² West-central Africa has been reviewed in this journal by P. de Maret (xviii, 3, 1977); West Africa by D. Calvocoressi and N. David (xx, i, 1979); Northern Africa by A. Close (xxi, 2, 1980); and Southern Africa by M. Hall and J. C. Vogel (xxi, 4, 1980).

³ Earlier reviews of these areas were by B. M. Fagan (x, i, 1969), J. E. G. Sutton (xiii, 2, 1972), R. C. Soper (xv, 2, 1974) and T. Maggs (xviii, 2, 1977). In the present article, brief but important reference is made to Ethiopia, in connection with work in northern Kenya (see p. 439), but the present writer was unable to obtain any information about Uganda.

⁴ For a discussion of such terms as ‘Earlier Stone Age’, ‘Middle Stone Age’ and ‘Late Stone Age’, see Kleindienst, M. R., ‘Questions of Terminology in regard to the study of Stone Age Industries in Eastern Africa: “Cultural Stratigraphic Units”’, in Bishop, W. W. and Clark, J. D. (eds.), Background to Evolution in Africa (Chicago, 1976), 821–859.Google Scholar On the use of ‘Neolithic’, see Onyango-Abuje, J. C., ‘Reflections on Cultural Change and Distribution during the Neolithic period in East Africa’, in Ogot, B. A. (ed.), Hadith, vi (Nairobi, 1976), 14–30Google Scholar; and Clark, J. D., ‘The problems of Neolithic culture in Sub-Saharan Africa’, in Bishop and Clark, Background, 601–24.Google Scholar The term ‘Iron Age’ is less problematical, but its subdivisions (‘Early’, ‘Middle’ and ‘Late’) present the same sort of problems as the Stone Age terms.

⁵ Hall, M. and Vogel, J. C., ‘Some recent radiocarbon dates from Southern Africa’, J. Afr. Hist, xxi, 4 (1980), 431.CrossRefGoogle Scholar

⁶ Phillipson, D. W., ‘Lowasera’, Azania, xii (1977), 1–32CrossRefGoogle Scholar; he considers two other dates, GX-4345 and GX-4347, too young for their respective levels.

⁷ Barthelme, J., ‘Holocene sites north-east of Lake Turkana: a preliminary report’, Azania, xii (1977). 33–41.CrossRefGoogle Scholar

⁸ Barthelme, pers. comm. Shell samples from site GaJj ii gave ‘modern’ dates which obviously do not refer to its occupation (GX-5477, HEL-1275).

⁹ Barthelme, pers. comm. Two charcoal samples sent to different laboratories (P-2610, SUA-637) were dated to the early second millennium b.c, which Barthelme considers to be more acceptable than a date in the ninth millennium b.c. from a shell sample (SUA-638).

¹⁰ Barthelme, , ‘Holocene sites’, 37–40.Google Scholar

¹¹ Ibid. 40; pers. comm. As only one date (from a bone sample) is available for this site, it should be regarded with caution, but it receives some confirmation from a site to the south-east, North Horr I, where stone bowls were dated to between the mid-third and mid-second millennia b.c. (c.f. Maggs, , J. Afr. Hist, xviii (1977), 169).Google Scholar

¹² Barthelme, pers. comm. The site also yielded a date in the seventh millennium b.c. (SUA-635); Barthelme regards this as too early.

¹³ Barthelme, pers. comm.

¹⁴ Ibid. GX-4641 is a ‘modern’ date.

¹⁵ Barthelme, , ‘Holocene sites’, 35Google Scholar; pers. comm. A date in the twelfth millennium b.c. (GX-5478) from Nderati Wells (GaJiI) relates to pre-ceramic microlithic material.

¹⁶ D. W. Phillipson, pers. comm.

¹⁷ Phillipson considers GX-5157 (G) too recent by a thousand years.

¹⁸ Noted by Phillipson and Barthelme.

