did. His quarrels with Belkho seem to have been in part due to his own

tactlessness and discourtesy, and in part to his inability to realise that the chief, for political reasons, did not desire him to go to the Sudan. Von Bary returned to Ghat, meaning to try once more to reach Nigeria as soon as he had picked up his stores and some more money, but his diary ends abruptly with the remark that he would be ready to start south again from there in fifteen to twenty days. He died within twenty-four hours of reaching Ghat, on 3rd October, 1877. He had spent a cheerful evening with Kaimakam,[26] and had gone to bed; at 6 a.m. he was breathing peacefully asleep; by ten o’clock he was dead. His death does not seem to have been quite natural. It remains one of the mysteries of the Sahara. Von Bary’s account of Air[27] is very incomplete and his observations are coloured by the hardships which he suffered. With the exception of certain botanical information and notes on one or two ethnological points, his descriptions contain little that had not already been made known by Barth.

Then began that competition among European Powers for African colonies which was soon to reach a critical stage. The Anglo-German Convention of 1890 had proposed to divide Africa finally, but before that date the French had seen one desirable part after the other fall to our lot. They determined before it was too late to take as much as possible of what still remained unallocated. Central Africa, east of Lake Chad, certain tracts of indifferent country on the western coast and the greater part of the Sahara were still unclaimed by any European Power. And so it was that in France the magnificent scheme was conceived of sending three columns from north, west and south to converge on Lake Chad, and formally to take possession of the lands through which they passed in accordance with the stipulations of the Congress of Berlin, where it had been laid down that territorial claims were only valid if substantiated by effective occupation. It was not till 1899, however, that the French plans reached maturity. Three expeditions duly set out from the Congo, the Western Sudan and Algeria to cross Africa and meet on Lake Chad. Their adventures constitute one of the most romantic chapters in Colonial history. The western column, at first under Captain Voulet, who was accompanied by Lieut. Chanoine and others, marched from the Niger along the northern edge of the Nigerian Emirates. Mutiny and murder among the European personnel were experienced. French politics at home, where the Jewish question had become acute, were responsible for all manner of delays; the command changed hands repeatedly. But the northern column and the Congo party were equally delayed; not until a year after the date fixed for the rendezvous on the lake did the three expeditions meet. The military escorts were united under Commandant Lamy, and gave battle to the forces of Rabah, one of the Khalif’s generals, who had crossed half Africa to carve out for himself a kingdom in Bornu and Bagirmi after the _débâcle_ of the Mahdia on the Upper Nile. Lamy defeated him and annexed French Equatorial Africa.

Of these three expeditions, the northern column, known as the Foureau-Lamy Mission, had passed through Air on its way south. The Europeans who accompanied it were in 1899 the first Frenchmen to enter the country and to carry out the plan originally contemplated by Flatters in 1881. The annexation of Air by France may be counted from this date.

The Foureau-Lamy Mission[28] entered the borders of Air from Algeria at the wells of In Azawa; their heavy losses in camels obliged them to abandon large quantities of material, but they eventually reached Iferuan in the Ighazar, not far from T’intaghoda. Here the camp of the expedition was attacked in force by the Tuareg, who were only driven off with great difficulty. The situation was critical. The whole country was hostile to the French; they were so short of camels that on the stage south of Iferuan to Agellal they had to move their baggage in small lots, marching their transport forwards and backwards. Their destiny hung in the balance when friendly overtures were made to them near Auderas by a Tuareg of considerable note, Ahodu of the Kel Tadek tribe, whose fathers and forefathers for five generations had been keepers of the mosque of Tefgun near Iferuan. Ahodu’s political sense has rarely been at fault, either then or since; he saw that the only end possible for his people from protracted hostilities with the Europeans was disaster. He promised the French peace while the column remained in Air. It reached Agades in safety, and the Sultan was obliged to hoist the French flag and provide transport animals and guides. No attack was made near the town, thanks to the efficacy of Ahodu’s presence, but his powers of persuasion were insufficient when the column marched out into the barren area further south. The guide purposely misled the expedition and it nearly perished of thirst, succeeding only with great difficulty in returning to Agades. It eventually started once more and reached the south, where its story ceases to concern the exploration of Air.

