The date palm is a comparatively late arrival in Air, where it was introduced from the north. The trees are a cross of the Medina and Fezzan varieties. As elsewhere in North Africa, each tree is an immovable asset like a house, and often does not belong to the same man as the ground on which it is grown.
At the foot of the palms were numerous gardens growing vegetables and grain. The fort had a wonderful kitchen garden with all sorts of melons, gourds and welcome European green food. The French officer in command of the post used to declare that Iferuan was the Switzerland of the Sahara, and the cool climate seemed to justify his praise. The Tuareg buildings had nothing remarkable about them with the exception of the large mosque of Tefgun not far away, and the khan or caravanserai built on the Arab plan. The Sudanese habit of making large clay amphoræ and baking them in situ, for the storage of wheat and millet grown in the gardens, has been adopted in Iferuan, and to my knowledge not elsewhere in Air.
Although the open desert on the way to Ghat is not reached much before In Azawa, several days further north, now, as in the past, Iferuan is the last permanently inhabited point in Northern Air. Between these points the mountain mass of Fadé has first to be crossed; it contains several watering-points and some pastures, and huts were occasionally built at a pool called Zelim, but they had no permanence. The mountains and the watering-places have long since been abandoned by their old owners, the Ifadeyen and Kel Fadé and now belong to the Ikazkazan and Kel Tadek tribes.
At Iferuan several important roads meet. The road from Air to Tuat and to Ghat, which is the main north and south caravan track across the Central Sahara, and the Haj road from Timbuctoo to Cairo, all three have a stage in common from Iferuan to In Azawa. The Haj road used to leave the Niger at Gao and enter Air at In Gall, whence it skirted the western edge of the plateau and then turned into the mountains to Iferuan: after passing In Azawa and Ghat it ran through Murzuk, Aujila and Siwa to Cairo. From Iferuan there are also several roads to the west, while the northern of the two alternative eastern roads across the desert to Kawar equally started from there, running, as already stated, by way of Taghmeurt n’Afara, Agamgam and Ashegur.
In seeking to identify Air with the Agisymba Regio of the Roman geographers, Duveyrier presumed that the Fezzanian Garamantes were in the habit of visiting the plateau in ox-drawn chariots or wagons. If they had, in fact, done so, it is logical to suppose the road they used would have come to Iferuan or one of the Ighazar villages. Indeed he states that he heard rumours of a direct road from Murzuk or Garama to Air, a “Garamantian way” which passed through a place called Anai, where there were rock drawings similar to those found in Algeria and Tripolitania. This Anai was south-west of Murzuk and must not be mistaken for the better known Anai of Kawar, which is north of Bilma on the Murzuk-Chad road.
I was at particular pains to inquire into the existence of this road from all the most prominent guides and personages in Air whom I could find. It would have been peculiarly interesting to establish its existence, for Duveyrier says, “La voie, avec ses anciennes ornières, est encore assez caractérisée pour que les Tebou, mes informateurs, qui en arrivaient, n’aient laissé dans mon esprit aucun doute à ce sujet.”[277] Other writers, presumably on his authority, have added that where this road crossed the sand, stone flags were laid for the wheels to pass over. Duveyrier’s informers stated that the petroglyphs at Anai represented ox-drawn vehicles, and that the road also passed by way of Telizzarhen, where Barth discovered the famous rock drawings depicting men with animal heads.[278] While the broad valley at T’intellust would afford easy passage for a wheeled vehicle, there is no way to the south for any but pack transport. There are no signs of any road for vehicles ever having existed either east or west of the Bagezan massif. The great Kel Owi road is only fit for pack animals; and although many parallel tracks are visible in the open country there are numerous defiles where a single path only a few inches broad occurs. I am convinced that wheeled transport could never have been used anywhere in Central or Southern Air. But, it may be asked, could chariots have arrived even as far as T’intellust or Iferuan? There are only three ways into the plateau from the north-east that are at all suitable even for loaded camels. They are (a) through the Fadé mountains to Iferuan, (b) by Fares water and the Agwau valley to the great north-east basin, and by Taghmeurt n’Afara to T’intellust. The first two are not practicable for wheeled traffic, and on hearsay evidence the third one is equally out of the question. I do not, therefore, think that wheeled transport could ever even have entered Air from the north or north-east, though wagons might, of course, have come as far as the borders of the mountains to points such as Fares or Agamgam, provided the surface of the desert were hard enough. This cannot be determined until Anai and the country between it and Air have been visited.
If any direct road between these areas ever existed, it is very unlikely to have run straight from Anai to T’intellust, as Duveyrier’s map shows. In my inquiries I heard in all of only four roads across the Eastern Desert: (a) the southernmost from Damagarim by Termit;[279] (b) the direct road to Fashi and Bilma from Southern Air, starting at Tabello; (c) the old Kel Owi Taghalam road from Agamgam to Ashegur, whence one branch goes north to Jado oasis and the other south to Fashi; and (d) a northern road from Fadé to Jado direct. Guides like Efale, who know every part of the Eastern Desert, state that there is no road from Air direct to Murzuk which does not go either by way of Jado or by way of the usual caravan road between Kawar and the Fezzan. The northernmost road from Fadé to Jado runs through two places called Booz and Ghudet, where water is found a short way below the surface; Efale travelled this way in his youth. He told me that it was known to and used by Tebu raiders to-day. But there are no deep wells on this track to be filled up to prevent raiders passing down the old Garamantian way, as Duveyrier implies was done. From Jado it, of course, is possible to reach Murzuk either by Anai or by joining the usual Chad road via Tummo. The existence of this northern Anai is certainly substantiated, and Jado, a Tebu oasis with a palm grove, is known to exist. It is called by this name among the Arabs, but Agewas by the Tuareg of Air and Braun by the Tebu themselves. The place has been reconnoitred by certain French officers, one of whom, a commandant of the fort of Bilma, I had the good fortune to meet. He was aware of the story of a flagged road, but after visiting Jado several times found no trace of any such track and did not believe in its existence. That the Garamantes and, indeed, other inhabitants of the Fezzan at one period in history used chariots drawn by oxen is quite likely, but it is highly improbable that they ever ventured so far afield in them as Air.
