CHAPTER IV

THE ORGANISATION OF THE AIR TUAREG

On 6th August, soon after noon, I marched out of Agades with twenty-six camels and eight men for Central Air. My two travelling companions had left the same morning with ten camels in the opposite direction, bound for a point called Tanut[107] near Marandet in the cliff of the River of Agades. Some men of the Kel Ferwan, who were camped under the cliff south of the river, had brought information concerning a lion. At Marandet, it appeared, a cow had been killed and the trail of the offending beast was plainly visible; notwithstanding, Buchanan was unable to secure this lion or any specimen, or even a skull, so it proved impossible to classify the animal.

Circumstantial evidence goes to show that the lion still exists in Air, but is nevertheless very rare. In the Tagharit valley, a few miles north of Auderas, there is a cave in the side of a gorge which a large stream has cut through a formation of columnar basalt: a pink granite shelf makes a fine waterfall in the rainy season with a pool which survives at its foot all the year round. A lion used to live in this den until recent years, when it was killed by the men of Auderas because it had pulled down a camel out of a herd grazing in the neighbourhood. The carcase had been dragged over boulders and through scrub and up the side of the ravine into the lair; a feat of strength which no other animal but a lion could possibly have accomplished. When I came to the overhanging rock the ground was fœtid and befouled, and the skeleton of the camel was still there and comparatively fresh. One of the men of Auderas who had been present at the killing secured a claw as a valuable charm; another had apparently been severely mauled in the shoulder. They had surrounded the “king of beasts,” as the Tuareg also call him, and had attacked with spears and swords. There was no doubt of the animal having been a lion.

The cave in the Tagharit gorge is a short distance from the point[108] where Barth[109] saw “numerous footprints of the lion,” which he conceived to be extremely common in these highlands in 1850, albeit “not very ferocious.” In 1905 a lioness trying to find water fell into the well at Tagedufat and was drowned; her two small cubs were brought into Agades, and one of them was afterwards sent to the Jardin des Plantes in Paris.[110] This lion, however, may not have been of the same variety as the Air species, for the latter is said never to have been scientifically examined.

The Air lion has been described as a small maneless animal like the Atlas species, though von Bary, who, however, never himself saw one, heard that it had a mane. He confirms the report that the animal was common, as late as in 1877, especially in the Bagezan massif, where it used to attack camels and donkeys.[111]

The advent of the rains during the latter part of July made travelling through Air in many ways very pleasant. But there were also disadvantages. With the first fall of rain the flies and mosquitoes came into their own again. The common house-flies were especially trying during my journey north of Agades. They infested the country miles from any human habitation or open water.

PLATE 13

[Illustration: AUDERAS VALLEY LOOKING WEST]

[Illustration: AUDERAS VALLEY: AERWAN TIDRAK]

South of Agades the rains proved a terrible burden. The combined onslaughts of flies, mosquitoes and every other form of winged and crawling insect made life intolerable for Buchanan’s party; meals had to be eaten under netting and naked lights were rapidly extinguished by incinerated corpses. Camels got no rest. Even the hardened natives had recourse to any device to snatch a little sleep. They went so far as to make their beds in the thorny arms of small acacia trees in order to escape the plague. The alluvial plain of the River of Agades had become so soft as to be almost impossible to cross. Mud engulfed the camels up to their bellies. The drivers used to unload them and push them bodily over on to their sides at the risk of breaking their legs in order to let the brutes kick themselves free. The several stream beds of the system, even if not too swollen to be completely unfordable, had such perpendicular banks where the water had cut its way down several feet below the surface of the ground that they became formidable obstacles. The constant threat of rain made long marches impossible, though it was abundantly clear that the longer the time that was spent in the valley the worse the ground would become. Buchanan was rewarded for his disappointment at not finding a lion by securing near Tanut two fine specimens of ostrich and an ant-bear. He also reported the existence near Marandet of a cemetery in the bank of a stream bed. It was unfortunate that he had not time to examine this site, as it seems to be an example of urn burial, probably of pre-Tuareg date.[112]

Half a day’s march from Agades brought me to the village of Azzal on the valley of the same name, the lower part of which is called “Telwa,” the most convenient name for the whole of this important basin. Azzal and the neighbouring Alarsas[113] are small settlements with a few date palms and some gardens. They were formerly inhabited only by serfs engaged in cultivating the gardens which supply Agades with vegetables. After the 1917 revolution in Air the noble population of certain villages in the Ighazar, which was evacuated, settled there temporarily under their chief, Abdulkerim of T’intaghoda. They were living in straw and reed huts, hoping in the course of time to return north and resume possession of the more substantial houses in their own country. During my stay in Air several families did, as a matter of fact, go back to Iferuan and Seliufet. But the presence of the remainder of these Kel Ighazar in the south is somewhat anomalous, as the country from the earliest times has been almost exclusively inhabited by servile people. The area, extending over the foothills of the main plateau, is not yet, properly speaking, Air, in the sense in which the name is used by the Tuareg. Like the desert further south, it is called Tegama.

After following the Telwa for a short distance the track crosses to the left bank and winds over low bare hills and torrent beds. A little before reaching Solom Solom there is a wooded valley which the road leaves to cross a stretch of higher ground by a small pass covered with the remains of stone dwellings, the site, I presumed, of Ir n’Allem. The track is evidently very old at this point, for in places it has worn deep into the rock. The country is wild and picturesque, but the earth-brown hills are fashioned on a small scale. The district used to be infested by brigands who preyed on the caravans bound to and from Agades.

The southern part of my journey followed the usual route, though Barth on his expedition from T’intellust to Agades travelled both there and back by an alternative track rather further east in the Boghel valley and via Tanut Unghaidan, which is not far from Azzal, where he rejoined the more habitual way.

