As this article was revised for publication in mid-1974, no account has been taken of a number of important publications on African, or West African, slavery which have since appeared.

¹ See Hill, Polly, Rural Hausa: A Village and a Setting (Cambridge, 1972) for background information on rural Hausaland and a map of the region.CrossRefGoogle Scholar

² See note 24.

³ Among the most notable historical sources are: Clapperton, H., Journal of a Second Expedition into the Interior of Africa from the Bight of Benin to Soccatoo (1829, reprinted London 1966),CrossRefGoogle Scholar which provides more detail about farm-slavery in Hausaland (Sokoto) than any other source save Baba of Karo (listed below); Barth, H., Travels and Discoveries in North and Central Africa, 1849–1855 (London, 1857, American ed. 1859; reprinted London, 1965);Google Scholar three works by Robinson, C. H., Hausaland (London, 1896)Google Scholar, “The Slave Trade in the West African Hinterland,” The Contemporary Review, (05, 1898),Google Scholar and Nigeria: Our Latest Protectorate (London, 1900; reprinted New York, 1969)Google Scholar, who describes the slave markets at Kano and Zaria, while hardly touching on farm-slavery; Nachtigal, G., Sahara und Sudan, Ergebnisse sechs jahriger Reisen in Afrika (Berlin and Leipzig, 3 vols 1879, 1881, 1889),Google Scholar whose travels were mainly outside Hausa- land; P. Staudinger, lm Herzen der Haussaldnder (Berlin, 1889),Google Scholar who exaggerates the incidence of rural slavery; Smith, Mary, Baba of Karo: A Woman of the Muslim Hausa (London, 1954),Google Scholar the recorded autobiography of a woman of rural Zaria (born 1885–8) who had vivid childhood memories of farm-slavery; and Lugard's, F. D.Memorandum No. 22, 1906,Google Scholar see note 24. Recent publications include: Smith, M. G., The Economy of Hausa Communities of Zaria (London, 1955);Google ScholarA. G. B., and Fisher, H. J., Slavery and Muslim Society in Africa: The Institution in Saharan and Sudanic Africa and the Trans-Saharan Trade (London, 1970);Google Scholar and Myers, A., “Slavery in the Hausa-Fulani Emirates,” in Aspects of West African Islam, McCall, D. F. and Bennett, N. R., eds. (Boston University Papers on Africa, V, 1971).Google Scholar

⁴ The last section of this article mainly relates to the fieldwork concerned with the transitional period. Other field material is incorporated in this section.

⁵ It must be remembered that “as much in the local African Muslim tradition as in the orthodox Muslim heritage” slavery had been “a very deeply entrenched institution, accepted alike by African believers and learned lawyers as a matter of course.” Fisher, , op. cit., p. 5.Google Scholar “No serious argument was ever put forward for the abolition of slavery in ancient Greece and Rome … on moral or any other grounds; this was also the case in India, China and the Islamic world.” Finley, M. I., "Slavery" in International Encyclopedia of the Social Sciences (1968).Google Scholar

⁶ Barth, , op. cit., Vol. II, pp. 143–4,Google Scholar thought that the numbers of slaves and free people in “Kano province” might have been about equal, but his estimates should not be taken seriously since he believed that slaves were “very rarely allowed to marry.”

⁷ Robinson, , op. cit., 1896, p. 127,Google Scholar estimated that about a third of the Hausa- speaking population was slaves; since he spent only five months in Hausaland, mainly in Zaria and Kano cities, his estimate (though constantly quoted) is worthless.

⁸ They were notably high in parts of Sokoto, as well as in Bauchi (scarcely Hausaland proper and an important slave-source). Smith, M. G. has rather recklessly stated that the numbers of slaves and free persons were probably about equal in Zaria Emirate: “The Hausa System of Social Status,” Africa (07, 1959) p. 242.Google Scholar

⁹ This figure (which may be too low) is much higher than early colonial estimates, which were based on most defective enumerations for tax purposes.

¹⁰ While there were large numbers of official and domestic slaves in the few big cities, it is likely that there were few non-farm slaves elsewhere.

