources, drawing on reports from contemporaries, as well as Hole, a key official under Jameson (1926, 308; Keppel-Jones, 1983, 271), give between 500 and 600 Ndebele casualties; one white man was injured and six wounded; a number of African auxiliaries were killed.

Advancing further, Company forces were attacked on 1 November 1893 at Bembesi (Egodade, Gadadi), about 30 km from Bulawayo. Hole (1926, 309) recorded 7,000 Ndebele men involved and very heavy casualties: the Imbezu (Imbizo) regiment alone was estimated to have lost 500 out of 900. Subsequent sources estimate Ndebele deaths at 800. If it is correct that one regiment suffered so badly, then it may have been more (Clarke and Nyathi, 2010, 81). While the Imbizo were at the forefront to the attack, other regiments were involved. Again, the Maxim guns were, together with rifle fire, the overwhelming factor in the battle; colonial forces lost four killed and seven wounded. ‘The carnage lasted not much more than an hour, and it decided the fate of Lobengula’s kingdom’ (Keppel-Jones, 1983, 274). These two battles were among the first where Maxim guns were used – but only by one side. They were further deployed in a smaller encounter by the southern colonial force from Bechuanaland but with lesser effect.

Lobengula and the inhabitants of Bulawayo retreated northwards and torched their settlement. On 3rd November 1893, F. R. Burnham, an American fighting with the Company forces, saw hundreds of ‘beautifully woven Matabele huts’ go up in flames (Ranger, Bulawayo Burning, 2010, 14). The ammunition store blew up and the material culture of the capital was largely destroyed. There were probably no deaths in the occupation of Bulawayo, but a column was sent northwards to capture the king. It moved slowly and a section under Allan Wilson went ahead, without a machine gun, to encounter the Ndebele army at Pupu about 150 km north of Bulawayo on 4th December.

With the king probably close at hand, the Ndebele attacked Wilson’s patrol of 34 men. They were able to defend themselves effectively with rifles and pistols for much of a day until all were killed. Welsh missionary Bowen Rees, at Inyathi, recorded that ‘a great number of the Matabele perished in the same battle’. Mtshana Khumalo, a leading Ndebele general, told Rhodes about the encounter a few years later at a Matopos indaba. Perhaps 300 to 400 Ndebele were killed (O’Reilly, 1970, 100) and there were subsequent skirmishes. Lobengula, who had made a further attempt to offer peace, died, probably soon afterwards (Clarke and Nyathi, 2010, 90-3).

It seems likely that at least 2,000 Ndebele men were killed in these major battles. If other smaller incidents and later deaths are included, it is possible that considerably more died. Moreover, thousands suddenly migrated northwards with some cattle, but with inadequate food supplies. This was tsetse fly country. Clarke and Nyathi (2010, 93) report a tradition that both cattle and people were dying and colonial sources recorded that ‘patrols are continuing to seize large numbers of cattle from the Matabele, the followers of Lobengula are dying of smallpox and starvation, and the Matabele are being prevented from sowing until they surrender their arms’ (Clarke and Nyathi, 2018, 90). Selous (1896, 46) wrote of this time: ‘Short of food, and living like wild beasts in the rocks and forests, with all the bitter discomfort which such a life entails even on savages during the rainy season in a sub-tropical country, … [they] saw their women and children sicken and die day by day’. Ndebele soldiers in turn tried to appropriate food from people who lived on these northern fringes of the former kingdom, creating further insecurity (Alexander et al, 2000). It is difficult to put a number on additional deaths but such records suggest these were at least similar to those in military encounters.

The new Company town of Bulawayo was quickly laid out very close to the old. Though it was not an ideal site, because of lack of water, Rhodes and Jameson wished to underscore the conquest. Rhodes encouraged Jameson to behave ‘as the conqueror he was’ and ‘parcel out Ndebele lands and cattle without waiting for permission from London’ (Rotberg, 1988, 448). More than 1,000 large farms were pegged around Bulawayo, an area around the size of Wales, and although most were not immediately occupied, many Ndebele returning to their homes after the war were informed that they were trespassing on European farms (Keppel-Jones, 1983, 391). Rotberg (1988, 448) wrote that ‘comparatively quickly the defeated Ndebele were shunted onto outlying, badly watered, unsuitable lands and condemned to the kind of penury which would make revolt imperative. Whites occupied the whole of the rich terrain around Bulawayo…Africans suddenly found themselves dispossessed, subject to white landlords (because few moved readily into the reserves) and reduced in every imaginable status, income, and attribute’. By no means all were displaced and most of the new landholders did not immediately begin to use their land, but control of resources in the old heartland had changed rapidly.

The core of the Ndebele kingdom around Bulawayo was good cattle country, free of tsetse fly and trypanosomiasis, and cattle were at the heart of the economy.  Estimates vary of their livestock holdings before the war of 1893 and it is not always clear which specific area and people are included.  David Carnegie of the London Missionary Society, based at Hope Fountain, close to Bulawayo, estimated that they held about 280,000 cattle (Ranger, 1967, 37-8, 106). He was not particularly sympathetic to Ndebele independence, so unlikely to exaggerate. Frank Sykes, who worked as an ox-wagon driver in the area, later fought in, and wrote a book about, the 1896 campaign, suggested about 250,000. It is probable that there were at least 200,000 cattle in the Ndebele heartland. With perhaps 100-120,000 people, this is not an unlikely figure compared to similar African societies.

Sabelo Ndlovu-Gatsheni (Ndebele Nation 2009) shows that the looting of cattle began with the invasion of the Ndebele state in October 1893 and was systematised after their defeat in November. There were direct seizures by the British South Africa Company and by the volunteers who fought for it in the war. ‘Loot kraals’ were established in Bulawayo and elsewhere. At the same time livestock were taken by the former amahole, some descended from pre-Ndebele societies, as also by Shona people beyond the peripheries of Ndebele state. ‘Unscrupulous traders’ took cattle as far as Botswana and South Africa where they were sold (Keppel-Jones, 1983).

Ndlovu-Gatsheni (2009, 148-9) has used Zimbabwean archives to detail the role of specific individuals: Schultz took over 200 head of cattle from Baleni’s area near Shiloh; Dawson, nearly five hundred; Goold-Adams of the Bechuanaland Border Police used patrols to loot about 2,000 in the western and northern parts of Matabeleland in February 1894. The Matabeleland Times reported in May that a detachment of about 25 British South African Police with a Maxim gun left ‘to suppress those refractory natives who refuse to give up their cattle and arms’. In June 1894 Colenbrander looted perhaps 3,000 cattle and forcibly recruited 1,000 labourers. Over five thousand Ndebele cattle were auctioned by Napier, Weir and Slater in June 1894 and the British South Africa Company supplied Combrinck, the largest butchers in Cape Town, at cheap prices in July.

The Company used the fiction that all cattle had belonged to Lobengula and that therefore all the remaining cattle belonged to the Company by right of conquest. In July 1894, a Commission headed by Joseph Vintcent, newly-arrived judge from the Cape, was appointed to demarcate reserves for the Ndebele and to arrange the distribution of remaining cattle among them. In 1895, the Vintcent Commission reported that they had counted about 74,000: 55 per cent or about 41,000 were earmarked for Ndebele private owners and 33,000 (45 per cent) were to be retained by the Company potentially for distribution or loot. They were registered and branded.

Carnegie thought that the Ndebele had lost 200,000 cattle and Keppel-Jones (1983, 398) accepted that, while the Company estimated 125,000, the first figure ‘could be nearer the mark’. Some who previously had little control over cattle, only ‘milking rights’, probably benefitted but on the whole numbers were drastically reduced especially for those who held political power beforehand. The 33,000 retained for company use, temporarily in the possession of African people, were not secure. Each Native Commissioner was expected to send 50 head from his district monthly for the use of the Company, and this number was considerably increased at times. ‘Thus the natives speedily understood that their cattle — the food of their children — were fast disappearing, and as far as they knew would soon all be gone’ (C. 8547, 31-2, Carnegie to Martin).

Taxes had to be paid and in the early years these were sometimes in a kind of tribute of grain or livestock (Gann, 1965,124). Younger men, including some from the amahole or those who served with the colonial forces, were appointed as ‘Native Police’ and they provoked particular bitterness. Sykes (1897, 223) quoted Ndebele military leader Sikhombo as complaining: ‘the young men … left their kraals, enlisted, and came back the masters of their fathers, and their indunas outraged the women, stole the cattle, and lashed their betters, without rhyme or reason. These men were the chief cause of the mutiny’.

In February 1896, the devastating African rinderpest panzootic hit Zimbabwe. A strategy then believed to halt its progress was to slaughter infected herds before it spread from them. Hole (1926, 348-9) recorded that the ‘Veterinary Officers of the Government advocated the destruction of all teams and herds of cattle in which the infection showed itself, and unfortunately this advice was followed. Thousands of healthy cattle were shot, including many of those which had been allotted to the natives under the recommendations of Mr Vintcent’s Commission’. Sykes (1897, 8) explained that the shooting of cattle to contain the outbreak, ‘impressed upon the indigenous peoples that the white man had returned their cattle only to kill them “as an act of spite”’.

Blake (1977, 123), a conservative historian, wrote: ‘By 1897 there were less than 14,000 head of cattle in African possession in the whole of Rhodesia. Four years earlier there had been over 200,000 in Matabeleland alone. This is a measure of the catastrophe’. He probably exaggerated the losses, because some of those further from white settlements and transport routes were able to isolate their livestock from looters and rinderpest. But for at least four years, a society in which cattle had been central experienced a shattering loss. This had a major impact on their most important source of protein, in the shape of soured milk, and on meat supplies, at a time when diets were fragile due to war and the disruption of the agricultural cycle.