At Dabaga my road from Solom Solom rejoined the Telwa valley and crossed the stream bed after a short descent into a basin covered with dense thickets of dûm palms and acacias. The trees were filled with birds. The river was in full flood, over a quarter of a mile wide and some two feet deep—an imposing stream draining south-western Bagezan and Todra into the River of Agades. I was luckily able to cross it with my laden camels, but some travellers only a little behind me were held up for several days by the floods which followed the heavy rain in Central Air. Travelling at this season of the year is slow, as camp must be pitched before the daily rains begin, usually soon after noon. On the other hand, it is very convenient to be able to halt anywhere on the road regardless of permanent watering-points; for every stream bed, even if not actually in flood, contains pools or water in the sand.

Climatically Air is a Central African country. It is wholly within the summer rainfall belt, the northern limit of which coincides fairly accurately with the geographical boundary of the country at the wells of In Azawa. The rains usually commence in July, and last for two months, finishing as abruptly as they have begun. Within the limits of the belt, the further north, the later, on the whole, is the wet season, though great irregularities are observed. In Nigeria the rains fall during May and June, at Iferuan they occur in August and September.[114] They are tropical in their intensity, and in Air nearly always fall between noon and sunset.

During my stay at Auderas there were a few days when the sky was overcast for the whole of the twenty-four hours, with little rainfall; the damp heavy feeling in the air reminded one of England, as the atmosphere was cold and misty. On one particular day it rained lightly and fitfully for fourteen hours on end with occasional heavy showers. Such phenomena, however, are rare. Precipitation follows a north-easterly wind and usually lasts three or four hours; as soon as the westerly wind, prevalent at this season, has sprung up, the nimbus disperses rapidly, leaving only enough clouds in the evening to produce the most magnificent sunsets that I have ever seen.

In 1922 the rainy season at Auderas was virtually over by the 10th September, though it continued a little later in the north. The rains were followed by a period of damp heat, and then by some days when the ground haze was so thick that visibility was limited to a few hundred yards. Until recent years there seems to have been a short second rainy season in the north of Air coinciding with the first part of the Mediterranean winter precipitation. In November near Iferuan I experienced several days on which rain appeared to be imminent, but none fell. Natives told me that up to three or four years previously they had often had a few days’ rain in December and January. In 1850 the last rain of the summer season, which, exceptionally, had begun as far north and as early as 26th May at Murzuk, was recorded on 7th October, but in November and December after a fine period the sky had again become overcast, and a few drops of rain actually fell in Damergu on 7th January, 1851. The cycles of precipitation in the Sahara are constantly varying and data are as yet insufficient to permit any conclusion. It would be quite incorrect, from the accounts of the last ten years alone, to suppose that the rainfall had markedly diminished, or that the second rainy season had disappeared.

During the rains the larger watercourses meandering among the massifs of the country often become impassable for days on end, which is inconvenient, for in ordinary times they are the channels of communication. Owing to the lack of surface soil and vegetation on the as yet undisintegrated volcanic rock, streams fill with surprising rapidity during the rains and are very dangerous for the unwary traveller. The great joy of these weeks was the freshness of the air after the intolerable heat of June and July, especially in the plains. With the rain too came the annual rebirth of plant life, which made one’s outlook very sweet. In European spring-time Nature awakes from winter sleep, but in Africa a new world, fresh and green and luxurious, is born after the rains out of a shrivelled corpse of sun-dried desert.

At Dabaga I was persuaded to forsake the caravan road which continues up the Telwa and take a riding road by Assa Pool and the T’inien mountains. Difficulties began at Assa, when I tried to pitch my tent on rocky ground, with the result that it was almost impossible to keep it erect in the rain squalls which followed. The evening, after the rain, was unsatisfactory. I wounded two jackals at which I had shot, but did not kill either. I missed several guinea-fowl and only secured a pair of pigeons among the dûm palms of the valley. Also, there were many flies. However, I made the acquaintance of one of the greatest guides in Air, Efale, who overtook me on his way north, and camped near me. He talked volubly that night. Next day, after dropping sharply into the T’inien valley by a narrow defile, the road became frankly devilish. At the bottom of the steep sides the soil is impregnated with salt, which effectually prevents anything growing. There are a number of circular pits where the sandy salt, called “ara” or “agha,” is worked. The mixture is dried in cakes and sold in the south for a few pence. It is only fit for camels, which require a certain amount of salt every month, more especially after they have been feeding on fresh grass. “Ara” can only be used for human food if the sand has been washed out and the brine re-dried.

After leaving Assa the vegetation had almost entirely disappeared. Low gravel-strewn hills on the right obscured the view to the east. The T’inien valley soon made a right-angle turn to the north, closing to a narrow cleft, which became even rougher. The track was a series of steps between huge granite and quartz boulders, among which the camels kept on stumbling. Their loads required constant readjustment and there was no room to kneel them down. The way was really only fit for unloaded camels or riders on urgent business. There had not been a tree or bush for hours. We climbed some 600 feet in about a mile, almost to the very top of the jagged peaks on the left that marked the summit of the T’inien range. By 11.15 a.m. I was beginning to despair of finding a camp site before the rain was due, as I foresaw a similar unpleasant descent on the other side of the col which had so long been looming ahead. Then as I reached the gates of the pass a view over the whole of Central Air suddenly burst upon me in such beauty as I can never forget.

The ground sloped imperceptibly to the east. It fell away only about a hundred feet to the north, where a row of small crags, the continuation of the T’inien range, cut off the western horizon. Straight in front in the distance, piled mass upon mass, the blue mountains of Central Air rose suddenly out of the uplands, soaring into the African sky. Between the bold cliffs and peaks of the Bagezan mountains and the long low Taruaji group to the right, a few little conical hills of black rock broke the surface of the vast plain which rolled away to the east. From so great a distance the plain seemed tolerably smooth, veined like the hand of a man with watercourses winding southwards from the foot of the mountains. Black basalt boulders covered the flat spaces between lines of green vegetation and the threads of white sand, where the stream beds were just visible. Over the whole plain the new-born grass was like the bloom on a freshly-picked fruit. To the south-east stood the blue range of Taruaji itself, flat-topped and low on the horizon. Either side of the hills the curve of the world fell gently away towards the Nile.

