The ruined buildings of Mashonaland and Matabililand have excited in recent years an amount of interest and curiosity which is disproportionate to their number, size, and beauty, but by no means disproportionate to their value as being the only record, scant as it is, we possess of what has been deemed an early South African civilization. I will describe in the fewest words such of these buildings as I saw, leaving the reader of archÆological tastes to find fuller details in the well-known book of that enterprising explorer, the late Mr. Theodore Bent. Some short account of them seems all the more needed, because the first descriptions published gave the impression that they were far more considerable than they really are.

Scattered over the plateau of southern Mashonaland and Matabililand, from its mountainous edge on the east to the neighbourhood of Tati on the west, there are to be found fragments of walls built of small blocks of granite resembling paving stones (usually about a foot long by six inches high), but often larger, not cut smooth, but chipped or trimmed to a fairly uniform size. These walls are without mortar or other cementing material, but the stones are so neatly set together, and the wall usually so thick, that the structure is compact and cohesive. The walls are mostly thinner at the top than at the base. The only ornamentation consists in placing some of the layers at an acute angle to the other layers above and below, so as to produce what is called the herring-bone pattern. Occasionally a different pattern is obtained by leaving spaces at intervals between the horizontal stones of certain layers, making a kind of diaper. In some cases this ornamentation, always very simple, occurs only on one part of the wall, and it has been said that it occurs usually if not invariably on the part which faces the east. I heard of ten or twelve such pieces of wall in different parts of the plateau, and saw photographs of most of these. Probably others exist, for many districts, especially in the hills, have been imperfectly explored, and trees easily conceal these low erections. One was described to me, where the walls are the facings of seven terraces, rising one above another to a sort of platform on the top. This I have not seen; but it is probably similar to one which I did see and examine at a place called Dhlodhlo, about fifty miles south-east of Bulawayo. This group of ruins, one of the most interesting in the country, stands high among rocky hills, from which a superb view is gained over the wide stretches of rolling table-land to the north and north-west, a charming situation which might have attracted the old builders did they possess any sense of beauty. On a low eminence there has been erected such a wall of such hewn, or rather trimmed, stones as I have just described. It is now about twenty feet in height, and may have originally been higher. On the eastern side this wall consists of three parts, each about six feet high, with two narrow terraces, each from five to six feet wide, between them, the second wall rising from the first terrace, and the third or highest wall from the second terrace. On this side some of the stone courses have the simple forms of ornamental pattern already mentioned. On the opposite, or western and north-western, side only one terrace and a low, unornamented wall of trimmed stones are now discernible. To the north, still within what seems to have been the main inclosing wall, are small inclosures built of trimmed stone, which may have been chambers originally roofed with wood or bushes. At the top of the highest wall there is at the north-north-west end a small level platform of earth or rubble, which seems to have been filled in behind the terraced walls. This platform is approached by a narrow passage between walls of trimmed stone, at one point in which there appears to have been a sort of narrow gateway barely wide enough for two persons to pass. There is no trace of any stone building on the top of the platform, and the remains of clay huts which one finds there may well be quite modern. To the south of this principal structure there is a second small hill or boss of granite, protected on three sides by steep sheets of granite rock. Its top is inclosed by a low wall of trimmed stones, now in places quite broken away, with no trace of any stone building within. All round on the lower ground are large inclosures rudely built of rough stones, and probably intended for cattle-kraals. They may be quite modern, and they throw no light on the purpose of the ancient buildings. Nor is much light to be obtained from the objects which have been found in the ruins. When I was there they were being searched by the Mashonaland Ancient Ruins Exploration Company, a company authorized by the British South Africa Company to dig and scrape in the ancient buildings of the country for gold or whatever else of value may be there discoverable, an enterprise which, though it may accelerate the progress of archÆological inquiry, obviously requires to be conducted with great care and by competent persons. So far as I could observe, all due care was being used by the gentleman in charge of the work at Dhlodhlo; but considering how easy it is to obliterate the distinctive features of a ruin and leave it in a condition unfavourable to future examination, it seems desirable that the company should, as a rule, await the arrival of trained archÆologists rather than hurry on explorations by amateurs, however zealous and well intentioned. Of the objects found, which were courteously shown to me, some are modern, such as the bits of pottery, apparently Indian or Chinese, the bits of glass, the bullets and fragments of flint-lock muskets, a small cannon, and an iron hammer. These are doubtless of Portuguese origin, though it does not follow that any Portuguese expedition ever penetrated so far inland, for they may have been gifts or purchases from the Portuguese established on the coast four or five hundred miles away. So, too, the silver and copper ornaments found, and some of the gold ones (occasionally alloyed with copper), which show patterns apparently Portuguese, may be recent. There are also, however, some gold ornaments, such as beads, bangles (a skeleton was found with bangles on the legs and a bead necklace), and pieces of twisted gold wire, which may be far more ancient, and indeed as old as the structure itself. A small crucible with nuggets and small bits of gold goes to indicate that smelting was carried on, though the nearest ancient gold-workings are six miles distant. Probably here, as at Hissarlik and at Carthage, there exist remains from a long succession of centuries, the spot having been occupied from remote antiquity.[6] At present it is not only uninhabited, but regarded by the natives with fear. They believe it to be haunted by the ghosts of the departed, and are unwilling, except in the daytime and for wages paid by the Exploration Company, to touch or even to enter the ruins. They can hardly be persuaded even to relate such traditions as exist regarding the place. All that has been gathered is that it was the dwelling of a line of mambos, or chiefs, the last of whom was burned here by Mosilikatze, the Matabili king, when he conquered the country sixty years ago. (The place does show marks of fire.) But the buildings were here long before the mambos reigned, and who built them, or why, no one knows. The natives come sometimes to make offerings to ancestral ghosts, especially when they ask for success in hunting; and if the hunt be successful, strips of meat are cut off and placed in cleft sticks for the benefit of the ghosts.