Socially and politically this also fragmented the society. Ndlovu-Gatsheni (2009, 148) contests views that the impact on the Ndebele was limited: ‘Cattle played a fundamental role in sustaining Ndebele life. Cattle sustained the institution of amabutho, they enhanced the legitimacy of the kingship through the king’s powers to distribute cattle to his subjects, they enhanced the client-patron relationships in the state, and they played a fundamental role in the Ndebele religious system. Added to this, cattle were a source of national wealth and determined status of individuals in the Ndebele state. Some tributary communities were loyal to the Ndebele State because they were given cattle. The looting of the cattle was therefore a blow to everything in the Ndebele way of life.’

After 1893, the Ndebele were deprived of their king, his capital and significant parts of their homeland. Some were initially able to stay on appropriated land and or to move back to the large farms carved out for settlers as only a limited number were used by colonists for agriculture. To compound the loss of cattle, some of those on farms had to provide labour and pay tax; forced labour was also mobilised for the mines.

Of the conquest as a whole, Verschoyle, pro-imperial clergyman and editor, wrote: ‘Besides Wilson and the men who died with him there were very trifling losses, and the expenses amounted to not much more than £100,000 – an incredibly small price to pay when one remembers what the very similar Zulu war in 1879 had cost the British Empire both in blood and treasure’ (Vindex, 1900, 326). Rhodes emphasised that he had paid for the conquest himself.

The war of 1893 cast a long shadow in the loss of lives, land and livestock as well as the disruption of agriculture. The arrival of rinderpest and poor harvests in some areas created a new crisis of subsistence in the early months of 1896 when famine was again reported. Depending on the specific timescale and area involved, it is likely that there were 4-5,000 deaths from war and war-related famine and disease among the core Ndebele nation, estimated at about 100-120,000 people, from 1893 to early 1896. It is possible that more women and children died – and there may be archives that can inform a fuller assessment (Iliffe, 1990). Those who lived further from the heartland, including the former subjects and clients of the Ndebele, probably did not suffer so severely.  Rhodes had alternative options at the time both to negotiate with the Ndebele in 1893 and to curtail the looting of livestock.

In March and April 1896, Ndebele groups attacked African police and settlers around Bulawayo of whom about 150 were killed. In June, settlers were attacked in areas near Salisbury, with about 120 deaths. Military encounters continued for about 15 months, and involved many African societies in the areas called Matabeleland and Mashonaland by the new colonial state. Both of these areas included diverse political and ethnic communities. There was not a single co-ordinated strategy by Ndebele-speaking people and this applied even more so to the many small chieftaincies in the eastern part of the colony called Mashonaland, where political authority was more devolved. This discussion does not address the complexities of African societies, nor the extensive historical writing about which African groups rebelled, why they did so, and the significance of African religious movements in the war. It focusses very largely on the scale, nature and cost of colonial violence.

Beach (1990b, 55) suggests total Ndebele and Shona losses in 1896-1897, ‘conservatively estimated at five thousand, are unlikely to have doubled that, at least as a result of battle’ – that is 5-10,000 deaths in these protracted military conflicts. Gann (1965, 140) gives 8,000 but does not cite a source or enlarge on the number. Both were thorough historians, Beach more sympathetic to African societies and Gann to settler. Hole (1926, 137) notes that 372 settlers lost their lives in total. The settler figures were comparatively high in relation to the small population, but not higher than those inflicted on the British armies when they invaded Zululand in 1878-9 nor when they invaded the Orange Free State and Transvaal in 1899-1900.

Assessment of overall deaths and the character of violence must again take into account losses from starvation and disease. The Company’s forces, and the imperial detachments that joined them in 1896, deliberately pursued a scorched earth strategy. There are many records of the burning of huts and the destruction of crops and grain stores. I will return to these figures and the deaths from famine.

In the following sections, some attempt is made to build up information on deaths by recording episodes of conflict, and estimates made at the time. I have not seen this done in published material. This exercise is by no means complete, but it suggests deaths in conflict were likely to be at the upper end of Beach’s estimate, 8,000 or more, and those from famine at least as much. Adding 4-5,000 from the earlier war and its aftermath, it is likely that African deaths in the military conquest of Zimbabwe over these four years were over 20,000 and perhaps closer to 25,000. These figures should be tested, and there is enough material to develop a fuller picture.

Company and imperial forces destroyed villages, grain stores and crops as a punitive measure and to undermine African capacities to resist. Much was also captured to provision troops and especially horses that were critical to the campaign (Baden-Powell, 1896, 24). It was difficult to bring grain to Bulawayo, because the railway still reached only to Mafeking (now Mafikeng or Mahikeng), about 10 days’ journey by mule coach and more by ox-wagon. Rinderpest had reduced access to oxen. Some supplies were brought to Salisbury via Beira, similarly a long overland route unsuitable for grain. Thus, the bulk of supplies for Company and imperial forces were appropriated from Africans and, in this way, Zimbabwean people were forced to fund their own conquest.

This can be roughly calculated. Plumer (1897, 212), commander of the Matabeleland Relief Force, estimated that their horses, largely dependent on grazing the natural veld, had on average 6lbs (pound weight) of grain per day. He took with him 1,150 horses from South Africa, was in the field in Zimbabwe for about seven months, and ended with about 700. This would have required about 1,000,000lbs or 5,000 200lb bags (900 horses x 6lbs x 200 days; the 200lb bag was the general bulk measure for grain in settler South Africa). For much of this period, the cost was about three times usual prices, £12-13 a bag, or roughly £50,000. Plumer led only one detachment and the quantities should be multiplied by at least four to give some rough sense of the value of grain appropriated for horses. This does not include the amount destroyed, nor that used for people and transport animals such as mules and donkeys.

The nature of violence is also at issue. In Company and settler writing about this war, as well as in memoirs by British soldiers, there was great emphasis on ‘murder’ by Africans, especially of white women and children. This term is not used when Africans were killed, even in relation to the far greater number of African women and children who succumbed. In Selous’s lists (1896), he records 23 women and children out of 155 Europeans killed before the main military engagements with the Ndebele and 4 in the additional list of 56 who were missing – about 27 out of 211. Hole (1926, 357) reported that official records showed ‘few’ women and children were killed in March/April 1896 – 16 out of 143. (Both figures exclude combatants.)

During the first indaba in the Matopos, in August 1896, Rhodes responded to Somabhulana’s forceful statement of Ndebele history, and criticism of colonialism, by asking ‘why did they kill women and children?’ (Stent, 1970, 51; Keppel-Jones, 1983, 500). Somabhulana countered by asking ‘who commenced the killing of women’ and gave the example of four African women shot by tax collectors. Colenbrander, translating, told Rhodes that this was true and that he should drop the topic if he wanted to secure a peaceful outcome. Rhodes did so.

Nevertheless, Selous (1897, 30-1) reflected broader white sentiments in claiming that the Ndebele attacks ‘excited a desire for vengeance, which could only be satisfied by a personal and active participation in the killing of the murderers. I don’t defend such feelings, nor deny that they are vile and brutal when viewed from a high moral standpoint …passions which can only be understood by those Europeans who have lived through a native rising, in which women and children of their race have been barbarously murdered by savages; by beings whom, in their hearts, they despise; as rightly or wrongly they consider that they belong to a lower type of the human family than themselves. I offer no opinion upon this sentiment, but I say that it undoubtedly exists, and must always aggravate the savagery of a conflict between the two races; … the murder of white women and children, by natives, seems to the colonist not merely a crime, but a sacrilege, and calls forth all the latent ferocity of the more civilised race’. In his view, this produced a ‘war of retaliation, … waged with … merciless ferocity’.

Settlers and soldiers – though by no means of one view – generally held strongly racialized ideas about those whom they were conquering. Selous (1896, 66-7), who guided the original pioneer column in 1890 and fought in both the wars, articulated an extreme version of Social Darwinism: if African people are not ‘reduced to a state of submission’, he wrote, then they must be ‘displaced’, must ‘go’ or ‘die’. He wrote:

‘the whole question of the colonisation by Europeans of countries previously inhabited by savage tribes must be looked upon from a broad point of view, and be judged by its final results as compared with the primitive conditions it has superseded … Just as in the establishment of the white man’s supremacy in the Cape Colony, the aboriginal black races have either been displaced or reduced to a state of submission to the white man’s rule at the cost of much blood and injustice to the black man, so also will it be in Matabeleland, and so must it ever be in any country where the European comes into contact with native races, and where at the same time the climate is such that the more highly organised and intelligent race can live and thrive, as it can do in Matabeleland; whilst the presence of valuable minerals or anything else that excites the greed of the stronger race will naturally hasten the process. Therefore Matabeleland is doomed by what seems a law of nature to be ruled by the white man, and the black man must go, or conform to the white man’s laws, or die in resisting them… the inexorable law which Darwin has aptly termed the “Survival of the Fittest”’.

A military intelligence report from February 1897, when cave shelters were being blown up and villages destroyed in Mashonaland, justified these strategies: ‘It seems to me that the only way of doing anything at all with these natives is to starve them, destroy their lands and kill all that can be killed’ (Ranger, 1967, 295).