I camped a mile or so north of the pass in a valley below the precipitous cliffs of a rock called Okluf, which has a castellated crown several hundred feet high. The rocks shone blue-black, with their feet in a carpet of green that seemed too vivid to be real. There were plenty of guinea-fowl and many other birds in the palm woods and thorn groves, and such grass as I thought only grew in the water meadows of England. I shall never forget the beauty of Central Air on that noonday in the rains, though I have it in me to regret the fiendish temper in which the day’s march had left me. The flies in the evening and the fast-running things upon the ground at night only made it worse. I had hurriedly and laboriously pitched a tent, and it never rained after all.

PLATE 14

[Illustration: MT. TODRA FROM AUDERAS]

On the following day I ascended the T’ilisdak valley which flows into the Telwa, and reached Auderas village, where some huts had been prepared for us by the chief Ahodu, a man who soon became my most particular friend. The T’ilisdak valley is renowned for its excellent grazing and for some mineral springs where men, camels and herds go after the rainy season to take a “cure” of the waters.[115] Near Okluf there are the remains of several hut villages, and some with stone foundations of a more permanent character. They belong to a servile tribe of Southern Air called the Kel Nugguru, who at present are living somewhat further west.

Air proper may be said to begin at the head of the T’ilisdak valley. The part of the plateau I had traversed was therefore still in Tegama, which includes the whole area south of Bagezan and Todra as far as the River of Agades, as well as the Taruaji massif, but not the country east of the latter and of Bagezan. Most of the villages in Tegama have gardens, and some have groves of date palms. That they are inhabited by serfs is, of course, natural, since the cultivation of the soil, in the estimation of the noble Tuareg, is not a worthy occupation for a man. When, however, in a nomad society agriculture is relegated to an inferior caste of people, it is inevitable that the practice should undermine the older allegiances. It becomes possible for the settled and therefore originally the servile people to accumulate wealth even in bad times when the profit from raiding or caravaning is denied to the upper classes of Air. The social effects of the disruption caused by the 1917 revolution may be observed in the village organisations, where people of different tribes are now tending more and more to live in association under the rule of a village headman, who for them is displacing the authority of their own tribal chiefs. The village headmen, it is true, are sometimes themselves the leaders of the tribes in whose area the village is situated, but more often they are merely local men acting on the delegated authority of the tribal chief, who in Tegama is probably the head of an Imghad or servile tribe dependent in turn upon some noble tribe living in a different part of Air. But in time the population of a village may become known collectively as the people of such and such a place, and so reference to the old tribal allegiance of the inhabitants disappears.

Tuareg tribal names deserve close investigation. They are of two categories: those which begin with “Kel” (People of . . .) and those which begin with “I” or sometimes “A.” This “I” or “A” may be quite strongly pronounced, but often represents the so-called “neutral vowel,”[116] which is very difficult to transliterate. Thus the word “Ahaggar” might as correctly be written “Ihaggar”; the initial vowel indeed is so little emphasised that the French have come to write simply “Hoggar” or “Haggar.” On the other hand, in the name Ikazkazan, an Air tribe, the “I” is marked; in the Azger tribe, again, the Ihadanaren, it is so lightly accentuated that Barth writes “Hadanarang.” This point, however, is of little moment: what matters is the question of the type of prefix to the name. To simplify reference I propose to call these two types “Kel name” and “I name” tribes. After examining the two categories at length, a distinction seemed to me to stand out clearly; I believe it holds good among other Tuareg as well as those of Air. The primary tribal divisions have names of the “I” category, except in certain cases where they are nearly always known to have been forgotten; the subdivisions of these tribes have “Kel names.” The former are proper names; the latter are derived either from the place where the people usually or once lived, or from some inherent peculiarity. The word “Kel” is also used to cover generalisations of no ethnic importance: the “I name,” on the other hand, is scarcely ever geographical or adjectival. The generalisation will be clearer for a few examples, chosen among the Air tribes. The noble tribe called Imasrodang has for sub-tribes the Kel Elar, Kel Seliufet and Kel T’intaghoda, called after the villages where they lived in Northern Air. Again, the Ikazkazan have one section or group of sub-tribes called the Kel Ulli—the People of the Goats—who are themselves subdivided into other factions bearing “Kel names.”

Certain other “Kel names” like Kel Ataram or Kel Innek are often heard in Air, but are not proper names at all; they were erroneously regarded by Barth as tribal names, but simply mean the “People of the West” and the “People of the East” respectively, and have no inherent ethnic significance. In Air the former term logically includes, and is meant to include, the Arab as well as the Tuareg tribes of the west.[117]

So clear is this use of geographical “Kel names” that we shall find repeated instances later on of tribes who, having migrated from a certain area, retain their old names, though these are no longer applicable to their new ranges. Take, for example, the Kel Ferwan—the People of Iferuan, in North Air; they now live in the southern parts of the country. Or, again, there are two Kel Baghzen, called after a mountain group in Central Air; the one group is still in that area, the other, which once lived there, has since migrated to the country north of Sokoto.

In certain forms the word “Kel” corresponds to the Arabic word “ahel,” but the latter seems more usually employed in connection with wide geographical indications of habitat, without much ethnic significance, like Kel Innek. The use of this type of “Kel name” is the exception rather than the rule in Temajegh and has a colloquial rather than traditional sanction. The more common “Kel names,” on the other hand, are definitely individual tribal names, and refer to small areas. They are not by any means restricted to sedentary tribes.[118]

A third category of names commencing with the “Im” or “Em” prefix is regarded by Barth[119] as virtually identical with the “Kel” class, but this is not quite accurate. The “Im” prefix is used to make an adjectival word form of place names; the “Kel names” only become adjectival by prefixing “People of . . .” Thus “Emagadezi” would be more correctly translated as “Agadesian” than as the “People of Agades,” whose correct designation is Kel Agades. “I names” partake of neither of these characteristics. For the most part their significance remains unexplained. It follows that “Kel names,” although proper to the tribes that bear them, being descriptive or geographical, are certainly not so old as the individual and proper “I names.”

