Under such conditions it will naturally be asked, whence did the foreign intruders obtain their food supplies? The answer to this question is suggested in The Ancient Ruins, where it is pointed out (p. 208) that the auriferous reefs of the central Zimbabwe district, and generally of all the districts in immediate proximity to the fortified stations, show no traces of having ever been worked for the precious metal. “Possibly the reason for the ancients ignoring the gold-reefs of this district [Zimbabwe] lies in the fact that the country round about is exceedingly well suited for agricultural purposes, the soil being rich and water plentiful, and all vegetable growths prolific and profuse. The large population of ancients, together with the enormous gangs of slaves, would naturally consume a vast quantity of grain, and this necessity would create a large agricultural class, who, for their own safety and for the protection of their crops and fruits, would naturally carry on their operations within such an area as could be safeguarded by the fortresses of Zimbabwe.”

It might at first sight be supposed that the food supplies were drawn chiefly from the extensive agricultural settlements of the Inyanga territory, on the northern slopes of Mashonaland, which drain through the Ruenga and its numerous affluents to the right bank of the Zambesi. This Inyanga district may be roughly described, from the archÆological point of view, as an area of old aqueducts, of old terraced slopes, and of old ruins of a less imposing type than the Zimbabwe remains. In a notice of The Ancient Ruins of Rhodesia contributed to the Geographical Journal for April, 1902, I first drew attention to the surprising analogy, or rather identity, between these terraces and those of the South Arabian uplands visited by General E. T. Haig in the eighties. So close is the parallelism that Haig’s description might almost change places with Mr. Telford Edwards’ account of the Inyanga works quoted in The Ancient Ruins, p. 353 sq., as thus:—

But Mr. Hall, who visited the Inyanga territory in May, 1904, now finds that the terraced slopes,[4] the so-called “slave-pits,” and the other remains, although “old,” are not “ancient.” That is to say, they date not from Himyaritic times, but probably from the eleventh or twelfth century of the new era, when parts of Rhodesia were reoccupied by large numbers of Moslem Arabs from Quiloa and their other settlements along the east coast. Hence, although the terraced slopes still form a connecting link between South Africa and South Arabia, the South Arabia here in question is that, not of pre-, but of post-Koranic times.

Of course, the ruined houses and ruined aqueducts are too much obliterated to supply any clear points of comparison. But their mere presence, and especially the vast extent of ground covered by them, will suffice to confirm Mr. Telford Edwards’ estimate of the vast numbers of civilised peoples who inhabited the rich Inyanga valleys in prehistoric times, and whom we may now call SabÆans, MinÆans, and others Himyarites.

Were the houses still extant, we should expect to find them covered with the same decorative mural motives as are still seen both on the Zimbabwe monuments and on the public buildings of Sana, present capital of Arabia Felix. Manzoni, who visited this city three times between the years 1877 and 1880, figures a mansion six stories high, which is richly ornamented with two such motives—the chevron and the vertical block pattern—closely resembling those everywhere occurring on the more ancient Rhodesian walls. The chevron, which is seen both in single and double courses exactly as on the great walls of the Elliptical Temple, is absolutely identical, while the block design differs only in being quite vertical at Sana, whereas it is slightly tilted, or else two rows of blocks converge to produce the herring-bone pattern on the Rhodesian walls, as at Little Umnukwana and many other places. The reader will find Manzoni’s mansion reproduced in Mr. D. G. Hogarth’s The Penetration of Arabia, 1904, p. 198, and he will there notice that the various motives fill up all the space between two parallel horizontal lines, as is so often the case in Rhodesia.[5] Here, therefore, style, motive, general treatment, everything corresponds between the Rhodesian remains and the decorative fancies still flourishing in Sana, heir to the cultural traditions of the neighbouring Mariaba and of the other ancient Himyaritic capitals in South Arabia.

In The Gold of Ophir frequent reference is made to the relations, social and commercial, established between Palestine and Madagascar certainly as early as the time of Solomon, and possibly even during the reign of his father David. On this point I might have spoken even more confidently, for I have since received a communication from M. Alfred Grandidier, by far the greatest living authority on all things Malagasy, who calls my attention to the evidence supplied in his monumental work, Histoire Physique, Naturelle et Politique de Madagascar (1901), of intercourse between the Jews and the natives of Madagascar and neighbouring islands even in pre-Solomonic days. Documents are quoted to show that the Comoros, stepping-stones between Madagascar and Rhodesia, were peopled in the reign of Solomon “by Arabs or rather by IdumÆan Jews from the Red Sea,” and that the people of the great island preserve many Israelitish rites, usages, and traditions, cherish the memory of Adam, Abraham, Lot, Moses, Gideon, but have no knowledge of any of the prophets after the time of David, “which seems to show that the Jewish immigrants left their home at a very remote date, since if the exodus had been recent they could not have forgotten the great names posterior to the time of David.” Hence he concludes that “there is nothing surprising in the presence of an IdumÆan colony in Madagascar, for we know that from the very earliest times the Arabs of Yemen had frequented the East African seaboard at least as far as Sofala.” These words lend further support to my identification of Tharshish with Sofala, and in a note it is added that “the Jews and Arabian Semites were not the only peoples who had formerly commercial relations with the inhabitants of the African seaboard. From time immemorial these southern waters were navigated by the fleets of the Egyptians, probably even of the Chaldeans, Babylonians, Assyrians, Phoenicians, Tyrians” (op. cit., p. 96). And again at p. 100: “From the earliest times the Indian Ocean was traversed by Chaldean, Egyptian, Jewish, Arab, Persian, Indian, and other vessels.”[6]