Evidence suggests that Rhodes to some degree shared this approach. Lockhart and Woodhouse (1963) wrote a substantial and generally respectful biography of Rhodes, which they claimed was the first to look fully at his own and other private papers. It is not generally seen as one of the critical, anti-colonial texts (Maylam, 2005). Woodhouse, who completed the book, was a Conservative politician, MP for Oxford, fellow of Nuffield College and prolific author. Yet they (351-2) remarked on Rhodes’s ‘ferocity toward the Matabele in the campaign’ and his deep prejudices towards them, who ‘unlike children …may be shot when they get out of hand’. And it is they who record the report, later disputed, of Rhodes saying ‘you should kill all you can … it serves as a lesson to them when they talk things over at night …and they begin to fear you’ (Lockhart and Wishart, 1963, 352).

Addressing settlers and soldiers in Gwelo (Gweru), where he was based in May 1896, Rhodes was reported to say ‘that no time should be lost in thoroughly thrashing the natives and giving them an everlasting lesson’ (Evening Express, Wales, 8.5.1896). In a speech given in Bulawayo in early June 1896, Rhodes said ‘Now we shall have to hunt them in the bush, and in the stones, and in the kopjes …Their food supplies will fail, and their courage will disappear … this country will be the abode of a white race’ (Vindex, 1900, 484).

Rotberg (1988, 557) reinforces such material about Rhodes’s ‘initial and for many months predominant reaction to the rising [as] harsh and vindictive’, repeating the report ‘kill all you can’. A sympathetic observer who accompanied Rhodes on patrol from Gweru remembered how they had been burning villages and capturing cattle: ‘I have been out with Rhodes looting corn all the morning’. Grey, recently arrived to replace Jameson as Administrator, wrote in mid-June 1896: ‘We must go on hammering and hunting them’ until ‘we . . . thoroughly convince them that this country is to be the country of the white, and not the black’.

After 1893 the remaining Ndebele chiefs and military leaders had some command over fighting men, and despite the colonial view that ordinary people were keen to throw off their yoke, they found surprisingly widespread, though by no means universal, African support in 1896. Clarke and Nyathi (2010, 123) argue that Lozikeyi Dlodlo, Lobengula’s senior wife, played an important, inadequately recognised, role in both the ritual and military spheres. They suggest more generally that royal wives and women formed a significant network in the resistance to colonial rule. Beach (1986) notes that most of the Shona-speaking communities that had been absorbed into the Ndebele kingdom, fought with them, although there were also a number of independent Shona chiefs who allied at least for a time with colonial forces. So did a few Ndebele.

Following the Ndebele attacks in March 1896, most of the white population, together with some African allies, withdrew into laagers in Bulawayo, Gwelo (Gweru), Belingwe and Mangwe. The Ndebele had the advantage of initial surprise and did modify their tactics. They camped near Bulawayo, using rifles to snipe at colonial forces when they left the laager on patrol, and avoided direct confrontation. In this context, Maxim guns were valuable in ensuring that the laagers were well-defended and they – together with artillery – were used in the many ‘skirmishes’ that resulted as Company patrols sought out Ndebele contingents. But white mounted men with rifles, supported by African allies – particularly experienced and paid black South Africans – proved decisive. Unlike some South African chiefdoms in the nineteenth century, the Ndebele did not incorporate horses.

There were many more dispersed encounters than in 1893 and mounted platoons were regularly sent out from Bulawayo. As Selous (1896, 64) wrote in his revealing but deeply disturbing record ‘I have stated plainly that we fired on these [Africans] at sight, and that although they offered no resistance, but ran away as hard as they could, we chased them and kept on firing at them as long as we could see them, and this action may possibly be cited as an example of the brutality and inhumanity of the Englishmen in Rhodesia’.

In early April 1896, one Company contingent lost 7 men, 20 wounded and 33 horses in 6 hours of fighting (Selous, 1896, 122-3). But helped by a Maxim they estimated that they killed and wounded between two and three hundred. Another patrol, starting on 4th April, was attacked by about 2-300 Ndebele soldiers: ‘Colonel Gifford then opened on them with the Maxim at about 600 yards, and this quite quenched their military ardour’. During their retreat, a mounted platoon attacked, killing twenty or thirty. On 6th April, Gifford’s platoon fired on another group of Ndebele and Selous was told that ‘the patrol killed at lowest 200 of the enemy, and many more must have been wounded’. In an intense period of conflict, from 30th March to 9th April, it seems that about 600 Ndebele were killed in these and other sorties made from Bulawayo.

On 10th April, three African men accused of being rebels were arrested, quickly condemned to death, and hanged, their bodies left dangling from the branches of a tree in Bulawayo, as a warning. At least 6 more were hanged soon afterwards ‘tried in a somewhat rough-and-ready fashion’.

Rhodes arrived in Zimbabwe direct from England, landing at Beira to avoid an enquiry into the Jameson raid in Cape Town. He travelled via Salisbury, joining a relief force to Gweru laager in early April. There he took on the informal role of ‘colonel’, to provide leadership for disparate military groupings, and engaged in operations (Keppel-Jones, 1983, 456).

Bulawayo remained the centre of colonial military activities and Selous records a sequence of skirmishes at the Umguza river from the 19th of April that resulted in small losses for the Ndebele and Company troops. An encounter on 25th April, when the Ndebele attempted unsuccessfully to capture a Maxim, resulted in ‘no fewer than 74’ dead (Selous, 1897, 189). He thought that considerably more could have died as they were exposed to machine gun and artillery fire while retreating and colonial soldiers reported that they had seen many wounded. Selous thought that this was a serious reverse for the Ndebele and felt assured that Bulawayo was no longer under threat of attack. The Ndebele did not attempt to shut off the road south-west to Bechuanaland, apparently because they hoped that settlers would leave by that route.

Rhodes, based in Gweru, who was in regular telegraphic communication with Cape Town and Salisbury, was keen to keep overall control of the military campaign and did not wish to lose the initiative to imperial forces. Chamberlain nevertheless insisted on an imperial presence and it proved valuable in winning the wars. During April and early May, around 850 men under the British army officer Colonel Plumer, with over 1,000 horses, arrived in Bulawayo as the Matabeleland Relief Force. Grey, the new Administrator, as well as 9 Maxim guns, travelled with them – 3 other Maxims were sent by sea to Salisbury. Plumer (1897, 15) recruited his detachment in South Africa and included British South Africa Company police who had left a few months earlier with Jameson for the raid. The force was paid for by the Company at the eventual cost of £250,000.

Selous (1896, 192), fighting with Company forces, records that on 21 May 1896 a strong mounted patrol routed a Ndebele force at Tabas Induna, twelve miles from Bulawayo, drove them into the bush and killed those retreating: ‘when it was at last abandoned a long line of corpses marked the track where the whirlwind of the white man’s vengeance had swept along. Vae victis!—”woe to the conquered!”—woe indeed; for amongst the men who took part … were [those] determined to use their opportunity to the utmost to inflict a heavy punishment’. Selous continued with a detachment down the Insiza valley, met up with Rhodes, discussed tactics and looted livestock. He remarked on Queen Victoria’s birthday, 24th May, which he celebrated by ‘burning a large number of kraals’; ‘all the grain that could not be carried with us was destroyed as far as possible’ (Selous, 1896, 213).

Plumer (1897, 100) records a skirmish on 20th May in which an estimated 90-100 Ndebele died. In late May his imperial detachment scattered a Ndebele contingent with about 50-60 killed and a further encounter against an estimated 1,500 Ndebele resulted in 70-80 deaths (Plumer, 1897, 92-6). The added strength of the Matebeleland Relief Force enhanced the military effectiveness of the colonial armies. Both they and Company detachments were regularly involved as the Ndebele contingents tried to maintain positions close to Bulawayo. A number of incidents were reported where the latter suffered over 50 dead and at least two, on 25th April and 21st May, may have resulted in over 100 deaths. From mid-April to the end of May, perhaps over one thousand Ndebele were killed.

The pattern continued in June with a focus on Ndebele contingents near the Umguza river, close to Bulawayo. Kraals were burnt and grain destroyed. On 6th June about one thousand armed Ndebele were forced to retreat and, running into the bush, they were shot by men on horseback. Selous (1896, 224) records ‘in the chase which followed, a large number of them were shot down …I am of opinion myself that the Matabele lost more heavily on this occasion than at any other fight during the campaign, for the very reason that it was not a fight but only a pursuit in which the natives were killed as fast as they were overtaken’.  Baden-Powell (1901, 60) estimated that ‘at least fifteen indunas and two hundred men’ died. He photographed a dead man and took his knobkerrie. Plumer (1897, 118-9), who was not present, heard that they ‘inflicted a tremendous defeat on the rebels, who were scattered in all directions, some 300 being killed. It was the greatest loss inflicted on them in any engagement during the campaign’.

By early June, Rhodes was based in Bulawayo and General Carrington arrived to take overall command. Carrington sent out three strong detachments to clear rebels from to the east and north of Bulawayo. Plumer with a force of about 450 men and horses, as well as machine guns, moved along the Gwayi river valley north of Bulawayo. This was not a centre for Ndebele military operations and the people retreated. Sykes (1897, 99-106) recorded that anyone they saw coming from the settlements was immediately shot down and ‘hundreds of kraals were burned all along the valley of the Gwai on both sides of the river. Many of them were well stocked with grain — Kaffir corn, millet, and mealies’, ‘At every kraal where there was grain’, he continued, ‘the troopers filled their horses’ nose-bags with as much as could be carried away, the remainder being destroyed’ (Sykes, 1897, 106). He listed ‘all descriptions of native-made articles’ from household utensils, dried food and soured milk to ‘ostrich feather head-dresses and capes, spears, shields, carved wooden jars’ that were also burnt in this destructive expedition that affected thousands of people. Rhodes accompanied the detachment led by Nicholson, which had similar experience.

