These operations of Septimius Flaccus and Julius Maternus have been held to concern Air. The latter, ἀπὸ Γαράμης ἅμα τῷ βασιλεῖ τῶν Γαραμαντίων ἐπερχομένῳ τοῖς Αἰθιόψιν ὁδεύσαντα τὰ πάντα πρὸς μεσημβρίαν μησὶ τέσσαρσι ἀφικέσθαι εἰς τὴν Ἀγίσυμβα. . . .[287] It is important to try to identify the area, since it appears to be the most southerly point to which Roman geographical knowledge is recorded as having extended. Duveyrier, arguing, on what may in any case be a false premise, that because Pliny mentions no camels in Africa there were no camels, concludes with the fantastic statement that the Romans must have used wheeled transport on their expeditions, and that that is why the “Iter præter caput saxæ” played such an important part in their operations; but I have seen no evidence which might lead one to suppose that this route over the Hammada el Homra was fit for wheeled traffic. The Garamantes were said by Herodotus to have used wagons drawn by four horses.[288] From this Duveyrier concludes that at a later date oxen were substituted for horses, and that in virtue of a perfectly imaginary road from Murzuk by way of Anai Air must be the Agisymba Regio. He gives no convincing reason for the identification, but implies that by a process of elimination it must be so. The name Agisymba and Bagezan have been connected by displacing the terminal and initial syllables respectively of the two words, but undoubtedly it was not this so much as the existence of a Garamantian road which appealed to the learned author.
One of the principal objectives which I had in mind in visiting Air was to seek evidence of Roman penetration. In the course of their long historical knowledge and occupation of the Fezzan, it seemed natural for the Romans to have explored the Air road. But I found no remains, nor evidence whatsoever of their penetration, not even at points which had considerable strategic value. Some more fortunate traveller than myself may one day chance upon an inscription or a camp. Such a discovery in so vast and little known a land is quite conceivable, but up till now the weight of evidence is against the Romans ever having come to Air. There is a certain historical analogy in the fact that the Arabs never invaded the country either. Their influence on the Tuareg of Air was confined to an unenthusiastic conversion to Islam in comparatively recent times. On the other hand, the Arabs in the first century of the Hijra, like the Romans, seem to have descended the Chad road at least as far as Bilma, and again, Arab influence in Central Africa east of the lake is at least as strong as, and perhaps even greater than, the Western Arab-Moorish influence on the Upper Niger.
I am, however, much more inclined to regard Tibesti and not Air as the Agisymba Regio. We find the Arabs in the Fezzan evidently feeling the same necessity of expansion southwards along the Chad road as did the Romans. By 46 A.H. the Fezzan had already twice been conquered by the Arabs, first in 26 A.H., soon after the occupation of Egypt had been completed and the attention of Islam was turned to North Africa, and again when the inhabitants had cast off their servitude to the Arabs. Okba ibn Nafé was induced by this breach of faith[289] to leave his army, which was on its way to conquer Ifrikiya (Tunisia and Western Algeria), at Sert in the Great Syrtis, and to lead an expedition to reconquer the desert. He took Wadan and Jerma, near Murzuk, and the last strong places of the country, and asking what lay “beyond,” learnt of the “people of Hawar,”[290] who had a fortress on the edge of the desert at the top of an escarpment. It was said to be the capital of a country called Kawar, the name which is borne even to-day by the depression along which the main caravan road passes south through Bilma and other small villages, any one of which may have been their stronghold, which El Bekri[291] also calls Jawan. After a march of fifteen nights Okba came to this place and eventually captured it. At one moment his expedition nearly perished of thirst, but according to the story Okba’s horse found water in the sand and saved the column, wherefore the place was called Ma el Fares, the “Water of the Horse.” This point is now spelt Mafaras on the Murzuk-Kawar road in about Lat. 21° 15′ N.[292]