Since 1899, then, the fate of Air has been settled in so far as Europe was concerned, for it was recognised as lying within the French sphere; but the country was not effectually occupied until 1904, when a camel patrol under Lieut. C. Jean established a post at Agades. The post was evacuated for a short time and then reoccupied. The exploration of the mountains has proceeded slowly since that date. Sketch maps were gradually compiled in the course of camel corps patrols, and in 1910 the Cortier geographical mission published a very creditable map of the mountains,[29] other than the northern Fadé group, based on thirty-three astronomically determined co-ordinates supplementing the five secured by the Foureau-Lamy Mission. Chudeau in 1905 made a brief geological survey and published some notes on the flora, which remain uncatalogued to this day;[30] very complete collections of the fauna have been made by Buchanan[31] and examined in England by the British Museum (South Kensington) and by Lord Rothschild’s museum at Tring. The ethnology of the country is very superficially discussed in a book published by Jean; Barth’s account remains the one of value. The complete exploration of the mountains and detailed mapping still remain to be done as well as other scientific work of every description.

“Air” as a geographical term for the mountainous plateau does not signify exactly the same thing to the inhabitants of the country themselves as it does to us; properly speaking, it is applied by them only to one part of the plateau, for the whole of which the more usual name of Asben or Absen is used. The latter is probably the original name given to the area by the people of the Sudan before the advent of the Tuareg. It is now very generally used even by them: it is universal further south. Barth has speculated at some length upon the origin of the name Air or Ahir, to take its Arabic form, and concluded that the letter “h” had been deliberately added out of modesty to guard against the word acquiring a copronymous signification. But early Arabic geographers give the form as Akir and not as Ahir, so the laborious explanation of the learned traveller is probably unnecessary.

The boundaries of Air may be defined either as running along the line where the rocks of the area dip below the sands of the desert, or as following certain well-marked basins and watercourses of material size, where disintegrated rock or alluvium has covered the lower slopes of the hills. The mountainous area is some 300 miles long by 200 miles broad. It lies wholly within the tropics and is surrounded by desert or by arid steppe. Owing to the general elevation of the country the climate is quite pleasant.

In remote ages the rainfall of the Central Sahara was sufficient to create the deep and important river beds which compose the hydrographic system of this part of North Africa. Among these watercourses is one of great size, flowing from the Ahaggar massif towards Algeria, called the Ighaghar. Duveyrier has tried to prove that it was the Niger of Pliny, largely on the grounds that the root “Ig” or “Igh” occurs in both words and in Temajegh means “to run.” The effect of this identification, which is hard to accept, would be to make the classical ethnology of the Sahara less easy to follow, but it has little significance in considering Air, except in so far as it would tend to show that the geographical knowledge of the Romans did not extend as far south as the plateau. Complementary to the Ighaghar but flowing south from the Ahaggar massif is another equally great river,[32] which early in its course is joined by a large tributary from the Western Fezzan. At a certain point this valley is crossed by the roads from Air to Ahaggar and Ghat, branching respectively at the wells of In Azawa or Asiu. The eastern branch is the caravan road to Ghat from the Sudan, the western one finds its way to In Salah in Tuat and to Algeria. This bed runs south and south-west towards the Niger, which it must have reached at some point between Gao and Timbuctoo in the neighbourhood of the N.E. corner of the Great Bend which the French call “La Boucle du Niger.” This river of remote times must have been one of the great watercourses of Africa, extending from the head-waters in 26° N. Lat. to its mouth in the Bight of Benin on the Equator. It is not possible to say whether the interesting terrestrial changes which diverted the Upper Niger at the lagoons above Timbuctoo into the present Lower Niger, and which brought about the desiccation of the upper reaches, took place suddenly or gradually, but the latter is more probable, for a similar diversion seems to be going on in the Chad area. The lake, in reality an immense marsh and lagoon, is much smaller than when it perhaps included the depression noted by Tilho as extending most of the way to Tibesti; some of the waters of the Chad feeders are already believed to be finding their way in flood-time into the Benue, and it is possible that in the course of time a similar process to that manifested in the Niger area will take place; then Lake Chad will dry up into salt-pans like those at Taodenit. The Saharan river, which flows southward to the west of Air, bears various names. Its course has never been accurately determined, but its general direction is known. From Ahaggar to a point level with the northernmost parts of Air it is called Tafassasset. The T’in Tarabin channel from Ahaggar more probably drained into the Belly of the Desert than into this system, but the Alfalehle (Wadi Falezlez) from the Western Fezzan most certainly seems to be a tributary; there are various reasons why it ought not to flow towards Kawar, as used at one time to be thought. West of Air the main bed spreads out into a vast plain-like basin under the name of T’immersoi; further south it is called Azawak. In general I prefer to use the name T’immersoi for the whole until a better one is suggested.[33]

The T’immersoi forms a collector in the west of Air for nearly all the water from this group of mountains. Nowadays only a comparatively small amount ever reaches the basin, as much is absorbed by the intervening plain land of Talak[34] and the Assawas swamp west of Agades. The latter are local basins or sumps covered with dense vegetation where some of the most nomadic tribes in Air pasture their herds. Talak is visited by Tuareg from Ahaggar and from the west for the same purpose. It plays an important part in the economy of the country, for water is always to be found in the alluvial soil however dry the season in the mountains has been. Many of the wells have now fallen into disuse, but the output of those which remain is still plentiful. The last rocks of Air on the west disappear below the alluvium of T’immersoi and in the subsidiary basins of Talak and Assawas. The T’immersoi system therefore forms the western boundary of Air.

