¹ The Tswana, London, 1953, 41–2.

² London, 1938; especially 135–47, with brief additional notes in the 2nd ed., London, 1955, xi, xxxi.

^{page 113 note 1} At the time of investigation, the number of registered taxpayers in each tribe was (approximately): Kgatla, 4400; Kwena, 8200; Ngwaketse, 9200; Tlokwa, 440. On average about 55 per cent of them were married.

^{page 113 note 2} Cf. Schapera, Tribal innovators: Tswana chiefs and social change, 1795–1940 London, 1970, 137.

^{page 113 note 3} The data summarized below is given more fully in Schapera 1970: 138–9, 142, 173–4, 193, 207–8, Si8, 246.

^{page 114 note 1} Both here, and in later Tables, T is the total number of cases investigated, and N is the number of wives (or, where relevant, husbands) for whom bogadi was given (or who gave bogadi).

^{page 114 note 2} It was also possible for people to marry by “civil” rites in a magistrate's office. But, in the groups studied, only one couple had done so (among the Ngwaketse). I include them in the “Xtn” category.

^{page 115 note 1} Among the Kgatla most men, in a casual sample of 47, were 20–25 years old when their age-set was created; the median age was 22. I do not have comparable data for other tribes.

^{page 115 note 2} Some men who had married more than once did not give bogadi for each wife; but even if they had given it for only one I have included them among the givers. The Table, it should be added, focuses upon husbands, and not (as in Tables I and II) upon total numbers of marriages.

^{page 116 note 1} An examination of church marriage registers among the Kgatla showed that, of 95 men whose ages were known precisely, the great majority (72) had married at the ages of 24–30, with 28 as the median. Only four were less than 24 years old when they married. The median age of inititation, as already noted (see p. 115), was 22.

In each of our four tribes, it is worth adding, a new age-set was created in 1922. Data I obtained in 1943 (21 years later), during an extensive inquiry into the extent of labour migration, showed that the following numbers of men belonging to that particular age-set were still unmarried: Kgatla, 13 out of 75 (= 17%); Kwena, 6 out of 47 (= 13%); Ngwaketse, 8 out of 65 (= 12%); Tlokwa, 14 out of 69 (= 20%).

^{page 118 note 1} London, 1938, 143–5.

^{page 118 note 2} Cf. Schapera, Tribal innovators, London, 1970, 139, 199.

^{page 118 note 3} Cf. Schapera, A handbook of Tswana law and custom, London, 1938, 133, 149.

^{page 119 note 1} Schapera, The political annals of a Tswana tribe, Capetown, 1947, 106.

^{page 121 note 1} 4 h/c & 2 sheep were given in each of eight marriages; 6 h/c & 2 sheep in each of four others; and, in the remaining two, 2 h/c & 2 sheep, and 10 h/c & 4 sheep, respectively.

^{page 121 note 2} In 5 of the Manare marriages, both cattle and sheep were given. In the court cases, cattle & sheep were given in 6 instances, and sheep alone in 3 others. In all other instances only cattle were given.

^{page 122 note 1} The payment offered was 1 heifer and 4 sheep; the payment accepted was 3 heifers. This suggests that here, at least, a sheep did not equal a bovine for bogadi purposes.

^{page 122 note 2} An “animal unit” was defined as “one bovine, one donkey, or five sheep and goats”.

^{page 122 note 3} Cf. also the case-histories of Maribe Mokotedi, Thebe Sejwe, and Malefshane Phiri (p. 117).

^{page 122 note 4} Schapera, Native land tenure in the Bechuanaland Protectorate, Alice, South Africa, 1943, 220.

^{page 122 note 5} The census was confined to people actually possessing livestock, and therefore omitted all those who did not; and the figures given here were specially singled out by me because they related to men of whose bogadi payments (or default thereof) I had definite information. The “animals” excluded small stock, which Tlokwa did not give as bogadi.

^{page 124 note 1} Cf. Schapera, The Tswana, London, 1953, 10.