^{page 317 note 2} The Bakitara or Banyoro: John Roscoe, Cambridge, 1923, pp. 18–20 and chap, x, passim.

^{page 318 note 1} The phrase is J. J. Maquet's: he uses it to characterize the even more rigorously stratified interlacustrine Bantu kingdom of Ruanda (Le Système des relations sociales dans le Ruanda ancien, Tervuren, 1954, chap. viii).

^{page 319 note 1} As our interest here is structural rather than cultural I do not give detailed accounts of Nyoro clan histories or totemic avoidances. A number of clans, together with their ‘totems’ (which include cattle of various kinds and conditions, animals and plants of different species, and sundry human artifacts), are listed in Roscoe's book (op. cit., chap. II). There are, however, some inaccuracies in Roscoe's list, and his division of the clans into pastoral, ‘freemen's’ and serfs' clans is not acceptable. I was unable to find any evidence for Roscoe's notion of an intermediate class of freemen, and I suspect that it derives from his over-formalized interpretation of the fact, conveyed to him by interpretation into Luganda, that in Bunyoro it has always been at least theoretically possible for persons of humble origin to rise to high political status. In fact the traditional but now hardly viable distinction between the pastoral people (bahuma) and the ‘serfs’ (bairu) is not correlated with any division of the clans, but rather cuts across it: many clans have both bahuma and bairu branches. This is why in Roscoe's list a number of clans appear in more than one of his three classes.

^{page 320 note 1} In fact if a Munyoro were questioning, say, a possible son-in-law he would first ask him what his clan avoidance object is: ‘What do you avoid (zira)?’ he would say.

^{page 321 note 1} Banyoro often asked me what my totemic avoidance was, that is, what clan I belonged to. As the reply that clans were no longer recognized in England was quite unacceptable and indeed cast grave doubts on my good faith, I soon acquired honorary membership in a Nyoro clan. This in itself may be regarded as in some degree corroborative of the view that for Banyoro clans are not kinds of descent groups, however large, but rather categories into which all people everywhere are divided. Banyoro were surprised and sceptical when told that certain important Nyoro clans were not represented in Europe.

^{page 321 note 2} The royal Babito clan forms the one exception to the rule of clan exogamy: Babito can marry their clan ‘sisters’. They would not, however, marry a sister born of the same ‘mother’ (i.e. two Babito whose mothers were of the same clan could not marry) nor, as a rule, would two very close agnates marry.

^{page 322 note 1} ‘Oruganda orukaguza nk'oku ruganjana, torukaguza nk'oku ruzarwa.’

^{page 324 note 1} I should say here that though I believe that what I have written is true for the greater part of Bunyoro it is not true for every part of Bunyoro. There are some isolated settlements on the Lake Albert shore in which the residential groups are still definitely structured in terms of unilineal descent. These are, however, quite atypical of Bunyoro as a whole, and I do not describe them here.

^{page 327 note 1} The restricted rights held by a wife over her domestic utensils, &c., form a partial exception to this statement. But these rights are in a sense conditional, since she acknowledges her husband's authority and is dependent for her status on him, and she retains these rights only so long as she remains his wife.

^{page 328 note 1} The Nyoro verb is kuhhesa, and the English ‘fine’ is only a vety rough translation of it. It involves the making of a payment, traditionally of meat and beer, to the offended party. These comestibles are then jointly consumed by the parties to the dispute and by any relatives and neighbours who may have adjudicated upon it.

^{page 332 note 1} This proverb and the preceding one are quoted in an article by Davis, M. B., ‘Lunyoro Proverbs’ Uganda Journal, vol. viii, No. 3.Google Scholar

^{page 332 note 2} Classically expounded by Radcliffe-Brown, A. R. in ‘The Mother's Brother in South Africa’ in Structure and Function in Primitive Society, London, 1925, chap. 1.Google Scholar

^{page 333 note 1} I use the term ‘group’ here somewhat uncritically, as a matter of convenience. Close agnates do, in certain contexts, act and see themselves as forming groups in some sense of the term. This is especially so in the context of marriage and affinal relationships.

^{page 335 note 1} There is plenty of evidence that this is so. Thus a man announcing himself, for example, to a chief would, in order fully to identify himself, name both his own and his mother's clan. He might say, for instance, ‘Ndi Muhinda mwihwa wa Babyasi’, ‘I am a Muhinda man and “sister's son” to the Babyasi clan’. And the Nyoro author H. K. Karubanga, writing of the blood pact (mukago), says that after the ceremony each party to it ‘goes and tells his own ruganda and the ruganda of his banyinarumi’ (Buky Nibwira, Nairobi, 1949, p. 35). The point is tha here both one's own people and one's mother' people are seen as distinct unitary groups.

^{page 335 note 2} In attempting to interpret Nyoro usages I found Sol Tax's discussion of the Omaha and Crow system of cross-cousin terminology (in The Social Anthropology of North American Tribes, Chicago, 1937) particularly helpful.

^{page 338 note 1} This ‘merging’ is not thought of as occurring at the great-grandparent—great-grandchild level. A man calls his great-grandchild katarabanyuma (literally ‘that which may not pass behind one's back’), and a great-grandchild is subject to certain prohibitions in his relations with his great-grandfather, just as a son is in relation to his father. Thus, in conformity with Radcliffe-Brown's principle, ego is ‘merged’ with his grandfather, ego's father with his grandfather.

^{page 339 note 1} Male Banyoro often refer to the preparation (where a relationship to a woman is concerned) or sharing (where a relationship to a man is concerned) of food when they wish to indicate that there is, or should be, close and friendly attachment between kinsfolk.