¹¹ Finley, M. I., “Was Greek Civilization based on Slave Labour?” in Slavery in Classical Antiquity, Finley, M. L., ed. (Cambridge, 1960) p. 60.Google Scholar

¹² The author is grateful to Professor M. I. Finley for this and other information.

¹³ F. baiwa, pl. bayi, bauta being slavery. The usual term for slave owner was ubangiji, defined by the renowned lexicographer G. P. Bargery as “the owner of a slave or any other object.”

¹⁴ As will become apparent, any words involving notions of “serfdom” or “labouring” would be even less acceptable.

¹⁵ Encyclopedia, , op. cit., p. 307.Google Scholar

¹⁶ Just before the British arrived “a distinction was still being made between Unbelievers and Muslims in the matter of enslavement. It cannot be shown conclu- sively that illegitimate raids were not carried out by the emirates but the law which frowns on this was by no means a dead letter.” Adeleye, R. A., Power and Diplomacy in Northern Nigeria, 1804-1906 (London; 1971).Google Scholar However, Smith, M. G. reports (1955, op. cit., p. 102)Google Scholar that his informants in Zaria Emirate agreed that slave raiders made little distinction between pagan and Muslim.

¹⁷ Although a few travellers, notably C. H. Robinson, vividly depicted the scene in slave markets, there is a dearth of information on slave dealers. Alhaji A'ala of Kazaure (note 48) said that those dealers he had known were mainly "rich Fulani" (both farmers and pastoralists), and that there were men known as 'yan tugu who bought slaves on credit in Kano city (some sixty miles away), reselling them from their houses. Far more slaves were sold from houses than in open market places.

¹⁸ Thus Hunwick, J. O. in “Songhay, Bornu and Hausaland in the sixteenth century“ in History of West Africa, Vol. I, Ajayi, J. F. A. and Crowder, M., eds. (London, 1971) states (p. 216)Google Scholar that “enormous numbers of slaves had been obtained from the southlands” by Kano in the fifteenth century.

¹⁹ As this article is concerned with the institution of farm-slavery as such (not with the supply of farm-slaves), as present-day informants speak very vaguely about slave origin (perhaps denoting all “pagans” as Gwarawa, an ethnic term), as many settled farm-slaves were unaware of their origin and as no writer has yet summarised the fragmentary material relating to slave raiding and the geographical and ethnic origin of slaves, so no attempt is made to deal with ethnic origins here. See Smith, Mary, op. cit.,Google Scholar for a vivid account of slave raiding in Zaria Emirate. The purpose of much raiding was the need to acquire tribute slaves for payment to Sokoto.

²⁰ Finley, , op. cit., p. 308.Google Scholar

²¹ Under Muslim law the number of wives was limited to four, but a man might have any number of slave concubines.

²² This (somewhat notional) formula was provided by informants and appears in the literature, e.g., in Lugard's, Memorandum No. 22, op. cit., p. 306.Google Scholar

²³ The number of courts in relation to the population was small. Moreover, many important matters were (and are) customarily settled outside courts, both for convenience and to avoid court charges.

²⁴ For the legal position see Ruxton, F. H., Maliki Law (London, 1916),Google Scholar published by order of Lugard. The practical working of the ransom system was out-lined by informants and is dealt with in Lugard's various publications, notably his Political Memoranda, 1906. Lugard's Memorandum No. 22, “The condition of slaves and the native law regarding slavery in Northern Nigeria,” was based on material provided by Residents in reply to circular questions and, though a valuable source, so far as is known has not been cited previously. The importance of the 1906 edition of Political Memoranda, which was a confidential document and is consequently very scarce, has been recently emphasised by Kirk-Greene, A. H. M. in his introduction to the third edition (London, 1970)Google Scholar of the second revised version (1919) of the memoranda, which excluded the old Memorandum No. 22 altogether. The full title of both works, which are always referred to as Political Memoranda, is Instructions to Political and other Officers, on subjects chiefly Political and Administrative; the 1906 version was printed by Waterlow,Google Scholar and Lugard originally sought to prevent the Colonial Office from seeing a copy.