On July 5 1896, colonial forces attacked Intaba zikaMambo a rocky area about 80 km north east of Bulawayo, where some of the remaining Ndebele army had retreated. Plumer was in command of 750 white and 200 black troops with Maxims and seven-pounder guns. Rhodes accompanied the column. ‘Everywhere it was a bloody fight, often at close quarters’ (Keppel-Jones, 1983, 459). Plumer lost about eighteen men killed and fourteen wounded and estimated Ndebele losses at about a hundred. Another account (Sykes, 1897, 147-8) describes fires being lit at the mouths of caves into which people had retreated and vegetation dropped into flames in order to smoke them out. Six hundred women and children were persuaded to give themselves up. Subsequently ‘as the suffocating smoke penetrated into the recesses, there was a general rush of concealed rebels to escape, and no sooner did they appear, amidst the flames and smoke, than they were shot down’. These appear to be deaths additional to those in Plumer’s report and Sykes (1897, 157) estimated 200 killed and 200 injured. He and Plumer record loot of about 1,000 cattle and 2,200 goats and sheep. Stent (1970, 22, 24), a journalist with Rhodes, probably exaggerated in remembering ten thousand head as loot that ‘seemed to cover the veld for miles’.

Stent (1970, 23) thought that the battle at Intaba zikaMambo shocked Rhodes not least because he saw the costs to colonial troops when the Ndebele had well-entrenched defensive positions. ‘There had been a good many natives killed too. But the death of the rebels who had murdered white women and children did not come home to us as the death of these eighteen of our own kind and colour’. Stent asserts that this prompted Rhodes to think about making peace.

The Imperial authorities tried to offer a qualified form of amnesty but it was clearly not trusted (or known). Instead colonial forces mounted a similar attack on the last significant Ndebele stronghold in the Matopos, south west of Bulawayo, with over a thousand troops. They joined battle on 20th July and, according to a prisoner, the Ndebele losses were heavy: a ‘large number of their best men killed, including five chiefs and Nuntwani, their general, severely wounded in the leg’ (Keppel-Jones, 1983, 461).

Nevertheless Plumer noted the change in Ndebele strategies, and concluded (1897, 169): ‘If they adhered to these tactics we have a long and difficult task before us …the Matabele … with their mobility and knowledge of all the intricacies of the koppies would have no difficulty in avoiding any large column, while they watched for an opportunity of swooping down upon any patrols or small parties that might become detached from the main body’. He was joined by Laing and his Company contingent from Belingwe, who was attacked at Inugu. Although this position was defended with machine guns, artillery and rifles, they lost 4 whites, 6 severely wounded and 27 Africans allies killed or missing (Laing, 1897, 296). Laing did not estimate Ndebele losses in his book but a report in a British newspaper recorded 90 (Evening Express, Wales, 24.7.1896, from High Commissioner, Cape Town).

Plumer nevertheless stayed on the offensive, taking on some leading Ndebele forces on the peripheries of the Matopos in early August. Sykes (1897, 188-90) recorded a number of Ndebele killed in the Mtshabezi valley on 1st August and the destruction of granaries belonging to Somabhulana’s people on the 2nd. Sikhombo’s forces were attacked (Sykes 1897, 195-7) and although seven whites and 2 Africans troops were killed, they inflicted severe casualties with Maxim guns. Plumer (1897, 179) estimated that ‘from the reports of the different officers engaged it is considered their numbers were probably nearly 3000, and their losses about 200’. On the 8th they opened fire on Umlugulu with artillery and found 13 bodies. Plumer noted that the battle with Sikhombo was the ‘severest engagement’ in the Matopos, ‘the rebels fighting with great determination and, in some instances, with desperate courage’. It seems that the Ndebele lost at least 400 men in battle between 20th July and 8th August in the Matopos; other contingents were involved and the burning of villages continued throughout (Plumer, 1896, 186). On 6th August Carrington and Rhodes arrived and held a parade of colonial troops.

As mentioned, Bulawayo was not the only centre of operations. Laing (1897, 133-4) wrote about his sorties from the Belingwe laager, about 150km to the east, before he came to the Matopos. He records destroying the kraals of a chief whom he called amahole, probably Karanga-speaking, and attacking Selemba where the ‘enemy’s loss was not estimated; it would have taken too long to find out in the thick bush, but it was pretty severe’. Elsewhere an old woman was recorded as saying that Laing’s troops shot their leaders, ‘scattered everybody and destroyed all our kraals’ (Laing, 1897, 224-5). After another encounter, 40 were seen dead. His column discovered a ‘large quantity of grain, hidden away in the thickest part of the bush…stored in grass bags, each of which contained about two ordinary sacks full of grain…  This was a most welcome and valuable discovery in more ways than one, because it not only enabled us to replenish our stock of food for the horses and mules, but it reduced the rebels’ store to a very considerable extent’ (Laing, 1897, 248-9). In June he attacked a pro-Ndebele Rozvi chief and became concerned that his contingent was taking so much loot that they would stop fighting in order to get it home.

These battles in June and July, with at least five major incidents, one with perhaps 300 deaths, and many smaller skirmishes, probably resulted in over a thousand deaths, as well as much looting and destruction of food. At this time, Carrington, head of the imperial troops, reported that a patrol to Nyamandlovu, the district to the west of Inyathi, captured all the grain they it could take – 72,000 lbs (360 200lb. bags or about 12 wagon loads) (Keppel-Jones, 1983, 519, note 30). Nyamandlovu and Inyathi were both subject to intense conflict and appropriation. Iliffe (1990, 24-5) argues that ‘famine became the core of the Company’s strategy’. Famine was taking its toll. In August 1896, a captive taken in the Matopos said ‘many women and children had already died for want of food’.

To summarise: it may be reasonable to estimate that roughly 600 died in conflict in early April, 1,000 from mid-April to the end of May and probably more than that from June to early August. If Laing’s sorties are added it is likely that over 3,000 were killed in conflict. There were other patrols that have not been included.

In June, some Shona chiefs had, surprisingly to the Company, rebelled. Rhodes was counting the costs of the campaign against the Ndebele both in money and men. Carrington and Plumer thought it would be costly to lay siege to so large an area as the Matopos and starve the remaining African contingents out. Ndebele leaders, who represented disparate groups were receptive when Rhodes put out feelers for a meeting. They knew they could not win and were stalked by famine, a point emphasised by Rhodes in the Indabas starting on 21st August. Much was, and is, made of Rhodes’ courage in attending these, but the Ndebele leaders were also highly vulnerable.

The war in Matabeleland did not cease during the negotiations. In September 1896, for example, Baden-Powell (1901, 287ff), who had led a patrol in the Matopos, went north of Bulawayo to track down Ndebele forces who were thought to be retreating to the Somabhula forest. On his route he became entangled in military action against Uwini (Whinya) – a Shona-speaking chief in the Ndebele kingdom who occupied a hillside village near the Vungu river valley (Clarke and Nyathi, 2010, 172 ff.; Ranger 1967, 159). Native Commissioner Gielgud thought this remained an area of resistance and captured Whinya, took prisoners and appropriated 72 bags of grain. Whinya’s status is unclear, but Baden-Powell, who arrived at this point, chose to believe that the chief was a key religious leader whose death might encourage a more general surrender. The American scout Burnham had recently assassinated a man, who he asserted to be a religious leader, near the Matopos, with the same intent (Ranger, 1967, 183ff.)  Burnham is also reputed to have enthused Baden-Powell about the ideas and practices of scouting.

Baden-Powell (1901, 297-9) subjected Whinya to a quick trial and then had him executed near his former stronghold in front of ‘all the natives in camp, both friendlies, refugees, and prisoners’ for the ‘moral effect’. On this occasion, there were repercussions. The High Commissioner in Cape Town, concerned about the conduct of the war, especially after the recent execution of Shona chief Makoni (see below), had instructed Carrington, head of imperial troops, that African prisoners should not be subject to such summary punishment. Baden-Powell (1901, 290), suggests that he did not know this, but Carrington was ordered to initiate a court martial. Sykes wrote at the time that ‘the ferocity exhibited on several occasions by the captors towards their victims was anything but an edifying spectacle’ and he described an incident where a trooper put a noose around a captive’s neck and made him run behind his horse, ‘with no other motive than sheer brutality’, till he died. Keppel-Jones (1983, 463) records other acts of summary justice, including the shooting of a woman accused of spying.

Some of Whinya’s men tried to hold out and Baden-Powell prevented their access to water. Two, perhaps more, were shot when they tried to quench their thirst and this strategy persuaded some to surrender. ‘Large stores of grain’ were captured (Baden Powell, 1901, 301). There was acute food shortage in the area and Baden-Powell also captured women and children in order to prevent them acting as carriers of grain to African fighters. For the rest of September, this patrol moved along the Vungu and Shangani rivers, and into the Somabhula forest where they believed that Ndebele soldiers had taken refuge. There were few encounters: a village was burnt, 20 Africans were killed in one incident and they blew up a cave shelter. Baden-Powell’s court martial took place in Gweru in late October 1896; the British military court exonerated him.

The region called Mashonaland was divided into a number of relatively small independent chieftaincies, which in Company thinking were not seen as a military threat. Detailed historical research explains why many rebelled, beginning on 18 June 1896 in the Hartley (Chegutu) area, south-west of Salisbury (Harare) (Ranger, 1967; Beach, 1971, 1979, 1986; Cobbing, 1977). About 120 whites, including several women and children, were killed in the first few days and the road from Salisbury to Bulawayo temporarily blocked.