Three hypotheses have been advanced regarding the Dhlodhlo building. One regards them as a fortress. The objection to this is that the terraced and ornamented wall is so far from contributing to defence that it actually facilitates attack; for, by the help of the terraces and of the interstices among the stones which the ornamental pattern supplies, an active man could easily scale it in front. Moreover, there is hard by, to the north, a higher and more abrupt hill which would have offered a far better site for a fort. The second view is that Dhlodhlo was a mining station, where slaves were kept at work; but if so, why was it not placed near the old gold-workings instead of some miles off, and of what use were the terraced walls? The inquirer is therefore led to the third view—that the building was in some way connected with religious worship, and that the ornament which is seen along the eastern wall was placed there with some religious motive. There is, however, nothing whatever to indicate the nature of that worship, nor the race that practised it, for no objects of a possibly religious character (such as those I shall presently mention at Zimbabwye) have been found here.

I visited a second ruin among the mountains of Mashonaland, near the Lezapi River, at a place called Chipadzi's grave, a mile from the kraal of a chief named Chipunza. Here a rocky granite kopje, almost inaccessible on two sides, is protected on one of the other sides by a neatly built wall of well-trimmed stones, similar to that of Dhlodhlo, but without ornament. The piece that remains is some fifty yards long, five feet thick at the base, and eleven feet high at its highest point. It is obviously a wall of defence, for the only erections within are low, rough inclosures of loose stones, and three clay huts, one of which covers the grave of Chipadzi, a chief who died some twenty years ago, and who was doubtless interred here because the place was secluded and already in a fashion consecrated by the presence of the ancient wall. That the wall is ancient hardly admits of doubt, for it is quite unlike any of the walls—there are not many in the country—which the Kafirs now build, these being always of stones entirely untrimmed and very loosely fitted together, though sometimes plastered with mud to make them hold.[7] There is nothing to see beyond the wall itself, and the only interest of the place is in its showing that the race who built Dhlodhlo and other similar walls in Matibililand were probably here also.

Much larger and more remakable is the group of ruins (situated seventeen miles from Fort Victoria, in southern Mashonaland) which goes by the name of the Great Zimbabwye. This Bantu word is said to denote a stone building, but has often been used to describe the residence of a great chief, whatever the materials of which it is constructed. It is a common noun, and not the name of one particular place. Europeans, however, confine it to this one ruin, or rather to two ruined buildings near each other. One of these is on the top of a rocky and in parts precipitous hill, the other in a valley half a mile from the foot of the hill.

The first, which we may call the Fort, consists of a line of wall, in parts double, defending the more accessible parts of the eastern and south-eastern end of the hill or kopje, which is about 500 feet high, and breaks down on its southern side in a nearly vertical sheet of granite. The walls, which in some places are thirty feet high, are all built of small trimmed blocks of granite such as I have already described, without mortar, but neatly fitted together. They are in excellent preservation, and are skilfully constructed in a sort of labyrinth, so as to cover all the places where an enemy might approach. From the openings in the wall, where doors were probably placed, passages are carried inward, very narrow and winding, so that only one person at a time can pass, and completely commanded by the high wall on either side. Everything speaks of defence, and everything is very well adapted, considering the rudeness of the materials, for efficient defence. There is no sort of ornament in the walls, except that here and there at the entrances some stones are laid transversely to the others, and that certain long, thin pieces of a slaty stone, rounded so that one might call them stone poles—they are about five to seven feet long—project from the top of the wall. Neither is there any trace of an arch or vaulted roof. None of what look like chambers has a roof. They were doubtless covered with the branches of trees. Very few objects have been found throwing any light on the object of the building or its builders, and these have been now removed, except some small pieces of sandstone, a rock not found in the neighbourhood, which (it has been conjectured) may have been brought for the purposes of mining.