²⁵ Oral information.

²⁶ Clapperton, (op. cit., 1829, p. 213)Google Scholar noted that slaves in Sokoto were allowed to enclose a part of their master's land for their own use, being free to dispose of the grain as they wished. See also Memorandum No. 22.

²⁷ See Hill, , op cit., pp. 240–1,Google Scholar and below.

²⁸ Slaves could not be prayer-leaders (imams) or judges (alkalai), nor were any occupations reserved for them; but, since slave owners were enjoined to provide them with Qur'anic instruction, they might become very learned.

²⁹ Memorandum No. 22, p. 298.Google Scholar

³⁰ Known as sarkin gandu (literally, “head of gandu”).

³¹ A word which referred to the farmland itself as well as to the working relationship between the slave owner (or father) and his slaves (or sons). See Hill, , op. cit., ch. IIIGoogle Scholar “Fathers and sons in gandu.”

³² See Clapperton, , op cit., p. 213,Google Scholar and Memorandum No. 22.

³³ There is much detail on varying practices in Memorandum No. 22.

³⁴ A. G. Hopkins' unjustified generalisation that West African slaves were mainly used “to provide foodstuffs for leading state officials, for their immediate circle of dependants, and for the army” is altogether inapplicable to Hausaland. Hopkins, A. G., An Economic History of West Africa (London, 1973), p. 24.Google Scholar

³⁵ Hausaland has long comprised a number of emirates, of varying sizes, owing a common allegiance to the Caliph (Sultan) of Sokoto. The emirs live in capital towns (birni, pi. birane).

³⁶ In certain emirates, notably Kano (with its population of several million), most District Heads (hakimai), the highest level of “chief” under the Emir, lived in the capital in precolonial times. See Hill, Polly, “The relationship between city and countryside in Kano Emirate in 1900,” West African Journal of Sociology and Political Science (10, 1975).Google Scholar

³⁷ See note 140.

³⁸ See Hill, , op. cit.Google Scholar

³⁹ As noted long ago by Clapperton, , op. cit.Google Scholar

⁴⁰ It was the failure to realise this which made Lugard and his successors so scared of the possibility that wholesale manumission would result in large-scale prostitution: that “the great cities would have been filled with vagrants, criminals and prostitutes” (Memorandum VI, 1919, p. 223)Google Scholar was a constant refrain.

⁴¹ Op. cit., 1968, p. 310.Google Scholar

⁴² Present-day informants constantly emphasised this matter, dealt with by Lugard, in Memorandum VI on “Slavery (Forced Labour, &c)” (1970), p. 247,Google Scholar a revised version of his Memorandum No. 6 of 1906.

⁴³ See note 3.

⁴⁴ Op. cit., p. 300.Google Scholar Thus from Kano it was reported that “no serf [i.e., slave] or child of a serf, ever enjoys immunity from sale, however long in the services of his master”; and from Zaria informants asserted “most positively” that masters had “an invariable and inalienable right to sell any slave at any time, even those born in his house.”

⁴⁵ The law recognised no distinctions between slaves based on generation or any other attribute.

⁴⁶ See, for instance, 1960, op. cit., p. 253.Google Scholar

⁴⁷ Apart from the Fulani ruling class, which had assumed a position of political dominance over the Hausa following the holy war (jihad) called in 1804, and which has effectively lost its Fulani identity, owing to intermarriage with Hausa, adoption of the Hausa language and so forth, in many country districts in Hausaland, including northern Zaria in particular, there are large numbers of sedentary and semi-sedentary farmers and pastoralists of pure Fulani stock who live in separate communities. Owing to definitional difficulties in census-taking and to the prestige supposed to be derived from dubbing oneself Fulani, the geographical distribution of these sedentary and semi-sedentary Fulani is not accurately known. This leads many writers to denote the rural population of Hausaland as Hausa-Fulani, rather than Hausa, though this is rather misleading as in many sections of rural Hausaland there are no separate Fulani communities, and local institutions are culturally Hausa.