There are examples of tribes which have lost their “I names” and are only referred to by a “Kel name,” though in many cases this is more apparent than real. When a tribe with an “I name” increases until the point is reached where it subdivides, one of the subdivisions retains the original “I name,” the remainder take other and, usually, geographical appellations. This process might be shown graphically:—

Original I name tribe. | +-----------------+-----------+------------+-----------+ | | | | I name sub-tribe Kel name Kel name Kel name (as above) sub-tribe sub-tribe sub-tribe \__________________________________________________________________/ Collective Kel name often the same as one of the sub-tribe Kel names if the latter has come to play a preponderating part in the group.

This difference of nomenclature has a definite bearing on the difficulties of co-ordinating sedentarism and nomadism in one people, which must have occurred to everyone who has studied the problem in administration. The exact relations between a village headman, the tribal chiefs of the persons who are living in his village and the tribal chief of the area in which the village is situated cannot be defined. One set of allegiances is breaking down and another has not yet been completely formed. This was already going on in Air when the position was complicated by the advent of a European Power demanding a cut-and-dried devolution of authority, and tending to encourage sedentary qualities in order to prevent raiding. These problems in Air to-day are almost insoluble, but they are of an administrative rather than of an anthropological order.

Auderas at the present time is probably the most important place in Air after Agades. As an essentially agricultural settlement it is an excellent example of the village organisation. The valley of Auderas lies about 2600 feet above the sea. Seven small valleys unite above the village and two affluents come in below, draining the western slopes of Mount Todra and a part of the Dogam group. The main stream eventually finds its way out into the Talak plain[120] under various names. The sandy bed of the valley near the village contains water all the year round. Both banks are covered with intense vegetation, including a date-palm plantation of some thousand trees. Under the date palms and amongst the branching dûm-palm woods, where the thickets and small trees have been cleared or burnt off, are a number of irrigated gardens supplied with water from shallow wells. Some wheat, millet, guinea corn and vegetables are grown with much labour and devotion. Onions and tomatoes are the principal vegetables all the year round, with two sorts of beans in the winter. Occasionally sweet potatoes and some European vegetables like carrots, turnips and spinach are grown from seeds which have been supplied by the French. Pumpkins do well and water melons are common. There is also a sweet melon. Three different shapes of gourds for making drinking and household vessels are cultivated. Cotton is found in small quantities, the plant having probably been imported from the Sudan. Its presence in Air is interesting, as in 1850 Barth had placed the northern limit of Sudan cotton in the south of Damergu. The cotton plant does very well when carefully irrigated and produces a good quality of fibre. Two samples which I brought home from Air were reported on respectively as: “good colour, strong, fairly fine 1³⁄₁₆ staple,” and “generally good colour, staple 1³⁄₁₆-1¼ inches, strong and fine”; the materials were respectively valued at 20·35 and 21·35 pence per pound when American May Future Cotton stood at 17·35 pence (May 1924).[121] The Tuareg spin their cotton into a rough yarn for sewing or making cord, but in Air they do not seem to weave. The indigo plant grows wild in Air: it is not cultivated, nor is it used locally for dyeing.

The gardens require much attention and preparation. The ground is cleared and the scrub burnt off as a top dressing. The soil is then carefully levelled by dragging a heavy plank or beam forwards and backwards by hand across the surface. The area is divided up into small patches about six feet square with a channel along one side communicating with a leat from an irrigation well. These wells are usually unlined and shallow, with a wooden platform overhanging the water on one side; on this a rectangular frame is set up with a second cross member carrying a pulley over which a rope is passed. An ox or a donkey pulls up the big leather bucket by the simple process of walking away from the well, returning on its tracks to lower it again. The bucket is a tubular contrivance, the bottom of which is folded up while the water is raised; when it reaches the level of the irrigation channel, a cord is pulled to open the bottom of the leather tube and the water allowed to run out. The other end of this cord is attached to the animal, and the length is so adjusted that the operation is performed automatically each time the bucket comes to the top. The pole and bucket with a counterweight and the water wheel are not known in Air for raising water; nor are any dams constructed either to make reservoirs in ravines or to maintain a head of water for flow irrigation in the rainy season. Each little patch in the gardens is hoed and dressed with animal manure. The seed is planted and carefully tended every day, for it is very valuable. Barth records seeing at Auderas a plough drawn by slaves. This was clearly an importation from the north; the plough is not now used anywhere in the country, which at heart has never been agricultural.

PLATE 15

[Illustration: GRAIN POTS, IFERUAN]

[Illustration: GARDEN WELL]

As in the south, millet and guinea corn are sown during the rains, but they usually require irrigation before they reach maturity. In certain areas rain-grown crops could be raised most years. In the past a fair amount of cereals seems to have been produced in this way; to-day the Tuareg are too poor to risk losing their seed in the event of inadequate or irregular rainfall. Although the wheat grown in the Ighazar used nearly all to be exported to the Fezzan, where it was much in demand on account of its excellent quality for making the Arab food “kus-kus,” Air at no time has produced enough grain for its own consumption. In the economics of Air necessary grain imports are paid for by the proceeds of wheat sales or live-stock traffic with the north, and by the profits of the trade in salt from Bilma; these provide the means of purchasing the cheaper millet and guinea corn of Damergu. Any additional surplus, representing annual savings, is invested in live-stock, especially camels, within the borders of the country.

The breakdown of the social organisations of the Tuareg in Air compelled numbers of nobles out of sheer poverty after they had lost their camels and herds to cultivate the soil; before the war not even the servile people were very extensively so employed if they could find enough slaves to do the work.