The other building is much more remarkable. It stands on a slight eminence in the level ground between the hill on which the Fort stands and another somewhat lower granite hill, and is about a third of a mile from the Fort. It consists of a wall, rather elliptical than circular in form, from thirty to forty feet high, fourteen feet thick near the ground, and from six to nine thick at the top, where one can walk along a considerable part with little difficulty. This wall is built of the same small, well-trimmed blocks of granite, nicely fitted together, and for more than half the circumference is in excellent preservation, although shrubs and climbing vines have here and there rooted themselves in it. The rest of it is more or less broken, and in one place quite overthrown. There are two gates, at the west and the north. The wall is quite plain, except for about one-third (or perhaps a little less) of the outer face, where there is such an ornament as I have already described, of two courses of stones set slantingly at an acute angle to the ordinary flat courses above and below. These two courses are the fifth and seventh from the top. In the space surrounded by the wall, which is about three-quarters of an acre, are some small inclosures of trimmed stone, apparently chambers. There is also a singular wall running parallel to the inner face of the great inclosing wall for some twenty yards, leaving between it and that inner face a very narrow passage, which at one point must have been closed by a door (probably of stone), for at that point steps lead up on either side, and hollow spaces fit for receiving a door remain. At one end this passage opens into a small open space, where the most curious of all the erections are to be found, namely, two solid towers of trimmed stones. One of these is quite low, rising only some five feet from the ground. The other is more than forty feet high, overtopping the great inclosing wall (from which it is eight feet distant) by about five feet, and has a bluntly conical top. It reminds one a little of an Irish round tower, though not so high, save that the Irish towers are hollow and this solid, or of a Buddhist tope, save that the topes which are solid, are very much thicker. There is nothing whatever to indicate the purpose of this tower, but the fact that the space in which it and the smaller tower stand is cut off from the rest of the enclosed area by a pretty high wall seems to show that it was meant to be specially protected or was deemed to be specially sacred.

Outside the main inclosing wall are several small inclosures of irregular shape, surrounded by similar walls of trimmed stones, but all low and broken and with nothing inside. One of these joins on to the main wall of the great inclosure.

This is all that there is to see at Zimbabwye. What I have described seems little, and that little is simple, even rude. The interest lies in guessing what the walls were built for, and by whom. Comparatively little has been discovered by digging. No inscriptions whatever have been found. Some figures of birds rudely carved in a sort of soapstone were fixed along the top of the walls of the Fort, and have been removed to the Cape Town museum. It is thought that they represent vultures, and the vulture was a bird of religious significance among some of the Semitic nations. Fragments of soapstone bowls were discovered, some with figures of animals carved on them, some with geometrical patterns, while on one were marks which might possibly belong to some primitive alphabet. There were also whorls somewhat resembling those which occur so profusely in the ruins of Troy, and stone objects which may be phalli, though some at least of them are deemed by the authorities of the British Museum (to whom I have shown them) to be probably pieces used for playing a game like that of fox-and-geese. The iron and bronze weapons which were found may have been comparatively modern, but the small crucibles for smelting gold, with tools and a curious ingot-mould (said to resemble ancient moulds used at tin workings) were apparently ancient.

What purpose were these buildings meant to serve? That on the hill was evidently a stronghold, and a stronghold of a somewhat elaborate kind, erected against an enemy deemed formidable. The large building below can hardly have been a place of defence, because it stands on level ground with a high, rocky hill just above it, which would have afforded a much stronger situation. Neither was it a mining station, for the nearest place where any trace of gold has been found is seven miles away, and in a mining station, even if meant to hold slave workers, there would have been no use for a wall so lofty as this. Two hypotheses remain: that this was the residence of a chief, or that it was erected for the purposes of religious worship. It may have been both—a palace, so to speak, with a temple attached. The presence of the inner inclosure, guarded by its separate wall, and with its curious tower, is most plausibly explained by supposing a religious purpose; for as religion is the strangest of all human things, and that in which men most vary, so it is naturally called in to explain what is otherwise inexplicable.

What, then, was the religion of those who built this shrine, if shrine it was? The ornamentation of that part of the outer wall which faces the rising sun suggests sun-worship. The phalli (if they are phalli) point to one of the Oriental forms of the worship of the forces of nature. The birds' heads may have a religious significance, and possibly the significance which it is said that vultures had in the Syrian nature-worship. These data give some slight presumptions, yet the field for conjecture remains a very wide one, and there is nothing in the buildings to indicate the particular race who erected the Fort and the Temple (if it was a temple). However, the tower bears some resemblance to a tower which appears within a town wall on an ancient coin of the Phoenician city of Byblus; and this coincidence, slight enough, has, in the dearth of other light, been used to support the view that the builders belonged to some Semitic race.