My statements regarding the long-standing relations of the Northern Semites with the peoples of Madagascar and South Africa as far as Sofala are thus fully supported by the greatest authority on the subject. But there are some minds so constituted that they seem incapable of accepting a new revelation. They can do nothing but stare super vias antiquas, and will strain every nerve to minimise the force of facts and arguments pointing at conclusions which run counter to their deep-rooted prejudices. I here reproduce the famous “Zimbabwe Zodiac” (Fig. 2.), which was found near Great Zimbabwe, and shows the twelve signs of the Zodiac carved round the rim, as described by the late Dr. Schlichter in the Geographical Journal for April, 1890. This specialist tells us that “the signs coincide in every respect with other finds which Bent and others have made in Zimbabwe. One of the pictures is an image of the sun analogous to the sun-pictures which Mauch and Bent found on the monoliths of Zimbabwe, and analogous also to finds in Asia Minor which belong to the Assyro-Babylonian period.” But a writer in the Guardian attempts to destroy the significance of this document by asserting that the Zodiac or its nomenclature is of Greek origin and consequently of no great age. Now the Hon. Emmeline M. Plunket has recently (1903) published a work on Ancient Calendars and Constellations, in which she maintains that the Babylonian Calendar, with its Zodiacal signs, dates from 6000 b.c., that is, about 8,000 years ago. It is true that this estimate is not clearly made out. But on the other hand, the reader may be assured that Miss Plunket does not hold by the “Greek” theory. Nor does F. Delitzsch, who reminds us that “when we distinguish twelve signs of the Zodiac and call them Ram, Bull, Twins, etc., in all this the Sumero-Babylonian culture is still a living influence down to the present day.”[7] Nor does Sayce, who points out that the Babylonian account of the Flood occurs in the eleventh book of the epic of Gisdhubar corresponding approximately with the eleventh sign of the Zodiac, at that time Aquarius, just as the fifth book records the death of a monstrous lion by Gisdhubar, answering to the Zodiacal Leo and so on. He further observes that “the Zodiacal signs had been marked out and named at that remote period (certainly before 2000 b.c.), when the sun was still in Taurus at the beginning of spring,”[8] and, let me add, when the Greeks had not yet been heard of, but when the great Gnomon, or Conical Tower, had possibly already been erected by the Semitic builders of Great Zimbabwe.

That this and the numerous other conical towers still standing amid the crumbling ruins of Rhodesia are all cast in a Semitic mould will be at once seen by comparing them with the conical tower of a temple, figured on a medallion found at Byblos in Phoenicia and here reproduced (Fig. 1.). The comparison may also be extended to the two embossed cylinders—one from Great Zimbabwe, the other from the Temple of Paphos, in Cyprus, here also reproduced (Figs. 3 and 4) from Bent’s Ruined Cities, pp. 170, 171. These two objects, so strikingly similar in general design, reminded Bent of Herodian’s description of the sacred cone in the great Phoenician Temple of the Sun at Emessa, in Syria, which was adorned with certain “knobs or protuberances,” a pattern supposed by him to represent the sun, and common in phallic decorations, such as are constantly turning up with every shovelful of dÉbris removed from the Zimbabwe Temple Enclosures.

But although thousands of stones have been washed and carefully examined for inscriptions, none have so far been discovered. As the inscription which stood originally above the gateway of Great Zimbabwe, as reported by the Arabs to the Portuguese pioneers early in the sixteenth century,[9] has since disappeared, there are no known written documents connecting these monuments with South Arabia or Phoenicia, except a few scratches on the rim of an earthenware vessel figured by Bent and by him supposed possibly to be of Himyaritic type.[10] As, on the other hand, South Arabia is covered with Himyaritic rock inscriptions, some of considerable length and hitherto reputed to be of great age, their absence from Rhodesia has naturally caused surprise. This negative argument has even by some of my critics been allowed to outweigh the overwhelming positive evidence derived from the monuments themselves, from the hundreds of old gold-workings already described or recorded, from the multitude of objects—phalli, birds, conic towers—which have been found in the ruins, and are, beyond all doubt, intimately associated with Semitic religious observances. But I think it may now be shown that this “negative argument” is no proof at all of non-Semitic origins, but, on the contrary, affords strong indirect evidence of the great antiquity of these Semitic remains in Rhodesia.

It is to be noticed, in the first place, that although the Phoenicians are believed to have migrated from the Persian Gulf to the Mediterranean about three millenniums before the New Era, no Phoenician inscriptions have yet been anywhere discovered in the Mediterranean lands older than about the seventh or the eighth century b.c. Before that time the Phoenicians, like the kindred Canaanites and Israelites, were rude, uncultured peoples, with no knowledge of letters, except, perhaps, of the hieroglyphs, cuneiforms, and other scripts of their Egyptian, Assyro-Babylonian, Hittite, and Cretan neighbours. Even the Moabite Stone, if it be genuine, is post-Solomonic, since its reputed “author” was the Moabite king, Mesha, contemporary of Jehoram of Israel and Jehoshaphat of Judah. How, then, could the unlettered Jews and Phoenicians of the time of David, Solomon, and Hiram leave any written records of themselves in Rhodesia? After that epoch the intercourse with South Africa was interrupted, because “Jehoshaphat made ships of Tharshish to go to Ophir for gold; but they went not; for the ships were broken at Ezion-geber” (1 Kings xxii. 48). And then the star of Jacob waned, and the scattering of the Ten Tribes of Israel was presently followed by the dire calamities that fell upon Judah, and put an end for ever to all further quest of treasure in the Austral seas.

In the second place I find that Semitic students are gradually coming to the conclusion that the age of the South Arabian rock inscriptions has been greatly exaggerated, especially by Glaser, whose authority was at first naturally accepted almost without demur. The language is, no doubt, Himyaritic, that is to say, the oldest known form of Arabic. But that language survived for many centuries after the New Era in the Axumite empire, Abyssinia, where it is called Geez, and in Yemen till some time after the Mohammedan irruption, and is still current in the island of Sokotra, and in the Mahra district east of Hadramaut, where it is called Ehkili. Hence the language of the inscriptions is no test of their antiquity, though many afford intrinsic evidence that they date certainly from at least a few hundred years before the New Era. The subject is at present sub judice, and no more can be said until the full results are known of the extensive researches now in progress throughout Yemen. Here a large number of agents of the French MinistÈre de l’Instruction Publique have been at work since the year 1901, and thousands of impressions or rubbings have already (1903–4) been received in Paris. Some have even begun to appear in the Nouveaux Textes YÉmÉnites, edited by M. Derenbourg, and several of the inscriptions are stated to be in a hitherto unknown alphabet quite different from that of the Himyaritic document which forms the frontispiece of the Gold of Ophir. Great revelations may therefore be pending; but, meanwhile, so much may, I think, be safely inferred, that the Himyarites who first arrived in Rhodesia, worked the mines, and built the monuments, some dating from apparently 2000 b.c., had little or no knowledge of letters, or at least had not yet begun to cover the rocks of their South Arabian homes with well-formed and carefully constructed inscriptions. Thus is also explained the absence of all such documents from their new homes in Rhodesia, where one may now almost venture to predict that none will ever be found. Nothing can be inferred from the vanished inscription over the Great Zimbabwe gateway, since the gold-workings appear to have been resumed for a time by the later (post-Mohammedan) Arabs, who were fond of decorating the faÇades of their mosques and other public buildings with the ornamental but relatively recent (eighth century) Cufic characters.

Mention should perhaps here be made of Professor Gustav Oppert’s Tharshish and Ophir (Berlin, 1903), in which the learned author claims to offer “a final solution” of the problem. But he leaves the question exactly as it stood over three decades ago, is still lost in the tangle of time-worn etymologies, and takes no notice at all of the revelations made by Messrs. Hall and Neal in the Ancient Ruins. The vast body of archÆological evidence derived in recent years from the Rhodesian remains is thus completely ignored, and fresh light excluded from the only source whence it might have been drawn. On the other hand, Professor Oppert, rather than admit a Tharshish in the Indian Ocean, suggests that the Tharshish of Kings and Chronicles either means “the sea,” possibly the origin of the Greek word [Greek: thalatta] itself, or else was by the authors of those books foisted into the texts instead of Ophir. Hence where Tharshish occurs as the objective of Solomon’s gold expeditions we are to read Ophir, although the original Ophir is allowed to have been where I place it on the south coast of Arabia. Now the Greek word [Greek: thalatta] is Homeric, and when the Homeric poems were first sung there were no Greeks in the Indian Ocean. Hence, even if the wild etymology could be admitted, it would not serve, and this essay cannot be accepted as “a final solution of the old controversy.”[11] It is pleasant to be able to add that my solution has been accepted as final by some of Professor Oppert’s fellow-countrymen—the editor of the Coloniale Zeitung amongst others—who declares that “the problem seems now really solved.”[12]

Let me conclude with a question. Those who still reject my solution, who cast about for the gold of Ophir all over the Indian Ocean—Egypt, Arabia, Persia, India—anywhere except South Africa, what do they propose doing with the hundreds of old Rhodesian workings, which are known to have yielded at least £75,000,000 in their time, and with the stupendous Semitic monuments connected with these workings, of which Mr. Hall here presents the public with scores of photographic reproductions, drawn exclusively from the central Great Zimbabwe group? Where does India, the spoilt child of the etymologists, stand beside these remains, which betray such undoubted evidence of their South Arabian origin?