Neither the advent of a European Power nor the subsequent changes in the social structure of the country has had very much effect on the position of slaves in Air. Of these there are two categories,[122] the household slave and the outdoor slave, and both of them are chattels in local customary law. The former are called “ikelan,” the latter “irawellan,”[123] or alternatively “bela,” “buzu” or “bugadie,” which, however, are not Temajegh words, but have been borrowed from the south. The term “irawel” is also used generically to cover both categories of slaves, although it primarily refers to the latter. In the use of this word Barth[124] makes one of the few mistakes of which he has been guilty, where he states that the most noble part of the Kel Owi group of tribes in Air is the “Irolangh” clan, to which the Amenokal or Sultan of the Kel Owi belonged. The paramount chief of his day, Annur, belonged to the Kel Assarara section of the Imaslagha tribe, which is probably the original and certainly one of the most noble of the Kel Owi, for it includes the Kel Tafidet, who gave their name to the whole confederation. The traveller’s mistaken reference to Irawellan or Irolangh is probably due to his having been informed by a member of some non-Kel Owi tribe that Annur and all his people were “really Irawellan,” or servile people. Such abuse of the Kel Owi is common among the other Air Tuareg. It is certainly not justified in fact, and is due to the contempt in which an older nobility will always hold more recent arrivals.[125]

The negro slaves, the Ikelan, are primarily concerned with garden cultivation, and are consequently sedentary. One half of the produce of their labour goes to their masters and the other half to support themselves and their families. Ikelan also perform all the domestic duties of the Tuareg to whom they belong, and herd their masters’ goats and sheep if they happen to be living in the same neighbourhood. A certain proportion of the offspring of the flocks is also given to the slaves. Since, primarily, they are cultivators of the ground, they do not move from place to place with their owners. They consequently often escape domestic work and herding. Despite their legal status they are in practice permitted to own property, though, if their masters decided to remove it, they would be within their rights to do so. In other words, the theoretical status of slavery which makes it impossible for a chattel to own property has been considerably modified, and not as a consequence of the altered conditions, or of the legislation of a European Power, but because slavery among the Tuareg never did involve great hardship. Their slaves, furthermore, always had the hope of manumission and consequent change to the status of Imghad or serfs, a rise in the social scale which, in fact, often did occur. It was in slave trading and not in slave owning that the Tuareg sinned against the ethical standards which are usually accepted in Europe, and obtained so unenviable a reputation last century.

Herding live-stock, and especially camels, is the primary function of the outdoor slave or Buzu. Though often also a negro, he is considered to possess a somewhat higher status than the Akel, for he does not as a rule work in the house or village. The Buzu’s work, if on the whole less strenuous than that of the tiller of gardens, is felt to be more manly because he is associated with camels. He travels with nobles or Imghad, to either of whom he may belong. He does all the hard menial work on the march. He is responsible especially for herding the camels at pasture and for loading and unloading them each day on the road. Such duties as filling water-skins, driving camels down to water, feeding them on the march and making rope for the loads, all fall to his lot. The Buzu may even accompany his master’s camels on raids or act as personal messenger for his lord. When the camels are resting he spends his days watching the grazing animals, or looking after any other herds which his master may own in the neighbourhood. On the whole I have found the Buzu a remarkably hard-working person. He is almost useless without his master to give him orders and to see that they are carried out, but ready to undertake any exertion connected with his work, which he regards as his fate, but not his privilege to perform without complaint.

It is difficult to determine whether there is any racial difference between the Buzu class, the tillers of gardens, and the ordinary household slaves. The first are more respected than the last, which may mean that they are more closely related in blood to their masters. The practice of concubinage, though not very widespread, has probably created the caste, and from them, in time, a certain proportion of the Imghad. While theoretically the children of a slave concubine and a Tuareg man ought to be “ikelan” like their mother, in practice they tend to rise into the superior caste of the Buzu, and eventually in successive generations to Imghad. In Air at least the general tendency is for the old-established caste distinctions to become more elastic and for the ancient order to pass away. Although the events of the last twenty years have contributed greatly to this change, the strongest factor has certainly been the increasing wealth of the Imghad, but another reason is probably that many Imghad tribes in Air were themselves originally Imajeghan before their capture in war or their subjugation by some means. Consequently with the dissolution of tribal allegiances in Air and enhanced prosperity they have tended to revert to their former status. They cling so tenaciously to nobility of birth that, rather than accept the logical results of inferiority consequent upon defeat in war, the people collectively combine to admit the fiction of servile people possessing dual status.

The presence of more than one racial type among the Imghad has led certain travellers to make quite unjustifiable generalisations about this section of Tuareg society. There have also been advanced numerous and most unnecessarily complicated theories to account for the division of the race as a whole into these two castes. The problem is really much simpler. Although by no general rule can it be said that the Imghad originally belonged to this or to that people, they are all clearly the descendants of groups or individuals captured in war and subsequently released from bondage to form a caste enjoying a certain measure of freedom, and having a separate legal or civil existence under something more than the mere political suzerainty of the noble tribe which originally possessed them. In this first stage, the noble tribe represents the original pure Tuareg race, while the oldest Imghad are the first extraneous people whom they conquered, in some cases perhaps as early as in the Neolithic ages. “It is necessary,” says Bates,[126] with great justice, “to state emphatically that the division into Imghad and Imajeghan is so ancient that the Saharan Berbers preserve no knowledge of its origin.” This antiquity may be held to account for the complete national fusion which has taken place among the two castes: nearly all Imghad would utterly fail to grasp a suggestion that they were not to-day as much Tuareg as their Imajeghan overlords, however they may dislike and abuse the latter. As time went on more and more Imghad were added to the race, each group being subject to the noble tribe responsible for its conquest. The possibility of a group of people becoming the Imghad of an Imghad tribe was precluded by the relations obtaining between serfs and nobles, whereby it is the sole prerogative of the latter to wage war or make peace. Should an Imghad tribe capture slaves in war they could not be manumitted except by the Imajegh tribe, the lords of the victorious Imghad; and by the act of manumission the newly-acquired slaves would then become the equals of their Imghad conquerors under the dominion of the Imajeghan concerned.