A laager was formed in Salisbury, with smaller contingents stationed at Umtali (Mutare), Mazoe and elsewhere. A similar pattern evolved as colonial patrols attacked different African chieftaincies and communities who in this case were more dispersed. There were few offensive actions by Africans, the rebellion did not involve all of the chieftaincies in the area, nor did they act simultaneously (Beach, 1979, 1986). For the most part Africans were trying to defend themselves against colonial troops. Rhodes did not think it was necessary to make peace with the chiefs in Mashonaland and ‘they were forced into capitulation by the harshest punitive measures’ (Davidson, 1984, 309). Such measures preceded the uprising and had been one of its causes: J.S. Brabant, chief Native Commissioner for Mashonaland, employed strong arm tactics in collecting taxes in 1895, including flogging, burning down villages and the confiscation of livestock (Rotberg, 1988, 553).

Alderson (1898, Appendix C I-III), head of the imperial Mashonaland Field Force, records some of the patrols from Salisbury both before and after he arrived. At the start they were small contingents, usually under 50 men, of locally raised units such as the Salisbury Field Force: 28 were sent out in two months from mid-June to mid-August. The first sorties were largely to bring in whites and African allies or to investigate deaths, bury bodies and to get forage from settler farms. But at least 16 incidents were recorded with 7 white and 4 black killed on the colonial side. About 223 African deaths were specifically recorded, 137 in one sweep through the Hartley Hills, seen as a rebel stronghold, from the 19th-28th July.

In July 1896 patrols were sent out from Salisbury to Enkeldoorn (Chivhu) where they were joined by Afrikaner settlers (Beach 1971, 396-7). De Moleyns, commanding a new force in Mashonaland, led them against chief Chesumba, about 12 miles from Salisbury, who had retreated to a koppie with boulders and some stone barricades. De Moleyns called out for the chief to surrender but instead was fired at, with the loss of two police. He then used dynamite – a material that was available from mining operations: 111 men and 500 women and children gave themselves up. Deaths from what may have been the first use of dynamite are not recorded.

As the scale of patrols increased, Alderson, using records from before he arrived, reported four further encounters where: ‘a good many rebels’ killed; ‘a good many dead bodies seen’; ‘many others believed killed’; and ‘some rebels’ killed. If these are added to the numerical records, there were perhaps 400 African deaths. From the start kraals were burnt, the biggest number recorded being 20 in one sortie, and livestock taken. As noted, kraals could vary from extended families to small villages of 100 huts, so that these could involve hundreds of dwellings. On five occasions, 4 to 6 wagon loads of grain was taken. A contingent from Mazoe in June was involved in a 12-hour battle in which 3 were killed on the colonial side and ‘a number’ on the African (Jones, 1953, 140); a column from Bulawayo of over 300 in July burnt 20 kraals, captured grain and killed a ‘considerable’ number. It may be fair to assume, that at least 500 were killed before the engagement of imperial troops.

When Alderson’s imperial Mounted Infantry and Mashonaland Field Force arrived at the end of July, colonial military intervention intensified. One of his first ventures was to attack a leading chief, Makoni, thought to be a key resistance leader (Alderson, 1898, 90ff.). He first bombarded the settlement with artillery, then opened up with a Maxim gun and when Makoni and his people retreated into caves and rock shelters, set fire to the village so that it was completely burnt out. He wrote: ‘We had walked into and destroyed the kraal of the biggest Chief in Mashonaland, inflicted on him a loss of some two hundred men, including his principal witch doctor and ten of his counsellors, made him go to ground like a rabbit, and, greatest blow of all, had captured 355 head of his cattle, and 210 goats and sheep’. Many were wounded while Alderson lost 3 men killed. To stamp his authority, he burnt several kraals on the way back. For the next week his column undertook a sweep of Marandellas (Marondera) to the south-east of Salisbury, burning homesteads and collecting grain. Another patrol burnt kraals in Mazoe. Systematic burning and looting of almost every village and homestead they passed became daily practice for Company and imperial forces.

In July Carrington ordered a section of Plumer’s troops in Matabeleland, under Watts, to assist in the suppression of the rebellion in Mashonaland (Plumer, 1897, 221ff.). They had their first serious encounter at the end of July near Marondera store where they killed an estimated 50-60. They joined a further attack on Makoni at the end of August (Alderson, 1898, 230). Makoni offered to surrender in return for amnesty, and Company officials, as well as Rhodes himself, were disposed to accept this as a way of neutralising an important chief, but the imperial authorities wanted unconditional surrender. His stronghold was located near the main road from Salisbury to Mutare and seen to threaten transport routes.

Arriving at Makoni’s, the huts and granaries already burnt, Watts fired shells into cave mouths where people had taken refuge. His column advanced the use of dynamite by dropping lighted sticks from above into the caves. These were seldom completely blocked off at the top and ‘it was usually possible to find some crevice’ (Ranger, 1967, 276-7). The effect produced after a few days of dynamiting was ‘terrible … and the stench from the dead bodies was over-powering’. Makoni was captured, Watts court-martialled him in rapid trial without authority, and had him shot in public. This, as noted, preceded the shooting of Uwini (Whinya) on Baden-Powell’s orders; the High Commissioner ordered an enquiry which exonerated Watts.

Intended as a warning to the other chiefs, it seems to have had the opposite effect because the message they took was that there was no point in surrender, which would likely lead to death. Some women and children came out of the cave after the first dynamite was lobbed inside, and some men escaped. It is not clear how many were killed; one source suggests about 60 (Keppel-Jones, 1983, 490).

Alderson divided his imperial force into four units of about 120 men each, who made a sequence of destructive patrols. I will put the names of chiefs attacked in inverted commas as I have not yet had a chance to verify them and their identity. For example, at the end of August (Alderson, 1898, 147-151) an expedition to ‘Untegeza’, in the eastern highlands, burnt a village and took its grain on the way. On reaching the chief’s area, in mid-September, some attempt was made to parley but the chief feared for his life and sent a gift of cattle and two sovereigns: ‘the kraal was absolutely destroyed, and every particle of grain from it and the surrounding kraals was carried away or spoilt. A good few prisoners were taken, and those that were not killed or taken prisoner cleared away in the night’.

Similarly, at the villages of ‘Gona’, they fought for a few days until they destroyed the settlement and carried off the grain. The same pattern took place at ‘Simbanoota’s (Alderson, 1898, 159) and when they marched to ‘Mzimilima’’s, they found the homesteads deserted so took ‘a fair bag of grain and cattle’. MacMahon, one of the section leaders, moved into Mazoe and attacked ‘Gaderra’ with artillery, rifles and dynamite. They burnt kraals, including ‘Nyanda’’s, mined caves, which were blown to pieces, ‘captured large amount of various loot’ and ‘a number of natives [were] known to have been killed’ (Alderson. 1898, 291).

Alderson wanted to complete the process of scorched earth, posting small contingents of colonial forces wherever they had destroyed villages and cleared the people, in order to prevent any from returning and sowing crops or using their old rock shelters. This was done in a few areas, but there were not enough troops for systematic control of countryside in this way.

There is insufficient numerical information in his book to estimate deaths effectively and it is unlikely that 200 were killed in every incident, as recorded for the first attack on Makoni. But deaths are commonly mentioned and in only one of the sorties was the settlement and rock shelter deserted. It may be fair to assume that if the attacks on Makoni are included, at least another 500 were killed in six sorties mentioned here in August and September.

I have not yet been able to compile a record of all similar incidents that continued for over nine months until July 1897 and the sources less often mention estimates of casualties. But other examples can be cited. In October a chief who was not part of the rising was attacked in error, and when he tried to escape, was shot along with ten of his men (Beach, 1971, 301). Further attacks in Mazoe destroyed cave shelters.

In early October, a devastating attack took place on Manyepera and his people, when they took refuge in caves in Marendellas (Ranger, 1967, 276-7, now Marondera). Company troops saw smoke coming out of a narrow crack in the roof and by removing stones could see the light of fires. They tried throwing down artillery shells and then smoking people out by pulling down nearby huts and using them in fires near the entrance. Receiving a consignment of dynamite they dropped lighted sticks down the cracks. Sixty women and children came out after this first explosion. The contents of several cases of dynamite were then inserted and the explosion ‘rent the cave from end to end’; only 2 people survived. ‘This terrible encounter’, Ranger (1967, 277) notes, ‘became the pattern for many attacks on Shona strongholds, despite humanitarian outcry in England’. It certainly was repeated, but the strategy was already well-established. Alderson (1898, 291) recorded: ‘Killed 15 natives. Many more believed to be killed’. The caves and shelters were not inspected for corpses.

Mashiangombi (Mashayamombe) was both a major chief in Mashonaland and a key figure in the resistance. His substantial settlement included ‘numerous small kraals …scattered along the broken granite kopjes on both sides of the Umfuli river. They contained 25 to 100 huts each’ with nearby ‘caves and recesses, partly natural but partly supplemented with artificial stone walls, thus making an excellent refuge and fortress in the case of attack’ (Ranger, 1967, 281ff). His settlement was first attacked in July 1896 when twenty people were killed and 500 head of cattle taken. A further expedition in August burnt kraals, including the chief’s. In October, Alderson (1898, 292) himself led a contingent of over 500 colonial troops and African allies, which burnt the whole of the remaining settlement, a ‘very large number of kraals’, and all the caves that could be located were blown up. Mashayamombe escaped and, seen to have successfully survived three attacks, remained a focus for resistance. Alderson noted that ‘enemy’s loss difficult to estimate owing to thick bush and numerous caves but must have been considerable’. He was later blamed for failing to deliver a knock-out blow but his account of this attack was devastating. In February 1897, crops that had been planted were destroyed in the middle of the growing season.