¹⁹ Phillipson, pers. comm.

²⁰ Phillipson, D. W., ‘The excavation of Gobedra rockshelter, Axum: an early occurrence of cultivated finger millet in northern Ethiopia’, Asania, xii (1977), 58, 79–81.Google Scholar In a personal communication, Phillipson has referred to Bulliet, R. W., The Camel and the Wheel (Cambridge, Mass., 1975)Google Scholar, for arguments pertaining to the likelihood of an early date for camels in the region. A date for an earlier occupation at Gobedra was reported by Maggs, , J. Afr. Hist, xviii (1977), 168, 186.Google Scholar

²¹ Lynch, B. M. and Robbins, L. H., ‘Cushitic and Nilotic Prehistory: new archaeological evidence from north-west Kenya’, J. Afr. Hist, xx, 3 (1979), 319–28.CrossRefGoogle Scholar

²² Lynch, B. M. and Robbins, L. H., ‘Namoratunga: the first archaeo-astronomical evidence in sub-Saharan Africa’, Science, cc (19 05 1978), 766–8.CrossRefGoogle ScholarPubMed

²³ Lynch, and Robbins, , ‘Cushitic and Nilotic Prehistory’, 323, n. 21.Google Scholar

²⁴ Ibid. 324–8. The excavators note that the date UCLA-2124K has a large standard deviation, and they are uncertain what is dated by GX-5041. From a nearby site, pottery in the same tradition has been associated with a date in the fifth century a.d. (N-909) already reported (Sutton, , J. Afr. Hist, xiii (1972), 19).Google Scholar There is linguistic evidence that proto-Masaian may have been spoken west of Lake Turkana around the fifth century a.d.: Ehret, C. et al. , ‘Some thoughts on the early history of the Nile-Congo watershed’, Ufahamu, v (1975), 85–112Google Scholar; I owe this reference to Dr P. Robertshaw.

²⁵ Other radiocarbon dates from Lokabulo rockshelter sites include GU-1232, which the excavator regards as too young to be trustworthy, and GU-1302 and GU-1238, both of which give too recent an age for the levels from which the samples were collected, probably due to contamination (N. David, pers. comm.).

²⁶ P. Robertshaw, pers. comm.

²⁷ Ibid.

²⁸ Ibid.

²⁹ N. David, pers. comm.

³⁰ Other dates from the same site (GU-1231, GU-1236) are from samples thought to be contaminated by charcoal from recent forging and/or smelting activities (David, pers. comm.).

³¹ Robertshaw, pers. comm.

³² Ibid.

³³ N. David, pers. comm.

³⁴ Angela, Close, ‘Current Research and recent radiocarbon dates from Northern Africa’, J. Afr. Hist, xxi, 2 (1980), 145–67.Google Scholar

³⁵ Jarrige, J. and Meadow, R. H., ‘The antecedents of civilization in the Indus Valley’, Scientific American, no. 243, 2 (1980), 122–33.CrossRefGoogle Scholar

³⁶ Bower, J. R. F., ‘Final report: archaeological study of environment and culture, Serengeti National Park, Tanzania’ (unpublished), 1–6.Google Scholar

³⁷ Bower, J. R. F. and Nelson, C. M., ‘Early pottery and pastoral cultures of the Central Rift Valley, Kenya’, Man, n.s., xii (1979), 556, 565, n. 3.Google Scholar

³⁸ Anthony, B., ‘Excavations near Elmenteita, Kenya’, Palaeoecology of Africa, iii (1967), 47–8;Google Scholar Isaac, G. L., Merrick, H. V. and Nelson, C. M., ‘Stratigraphic and archaeological studies in the Lake Nakuru Basin, Kenya’, Palaeoecology of Africa, v (1971), 225–32.Google Scholar

³⁹ These dates have already been reported in this series: Sutton, , J. Afr. Hist, xiii (1972), 3, 5, 19Google Scholar; Soper, , J. Afr. Hist, xv (1974), 179.Google Scholar

⁴⁰ Most of the evidence has been reviewed by Bower and Nelson (‘Early Pottery’); other evidence is cited in articles referred to below. On the basis of finds at Lukenya Hill, east of Nairobi, in 1979, Nelson has argued that cattle were kept there as early as the twelfth or even the twentieth millennium b.c. (cf. Richard, Leakey, The Making of Mankind (London, 1981), 197)Google Scholar; further details are not available at the time of writing.