The Imghad of Air may be divided into three categories whose history is so intimately bound up with the noble tribes that it cannot be considered separately. There are the Imghad whose association with their respective Imajeghan dates from before their advent to Air; their origin must be looked for in the Fezzan or elsewhere at some very early date. Secondly, there are the Imghad who were the original inhabitants of Air before the Tuareg came, and who by some agreement at the time, like the traditional one of Maket n’Ikelan,[127] were not enslaved but allowed to continue living in the country side by side with the new arrivals in a state of vassalage or semi-servitude. Lastly, there are the Imghad who are either Arabs, Tuareg of other divisions, or negroids from the south captured in the course of raids from Air, in some cases as recently as a generation ago. With these different origins it is not surprising to find among the Air Imghad both a strongly negroid type, a non-negroid and non-Tuareg type, and a type showing the fine features and complexion characteristic of the Imajeghan themselves. The first type is the pre-Tuareg population of Air. It is the most common, if only for the reason that negroid characteristics always appear to be dominant in the cross-breeding which ensued. The second type represents the Arab or Berber element acquired by conquest. The third type represents the subjugated groups of Imajeghan of other divisions.[128] Of the latter category are, for instance, the Kel Ahaggar, Imghad of the Kel Gharus, who were originally nobles from the great northern division of the Tuareg. Many of the Kel Ferwan Imghad are believed to be Arabs or Tuareg of the west, captured comparatively recently on raids into the Aulimmiden territory. The Kel Nugguru are the freed slaves of the Añastafidet, the administrative head of the Kel Owi confederation: they have become so prosperous that they are now laying claim to be of noble origin, a pretension which no right-minded Imajegh in Air will admit for a moment. But it is almost impossible nowadays to trace the history of each Imghad tribe in detail. Generally, in the absence of more precise data, it may be assumed that those Imghad tribes which have “I names” are the oldest; for here the process of assimilation to the mass of the Tuareg race is most complete, either on account of the length of their mutual association or owing to the fact that they were originally themselves of the same race; the “Kel name” Imghad, on the other hand, are probably more recent additions.[129]

The confusion reigning on the subject of the “Black” and “White” Tuareg in the minds of the few people in Europe who have ever heard of the race is due to the practice in the north of the servile wearing a white, and the nobles a black, veil. But a “Black” Tuareg, being a noble, will, in the vast majority of cases, have a much fairer complexion and more European features than a “White,” or servile Tuareg. In Air the colour of the veil affords no means of distinguishing the caste of the wearer. The best veils, being made in the south, are consequently cheaper in Air than in the north, and this is probably the reason why Imajeghan and Imghad alike in Air wear the indigo-black Tagilmus. When a white veil is seen, it usually means that the wearer is too poor to buy a proper black one and has had to resort to some makeshift torn from the bottom of his robe.

Slaves, domestic or pastoral, do not wear the face veil at all. This is the essential outward difference between them and the Imghad. The latter, whatever their origin, are considered to be a part of the Tuareg people; the former cannot be so, for they are simply accounted to belong, as camels do, to the People of the Veil.

The exact status of the Imghad, or “meratha” (merathra) as they are called by the Arabs in Fezzan, is somewhat difficult to define. There is no adequate translation in any European language of the word “amghid.”[130] The process of their original enslavement and subsequent release to form a category of people who have achieved partial but not complete freedom has, I think, no parallel in Europe except in a modified form in the state of vassalage. Yet, as “servile” conveys too narrow and definite a relationship, so “vassal” is certainly too broad a term. In the state of servility or, to coin a word, “imghadage” to which the pre-Tuareg inhabitants of Air appear to have been reduced, the process of enslavement and release may be said to have taken place only as a legal fiction, and not, if the tradition is to be accepted as accurate, in real fact. The general practice seems to have been that when large groups of people were subjugated or captured in war they were simultaneously released into the state of imghadage, but when individuals or a few persons were acquired by force or by purchase, they were only manumitted in the course of time, if at all, and incorporated at some later date into an Imghad tribe or village already in existence.

In contradistinction to slaves, the Imghad are not bound individually, but collectively, and not to individuals, but to a noble tribe or group of tribes. They are in no sense considered to be the property of the latter; but the relationship is closer than that of suzerain and vassal. It is not within the power of an Imghad tribe to change its allegiance, since in the first instance its members were theoretically at least the property of its overlord tribe; they owe their separate existence to an act of manumission freely and voluntarily accomplished. A change of allegiance could occur only if a servile tribe were captured in whole or in part; it follows that when this has occurred one servile tribe might owe allegiance in several parts to different noble groups.[131] The bond between them consists of the right of the responsible noble tribe alone, and therefore of its chief, to administer justice among the dependent Imghad, either in small cases by tacitly confirming the verdict of their own headman, or in more weighty matters by express reference. The Imghad tribe may be fined or punished collectively by their lords, and would have no right to appeal to the Amenokal without permission. For the Amenokal to interfere on behalf of an Imghad tribe would constitute a breach of tribal custom and ensure a rebuff, if not worse. A certain proportion of the marriage portions payable in the Imghad tribes goes to their Imajeghan, who have the right to give or withhold consent to these contracts. One of the functions of the Imghad is to take complete charge of and use the camels of their lords for long periods or to trade with them on their behalf. In such cases the Imghad act as the agents of the nobles, each one of whom has a right to ask the servile tribe as a whole to undertake these duties. But such obligations are imposed collectively on the tribe and not on any one Imghad. It is the custom to share the offspring of the camels thus herded in equal shares, though in the event of any of the animals dying whilst under the charge of the Imghad, the latter are collectively responsible for making good the loss, save in extenuating circumstances. Conversely, the nobles are, in every case,[132] the protectors of their dependents. The relations between Imghad and Imajeghan are a mixture of those obtaining under the feudalism of Europe and the “client” system of Rome.

A consequence of the interruption of caravan traffic and the disappearance of one of the principal sources of revenue of the noble Tuareg is that the Imghad as camel herders, and generally speaking as the more laborious members of the community, have gained where the nobles have lost.

Prosperity is emancipating the Imghad, and is materially assisting the breakdown of social distinctions which in time will survive only in the philosophic contemplation of the Imajeghan dreaming idly of the return of better days. The Imghad tribes used to be the unquestioning allies of their overlords in war; their numbers contributed greatly to the strength of any Imajegh tribe. Though they might not make war on their own initiative, the Imghad carried and still carry weapons.[133] They used to go on raids with their masters, or, if the Imajeghan were busy elsewhere, represent them with their masters’ camels and the weight of their own right arms. But the chiefs of the Imghad were never more than subordinates, or at the most advisers to the nobles.

To-day this unquestioning subservience has almost disappeared and we even find Khodi, chief of the Kel Nugguru, disputing with the noble Ahodu the leadership of the village of Auderas. This issue was one of great importance in local politics and originally arose out of the disputed ownership of certain palms which had been given to Ahodu when he was installed as head of the village as a reward for service rendered by him to the Foureau-Lamy expedition. The village is on the edge of the Kel Nugguru country, while Ahodu in fact comes from a northern tribe, the Kel Tadek, who have no real concern with this district. The impossibility of reconciling the tribal and settled organisations was clearly demonstrated in every aspect of this controversy. Khodi, living as a nomad with his people and camels at some distance from the village, sought, without success, to govern the community through various representatives, while Ahodu, who had given up wandering, was suspended by the French during the settlement of the legal case, and sat in the village watching mistake after mistake being made. Under the old system Khodi could never have pretended to dispute with a noble the position of chief of a large village: in fact an Imghad tribe without a protecting noble overlord would have been unlikely to administer a village at all. Similarly among the Ahaggaren Imghad of the Kel Gharus, a man of servile origin, Bilalen by name, has come to share with T’iaman the lordship of a once noble people of the north, a position of such importance that he is regarded as one of the most influential chiefs in Air. Bilalen has only become associated with the Ahaggaren by marriage; he could never have achieved even this, much less could he have attained so powerful a following in the country, under the old _régime_.

[Illustration: THE GOVERNMENT OF THE AIR TUAREG

_Note._—The scheme is largely theoretical, as the Amenokal has rarely had much authority over any tribes except the People of the King. His authority over a part of the Aulimmiden has been even more nominal and has varied considerably from time to time.]

In addition to the social distinctions between nobles and serfs, the Tuareg attach great importance to tribal classification. Among the inhabitants of the mountains a man will describe himself as, say, “Mokhammad of the Kel Such-and-such of the Kel Owi,” or of the other category, which is called the “People of the King,” as the case might be. These two great tribal divisions (there were three before the departure of the Kel Geres for the Southland) will be referred to in detail when the history of the migrations of the Air Tuareg is considered. The divisions are absolute; a tribe either is of the Kel Owi or is not of the Kel Owi. There is usually never any doubt; the erroneous attribution of a man’s tribe to the Kel Owi confederation would provoke the indignant rejoinder that his clan were “People of the King” and did not “belong (_sic_) to the Añastafidet.” The distinction means all that the difference between an ancient landed nobility and a _parvenu_ commercial aristocracy denotes. Many of the older men of the “People of the King” go so far as to say that there are no nobles among the Kel Owi at all.[134] Apart from their slightly different ethnic origin, the principal reason why the Kel Owi have stood apart from the other tribes is that they possess an administrative leader of their own who represents the whole confederation; as they say, “he _speaks_ for them to the Amenokal at Agades.” He is called the Añastafidet, the Child of Tafidet. The non-Kel Owi tribes, on the other hand, have no single leader other than the king; in their case each tribal chieftain transacts the business of his own tribe with the former independently of the other chiefs. For them the Amenokal of Air assumes the dual function of nominal ruler of the whole country and of direct overlord of certain tribes.

In accordance with the democratic traditions of the Tuareg, the Añastafidet,[135] like the Sultan, is elected. He must be a noble, but need not always be chosen from the same family. He is elected for a period of three years, but his tenure of office is really dependent upon a yearly revision by the Kel Owi tribes when they concentrate in the autumn to go with the salt caravan to Bilma. The tribal groups mainly responsible for the choice are the Kel Tafidet and Kel Azañieres; the Ikazkazan, being the junior group of the confederation, have little voice. The Añastafidet’s badge of office[136] is a drum; he retains no authority on leaving office, though it entitles him to a certain degree of respect, and leads to his being consulted on State matters. In practice if the Añastafidet is reasonably capable he is confirmed in power for a succession of three-year periods. During the last fifty years there have been in all about six Añastafidets; one, I think the last holder of the office, is at present living at Zawzawa in Damergu. The Añastafidet’s official place of residence was at Assode in Central Air, but since the evacuation of the north he has been living at Agades in direct touch with the Amenokal. His principal duties are to represent the confederation at the Court of the Sultan and maintain the freedom of transit through Air and Damergu for caravans, on which the prosperity of the tribes depends. Trade with the north and the position of the Kel Owi in Air astride the great caravan road which passes from north to south, east of the Central massifs, have in effect combined to place the foreign relations of all the Air people with Ghat and the Fezzan in the hands of the Añastafidet, business with the potentates of the south, on the other hand, being, as has already been stated, in the hands of the Amenokal at Agades. The breakdown of the trans-desert traffic during the war deprived the Kel Owi of most of their prosperity and the Añastafidet of his work.

The Añastafidet was assisted in his duties by four agents, two of whom dealt with local business, while the other two lived in the Southland to assist the Kel Owi tribes in their transactions there. Neither the Añastafidet nor his agents ever seem to have received a salary, and the former at least was expected to give munificent presents, but no doubt their official positions brought perquisites which compensated for any outlay. As in the case of the Sultan, the importance of the Añastafidet’s office depends entirely on the personality of the holder. When von Bary visited the country, Belkho, chief of the Igermaden tribe, living at Ajiru in Eastern Air, thanks to his military prowess and political wisdom, was the _de facto_ ruler of the whole country. His relations with the Amenokal were strained, even though he had him more or less under his influence; the Añastafidet had become of so little moment that he is only once mentioned by this traveller.[137] In Barth’s day, when Air was under the domination of Annur, another Kel Owi chief of the same type, the Añastafidet was a mere shadow in the land.

The Añastafidet doubtless represents the surviving functions of a Kel Owi Amenokal. The restriction of his duties was probably the result of a compromise arrived at when the Kel Owi entered Air and found an Amenokal already established in the country, supported by the Kel Geres and the various tribes known as the “People of the King.” The more intimate inter-tribal relations between the various units of the Kel Owi confederation and the organisation of the “People of the King” will be referred to hereafter in detail.