The Romans seem to have had much the same experience as the Arabs, though we can identify the movements of the latter with greater certainty. The expedition of Septimius Flaccus and Julius Maternus started from Garama. Now an expedition from the Fezzan proper to Negroland would normally have proceeded along the Chad road, which runs south, and not in the direction of Air, which lies south-west. Furthermore, we have already seen that there is no direct road from the Fezzan to Air save by making a detour via Jado and crossing the worst part of the desert. Had the Romans intended to use the Air road to Negroland they would assuredly have started from Rapsa (Ghat) and not from Garama; alternately had they started from Garama and proceeded by way of Ghat, it is likely to have been mentioned, nor would the enterprise have been so directly connected with the Garamantes. After marching south from Garama the expedition reached the Agisymba Regio. But if the Air mountains are neither south of Garama nor on a direct road from that place, both these conditions do apply to Tibesti. This country lies due south of the eastern Fezzan and there is a direct road from Garama by way of Tibesti to Negroland, though it is not so well known as the main Chad road. The latter trade road, however, and the Tibesti mountains seem to fit the description of the course taken by the expedition sufficiently well, and clearly better than the Air road and plateau. The Romans, we are told, marched for three months to the south; it may be objected that this would be an inordinately long time to take on a journey to Tibesti and that Air, being somewhat further away from Garama, is the more probable. But expeditions may take longer or shorter times to traverse any particular desert road according to the difficulties encountered, the fighting sustained and the pasturage available on the way for the transport animals, and I do not think that any conclusion can be drawn from the reported length of the march. A period of three to four months might as easily bring one expedition from the Fezzan to Tibesti or to Air as it would be insufficient for another under different conditions but on the same road to get more than half-way.
PLATE 45
ASSARARA
If circumstantial evidence seems to point to Tibesti, there is also that of the place names given in the account. The Agisymba Regio contained the mountains of Bardetus, Mesche and Zipta. No similarity to these names can be found in Air, but in Tibesti the first may well be the area and massif round the village of Bardai, while Mesche may be a Latinised form of Miski, a valley and group south-west of Bardai. For Zipta I can offer no suggestion.
Like the Romans and the Arabs the modern Turks also penetrated Tibesti as a consequence of their occupation of the Fezzan in an attempt to stop the Tebu raiding. History is curiously consistent in that we have no evidence of the Arabs or the Turks having penetrated Air. The Romans, I assume, probably did not do so either.[293]
The Romans must have come into contact with the Tuareg in the Fezzan, where the latter, it might be assumed from Arab evidence alone, were early established if they did not actually constitute the majority of the original population. It is possible to trace in Roman records the names of certain well-known Tuareg tribes. The description which Corippus gives of the Ifuraces, the Ifoghas tribe of the Southern Tuareg, corresponds accurately with that of the present-day camel riders of the Sahara. In a description of an encounter with the Byzantine forces under John, the general himself cuts down a camel with his sword and the rider falls with the accoutrements and paraphernalia, which are those of a Tuareg on campaign or in battle to-day.[294] The activities of the Circumcelliones during the troubles described by Opatus[295] during the Donatist heresy in North Africa in the course of the fourth century A.D. remind one irresistibly of those of the Tuareg. These bands of marauders from the desert came into Southern Tunisia and Algeria on swift and remorseless errands of plunder for the greater glory of their heretical Faith. They lived in the barren hills of the outer waste and descended to burn churches, sack houses and carry off live-stock with such deadly efficiency and ease that the motive power of their organisation can only have come from a spirit which considers raiding a national sport. “When they were not resisted they usually contented themselves with plunder, but the slightest opposition provoked them to acts of violence and murder. . . . The spirit of the Circumcellians, armed with a huge and weighty club, as they were indifferently supplied with swords and spears, and waging war to the cry of ‘Praise be to God’ . . . was not always directed against their defenceless enemies, the peasants of the orthodox belief; they engaged and sometimes defeated the troops of the province, and in the bloody action of Bagai they attacked in the open field, but with unsuccessful valour, the advance guard of the Imperial cavalry.”[296]