⁴¹ Bower, and Nelson, , ‘Early Pottery’, 555–6.Google Scholar

⁴² Bower, , ‘Final report’, 3–4.Google Scholar

⁴³ Onyango-Abuje, J. C., ‘Crescent Island: a preliminary report on excavations at an East African Neolithic site’, Azania, xii (1977), i, 147–59.CrossRefGoogle Scholar

⁴⁴ Ibid. 156–7; cf. Phillipson, , J. Afr. Hist, xi (1970), 4.Google Scholar

⁴⁵ Bower, J. R. F., Nelson, C. M., Waibel, A. F. and Wandibba, S., ‘The University of Massachusetts Later Stone Age/Pastoral “Neolithic” comparative study in Kenya: an overview’, Azania, xii (1977), 134.Google Scholar The ‘young age’ from the collagen sample GX-4319 is not an isolated case. Bower and Nelson write: ‘…it has been our general experience that, when dealing with samples from Central Kenya no older than 3000 B.P., collagen usually yields dates between 300 and 1,800 years more recent than apatite, while apatite yields dates closely comparable with determinations from associated wood charcoal, after C-13 corrections have been calculated. Where apatite is obviously in error, it invariably gives a date of “modern”. In samples which lack collagen and are greater than 10,000 years in age, apatite often generates dates 2,500–3,200 years younger than associated wood charcoal’ (Bower and Nelson, ‘Early Pottery’, 559).

⁴⁶ Bower, et al., ‘Later Stone Age’, 134.Google Scholar

⁴⁷ There are marked discrepancies between dates run on apatite and collagen both for the Kansyore-type ware (GX-4421 (A), GX-4421 (C)) and for the Narosura ware (GX-4468 (A), GX-4468 (C)). For a possible explanation see footnote 45 and source quoted there.

⁴⁸ Bower, et al., ‘Later Stone Age’, 131Google Scholar; Bower, and Nelson, , ‘Early Pottery’, 555Google Scholar; the date GX-4464 (A) is too recent for the associated cultural material.

⁴⁹ Bower, et al., ‘Later Stone Age’, 131, 137.Google Scholar

⁵⁰ Ibid. 140–1; Bower, and Nelson, , ‘Early pottery’, 556.Google Scholar

⁵¹ Bower, et al. ‘Later Stone Age’, 131, 139.Google Scholar

⁵² Ibid. 133–4.

⁵³ Ibid. 129, 137.

⁵⁴ Ibid. 121–5; Bower, and Nelson, , ‘Early pottery’, 555–6.Google Scholar

⁵⁵ Bower, et al., ‘Later Stone Age’, 140, 145.Google Scholar

⁵⁶ Neville, Chittick, ‘Excavations at Deloraine Farm, Rongai’, Azania, xiv (1979), 162–3; pers. comm.Google Scholar

⁵⁷ Bower, et al., ‘Later Stone Age’, 133, 144Google Scholar; Bower, and Nelson, , ‘Early pottery’, 562.Google Scholar A ‘modern’ date, GX-4420 (C), is unsound (Ibid. 559).

⁵⁸ Bower, et al., ‘Later Stone Age’, 131, 144.Google Scholar

⁵⁹ Bower, and Nelson, , ‘Early pottery’, 564.Google Scholar

⁶⁰ Bower, et al., ‘Later Stone Age’, 134Google Scholar; Bower, and Nelson, , ‘Early pottery’, 555.Google Scholar

⁶¹ Ibid. 564.

