The resemblance between the round towers of Rhodesia and the primitive, conical, round minarets of the old mosques at Lamu, Malindi, Mombasa, and other places on the East African coast (dominated by Arabs for at least 2000 years) is very striking. Both may date from the pre-Islamic period. There are other analogies between the Rhodesian ruins and the ancient buildings of pre-Islamic Arabia which suggest, as the most probable explanation of this mystery in African colonization, that the ancient gold-miners, _phallus-_ and sun-worshippers, irrigators, terrace-cultivators of Matebele- and Mashonaland, were likewise Arabs from some part of western or southern Arabia who penetrated inland from Sofala attracted by the signs of gold. After a century or so of profitable gold-mining in a land which had only then a spare Bushman population, the Bantu hordes from the north descended on these Semitic colonies and eventually exterminated or drove away the Arabs, taking their place clumsily as gold-miners and builders. Although the Arabs never regained their position in the interior, they continued or resumed their occupancy of the south-east African coast-line down to the arrival of the Portuguese. Probably also the Tsetse fly, by its interference with the means of transport, was another deterrent factor in the history of this colonization which failed to spread. It is possible, nevertheless, that Madagascar and Bantu East Africa owe to these hypothetical, unnamed, prehistoric Arab colonizers not only the introduction (indirectly from India) of the edible banana or plantain, which afterwards spread right across the continent, but also the long-horned, straight-backed Egyptian ox, and the domestic fowl; hemp perhaps likewise, a “smoking mixture” which preceded tobacco by many centuries.