⁴⁸ Thus Alhaji A'ala of Kazaure had no knowledge of the word, although he himself had been a slave owner: in 1971 he talked very freely about slavery and was among those who emphatically insisted that cucanawa could be sold like anyone else.

⁴⁹ This was the invariable usage of the author's informants in Kano Emirate.

⁵⁰ Stenning, D. J., Savannah Nomads: A study of the Wodaabe Pastoral Fulani of Western Bornu Province, Northern Region, Nigeria (London, 1959), p. 66.Google Scholar

⁵¹ In Fulani rimda means “to be free” and rimdina is “to set free”; the usual Fulani word for slave is macudo, pl. macube.

⁵² See, e.g., Memorandum No. 22, 1906, op. cit., p. 302.Google Scholar It was reported that in Bauchi “many farm slaves became rich and owned many slaves of their own.”

⁵³ For cowry-storage, slaves (unlike free men) sometimes used huge funnel-shaped pots (kakakin kasa), perhaps some 5 to 6 feet high, which they buried in the ground for safety.

⁵⁴ In the present writer's experience there is no economic subject with legal aspects (not even usury) which informants are less inclined to discuss than inheritance—for practice usually departs from legal precepts. The notably free informant Alhaji A'ala (note 48) was therefore probably correct in stating that in actual practice slaves' property passed to their owners only when they had no children and that a rich slave's son might have been able to marry a free woman after inheriting property from his father.

⁵⁵ See Ruxton, , op. cit., p. 357.Google Scholar

⁵⁶ See Smith, Mary, op. cit., p. 39,Google Scholar for an account of slaves' fleeing on death: “The family was wailing because he was dead and the slaves saw their chance and ran away.”

⁵⁷ See Hill, , op. cit., p. 270, et seq.Google Scholar

⁵⁸ Although fraternal gandu was formerly quite common (see Hill, , op. cit., 1972),Google Scholar so that the land was often not divided, it is probable that slaves (unlike land) were so valuable that they were apt to be shared among the heirs.

⁵⁹ Farm maps and other information show that country people seldom follow the letter of the law when dividing property on death, although they are reluctant to admit this. See Hill, , op. cit., 1972, pp. 270et seq.Google Scholar

⁶⁰ Informants stated that only large markets, of which there were few, had slave sections.

⁶¹ Finley, , op. cit., 1960, pp. 70.Google Scholar

⁶² “In slave societies hired labour was rare and slave labour the rule, whenever an enterprise was too big for a family to conduct it unaided.” Finley, , op. cit., 1968, p. 310.Google Scholar As the institution of farm labouring was usually nonexistent in West Africa it is absurd to suggest that “The use of slave rather wage labour was a matter of deliberate choice on the part of African employers.” Hopkins, , op. cit., p. 24.Google Scholar

⁶³ As that most perceptive observer E. D. Morel noted: “Northern Nigeria … is a remarkably self-sufficing country, one part of it supplying the wants of another.” Nigeria: Its Peoples and its Problems (London, 1911, reprinted London 1968), p. 171.Google Scholar It is in any case a fallacy to suppose that slave holding was necessarily based on “production for export,” however that term might have been defined.

⁶⁴ Many countrymen joined long-distance donkey caravans and the more prosperous among them were especially apt to be slave owners. See Hill, , op. cit., 1972, pp. 243et seq.Google Scholar

⁶⁵ This conclusion derives partly from fieldwork and published sources (which are, however, seldom explicit on this matter), but also from the belief that inequality of wealth was even more pronounced then than it is today (see below).

⁶⁸ See McLoughlin, P. F. M., “Economic Development and the Heritage of Slavery in the Sudan Republic,” Africa (Oct., 1962).Google Scholar

⁶⁷ Smith, , op. cit., 1955, pp. 102–3.Google Scholar

⁶⁸ After his victory the first Fulani Emir of Zaria “allocated lands to the Fulani kin-groups who had helped him. … On these lands were built the rumada or rinjoji (sing, rinji. slave villages) of the families concerned …, ” Smith, ibid., p. 13. There is no Hausa word for a slave village, and it is possible that rinji and rumada derive from the Fulani word rumde, which is a wet season slave-farming village, ruma meaning “to pass the rains.” See also note 113. Rinji is now often employed by Hausa speakers to denote any large house in which many co-slaves resided.

⁶⁹ Lugard's Memorandum No. 6. 1906, p. 150. While emirs were permitted to continue paying tribute to their overlord, the Caliph at Sokoto, it was stated that this should no longer include slaves.

⁷⁰ Memorandum, VI, 1919, p. 221.Google Scholar See also Lugard, Lord, The Dual Mandate in British Tropical Africa (London, 1922, reprinted London, 1965).Google Scholar In Memorandum No. 6, 1906, Lugard had expressed the view (p. 140) that as the institution of slavery would die a natural death in the.course “of a limited number of years,” it was “the policy of Government not to hasten that event unduly, and to secure to the Mohammedan gentry the continued service of their existing household slaves during the period of transition.” Especially in the early stages, the colonial administration always tended to assume (though Memorandum No. 22, 1906 provided contrary evidence) that most slave owners were, in some undefined sense, “upper class”: “slave-owners belonged for the most part to the ore intelligent and influential class of native.” C. W. J., (Sir Charles) Orr, , The Making of Northern Nigeria (London, 1911, reprinted London, 1965), p. 203.Google Scholar

⁷¹ This term included both household and farm slaves. Lugard considered a policy of gradualness especially justified with the latter.

⁷² Memorandum No. 6. 1906, p. 136.Google Scholar

⁷³ This remarkable understatement was made in Memorandum VI, 1919, p. 221.Google Scholar

⁷⁴ Memorandum VIII, 1919, p. 279.Google Scholar

⁷⁵ Ibid., p. 142.

⁷⁶ Memorandum No. 6, 1906, p. 156.Google Scholar

⁷⁷ Ibid., p. 142.

⁷⁸ Ibid., p. 142.

⁷⁹ Memorandum X, 1919.Google Scholar

⁸⁰ Memorandum No. 6, 1906, p. 136.Google Scholar

⁸¹ Memorandum X, 1919, p. 356.Google Scholar

⁸² Lugard's exposition in Memorandum X (Ibid.) is very muddled on this point.

⁸³ Ibid., p. 356.

⁸⁴ Ibid., p. 356.

⁸⁵ Orr, , op. cit., 1911, p. 202. Ch. XI of this book is an apologia for slavery based on Lugard's arguments; it describes the institution as “moribund” (p. 206).Google Scholar

⁸⁶ Memorandum VI, 1919, p. 224.Google Scholar

⁸⁷ Ibid., p. 241.

⁸⁸ Ibid., p. 224.

⁸⁹ Ibid., pp. 220-21.

⁹⁰ Lugard, , op. cit., 1922.Google Scholar

⁹¹ Ibid., p. 368, fn.

⁹² Colonial Reports—Annual: Northern Nigeria (London, HMSO, Cd. 2238).Google Scholar

⁹³ Ibid. (Cd. 2684).

⁹⁴ Ibid. (Cd. 3729).

⁹⁵ These are deposited in the Nigerian National Archives at Kaduna.

⁹⁶ One bold author (Arthur Festing) of an Annual Report on Zaria Province for 1908 touched on the subject of farm-slavery “with great diffidence,” stating that the wisdom of “encouraging such wholesale freeing of slaves” was open to criticism because of the stagnation of agriculture and of “the labour market.” National Archives, Kaduna, SNP/7/819/1909.

⁹⁷ In a rare reference to taxation of farm-slaves, Lugard made the absurd sug- gestion that this should consist of sums deducted from the dues he wrongly supposed they paid to their owners. Colonial Reports-Annual: Northern Nigeria. Report for 1902 (London, HMSO, Cd. 1768).Google Scholar

⁹⁸ There is no means of telling whether official population figures, based on enumerations for tax purposes, included slaves or not.

⁹⁹ Working during World War II, C. K. Meek and Meyer Fortes were the first anthropologists to make systematic reference to Northern Nigerian departmental material of a socioeconomic type. See Land Law and Custom in the Colonies by Meek, C. K. (London, 1949)Google Scholar and Daryll Forde's contribution to The Native Economies of Nigeria, Vol. I, Perham, M., ed. (London, 1946), which incorporates Fortes' material.Google Scholar

¹⁰⁰ Memorandum X, p. 344.Google Scholar

¹⁰¹ Ibid.

¹⁰² Ibid., p. 351.

¹⁰³ As there had been no sense in which those hakimai who lived at Emirate capitals had owned the land in their Districts, so they could not have allocated it; nor had these officials demanded labour-services from free men on their own estates, which were not necessarily situated in their Districts and had been farmed by slaves.

¹⁰⁴ Thus Lugard in 1919: “The country is a mass of peasant occupiers, who, however, cannot alienate their holdings,” Memorandum X, p. 356. See note 27 above.Google Scholar

¹⁰⁵ Lugard most curiously identified “peasants” with ex-slaves. “Nothing would so effectively tend to emancipate the peasant class from the servile attitude of mind which long generations of slavery had induced in them, or better promote a sense of individual responsibility, than to become proprietors of their own fields.” Ibid., p. 346.

¹⁰⁶ McPhee described the 1910 Ordinance as a measure “riddled with imperfections”: McPhee, A. M., The Economic Revolution in British West Africa (London, 1926, 2nd edition, 1971), p. 180.Google Scholar The notion that all tenures were leaseholds “achieved nothing owing to the lack of an adequate survey” (Ibid., p. 181).

¹⁰⁷ The Hausa land-borrowing system, known as aro, involved the loan of land, at a nominal rent, for one season only, the parties being local farmers who were usually kin.

¹⁰⁸ Smith, , op. cit., 1960, p. 260.Google Scholar

¹⁰⁹ Ibid., p. 258.

¹¹⁰ As elsewhere in northern Nigeria, for instance Nupe. See Mason, M., “Captive and Client Labour and the Economy of the Bida Emirate: 1857-1901,” Journal of African History, 3 (1973).Google Scholar

¹¹¹ The existence of such a rent was reported by Cole, C. W., Report on Land Tenure, Zaria Province (Kaduna, 1949).Google Scholar

¹¹² See Smith, , op. cit., 1955, p. 105.Google Scholar

¹¹³ Interestingly, it was reported to Lugard that in certain parts of Zaria Province there were “whole towns peopled entirely by ex-slaves freed by the Fulani.” Memorandum No. 22, 1906, p. 308.Google Scholar See Dupire, M.Organisation sociale des Peul (Paris, 1970), pp. 430, 446,Google Scholar for references to slave villages (runde) attached to pastoral Fulani (Peul).

¹¹⁴ See Cole, , op. cit.Google Scholar Smith rightly observed that the British (including Cole) found “the logic behind the system of land tenure in Zaria somewhat obscure and puzzling.” Smith, , op. cit., 1960, p. 223.Google Scholar

¹¹⁵ Smith, , op. cit., 1955, p. 104.Google Scholar

¹¹⁶ “Gifts in kind periodically made as the declaration of allegiance. …” Smith, Ibid., p. 104.

¹¹⁷ I am grateful to Mr. A. H. M. Kirk-Greene for the information that galla derives from the Fulani word ngalla which is defined in F. W. Taylor's Fulani- English dictionary (1932) as “usufruct of a slave” (sing.), “revenue accruing from land” (pl.).

¹¹⁸ These enquiries were a byproduct of research mainly concerned with the socioeconomic organisation of farmers in present-day Hausaland, with special reference to economic inequality.

¹¹⁹ Arbitrarily denoted as Dorayi, the area is situated about three-quarters to three miles south of the ancient wall of Kano city, close to Abdullahi Bayero College; it includes all the houses of those who pay tax in the following five Village or Hamlet Areas, the latter being administrative subdivisions of the former: Dorayi Karama (V.A.), Ja'en Yamma (V.A.), Cikin Gari and Tudun Mazadu (both H.A.s within Ciranci V.A.), and Cikin Gari (H.A. within Gwazaye V.A.).

¹²⁰ See Hill, Polly, “Big Houses in Kano Emirate,” Africa (April, 1974)Google Scholar, for background material on Dorayi. This densely populated zone is entirely atypical: as aerial photographs and other information show, most Hausa farmers have free access to uncultivated land, as they do not in Dorayi.

¹²¹ In an area of dispersed settlement where Hamlet boundaries are functionless this is necessarily an arbitrary figure.

¹²² In 1972 about an eighth of all the farmland was owned by such farmers, some of whom had migrated from Dorayi to Kano city, retaining their farmland.

¹²³ Reticence about slave origin and ignorance (especially on the part of younger men) were among the factors that would have made it altogether impossible to collect this information from slave descendants themselves. I am profoundly grateful to Mahamman Lawal, for some 35 years the Hamlet Head of Cikin Gari, Ciranci, who provided most of the information on farm-slavery. Born about 1894, he died in 1974.

¹²⁴ Since wives remove to their husband's house on marriage, since many men are married to women from rural areas outside Dorayi, and since divorce is common, there is a continuous flow of women in and out of Dorayi, most of the continuity being provided by men.

¹²⁵ The whole area was mapped from an up-to-date aerial photograph on which farm boundaries could be clearly seen, and ownership of each plot was ascertained.

¹²⁶ No cadastral survey exists, and the only statistics of any practical value derive from the current tax-lists, earlier lists not having been preserved.

¹²⁷ By Fogel, R. and Engerman, S. (Boston, Mass., 1974).Google Scholar

¹²⁸ Dorayi is in Kumbotso District which, because of deaths, transfers, dismissals and retirements, has had as many as twelve District Heads since its creation in 1908.

¹²⁹ while there were certainly a few others, they have left no descendants in Dorayi.

¹³⁰ The slave owner, together with his slaves and their dependants, would have withdrawn there for safety in the event of war.

¹³¹ Dagatai (sing, dagaci).

¹³² Then as now, the high population density of Dorayi would not have been sustained had not huge quantities of organic household refuse been brought in, as manure, from Kano city.

¹³³ As most female slaves were married to male slaves, the impossibility of tracing all the former's male descendants does not affect the general conclusions.

¹³⁴ He also owned numerous other estates in the Emirate.

¹³⁵ Unlike most of the slaves owned by private farmers, many of the Emir's slaves had probably been captured recently, in raids in other Emirates. To distinguish them from longer-established slaves they were referred to as baibayi (lit. “pagans”); many of them could not speak Hausa—“they spoke by gesture as if they were deaf.”

¹³⁶ The private plots given to sons-in-gandu are still called gayauni today. See Hill, , op. cit., 1972.Google Scholar

¹³⁷ The sale of cooked meals is very common in Hausaland.

¹³⁸ They are no more apt to be of slave origin than anyone else.

¹³⁹ See Paden, J. N., Religion and Political Culture in Kano (University of California, 1973), pp. 244–7, for a list of Districts, many of which were created by the British.Google Scholar

¹⁴⁰ Such selling often occurred over several decades. One of the few Assessment Reports to mention farm slavery, but in retrospect in 1932, notes that farms on the main route from Kano to Zaria (which ran near the western boundary of Dorayi) had been prized by “wealthy merchants” who owned farm-slaves: “The abolition of slavery threw these estates … out of cultivation. The original owners are dead. … Yet the estates still remain, owned by aged widows and small traders who exist in the ruined mansions of their husbands and fathers which line the road to Goron Dutse [the westerly of the two Kano hills].” National Archives Kaduna, Kano Pro 5/24, No. 764. None of these estates now exists.

¹⁴¹ This is known because a list of all the absentee owners of mapped farms was compiled.

¹⁴² His father had been an imam in the city.

¹⁴³ Beza is a type of salt. This woman's former land is still spoken of as Gandun Mai Beza.

¹⁴⁴ One of these men (Babali dan Mai Kanwa, or Babali son of the natron trader), who bought cattle and slaves in Adamawa, ultimately received a prison sentence for slave-dealing, whereat many of his slaves departed.

¹⁴⁵ All statistics concerning the incidence of slave descent relate to married men and to descendants in the male line.

¹⁴⁶ This matter is of course entirely distinct from concubinage, the incidence of which is unknown.

¹⁴⁷ Hill, , op. cit., 1974.Google Scholar The existence of such houses among the Muslim Hausa, as distinct from the pagan Maguzawa (also Hausa-speaking people), had not hitherto been recorded.

¹⁴⁸ See note 119.

¹⁴⁹ See Hill, ibid.

¹⁵⁰ In recent decades Dorayi men of whatever origin have seldom migrated during their fathers' lifetimes.

¹⁵¹ This came as no surprise as even in a Katsina village (see Hill, , op. cit., 1972) where there were ample supplies of uncultivated land, a significant proportion of men were found to be landless.Google Scholar

¹⁵² Aro is a most important “safety valve” in overcrowded Dorayi where it involves acquaintances, not only close kin as is usual in Hausaland; any reliable man can obtain an aro plot.

¹⁵³ Land is so scarce that few fathers give gayauni to their sons, who in this respect are in a worse plight than the slaves had ever been.

¹⁵⁴ Farmers reported as owning significant acreages outside Dorayi were omitted.

¹⁵⁵ See Hill, , op. cit., 1972, for the methods of classifying married men according to their living standards; similar, though not identical, methods were followed in Dorayi. The proportions of “rich men” (masu arziki) but were 6% and 10% respectively for those of slave and free descent in the two subdivisions where most of the former lived.Google Scholar

¹⁵⁶ See Hill, ibid., for a detailed examination of the type of economic inequality usually prevailing in rural Hausaland, the existence of which is entirely consistent with a lack of “class stratification”—though conditions are rapidly changing in Dorayi owing to the rising price of land.

¹⁵⁷ Thus the 63 married men (excluding married sons) of slave descent, not considered poverty stricken, were considered better off than 84 impoverished men of free descent. Incidences of severe impoverishment for the two groups were 37% (slave descent) and 23% (free descent).

¹⁵⁸ In the northern Katsina village (Batagarawa) described in Hill, , op. cit., 1972 where land was not scarce, there appeared to be no significant relationship between standard of living and descent—whether free or slave.Google Scholar

¹⁵⁹ Whereas the ratio of married men aged about 55 and older who are “rich” is 22%, the corresponding ratio for the whole population of free descent is 15%. It is interesting that the age-structures of the populations of free and slave descent are quite similar, tending to confirm that the two groups have roughly equal rates of outward migration.

¹⁶⁰ prices today, though they are very variable, may be of the order of £60 an acre compared with say £5 to £ 10 an acre twenty years ago. In rural Hausaland generally prices are still very low: see Hill, , op. cit., 1972, pp. 239–40.Google Scholar

¹⁶¹ Communal labour systems, known as gayya (Ibid., p. 251), usually provided assistance on special occasions only, or involved special categories of people, such as affines or old men.

¹⁶² Ibid.

¹⁶³ Most of the vast numbers of labourers employed on the cocoa farms of Ghana are strangers, mainly from far away.

¹⁶⁴ See above.

¹⁶⁵ This is in contrast to former times, when barori (said to have been more numerous than today) were usually local people; such men were certainly not “serfs” but personal retainers or farm servants. It was reported that when their elderly masters became “retired farmers” they likewise retired. See ibid., p. 176.

¹⁶⁶ Age and wealth are the main determinants of the type of occupation pursued.

¹⁶⁷ The Village Heads resemble the Hamlet Heads in being local people. See also, note 131.

¹⁶⁸ See Hill, ibid. It is interesting that in Batagarawa (note 158) the question of slave origin appears to have had no significant influence on the choice of marriage partner.

¹⁶⁹ Nor in rural areas does the type of non-farming occupation determine a man's status, as is so commonly supposed, though there are small numbers of occupations, such as well-digging and butchering, which are considered inferior. This lack of hierarchicisation is, of course, partly associated with the fact that slaves and free men had pursued similar occupations. Ibid.

¹⁷⁰ There are many peripheral regions for which there is virtually no information on the socioeconomic organisation of rural economies.

¹⁷¹ Ibid., for references to this research.