⁶² Robert, Soper, ‘Iron Age archaeology and traditional history: Embu, Mbeere and Chuku areas of Central Kenya’, Azania, xiv (1979), 31–59.Google Scholar

⁶³ Masao, F. T., The Later Stone Age and the Rock Paintings of Central Tanzania (Wiesbaden, 1979), 194.Google Scholar At Kandaga, GX-3678 (last 200 years) dates microlithic material; at Kirumi Isumbira, GX-3682 comes from a disturbed context.

⁶⁴ Sutton, J. E. G., ‘Engaruka and its Waters’, Azania, xiii (1978), 37–70.CrossRefGoogle Scholar

⁶⁵ Phillipson, , J. Afr. Hist, xi (1970), 8–9.Google Scholar

⁶⁶ Sutton, , ‘Engaruka’, 38.Google Scholar

⁶⁷ Ibid. 59.

⁶⁸ Sutton, , J. Afr. Hist, xiii (1972), 8–9.Google Scholar

⁶⁹ Schmidt, Peter R., Historical Archaeology: a Structural Approach in an African Culture (Westport, Conn., 1978), 191.Google Scholar

⁷⁰ Prof. Schmidt has kindly provided these dates in advance of their publication.

⁷¹ E.g. Roland Oliver's review of Schmidt's, Historical Archaeology in J. Afr. Hist, xx (1979), 289–90Google Scholar; van Noten, F., ‘The Early Iron Age in the interlacustrine region: the diffusion of iron technology’, Azania, xiv (1979), 73.Google Scholar

⁷² The dates for Ise Dura were reported by Calvocoressi, and David, , J. Afr. Hist, xx (1979), ii.Google Scholar

⁷³ See Onyango-Abuje, J. C., ‘A contribution to the study of the Neolithic in East Africa with particular reference to Nakuru-Naivasha basins’ (unpublished, 1977)Google Scholar; Schmidt, , Historical Archaeology, 296Google Scholar; Mgomezulu, G. G. Y., ‘Food production: the beginnings in the Linthipe/Changoni area of Dedza District, Malawi’ (Ph.D. thesis, University of California: Berkeley, 1978).Google Scholar

⁷⁴ Bower, and Nelson, , ‘Early pottery’, 565.Google Scholar

⁷⁵ Phillipson, D. W., ‘Some Iron Agesitesin the lower Tana valley’, Azania, xiv (1979), 155–60;CrossRefGoogle Scholar pers. comm.

⁷⁶ H. N. Chittick, pers. comm.; see also Neville, Chittick, ‘An archaeological reconnaissance in the Horn: the British Somali Expedition, 1975’, Azania, xi (1976), 117–33.Google Scholar

⁷⁷ H. N. Chittick, pers. comm.; he thinks that this radiocarbon determination is a ‘further example of a sample from a low latitude which gives a result which is too early (cf. Kilwa and Manda)’.

⁷⁸ Chittick, pers. comm.

⁷⁹ British Institute in Africa, Eastern, Report for 1978–9 (London, 1980), 3–4Google Scholar; cf. Neville Chittick, Excavations at Manda, 1978 (Memoir no. 8 of the British Institute in Eastern Africa: Nairobi, in press).

⁸⁰ I am grateful to the Executive Director of the National Museums of Kenya, Mr R. E. Leakey, who kindly gave me copies of two unpublished reports by Dr T. H. Wilson on research by the Museums along the coast between 1978 and 1980.

⁸¹ Wilson, T. H., ‘The monumental architecture and archaeology north of the Tana River’ (11. 1978), 82–8).Google Scholar

⁸² Wilson, T. H., ‘The monumental architecture and archaeology of the central and southern Kenya Coast’ (02. 1980), 54–77.Google Scholar

⁸³ Ibid. 977–8.

