of Sofala, known to have been frequented by Arab traders in mediaeval times. It was at once assumed that they were of Sabaean or Phoenician origin and of great antiquity. The subject was voluminously but uncritically written up. See for instance Ancient Ruins of Rhodesia, by Hall and Neal, London, 1894; Monomotapa, by A. Wilmot, London, 1896, and The Ruined Cities of Mashonaland, by J. T. Bent, London, 1902.
The appearance of the structures suggested the form of ancient Arabian temples, and the locality was at once identified with the ubiquitous "land of Ophir" of King Solomon's voyages. Professor Müller (Burgen und Schlösser, II, 20), noted a resemblance between the Zimbabwe enclosure (20° 30′ S., 31° 10′ E.) and the temple at Marib, the capital of the ancient Sabaean kingdom of Southern Arabia. The whole argument was of course pure assumption, as there is no reference in ancient literature to any knowledge of the African coast within six hundred miles of the port of Sofala. Dr. David Randall-Maciver made a careful investigation of the ruins in 1905, and proved conclusively in his account of that work, Mediaeval Rhodesia, London, 1906, that the structures were the work of negroes, probably Kaffirs, of the so-called kingdom of Monomotapa. A piece of Nankin china of the late mediaeval period, found in the cement at the bottom of one of the structures, showed that they could not date earlier than the 14th or 15th century. They were enclosures for defence, rudely built of loose stone, and their supposed orientation was found to be inexact and probably accidental.
The service done by Dr. Maciver in disproving the antiquity of this Kaffir kraal did not, however, need to be supplemented by his denial (pp. 1–2) of the probability of Arabian trade far down this coast at a very early age. The Periplus mentions Rhapta, some distance south of the Zanzibar islands, as the last settlement on the coast; and Ptolemy describes Cape Delgado. Dr. Maciver may have known the Periplus only through the account given by Guillain in 1856 (Documents sur l'histoire, la géographie et le commerce de l'Afrique Orientale), but at all events he ignores the detailed account given in both those works, and in the Periplus the statement is definitely made that this whole coast (to about 10° S.) was "subject under some ancient right to the sovereignty of the power which held the primacy in Arabia;" that is, in the 1st century A. D. the right was still so ancient as to be beyond the explanation of the merchant who described it. The coast was frequented by Arab ships in command of Arab captains who knew the harbors, spoke the language of the natives and intermarried with them.
