In addition to these totals of lengths of passages cleared out or traced, there are many other passages still buried in débris, the outcrop of their side walls being seen here and there on the surface near several ruins. Many, of course, must be completely buried under the veld, for some were lately discovered at least 3 ft. below the surface, with native paths crossing them in all directions, while it is quite reasonable to suppose that with the great area of ruins yet unexplored very many more passages will yet be found, especially when it is recollected that the discovery of one buried passage has most frequently led to the discovery of several side passages.

Traces were found of two other passages leading from the base to the summit of the Acropolis Hill, and these remain unexplored, and each would be fully 900 ft. in length, while traces of several lines of passages are to be seen encircling at various heights the south, west, and north faces of the Acropolis Hill. These also at present remain unexplored.

There are many points of interest concerning these passages:—

(a) Passages were evidently constructed as part of the plan of the fortifications, but in some instances only as means of communication between certain buildings within the fortified area and for securing privacy. In the one class of passage buttresses and traverses are repeated with a marvellous redundancy; in the other class of passage not a single buttress or traverse is to be found.

(b) In passages leading from main ruins to exterior buildings the walls of the passage nearer the main ruins are better built, and the steps and floors are better constructed in the portions nearer the main ruins than are those of the more distant portions of the passage. So imperceptibly do the better-built portions merge into the less superior class of wall that it is extremely difficult to ascertain the exact point where the change in the quality of the construction takes place, though the difference in the class of building at one end of the passage and that of the other is most obvious. But though this difference in the construction of the passage walls is so apparent, there is no suggestion that portions were of a later period, for they are built upon one plan, have one line of direction, serve as a complete communication with one obvious and particular point, and one length of the passage without the other would be purposeless, so far as the intention of the builders may be gathered. With regard to the passages ascending the Acropolis Hill, the completeness of the plan of these passages is best seen from the summit of the hill or from the summit of Makuma Kopje on the opposite side of the valley, from which heights respectively a complete view of those passages in their entire length is to be obtained.

(c) Excepting some of the passages in the Elliptical Temple and a few others on the Acropolis, all the passages at Zimbabwe are exceedingly narrow and tortuous, many being only shoulder wide, while, owing to their winding lengths, it is not possible to see many feet on ahead. Such of these passages as have their floors below the levels of adjoining enclosures have in many places their side walls bulged by the weight of earth and débris behind into the passage-ways, and in some such instances the side walls have collapsed and blocked up the passages.

(d) Almost every passage appears to have originally been paved with blocks which were covered over with granite cement, but the cement, except in a few instances, has decomposed and been washed away by centuries of rains, though abundant traces of it remain.

(e) Sunken passages built very much below the levels of the ancient floors on either side of them are numerous. The best instances of sunken passages are the North-East Passage between the Elliptical Temple and the Valley of Ruins, also the North-West Ascent to the Acropolis (upper portion), and the sunken passage in the Eastern Temple on the Acropolis.

(f) The walls of the ascents to the Acropolis as originally built would have precluded any outsider from seeing, even if standing on an adjoining kopje, the movements of people passing up and down the ascents; and to-day as a native ascends these passages it is almost impossible to see him till he reaches the summit, except as he is passing gaps or walls which have become considerably dilapidated. Some of the outer walls of these ascents are still 10 ft. in height.

(g) The Elliptical Temple and the Western Temple on the Acropolis have each long and narrow and deep parallel passages on the inside of their main walls, and it is possible that the Pattern Passage served for a similar purpose at the Eastern Temple. The Parallel Passage in the Elliptical Temple communicated only between the North Entrance and the Sacred Enclosure where are the conical towers, and this passage has no communication with any other portion of the interior of the temple. Several of the known writers on these ruins, including Bent, have conjectured that these parallel passages in the temples were reserved for the use of the priests.

(h) Cliffs and large boulders have been frequently utilised to form lengths of passages. Instances of this practice are to be seen on the Acropolis in the Rock Passage of the South-East Ancient Ascent, Buttress Passage, North Passage, and elsewhere. In some instances the walls are made to go out of their line so as to include neighbouring boulders, the sole object, so it would appear, being to deprive any invading force of the vantage offered by the height of the boulders for an attack to be made on the passage.

(i) There are no evidences that any of the passages, except as stated later, were ever roofed. Possibly the winding stairs and the sunken passage in the Eastern Temple were originally covered over, as a great quantity of long, flat slate beams were found on their floors. It is believed that a single wall once crossed over the sunken passage in Platform Enclosure at about 15 ft. from its upper end, for when this passage was opened in 1902 slate beams were found at this spot, but at no other point in the passage. The passage through the main west wall of the Western Temple, which was blocked up by a Makalanga-built wall, of course, was covered over by the main wall, while the Covered Passage in the same temple remains intact as originally built. Moreover, the widths of many of the passages though narrow on their floors are wide at the summits of their side walls, and their irregular form precludes suggestion of any roofing having been placed over them, some being doubly as wide as the longest of the slate and granite beams found, beside which the general absence of long slate and granite beams on the floors of the passages would seem to further negative any such conjecture. The West Passage leading to the South Cave was not artificially roofed over, but the outer wall was raised up to the height of the boulder which overhangs the passage.