A series of further attacks took place in October and early November. At ‘Chena’s’ in the Hartley hills, kraals were burnt, caves destroyed, and ‘a considerable number of natives killed’ (Alderson, 1898, 208). Evans burnt kraals at chief ‘Cheri’ and was then killed in the attack on ‘Gatzi’, where troops nevertheless burnt the settlement and killed all in the caves, with 16 bodies counted and ‘many more believed to have been killed’ (Alderson, 1898, 292). Mapondera, another substantial chief, was attacked and his villages destroyed. In the face of this onslaught, Alderson heard that many had fled into ‘fly country’ – to the margins of settlement, beyond the reach perhaps of the colonial troops, but also away from food supplies.

This is not an exhaustive record of the Salisbury based patrols and there were others. In the first week of October 1897, Baden-Powell (1901, 364ff., before his court martial) moved eastwards from Matabeleland and attacked communities near Wedza mountain about 120 km south of Salisbury. When people retreated to shelters, Baden-Powell chose to interpret the response as an act of war. Having burnt vacated settlements as an ‘object lesson’, his troops proceeded to ‘freely help ourselves’ to the grain and livestock. ‘We began to hammer away with the 7–pounder, the Maxims, and Nordenfeldt, taking each koppie and its kraal in turn. Through the glass I could see the natives move from the kraals into the caves, and when we shelled these, we could see them stealing away through the rocks and bush, evidently anxious to make their escape’; when people tried to escape, they were shot down (Baden Powell, 1896, 380, 387).

The settlements on Wedza mountain were cleared and burned. Baden Powell (1901, 389-95) then blew up a stronghold with dynamite and celebrated ‘the complete destruction of the enemy’s villages and the clearing of their grain stores … the blazing evidences of it gleaming out their message to all the rebels for miles round’. Heading east, his patrol then looted and burnt the abandoned defensive village of ‘Monti’ (Mondi) – this located on a ‘bold, upstanding, solitary peak, a regular acropolis’. He continued with a series of attacks described in his narrative: ‘We helped ourselves to all the corn that we could carry, as well as to some little bits of loot, such as a Kaffir piano and some tambourines—the piano being a small flat board on which is fixed a row of iron tongues, and these when struck give each a different note of soft, metallic sound… Then we set the village in a blaze’.

‘In Monogula’s’, he wrote, ‘we placed thirty-four cases of dynamite, and at one grand burst blew up the whole koppie, so that where there had been hill there remained but a crater. Previous to demolishing the caves, we had of course removed, for our own use, the stores of grain which had been stowed away for the rebel garrison’ (Baden-Powell, 1901, 409-11, 428). Most had been able to escape but they found nine dead. More kraals were burnt and 26 killed en route to Enkeldoorn (Chivhu) on the way to Salisbury.

As noted, Rhodes himself had been involved in such practices since May in Gweru and he continued till the end of his stay in Zimbabwe. Alderson (1898, 242) received news on 31st October that Rhodes, then on his way from Bulawayo to Salisbury, when passing Enkeldoorn (Chivhu), had gone out with members of the garrison there, burnt ‘Singala’’s kraal, and killed about thirty. Rhodes wrote enthusiastically to Grey of this sortie: ‘we went out and destroyed his [chief Sango’s] kraal, killing a good many natives’ (Ranger, 1967, 285; Beach 1971, 348).

In these twelve specific attacks mentioned in October, many were killed but it is difficult to provide any clear estimates because those blown up in shelters and caves were not counted. Again, it may be legitimate to suggest at least five hundred, as well as devastation of the countryside. Even an incomplete record of the first few months of this punitive war suggests more than 1,500 were killed and tens of thousands of people affected by the destruction of villages.

Baden-Powell (1901, 447) met Rhodes in Salisbury on 22nd November and talked to him about ‘ways and means or plans of campaign … [Rhodes] full of restlessness and energy’. They both joined a hunt (for wild animals) and as they passed under a telegraph line Rhodes ‘at once went into particulars of that’. Rhodes, who had long been immersed in military planning in Matabeleland, was equally engaged at the time when villages were being burnt, caves blown up and grain stores destroyed in Mashonaland. In December, Baden-Powell travelled back with Rhodes by ship from Beira to Cape Town and then to England. At Port Elizabeth they were greeted by crowds and treated to a banquet for 500, as well as dinner at the club.

By the time he left, Rhodes and his military leaders mistakenly believed that the rebellion was at an end. Alderson (1898, 249) summarised the position in November, claiming that the major rebel chiefs around Salisbury had been defeated, their settlements destroyed, and the main routes to Mutare and Mozambique, as well as Bulawayo, had been secured. He and his contingent of imperial troops left in late November and December. The cost of keeping them longer was seen as too high, and they faced the wet, hot summer months. The Company wanted to assert its control over military as well as civil administration.

Beach (1986, 109) believes that by December, only limited areas were under colonial control, ‘in spite of frequent illusory successes such as that of Alderson at Mashayamombe’s’. He also sees this period as one when the Kaguvi spirit medium began to feature even more centrally in European minds as a major factor in the rebellion. Troops were sent out to capture him in January and this led to further destruction of a now empty settlement.

Beach underestimates the destruction but some Shona chiefs were remarkably persistent despite the devastation that they saw around them. In January 1897, de Moleyns destroyed kraals at ‘Sosve’, who had been attacked earlier; in February he captured 60-70 wagon loads of grain and in April a patrol with African allies took a stronghold at Shangwe after cutting off food and water supplies and killing perhaps 90 (Keppel-Jones, 1983, 514-5; Davidson, 1984, 311). In May 1897 a contingent attacked Chief Mashanganyika who, with his people, retired into caves. Over a week, all the caves were destroyed ‘with great loss to the occupants’.

In June a colonial force captured Kunzwi’s stronghold in a fortified koppie after two days of hard fighting. Major Gosling thought the engagement was ‘the most severe the police have as yet taken part in’. Ranger (1967, 305) agreed that this was the fiercest action of 1897, but provides no record of casualties. Although Kunzwi escaped, he surrendered in August.

In the final attack on Mashayamombe in July 1897, 130 men with a Maxim maintained a cordon, shooting anyone who attempted to escape. The chief was killed, trying to move between shelters, and the next day the caves were destroyed with dynamite. The losses were ‘heavy’ and 100 men and 320 women and children surrendered (Keppel-Jones, 1983, 516). The Kaguvi and Nehanda mediums were apparently in Mazoe district, where they were attacked on 23rd July and their supporters took to caves, which were blown up with many killed. In August, remaining huts at the Kaguvi medium’s settlement were burnt (Ranger, 1967, 300) and he surrendered. It was only in August that ‘Mangwende’, ‘Svoswe’, ‘Zwimba’, ‘Chiquaqua’ (Chikwakwa), ‘Chinamora’, ‘Seki’ and others surrendered. Nehanda, who had escaped, was captured in December, hung in Salisbury, with the Kaguvi medium and 25 men identified as rebels.

I have described briefly many incidents of highly destructive behaviour by the Company and imperial troops, including bombarding settlements with artillery, burning villages, shooting those who tried to escape, and dynamiting caves – with the probability of many deaths, perhaps hundreds, in some incidents. This record does not look systematically at patrols and attacks from November 1896 to July 1897. Further research would be needed to identify chiefs mentioned as well as to develop a fuller record. The frequency of military patrols is likely to have declined when the imperial troops left in early December. But it is likely that there were similar losses in these nine months as in the first five months from June to October, estimated as at least 1,500. Total deaths in Mashonaland were higher than those in Matabeleland (estimated above at over 3,000). Food was destroyed on a large scale. The new growing season started with the rains, generally around October, with grain only maturing some months after that, so that the many whose food supplies were captured or burnt from July 1896 to February 1897 were in a perilous position over a long period. Growing crops were also destroyed.

We cannot be certain of the losses in this protracted conflict, lasting in various phases for over 18 months from March 1896. This memorandum was researched and written in a very limited period, relying on published material, and it should be used as a springboard for a more thorough examination of the sources available. By no means all published material has been used, nor any government printed papers and archival sources, including military records. The most valuable sources for the sections on Zimbabwe have been detailed histories such as Keppel-Jones, Ranger and Rotberg as well the published memoirs of colonial participants themselves, such as Alderson, Baden-Powell, Laing, Plumer, Selous and Sykes. To my knowledge, little attempt has previously been made to discuss the scale of violence and losses directly due to colonial military intervention. The material seen so far is fuller for the six-month war in Matabeleland than the full year of campaigning in Mashonaland. It would certainly be possible to construct a far more detailed history of the campaigns, and assessment of losses, than offered here.

Iliffe, a leading British historian of Africa, is one of the few who has looked at the Zimbabwean archives with famine specifically in mind, but he has only one short chapter dealing with this period and he has not focussed on issues of scale. Nevertheless, he paints a bleak picture and insists that the famine of this period was not primarily caused by natural disasters such as drought, rinderpest and locusts, but ‘created by the violence of the rebellion and its suppression’ (1990, 23).