The upper part of the T’immersoi, where it is called the Tafassasset, is also the northern boundary of Air. The wells of In Azawa[35] and Asiu in this valley may be regarded as the point where the main roads from the north enter the extreme limits of the country. Further east on another road between Air and Ghat, von Bary fixed the boundary at the Wadi Immidir, which is in the same latitude as In Azawa.[36]

The eastern boundary of Air runs along the line where the last rocks of the group disappear below the sand of the steppe and desert, which extends from north to south between the mountains of the Fezzan and the fringe of Equatorial Africa, and from west to east between the mountains of Air and those of Tibesti with its adjacent massifs. This vast area is crossed by a few roads only, the most important ones being (_a_) the road from Murzuk along the Kawar depression to Agadem and Lake Chad, (_b_) and (_c_), the two principal tracks from Air eastwards to Bilma by Ashegur and Fashi respectively, and (_d_) the road from Zinder by Termit to Fashi and Kawar. Watering-points are very few, and the habitable oases can be numbered on the fingers of two hands; pasturage is everywhere scarce. This great waste is one of the most unknown parts of North Africa; its eastern portion along the Tibesti mountains as far north as the Fezzan may be said to be absolutely unknown except for two tracks to the mountains whither occasional camel patrols have passed.

Kawar and the other oases along the Chad road appear to be closed basins of the Eastern Saharan type. They seem to have no outlet towards the south either into the Chad or into the Niger systems. The desert east of Air, therefore, contains the eastern watershed of the T’immersoi basin, for the valleys of Eastern Air do not run into the desert as Chudeau has suggested,[37] but turn southwards on leaving the hills, in ill-defined depressions or folds which join the Tagedufat valley or one of the other channels flowing westwards in Tegama or Damergu. One valley to the south of Air, probably the Tagedufat itself, is stated to run all the way from Fashi across the desert.

The southern limits of Air may be placed along the Tagedufat basin, where the rocks of Air disappear below the sand dunes and downs of Tegama and Azawagh steppe desert. The valley is of some size and flows roughly N.E. and S.W. towards T’immersoi, but whether it actually joins this system or the Gulbi n’Kaba, which finds its way into Sokoto Emirate under the name of the Gulbi n’Maradi and thence into the Niger, is not certain. The former hypothesis seems more probable, but I was unable to follow the Tagedufat sufficiently far west to verify it, nor could I discover any data on the French maps;[38] local reports substantiate my supposition. Both systems in any case are in the Niger basin. Air is not on the watershed between Niger and Chad. The choice of the Tagedufat valley as the southern boundary of Air is made on geographical grounds. What may be termed the political boundary is rather further north along the line of the River of Agades.

PLATE 3

[Footnote 2: The geography of the Sahara as a whole is briefly treated in _Le Sahara_, by E. F. Gautier, Collection Payot, Paris, 1923, and with greater detail in _Le Sahara_, by H. Schirmer, Paris, 1893, but much recent work is not included in the latter.]

[Footnote 3: O. Bates: _The Eastern Libyans_ (Macmillan), pp. 48-9.]

[Footnote 4: Cf. Rohlfs, _Kufra_, Chap. VIII.]

[Footnote 5: “Alguechet” in Leo Africanus, Vol. III. pp. 802, 818, etc. (For particulars see beginning of Chap. IX.)]

[Footnote 6: Until motor-cars began to cross the Sahara further west.]

[Footnote 7: Bissuel, _Les Touareg de l’Ouest_, p. 63, says: “A plant called locally ‘Bettina’ and not the Alfalehle (Arabic: Falezlez) was used.”]

[Footnote 8: Gautier: _La conquête du Sahara_, Paris, 1922.]

[Footnote 9: See _Life of Charles de Foucauld_, by R. Bazin, translated by P. Keelan, and De Foucauld, _Dictionnaire abrégé Touareg Français_ (Dialecte Ahaggar), publié par R. Basset, Alger, 1918-20.]