The existence of a road between Air and the Fezzan may be admitted as possible, but only on condition that it is not made to run direct between these countries. South of Anai it would almost certainly pass through Jado, and thence may have reached the plateau either by Ghudet and Booz to a water-point called Temed[280] on the eastern edge of Fadé north of the Tamgak group, or else by Ashegur and Agamgam north-east of T’intellust. This is not the road of the Garamantes on Duveyrier’s map; and beyond this his story cannot be further substantiated. As against this line of argument it must be observed that Von Bary[281] during this stay in Air collected information which led him to believe that there was a road from Air to Jerma by way of Anai. It is implied that it went direct, but he was never able to learn any details and was probably influenced by Duveyrier’s statements. He heard that there were some traces visible, but found no evidence to confirm the report of flagstones, wheel-marks or sculpture along its course.
There is nevertheless one piece of evidence which militates in some measure against my belief that chariots never were seen in Air, and that is a rock drawing which I found in Air on a boulder in the Anu Maqaran valley just west of Mount Arwa. The drawing is reproduced in [Plate 41.] In the conventional manner adopted in these designs it represents oxen pulling four-wheeled vehicles. The identification of the ox is confirmed from the many other similar pictures of this animal on rocks in Air. The object behind it must apparently be a cart. The whiteness of the marks in the Anu Maqaran drawing appears to indicate that it is a comparatively recent production, although the colour and degree of patination of Saharan drawings are of course no real criteria, for weathering is notoriously uneven in its action. Near the drawing of the ox and chariot, but on a different boulder, was the magic square shown in the same figure. Both drawings were in a very sheltered place and seemed contemporary. The evidence of this picture of the chariot or wagon is too unreliable and slender to establish any theory, but it is certainly difficult to understand where the draughtsman obtained his idea except as a result of seeing chariots drawn by oxen, a condition which does not, I think, obtain in the Fezzan to-day. Wheeled vehicles have only been known in the Sudan since they were imported by Europeans during the last twenty years, and I am not aware that even those are ox-drawn. Furthermore, although the most puzzling point about the Anu Maqaran rock drawing is its apparent modernity, which is paradoxical in view of the disuse of wheeled vehicles in the Sahara, it is almost certainly older than this century. Yet the application of an ox to a cart is not likely to have been imagined by any Tuareg who had not seen an instance of it, and there seems to be no adequate reason for him to reproduce his knowledge on a rock in Air even if chance had taken him so far afield as the Mediterranean littoral, where he might have seen the equipage, unless it had in some way become associated with Air.
The identification of Air with the Agisymba Regio of the Romans has been accepted by many authorities other than Duveyrier. It raises the whole problem of the Roman penetration of the Sahara. They are known to have administered the Fezzan, and it is even pretended that they reached the Niger, but evidence on this point is more scanty. Doubtless as the exploration of the Central Sahara is carried out systematically further evidence of their penetration will come to light. I am, for instance, not aware that any remains have actually been found at Ghat, though the city, which was known to them as Rapsa, was almost certainly that place and was visited in 19 B.C. by Cornelius Balbus. The Roman remains discovered by Barth on the road from Mizda over the Hammada el Homra to Murzuk are better known. This route seems to have been opened about the time of the Emperor Vespasian, and to have rendered possible or at least easier the occupation of the Fezzan, which had, however, already been visited by military expeditions earlier than that reign. Pliny writes: “Ad Garamantes iter inexplicabile adhuc fuit. Proximo bello, quod cum Œensibus Romani gessere auspiciis Vespasiani Imperatoris, compendium viæ quatridui deprehensum est. Hoc iter vocatur ‘Præter caput saxæ.’” Evidently the road was called by the natives, even in those days, by the same name which it now possesses, for the Pass over the Red Rock Desert at 1568 feet above the sea is still known to the Arabs as “Bab Ras el Hammada.”[282] In about A.D. 100[283] Septimius Flaccus penetrated from the Fezzan into Æthiopia at the head of a Roman column; Julius Maternus marching from some point on the coast to Garama had joined forces with the Garamantes in order to proceed southward together against various Æthiopian bands. By this date, then, it is probable that an occupation of the Fezzan had been accomplished, for this alone would justify a further advance or punitive expeditions on such a scale against raiders from the south. Indeed, from the account given by Pliny[284] of Cornelius Balbus’ expedition of 19 B.C. to the Fezzan, it might be supposed that the occupation of Southern Tripolitania and the Central Sahara had taken place a century earlier. The identification of the cities conquered by Balbus has not been satisfactory except in the case of Cydamus, Cillaba or Cilliba, Tabudium,[285] Rapsa and Jerma, respectively Ghadames, Zuila,[286] Tabonie, Ghat and Garama; the last named being the capital of the Garamantes and of the whole Fezzan, a position which later passed on to other places and finally to Murzuk.