Hugh Marshall Hole (1926, 375-6), graduate of Balliol, and a key administrator under Jameson, praised the Company’s generosity, but noted that ‘for some months the Matabele suffered severely from famine’ and ‘there were many deaths from starvation’ in 1896. Carnegie, the missionary, who knew Matabeleland well, wrote ‘there never before was such a famine in this country and no one living outside it will ever know how much pain and suffering thousands of natives went through. We were in the midst of it and saw its ravages with or own eyes. We had to try and save people from dying of hunger whenever we could. … This was one of the sad results of the war’ (Jones, 1953, 127). Ranger (1967, 264) records another recollection that ‘an incredible number of deaths from starvation occurred’ and that ‘there remained not more than 5000 head of cattle …milk was practically unobtainable for the sustenance of Native child-life’.

Native Commissioner Gielgud wrote of deaths from famine in Inyathi in October 1896, and of a ‘very terrible’ famine in which ‘whole families’ were dying in January 1897. He estimated in April 1897 that a quarter of the people in Inyathi died of famine (Iliffe, 1990, 28). In his detailed demographic work, Beach (1990a, 50) gave a figure of about 6,000 in Bubi district, of which Inyathi was part, in 1898. The precise reference point of these observations is not clear and the archives are not presently accessible. If a quarter had died, there may have been 8,000 people before the famine with as many as 2,000 deaths.

Inyathi was north of, and relatively close to, Bulawayo and in easy reach for colonial forces, so that it may have suffered disproportionately. But such losses only need to have been replicated in a couple of other Matabeleland districts, some of which had larger populations, for the number of deaths to have been 6,000. As noted, a captive taken in the Matopos in August 1896 said ‘many women and children had already died for want of food’. This was to the south-west of Bulawayo and a large number of people, including soldiers, had moved there to find shelter amidst the rocky terrain. In October, people surrendering in Nyamandlovu were ‘in a horrible state of starvation’ (Iliffe, 1990, 26). Villages and grain stores had been systematically destroyed by Plumer in the Gwayi valley, to the north-east of Bulawayo, after the harvest in July.

Proponents of Rhodes’s generosity emphasise his contribution of £50,000 for the supply of grain after the Indabas in October 1896. Some of this was captured grain (Clarke and Nyathi, 2010, 175), most of it was for those in the Matopos and supplies that came by rail and wagon were largely restricted to Bulawayo because there was little transport to take them to outlying districts (Iliffe, 1990, 27). It certainly saved some but this was not a huge amount of grain; as noted, its value was probably similar to that appropriated to feed Plumer’s horses over 7 months. Iliffe, by chance, mentions an amount of 1,000,000 lbs of grain distributed by the Company, similar to that I have roughly calculated being appropriated for horses. The deaths described by Gielgud in Inyathi and to its north came after this distribution so that it was clearly inadequate for those areas (Iliffe, 1990, 28). Bodies were found on the veld near Bulawayo too. Blake (1976, 147) notes that the Company paid out £250,000 in compensation to the settlers in Matabeleland and £100,000 in Mashonaland: ‘the settlers were very generously treated’.

As noted above, estimates of deaths in a number of different incidents suggest that the Ndebele lost at least 3,000 men in conflict in 1896 and it is likely – if around 2,000 died of famine in one district – that the losses were at least twice that number. This would make perhaps 8-9,000 in all.

Iliffe (1990, 29) does not think that the people of Mashonaland suffered so severely from famine as the Ndebele; he notes that they had more experience in secreting their grain and that, as Beach also suggests, many were able to plant in the wet season at the end of 1896. Some may have been able to survive off gathered food, or by flight to areas that were not directly attacked. But Iliffe’s treatment of this area is brief and he does not begin to record the scale of destruction of villages and food supplies noted here. Many of the leading chieftaincies experienced attacks that diminished or destroyed their livestock, grain stores and crops. Even if some were able to plant again in 1896-7, after the colonial troops moved on, their grain would not be available before April/May 1897. In Mashayamonbe’s case, one of the biggest settlements, grain stores were destroyed in 1896 and growing crops were destroyed in February. Conflict in Mashonaland lasted longer and the scorched earth tactics by the Company and British forces seem, at least initially, to have been even more destructive.

I cannot give an estimate of deaths from famine and disease in Mashonaland at this point, but I am suggesting that here too, they are likely to have exceeded deaths in direct conflict and so that the overall mortality was probably similar to that in Matabeleland, at perhaps 8-9,000. In total, 15-20,000 African deaths in 1896-7 may not be unrealistic. To this should be added around 4-5,000 in 1893-5. It is likely that between 20-25,000 African people died as a result of these colonial wars. There may have been more, particularly from the famine in Matabeleland.

If the core Ndebele population is considered separately, estimated at the time to be around 100-120,000, then losses of 4-5,000 in 1893-5, and 8-9,000 in 1896-7, may amount to two successive blows of about 4 per cent and 8 per cent of the population. Who knows the cost in malnutrition resulting from the loss of so many of their cattle?  To those in the UK who may say: if there had been no rebellions there would have been fewer deaths, we must surely emphasise that the 1893 war was an invasion by Rhodes and Jameson, not a rebellion. And in relation to 1896, the colonial intrusion was sudden and demanding. What would they have expected British people to do if invaded and partly dispossessed?  Is the analogy that it would have been better to accept a German conquest in the Second World War?  To those who might argue that these numbers were not so great, it is worth recalling that the UK (including Ireland) lost about 800,000, under 2 per cent of the population, in the First World War (1914-18), and little of the fighting was on British land. Yet it was a searing experience for the society.

With respect to the Zimbabwean population as a whole, variously estimated at between 500,000 and 700,000 in the late nineteenth century, the percentage lost in these wars was perhaps closer to 3-4 overall. Conflict was concentrated in specific parts of the country, especially around the Ndebele heartland and Salisbury (Harare). As in some other parts of Africa, conquest could be deeply disruptive and was marked in places by demographic stasis, but over the longer term, the colonial period was more generally characterised by demographic growth, especially after the Second World War. This was also the case in Zimbabwe.

One justification used by the British South Africa Company for its actions was the historical brutality of the kings Mzilikazi and Lobengula in establishing a conquest state and the claim that they were saving the Shona from the Ndebele. The Company was able to win some African support in these wars, from Shona-speakers, from Botswana, from black South Africans and from a few Ndebele chiefs. But the picture is mixed. Some client chieftaincies on the peripheries of the Ndebele chiefdom, including Shona-speakers, supported the 1896 uprising and the question arises as to why many chiefs in Mashonaland also rebelled against Company rule in 1896-7 and why some sustained a dispersed resistance, at enormous cost in lives and resources, for over a year. As noted, there is a large historical literature explaining and debating this.

Perhaps H.C. Thomson (1898, 148-9), a correspondent for The Times, captured the issue most succinctly at the time when he heard that those resisting ‘preferred the Matibili rule to ours, because under them they were troubled but once a year, whereas now their troubles come with each day’s rising sun’. Though not unsympathetic with colonialism and the settlers, Thomson (1898, 10) believed that ‘Mr Rhodes’ record in Rhodesia has been written in blood, and cannot be obliterated by the assertion, however emphatic, of his adherents and himself that his motives have been disinterested and patriotic’. Ndebele raids did not cover the whole of modern-day Zimbabwe, some of the devolved chieftaincies reached accommodation with them and Beach (1986) shows in detail that the balance of power was changing as some of the Mashonaland chiefdoms acquired firearms.

Supporters of the statue also focussed on Rhodes as peacemaker in the 1896 Indabas and his generosity in spending £50,000 to distribute grain; some Ndebele, erstwhile enemies of the Company, saluted Rhodes’s funeral cortege when it passed to his grave site in the Matopos in 1902. But there were earlier opportunities to make peace and the Indabas started when Rhodes was counting the cost of the war. Although there were overtures to potential allies, he failed to seek peace with the chiefs of Mashonaland. The distribution of grain to the Ndebele in October 1896 saved some from starvation but famine preceded it and there are clear reports of acute famine after it. It did not stop sustained scorched earth strategies by Company and imperial troops in Mashonaland.

Ranger (1999) analyses in detail the relationship that developed between Rhodes and some Ndebele chiefs after their kingdom had been devastated. He apparently won some respect in the Indabas and they certainly articulated an understanding of Rhodes paternalistic authority. Rhodes allowed a limited number of conquered Ndebele people, deprived of their land, to settle on his newly acquired estate in and around the Matopos and arranged for a feast to be held. A relationship was formed and, when Rhodes died, the Ndebele were concerned that their access to this land would be curtailed. Faced with increasing settler demands, they invoked what they understood to be promises by Rhodes that they should retain their new settlements.

The forces of the British South Africa Company and the British government used extreme violence in Zimbabwe. These were brutal suppressions of people on their own land following an invasion. They were of no threat to Britain, to its empire, or to South Africa. Lobengula was seen as a threat to British interests in the neighbouring Bechuanaland protectorate, but he had little power to challenge colonial authority there. Up to 1893, he very largely protected whites in his kingdom. Lobengula, and other chiefs in Zimbabwe, were perceived by Rhodes to be a barrier to his unencumbered control of the Company’s ambitious territorial claims and to a settler colony. A more careful, less aggressive and less hubristic policy by Rhodes and Jameson may well have averted war.

Rhodes showed that it was possible to negotiate in August 1896 when he was concerned about the costs of a protracted war with the Ndebele and uncertain about the scale of the Mashonaland rising. He could have reached an accommodation earlier with the Ndebele; equally he and the imperial authorities could have ended the war against the Shona far earlier. His statements (see above) and actions at the time suggest he had no scruples about the violence used in this campaign: unbridled use was made of the Maxim gun: livestock and grain were looted on a large scale; fleeing Ndebele soldiers were shot; supposed rebels were sentenced and hung without due process of law; men, women and children in shelters in Mashonaland were blown up.