⁸⁴ Ibid. 43–5.

⁸⁵ Phillipson, D. W., The Later Prehistory of Eastern and Southern Africa (London, 1977) 146.Google Scholar Dates for cattle in the third century a.d. from Makwe in eastern Zambia, and in the second to fifth centuries a.d. from Broederstroom, west of Pretoria, further cast doubt on Phillipson's theory: see Mgomezulu, ‘Food Production’ Huffman, T. N., ‘African Origins’, S. Afr.J. Science, lxxv (1979), 233–7Google Scholar; but also Hall, and Vogel, , J. Afr. Hist, xxi (1980), 441, n. 58.Google Scholar

⁸⁶ Robinson, K. R., The Iron Age of Northern Malawi: an Archaeological Reconnaissance (Malawi Antiquities Department, publication no. 20: forthcoming); this expands on the preliminary report by Robinson in J. Afr. Hist, vii, 2 (1966).Google Scholar

⁸⁷ Robinson, Iron Age, and pers. comm.

⁸⁸ Phillipson, , J. Afr. Hist, xi (1970), 6.Google Scholar

⁸⁹ Robinson, Iron Age, and pers. comm.

⁹⁰ Mgomezulu, ‘ Food Production.’

⁹¹ The full pottery sequence as known at present consists of Nkope, Kapeni, Mawudzu, Nkudzi and modern Chewa.

⁹² H. de Villiers, ‘Report on human skeletal remains from the Linthipe/Changoni project, site DZ/26’, in Mgomezulu, ‘Food Production’.

⁹³ Mgomezulu, ‘Food Production’.

⁹⁴ Prof. J. D. Clark, pers. comm.; see also Mgomezulu, ‘Food Production’.

⁹⁵ Ibid., Robinson, K. R. and Sandelowsky, B., ‘The Iron Age of northern Malawi: recent work’, Azania, in (1968), 122–4Google Scholar; Robinson, K. R., Iron Age Sites in the Dedza District of Malawi (Malawi Antiquities Department publication, no. 16: Zomba, 1975), 12–15.Google Scholar

⁹⁶ Mgomezulu, ‘Food Production.’

⁹⁷ Robinson, K. R., The Nkhota-kota Lakeshore and marginal areas, Malawi: an archaeological reconnaissance (Malawi Antiquities Department publication, no. 19: Limbe, 1979).Google Scholar

⁹⁸ Ibid. 13.

⁹⁹ The research was directed by the present writer; analysis of the material recovered is in progress.

¹⁰⁰ Longwe ware was first described by Robinson, K. R., Iron Age Occupation North and East of the Mulanje Plateau, Malawi (Malawi Antiquities Department publication, no. 17: Limbe, 1977).Google Scholar Pta-1798 was communicated to me by Robinson in August, 1978. Mr Yusuf Juwayeyi has informed me that Longwe pottery has been observed in the Malowa/Milolongwe area, some 30 km west of the type site. The Longwe dates from this area are awaited with interest. The Kasuku Village site (JkJe4) is the westernmost site for Longwe ware.

¹⁰¹ Maggs, , J. Afr. Hist, xviii, 2 (1977), 176.Google Scholar

¹⁰² N. M. Katanekwa, pers. comm.

¹⁰³ Idem, ‘Some Early Iron Age sites from the Machili Valley of south-western Zambia’, Azania, xiii (1978), 149–50.Google Scholar

¹⁰⁴ Ibid. 159 n., and pers. comm.

¹⁰⁵ Katanekwa, , ‘Some Early Iron Age sites’, 158.Google Scholar

¹⁰⁶ Katanekwa, pers. comm.

¹⁰⁷ J. H. Robertson, pers. comm.

¹⁰⁸ Idem, ‘Zambia’, Nyame Akutna, xii (1978), 45–6.Google Scholar

¹⁰⁹ F. B. Musonda, pers. comm.