ENTRANCES AND BUTTRESSES

When in 1891 Bent approached Zimbabwe through North Bechuanaland, Gwanda, Tuli, and Belingwe, he passed through the centre of that area in which the earliest of the many ancient ruins of Rhodesia are located. All the ruins he described or mentioned had rounded ends of walls and rounded buttresses, all angular features being conspicuous by their absence. This fact appeared to him so striking that he was constrained, after comparing these ruins with Zimbabwe, to believe that such rounded features belonged to the earliest period of Zimbabwe architecture. Fully a score of competent writers on our ruins, whose valuable and trustworthy contributions, based on personal examination of the same area, have been welcomed by the leading scientific associations of Great Britain and Germany, are also emphatic as to the rounded entrances and buttresses being one of the chief distinctive features of the earliest Zimbabwes. This is further demonstrated in the detailed descriptions of almost one hundred ruins within the same area which are given in The Ancient Ruins of Rhodesia, in the great majority of which ruins angular features, except in reconstructions of a later period, are altogether absent.

But the Great Zimbabwe being the finest type of that early class of ancient building, it may be interesting to know that Bent’s conclusion is thoroughly confirmed by these ruins.

All ends of walls which are still intact are rounded, there being only a few examples so far discovered of angular-ended walls.

The above figures show conclusively that these rounded features, excluding the ends of walls which are almost always rounded, are in a far greater proportion than 146 to 13 which are angular, and at least three of the latter, if not others, for reasons explained elsewhere, can be shown to have been erected at a much later period, one being built upon a floor of common Makalanga daga, and another débris containing ordinary Kafir articles of no very great age.

All the entrances in the main outer walls, save one, are rounded, the few angular entrances being found, with two exceptions, in slighter walls, mainly divisional, some of which were erected later possibly to suit the immediate convenience of later occupiers, for divisional walls had been removed, reconstructed, or entirely fresh ones erected in new directions in almost every ruin, and in some instances the foundations of the later walls cross at right or oblique angles over the reduced summits of older divisional walls.

Walls of the earliest period widen out as they near entrances. This feature is not present in plumb and angular walls of later construction.

There is no evidence whatever in the rounded entrances that they were ever covered over, but in two angular entrances on the Acropolis the butts of the broken slate lintels still remain in the side walls.

Although there are not sufficient proofs to enable one to definitely determine whether the rounded entrances as a rule were once covered over, some of the evidences to negative the covering in of rounded entrances may be noted:—

(a) Had such entrances been roofed in, the collapse of the lintels must have brought down far more of the walls than have fallen.

(b) The courses of the blocks at the necessary height above the floor of the entrances on either side do not always correspond.

(c) The top courses near the summit of the walls on either side of the entrances show distinct signs of curving inwards towards the entrances. This is particularly noticed on the east side of the north-west entrance to the Elliptical Temple.

(d) No splinters of slate or granite beams which could have been used as roofing were found in any of the very many rounded entrances.

(e) Two intact rounded entrances, one open up to the summit on either side to a height of 19 ft., one entrance being at the east end of Pattern Passage on the Acropolis.

No main entrance has buttresses on either hand on the outer side, possibly because these would have provided any attacking party with excellent shelter. All buttresses of such entrances are on the inside. Divisional entrances which have buttresses have them on the inside only.

The entrances through a wall of the earlier period are carried over the common foundation in the opening forming the steps, which were evidently constructed before the side walls were erected. These steps are large, broad, and high, and where intact look most imposing. Such entrances resemble stiles, as they are much higher than the levels of the floors on either side.

The entrances through an angular wall of a later period have steps which are not part of either side walls, but were built in after the entrance passage had been constructed, and these show poor workmanship and are very shallow, and recede only two to four inches. As the levels of the enclosures on either side have filled in over the original floors, such “cat-steps” have in some instances been built over the original large steps for the purpose of raising the floor of the entrances, seeing that the enclosures on either side had been filled in some feet above their original levels.

Directly opposite the main entrance of the “Outspan Ruins” is a large circular buttress, as if it were intended to divide any attacking party into small numbers.

CAUSE OF DILAPIDATION TO ENTRANCE BUTTRESSES

The entrance buttresses with portcullis grooves are in most instances comparatively small, some projecting only two to three feet towards the interior of the building, and these are built up against main and divisional walls, and are in point of construction altogether independent erections, there being no dovetailing or binding between the buttresses and the walls.

In some of the entrances the side lintels of slate, granite, and unworked soapstone beams have been found built into the portcullis grooves. In The Ancient Ruins of Rhodesia it was noticed that at several of the ruins therein mentioned stone side lintels were found in situ. The stone lintel posts in situ at Zimbabwe had not then been discovered. The tallest of such stone lintels at Zimbabwe is 8 ft. above the ground. The buttresses appear to have been built after the stone posts had been erected, for the walls at the sides of the lintel follow the irregularities of the side faces of the beams.

The great destruction which has occurred to these structures might possibly be accounted for by (1) the weight of the stone lintel on getting off the perpendicular, which would lever down the buttress into which it was built; (2) the foundations of buttresses are not so deep as those of the main wall up against which they were built; (3) when some later people, possibly natives, deliberately built up and blocked the entrances they might have used the blocks of these buttresses for their building material; (4) the passage-way between each pair of buttresses being so very narrow, damage could easily have been wrought by ordinary traffic; and (5) the main walls are much higher than the summits of the buttresses, and the walls on either side of the entrances being always more dilapidated on the summits, the falling of huge masses of masonry on to the buttresses immediately below might have effected their destruction.