The violence of the British South Africa Company was publicised in the press and Rhodes was criticised in Britain as well as the Cape (see above). Liberal politician Henry Labouchere, who published the weekly Truth, was a persistent detractor of ‘Mr Rhodes and his pernicious company, a wretched, rotten, bankrupt set of marauderers and murderers’; Liberal MP A.C. Morton asked in parliament if the government approved ‘of this murder to 3,000, or even 500 men, for the purpose of plundering and stealing their land’ (Davidson, 1984, 236). Further rebukes were made in parliamentary debates. Labouchere was a controversial figure, critical of homosexuality, Jewish people, and women’s rights as well as the excesses of imperialism. Every detail published in Truth may not have been correct. But there was a systematic attempt in publications supportive of Rhodes and imperialism, which were in the majority, to counter what influence such critics carried.

A photograph of the three African men, hung from a tree in Bulawayo in April 1896, with a row of white onlookers behind them, was used in the first edition of Olive Schreiner’s Trooper Peter Halket, published in Britain in February 1897. It coincided with the British Parliamentary Committee that enquired into the Jameson raid. Davidson (1984, 214-5) reproduced it in his biography of Rhodes and suggested that it left ‘a profound impression’ on readers. Massie (2016, 221) quotes the Guardian of that time: it ‘is very horrible …but we cannot blame the author or publisher for giving it here. It is only through such shocks that English people can be roused to a sense of the degradation which England is suffering in South Africa’.

Over the course of the previous year, however, reporting of the war in Zimbabwe was generally favourable to Rhodes. His actions both in the field and in negotiations with the Ndebele won support in Britain (Massie, 2016, 178-9). He also emerged largely unscathed, by some accounts victorious, from the Parliamentary Committee in February 1897; before one session, the Prince of Wales shook his hand (Massie, 2016, 212). Though he lost support of key British political figures over the Jameson raid, Rhodes was protected by the weight of conservative and pro-imperial British sentiment and interests.

In 1892, an honorary doctorate was conferred on Rhodes by the University of Oxford; this preceded the most controversial episodes of his career. When he planned a visit in 1899 to receive it, there was considerable discomfort, led by the Master of Balliol, with the support of about 90 academics including the university proctors. But Kitchener, who was to receive a similar award, threatened to withdraw if Rhodes was not honoured. The Provost of Oriel, the Vice-Chancellor and the majority of staff and students came out in favour; Rhodes received ‘an uproarious welcome’ (Massie, 2016).

Rhodes must be accorded responsibility for the deaths in Zimbabwe and for the character of extreme violence. He and his Company chose to colonise Zimbabwe by force. Leaving aside the legitimacy of the Rudd concession, they could have annexed Zimbabwe on this basis, appropriated less cattle and land, and exercised authority with more care. Even Milner, the High Commissioner from 1897, and Milton, who took over as administrator of Mashonaland – both sympathetic to Rhodes – were alarmed by the practices of the British South Africa Company in the early years. There were alternative models at hand in the Bechuanaland protectorate. A different approach may not have averted conflict entirely but it may have minimised the risk of warfare.

I am very grateful to Dr Stephen Massie, who worked for two months as research assistant in helping me to prepare the historical appendix to the Oriel Commission (see the College website) on which this memorandum is based. We did the research in a short period, under the restricted conditions of the coronavirus pandemic, with very limited resources. Thanks also to Dr Ken Wilson for valuable comments and references; I benefited greatly from his deep knowledge of Zimbabwe. Neither are responsible for my discussion.

E.A.H. Alderson, With the Mounted Infantry and the Mashonaland Field Force 1896 (1898)

Jocelyn Alexander, JoAnn McGregor and Terence Ranger, Violence and Memory (2000)

Robert Baden-Powell, The Matabele Campaign (Fourth edition, 1901, Gutenberg on line edition used; first published 1897)

D. N. Beach, ‘The Rising in South-Western Mashonaland, 1896-7’, unpublished Ph.D. thesis, University College of Rhodesia (1971)

D. N. Beach, ‘Chimurenga: The Shona Rising of 1896-97’ Journal of African History, 20 (1979), 395-420

D. N. Beach, War and Politics in Zimbabwe 1840-1900 (1986)

D. N. Beach, ‘Zimbabwean Demography: Early Colonial Data’, Zambezia, (1990a) XVII, 1, 31-83

D.N. Beach, ‘First Steps in the Demographic History of Zimbabwe: The Colonial Period from 1895-1922’ in Demography from Scanty Evidence edited by Bruce Fetter (1990b)

W. Beinart and S. Dubow, The Scientific Imagination in South Africa, 1700 to the Present (2021)

Robert Blake, A History of Rhodesia (1977)

Vivian Bickford-Smith, Ethnic Pride and Racial Prejudice in Victorian Cape Town (1995)

Colin Bundy, The Rise and Fall of the South African Peasantry (1979)

Colin Bundy, ‘More than a Game: Cricket, Imperialism and South Africa’, unpublished paper from the author

Marieke Faber Clarke with Pascal Nyathi, Lozikeyi Dlodlo, Queen of the Ndebele (2010, third extended edition 2018)

Julian Cobbing, ‘The Absent Priesthood’, Journal of African History 18, 1 (1977), 61-84

Apollon Davidson, Cecil Rhodes and His Time (1984)

Evening Express, Wales, references from Ken Wilson

John S. Galbraith, Crown and Charter: The Early Years of the British South Africa Company (1974)

L. H. Gann, A History of Southern Rhodesia: Early Days to 1934 (1965)

Michelle Gordon, Extreme Violence and the ‘British Way’: Colonial Warfare in Perak, Sierra Leone and Sudan (2020)

Hugh Marshall Hole, The Making of Rhodesia (1926)

Monica Hunter (Wilson), Reaction to Conquest (1964, first edition 1936)

N. Jones, Rhodesian Genesis (1953)

Arthur Keppel-Jones, Rhodes and Rhodesia: The White Conquest of Zimbabwe, 1884-1902 (1983)

David Tyrie Laing, The Matabele Rebellion, 1896: With the Belingwe Field Force (1897)

Phyllis Lewsen (ed.), Selections from the Correspondence of John X. Merriman 1890-1898 (1963)

R.E.R. Martin, C. 8547, British Parliamentary Papers, report on the Native Administration of the British South Africa Company (1897)

Stephen Massie, ‘The Imperialism of Cecil John Rhodes: Metropolitan Perceptions of a Colonial Reputation,’ unpublished Ph.D., Oxford Brookes University (2016)

Paul Maylam, The Cult of Rhodes (2005)

J.G. McDonald, Rhodes: a Life (1927)

Sabelo Ndlovu-Gatsheni, The Ndebele Nation: Reflections on Hegemony, Memory and Historiography (2009)

Farai Nyika and Johan Fourie, ‘Black Disenfranchisement in the Cape Colony, 1887-1909’, Journal of Southern African Studies, 46, 3 (2020), 455-469

A. Odendaal, The Founders: the Origins of the ANC (2012)

A. Odendaal et al, Cricket and Conquest (2016)

I. R. Phimister, ‘Rhodes, Rhodesia and the Rand’, Journal of Southern African Studies, 1, 1 (1974)

H. Plumer, An Irregular Corps in Matabeleland (1897)

John O’Reilly, Pursuit of the King: an Evaluation of the Shangani Patrol in the Light of Sources Read by the Author (1970)

Kenneth J. Panton, Historical Dictionary of the British Empire (2015)

Brian Roberts, Cecil Rhodes: Flawed Colossus (1987)

Terence Ranger, Revolt in Southern Rhodesia, 1896-97 (1967)

Terence Ranger, Voices from the Rocks (1999)

Terence Ranger, Bulawayo Burning (2010)

Robert Rotberg, The Founder: Cecil Rhodes and the Pursuit of Power (1988)

Stanlake Samkange, What Rhodes Really Said About Africans (1982)

F. C. Selous, Sunshine and Storm in Rhodesia (1896)

Walter Stanford, The Reminiscences of Sir Walter Stanford (2 vols., 1958, 1962), edited by J. W. MacQuarrie

Vere Stent, A Personal Record of Some Incidents in the Life of Cecil Rhodes (1924; 1970 edition used)

Frank Sykes, With Plumer in Matabeleland: An Account of the Matabeleland Relief Force during the Rebellion of 1896 (1897)

Mordechai Tamarkin, Cecil Rhodes and the Cape Afrikaners (1996)

H.C. Thomson, Rhodesia and Its Government (1898)

S. Trapido, ‘White Conflict and Non-White Participation in the Politics of the Cape of Good Hope’, unpublished Ph.D, University of London (1970)

Robert Turrell, Capital and Labour on the Kimberley Diamond Fields (1987)

Lawrence Vambe, An Ill-fated People: Zimbabwe before and after Rhodes (1972)

Russel Viljoen, ‘“Smallpox War” on the Kimberley Diamond Fields during the 1880s’, Kleio, 35 (2003), 5-18

Vindex (pseud. for John Verschoyle), Cecil Rhodes: His Political Life and Speeches, 1881-1900 (1900)

Brian Willan, Sol Plaatje (2018)

Francis Wilson, Labour in the South African Gold Mines 1911-1969 (1972)

Jonty Winch, ‘J.H. Hofmeyr and the Afrikaner Bond: A challenge to William Milton’s “Englishness” and the advance of the “coloured” cricketer’. Historia, 59,1 (2014)

William Worger, South Africa’s City of Diamonds (1987)