There is an interesting tradition regarding the subterraneous gallery recorded by El-Bekri,[135] who says that El-Kahina, the celebrated chieftainess of the Aures, having been besieged in this amphitheatre, which she had converted into a fortress, caused a passage to be excavated in the rock as far as Sallecta, large enough to permit several horsemen in line to pass along. By this means she obtained supplies of provisions, and everything she required.
The Sheikh et-Tidjani also says that when El-Djem was subsequently attacked by Yehia ibn Ishak el-Mayorki, Prince of the Balearic Islands, about the six hundredth year of the Hedjira, he was soon compelled to raise the siege after a most ignominious manner. The defenders, to show how well they were supplied with provisions, threw down fresh fish at their besiegers, which they had obtained by means of the subterranean passage to Sallecta.[136] These traditions linger amongst the people of El-Djem to the present day. Careful excavations in the basement of this structure could hardly fail to be richly rewarded.
It is by no means certain that this amphitheatre ever was completed. If we may judge from Bruce’s sketches, as well as the actual condition of the monument, it is doubtful whether the attic ever was decorated with pillars, though undoubtedly some of the pedestals of this order were placed in position. Some of the ornamental details also are in an unfinished condition. The keystones of the arches of the lowest order were probably all intended to be sculptured, as in the amphitheatre of Capua; but they are still in their original rough condition, with the exception of two, one of which bears the head of a human being, and the other that of a lion. Still, neither of these facts actually proves that the amphitheatre remained uncompleted; in many similar buildings it was never intended to finish all the details with minute care, and even in the Coliseum some of the capitals are but roughly sketched out.
The outside gallery on the ground floor, where most perfect, has been utilised by the Arabs as store-rooms for their corn and forage; some of the arches are converted into shops, and there is evidence that the upper galleries also have at some time or other been converted into dwellings, holes in the masonry for the reception of joists being visible in every direction.
Several inscriptions have been found here; the most important has been preserved in the enclosure of the Chapel of St. Louis at Carthage, and has been often quoted: the name of the town is twice mentioned in it, once as Thysdrus, and again as Thysdritana Colonia.[137]
A number of rude Arabic or Cufic inscriptions, accompanied by representations of swords and daggers, have been scratched on the exterior wall above the principal entrance, and one, which is certainly of Berber origin, may date from the era of El-Kahina.
I am not aware of any sculpture now existing which has been found at this place. Desfontaines purchased a small head of Diana in white marble during his visit in 1784, and he mentions a report current that several fine statues had formerly been carried off by English travellers.[138]
The stone of which the amphitheatre is built was obtained from Sallecta on the sea-coast: the Sallecti of the tables of Peutinger and the Syllectum of Procopius, the first resting-place of Belisarius in his march from Caput Vada to Carthage. The natives assured me that between this place and El-Djem the remains of the ancient paved road can easily be traced. The stone itself is of the youngest geological formation (Pliocene age), belonging to the raised coast-beaches found at from 200 to 600 feet above the present level of the Mediterranean. It is a somewhat fine-grained marine shell-limestone, with an admixture of siliceous sand full of fossil shells of considerable size, such as Pectunculus and Carduum, but no microscopic forms of shells are visible amongst the fine grains of shell-sand which make up the rock. Such a material is worked with the utmost facility; indeed, it may be cut with an axe, but it is not susceptible of being dressed with the same precision as more compact stone. The consequence is that the masonry is far inferior to the finest specimens of Roman work in Africa. Mortar has been plentifully used between the joints, and the stones are neither as large nor as closely fitted as usual; the average dimensions are—length, 37⅜ inches, and height of courses, 19⅝ inches.
Another feature of the construction of this building, never seen in others of the best period of Roman art, is the manner in which the appearance of nearly all the stones has been spoilt by triangular lewis holes being cut in their exterior faces, for the purpose of raising them into position. This gives the masonry a very slovenly appearance. The dimensions given by Bruce are:
| Ft. | In. | |
|---|---|---|
| Length of entire structure on major axis | 488 | 0 |
| „ „ minor axis | 400 | 0 |
| Length of arena, major axis | 213 | 0 |
| „ „ minor axis | 172 | 0 |
| Depth of foundations | 32 | 11 |
| Height of first stage to impost of arch | 21 | 11 |
| „ „ above impost | 15 | 2 |
| Height of second stage to impost | 21 | 11 |
| „ „ above impost | 16 | 4 |
| Approximate total height of building, including foundation | 183 | 7 |