Had we nothing but the ruined walls of Zimbabwye, Dhlodhlo, and the other spots where similar ruins have been observed, the problem would be insoluble. We could only say that the existing native races had at some apparently distant time been more civilized than they are now and capable of building walls they do not now build, or else we should suppose that some now extinct race had built these. But there are other facts known to us which suggest, though they do not establish, an hypothesis regarding the early history of the country.

In very remote times there existed, as is known from the Egyptian monuments, a trade from South-east Africa into the Red Sea. The remarkable sculptures at Deir el Bahari, near Luxor, dating from the time of Queen Hatasu, sister of the great conqueror Thothmes III. (B.C. 1600?), represent the return of an expedition from a country called Punt, which would appear, from the objects brought back, to have been somewhere on the East African coast.[8] Much later the Book of Kings (1 Kings ix. 26-28; x. 11, 15, 22) tells us that Solomon and Hiram of Tyre entered into a sort of joint adventure trade from the Red Sea port of Ezion-geber to a country named Ophir, which produced gold. There are other indications that gold used to come from East Africa, but so far as we know it has never been obtained in quantity from any part of the coast between Mozambique and Cape Guardafui. Thus there are grounds for believing that a traffic between the Red Sea and the coast south of the Zambesi may have existed from very remote times. Of its later existence there is of course no doubt. We know from Arabian sources that in the eighth century an Arab tribe defeated in war established itself on the African coast south of Cape Guardafui, and that from the ninth century onward there was a considerable trade between South-east Africa and the Red Sea ports—a trade which may well have existed long before. And when the Portuguese began to explore the coast in 1496 they found Arab chieftains established at various points along it as far south as Sofala, and found them getting gold from the interior. Three things, therefore, are certain—a trade between South-east Africa and the Red Sea, a certain number of Arabs settled along the edge of the ocean, and an export of gold. Now all over Mashonaland and Matabililand ancient gold-workings have been observed. Some are quite modern,—one can see the wooden supports and the iron tools not yet destroyed by rust,—and it would seem from the accounts of the natives that the mining went on to some small extent down to sixty years ago, when the Matabili conquered the country. Others, however, are, from the appearance of the ground, obviously much more ancient. I have seen some that must have been centuries old, and have been told of others apparently far older, possibly as old as the buildings at Zimbabwye. I was, moreover, informed by Mr. Cecil Rhodes (who is keenly interested in African archÆology) that he had seen on the high plateau of Inyanga, in eastern Mashonaland, some remarkable circular pits lined with stone, and approached in each case by a narrow subterranean passage, which can best be explained by supposing them to have been receptacles for the confinement of slaves occupied in tilling the soil, as the surrounding country bears mark, in the remains of ancient irrigation channels, of an extensive system of tillage where none now exists. The way in which the stones are laid in these pit-walls is quite unlike any modern Kafir work, and points to the presence of a more advanced race. Putting all these facts together, it has been plausibly argued that at some very distant period men more civilized than the Kafirs came in search of gold into Mashonaland, opened these mines, and obtained from them the gold which found its way to the Red Sea ports, and that the buildings whose ruins we see were their work. How long ago this happened we cannot tell, but if the strangers came from Arabia they must have done so earlier than the time of Mohammed, for there is nothing of an Islamic character about the ruins or the remains found, and it is just as easy to suppose that they came in the days of Solomon, fifteen centuries before Mohammed. Nor can we guess how they disappeared: whether they were overpowered and exterminated by the Kafirs, or whether, as Mr. Selous conjectures, they were gradually absorbed by the latter, their civilization and religion perishing, although the practice of mining for gold remained. The occasional occurrence among the Kafirs of faces with a cast of features approaching the Semitic has been thought to confirm this notion, though nobody has as yet suggested that we are to look here for the lost Ten Tribes. Whoever these people were, they have long since vanished. The natives seem to have no traditions about the builders of Zimbabwye and the other ancient walls, though they regard the ruins with a certain awe, and fear to approach them at twilight.

It is this mystery which makes these buildings, the solitary archÆological curiosities of South Africa, so impressive. The ruins are not grand, nor are they beautiful; they are simple even to rudeness. It is the loneliness of the landscape in which they stand, and still more the complete darkness which surrounds their origin, their object, and their history, that gives to them their unique interest. Whence came the builders? What tongue did they speak? What religion did they practise? Did they vanish imperceptibly away, or did they fly to the coast, or were they massacred in a rising of their slaves? We do not know; probably we shall never know. We can only say, in the words of the Eastern poet: