Colonial rule brought about significant changes in the regional economies of West Africa, which had already been influenced by Europeans through the slave trade and the succeeding period of legitimate trade. Colonial administration extended the cultivation of export crops, introduced new systems of taxation and currencies, reorientated trading networks, and introduced policies aimed at ending domestic slavery. Levels of population mobility also increased, especially in the agricultural sector, as the seasonal redistribution of labour became an integral part of the production of export crops. This paper looks at the origins and development of dry season labour migration from Sokoto in north-western Nigeria, a migratory system which is generally accepted as being a product of colonialism. The argument takes the opening years of colonial rule in Sokoto and tries to uncover the nature and type of migration which obtained at this time. It is suggested that early migration was influenced by the economy of the central Sudan of the late nineteenth century, and there was an overlap, rather than a sharp break, between pre-colonial and colonial economies. The likely influence of pre-colonial structures is especially pertinent in an area which was the centre of the Sokoto caliphate, the largest and most populous state in nineteenth-century West Africa. But, in the early 1930s, dry season migration changed as the volume of migrants increased and their destination altered; it is the identification of this point of change which distinguishes the early period of migration and the changing impact of colonial policies.

The causes of migration are a major theoretical issue in migration studies. Some writers, such as Udo, see migration as involving a permanent or semi-permanent change of residence. It is usually considered to be a rational, that is an economic, man's reaction to economic differentials between regions (Udo, 1978: 124). Hill adds that in rural West Africa the propensity to migrate does not necessarily correlate at all closely with population density as demographers and economists so often assume; it may be the result of lack of lucrative non-farming occupations, because farming is not regarded as ‘work’ by some West Africans (Hill, 1978: 25–6).

There are many variables which affect nutritional status (Bennett and Stanfield, 1972: 7–8). A number of recent nutritional surveys have tried to assess the influence of polygyny on the incidence of malnutrition, though to date no clear relationship has been found between the two. This may partly reflect a methodological problem: accurate determination of a woman's marital status is not easily achieved by questionnaire, and this is especially true in view of the developmental instability of polygynous households (Muhsam, 1956: 16). Yet questionnaires are frequently used in nutrition surveys and they can generate large quantities of data, although Johnston et al. have urged that studies be simplified in the interests of discovering some underlying structure (1980: 292). As a result of the massive bulk of disparate data, multivariate analysis has come to the fore in an attempt to establish significant correlations among the many variables which can be associated with malnutrition. But, as McDowell and Hoorweg point out (1975: 98), this method requires that an adequate research design be employed: it is not good enough to collect the data first and to seek the associations afterwards. The data to be collected must be carefully chosen so that pre-selected hypotheses can be tested. An anthropological perspective is very important in determining this choice of data. For example, although the term ‘polygynous’ describes a woman's marital status, it gives no indication of her position in a household relative to at least one other wife. For this reason it would be preferable to make the polygynous household, rather than the individual woman, the unit of study. Specific information could then be collected to clarify if there are constraints or advantages experienced by different wives within a polygynous household, or between wives in polygynous and monogamous households, in order to see whether these influence the nutritional status of their children.

This paper explores the nature of ethnic identity in the small town of Guider in northern Cameroon. It focuses on the process of ethnic identity change, specifically upon the incorporation into the Fulbe (sing. Pullo) ethnic group of individuals who originate outside it. Ethnic group change of this kind is not unknown in Africa (see Little, 1951; Richards, 1954; Banton, 1957; Wallerstein, 1960; Burnham, 1972; Salamone, 1975; Nicolas, 1975; Vaughan, 1981). The data in this study contribute to a growing body of literature concerned with ethnic boundary crossing, but, I shall argue, they suggest that we re-examine certain widely held assumptions about ethnicity and the process of ethnic assimilation.

To those outside the anthropological disciplines, the Zulu are probably the best known chiefdom of southern Africa. Zulu military prowess acted as a sharp break on the process of colonial expansion in the nineteenth century and led to a series of battles which have been immortalized on celluloid and in the more popular historical accounts of the era. Despite this, the Zulu have prompted a comparatively sparse ethnography that is based largely on the work of nineteenth-century observers and is less comprehensive than the literature on some other southern African communities.