[Footnote 10: Jean: _Les Touareg du Sud-Est_, Paris, Larose, 1909.]

[Footnote 11: Barth: _Travels and Discoveries in Central Africa_, London, Longmans, 1857-8, 5 vols.]

[Footnote 12: From “Litham,” لثام (root لثم), a veil.]

[Footnote 13: The slaves which they possess do not wear the veil. The slave is not a man but a chattel. As soon as a slave is freed and becomes a serf he wears the veil like the noble Tuareg.]

[Footnote 14: In the Air dialect this word is so pronounced. Variations in other dialects are referred to elsewhere. Imajeghan is the plural form of Imajegh. Temajegh is a feminine form of Imajegh.]

[Footnote 15: “Kel” means “People of,” “Tagilmus” is the name of the Veil in Temajegh, the language of the Tuareg.]

[Footnote 16: For an explanation of this term see Chap. IX.]

[Footnote 17: The term “Hausa” throughout this volume is not used in an ethnological sense. It is primarily a linguistic division which may or may not also have an ethnic significance.]

[Footnote 18: “Adghar” or “adrar” = mountain in Temajegh. This mountain group between Air and the Niger and south of Ahaggar has no name. It is called the “Mountain of the Ifoghas” (Adghar n’Ifoghas), while the people who live in it are known as the “Ifoghas of the Mountain,” to distinguish them from the Ifoghas tribe in Damergu and the Ifoghas tribe of the Azger.]

[Footnote 19: Leo Africanus: Hakluyt Society edition, Vol. I. p. 127, and Vol. III. pp. 798-9.]

[Footnote 20: Notably by M. Ch. de la Roncière: _Revue des Deux Mondes_, 1st February, 1923: “Tombuctou au temps de Louis XI.”]

[Footnote 21: M. de la Roncière in a private letter of July 1923 to the author.]

[Footnote 22: The edition I have used is a French one: Hornemann, _Voyage dans l’Afrique Septentrionale_, edited by my ancestor Rennell. Paris: Dentu, 1803.]

[Footnote 23: Denham and Clapperton: “Discoveries in North and Central Africa, 1822-4,” Murray, 1826.]

[Footnote 24: See Introduction to Richardson’s _Travels in the Great Desert of the Sahara_, London, 1847, and Barth, _op. cit._, Vol. II. pp. 219-20.]

[Footnote 25: Barth calls this area Fadeangh, a name not known to-day.]

[Footnote 26: The Governor appointed by the Turks.]

[Footnote 27: Von Bary’s Diary, “La dernier rapport . . . sur . . . les Touaregs de l’Air.” Edited by Schirmer; Paris, Fischbacher, 1898.]

[Footnote 28: _Documents Scientifiques de la Mission Foureau-Lamy_. Various fascicules.]

[Footnote 29: Carte de l’Air: Mission Cortier (2 feuilles), 1/500,000. Service Géogr. du Min. des Colonies.]

[Footnote 30: Chudeau and Gautier: _Missions au Sahara_, Paris, Armand Colin, 1909 (Vol. II., _Le Sahara Soudanais_, by Chudeau).]

[Footnote 31: Buchanan: _Out of the World North of Nigeria_, Murray.]

[Footnote 32: Where the words “rivers” or “watercourse” are used they must be understood to mean drainage channels which are dry most of the year.]

[Footnote 33: Gautier on his sketch map in _Le Sahara_ uses the name Tafassasset, which, however, is even more of a local name in the north than T’immersoi is in the south.]

[Footnote 34: In Temajegh “Talak” means “clay.” Cf. Chudeau: _Le Sahara Soudanais_, p. 63, etc.]

[Footnote 35: Meaning in Temajegh “of the Tamarisk.”]

[Footnote 36: Von Bary’s Diary, pp. 108-9. He joined the main road followed by Barth in the T’iyut valley.]

[Footnote 37: In the case of the Tafidet and other eastern valleys of Air, Chudeau, _op. cit._, p. 62. He supposed, as I think erroneously, that the Air group itself and not the desert was the eastern watershed of the T’immersoi basin.]

[Footnote 38: The country south of Air and north of the limit included in the maps published by the Mission Tilho of the area each side of the Franco-British boundary between Nigeria and the Territoires Militaires du Niger is hardly mapped at all.]

[Footnote 39: Chudeau, _op. cit._, pp. 263-4.]

[Footnote 40: _Vide_ Plates 23 and 39.]

[Footnote 41: _Vide_ Plates 13 and 14.]

[Footnote 42: In the case of Tamgak.]

[Footnote 43: Chudeau, _op. cit._, p. 57.]