The system by which the Kel Owi have an administrative leader who seems to have practically no warlike or judicial functions has in no way modified the tribal or social organisation of the confederation. As in the case of all the Tuareg tribes, other than those which have become entirely sedentary, the government of each unit, large and small, is patriarchal and similar to that of Bedawin tribes. The chief of a noble tribe is the leader in war and the dispenser of justice in peace. The functions are not necessarily hereditary. In council with the heads of families he exercises authority over the Imghad tribes associated with his clan, through the chiefs of these servile groups in the manner already described. The council of the heads of families is of great importance, but plays an advisory rather than an executive part. The heads of families rule their own households, including their slaves.

Within ill-defined limits, certain tribes are grouped together under a common leader known as the “agoalla” or “agwalla.” This usually occurs in the case of tribes which are nearly related to each other. Three groups in the Kel Owi division have already been mentioned; in two of these, the Kel Tafidet and Kel Azañieres, the office of “agoalla” is said[138] to be hereditary, but I have been unable to find any confirmation of this except in so far as the son of a man who, by his personal ascendancy, has secured control over more than one tribe, would probably more easily step into his father’s shoes than another person. The grouping of tribes may also occur for military reasons, but in such cases it has a tendency to be of a temporary character. It is best to assume that the tribe is the unit of Tuareg society and that the tribal chiefs are the elements of which their Government is constructed. “Agoallas” are an exotic form principally due to individual personality or temporary conditions prevailing over long-standing customs.

Tribes sometimes group themselves into temporary or permanent alliances. The former probably spring from military exigencies, the latter may be due to common origins in the recent past. Such aggregations as the Kel Azañieres and Kel Tafidet in the Kel Owi tribes are so obviously due to common tribal origins that they require no further examination. But the Kel Owi confederation in Air plays a far larger rôle than do mere tribal alliances. Here is no mere question of relationship or community of origin, but a more strict bond, which, however, cannot be defined. Such groups as these have been termed confederations, though the term is a little misleading, as no unity of government is implied. The origin of the confederation, which carries with it more moral than material obligations, is to be explained by the entry of the Kel Owi tribes into Air as a mass of people confronted by an already established hostile or at least jealous population of the same race as themselves. It followed that the new arrivals would tend to hold together and act with one another. The conditions of the confederation nevertheless have been such that the representative is only an administrative head and not a ruler. He is there to embody a common policy and to dictate one. Loose as these bonds have been they have served the Kel Owi in good stead, for their commerce has gained by co-operation at the expense of their rivals, the “People of the King,” who in the absence of any organisation have been forced to rely on the fickle ties of common jealousy. How far there are groups or confederations like the Kel Owi within the larger northern division of Azger or Ahaggar I cannot say, but the former are a confederation as the people of Air generally never have been.

Much has already been said of the status of the Tuareg men and their tribal organisation, but before it is possible to consider their family life, the method they follow in tracing their descent must be described. A man’s status, in Air, as elsewhere among the Tuareg, is determined by the caste and allegiance of his mother. Survivals of a matriarchal state of society are numerous among the People of the Veil. They colour the whole life of the race. A woman, they say, carries her children before they are born, and so they belong to her and not to the father. “After all,” as one of them said to me when we had been discussing this question for some time, “when you buy a cow camel in calf, the calf is yours and not the property of the man who sold the camel to you. It is the same with women,” he added; and he seemed to me to have some show of logic. Our medieval (and perhaps modern) lawyers would have said instead, “partus sequitur ventrem,” but he would have meant the same as my Tuareg friend. If a woman marries a man in her own tribe the children, of course, belong to that tribe, but if she marries away from her people they belong to her own, and not to her husband’s clan. In this case, were the husband to predecease his wife, the children and their mother would return to live with her tribe. If the father survives, the children usually go on living with him for a time, but as they belong to their mother’s tribe in any event, they eventually return there. Should inter-tribal hostilities break out they must leave their father and fight for their mother’s tribe, even against their own parent if need so be. Until this is understood the relationships of the Tuareg appear very puzzling to the traveller. When I first met Ahodu he informed me that he was of the Kel Tadek people, who are Kel Amenokal, but he had a half-brother and a paternal cousin who belonged to the Añastafidet. It appears that the fathers of Ahodu and Efale, the famous eastern guide, were brothers of a man in the noble Kel Fares of the Kel Owi confederation. Ahodu’s father took a wife from the Kel Tadek, so the son became a member of the latter tribe, whereas Efale’s father married within the confederation. The maternal allegiance is so strong that, though proud of his father’s repute as a holy man and representative of the fifth generation of keepers of the mosque of Tefgun near Iferuan, Ahodu used to speak of the Kel Owi in disparaging terms when comparing their recent origin with the antiquity of the Kel Tadek and the other “People of the Amenokal.”

The following examples of definite cases may assist in understanding the position:

1. A man of the noble Kel Tadek marries a woman of the noble Kel Ferwan. The children are Kel Ferwan, but will live with the father until his death or the divorce of the mother, when they return with her to her own tribe.

2. A man of the noble Kel Tadek married a woman of the Imghad of the Kel Ferwan. The children will normally be Imghad of the Kel Ferwan.

3. If a man marries a slave woman of another tribe, this woman has become the property of the husband’s tribe by his purchase or payment of the marriage portion, and the children belong to the father. This occurred in Ahodu’s case. One day the Kel Gharus came over and stole eight slaves belonging to the Kel Tadek, who proceeded to retake them. The slaves in question were Kanuri people of Damagerim. The Kel Gharus appealed to the religious court at Agades, which awarded four slaves to each tribe. Later two of those allotted to the Kel Gharus ran away to the Kel Tadek, who were allowed to keep them on the ground that they had been ill-treated by their former masters. One of these two women Ahodu married, and his son is considered to belong to his own clan and not to his wife’s former tribe. In this case Ahodu nevertheless had to pay some compensation to the former masters of his wife.

The derivation of tribal allegiance through the female line has carried in its train the consequence that a man or woman’s social status is always determined by that of the mother. But the restricted number of noble women, the deference and respect paid to them, and the impossibility of taking them as concubines have combined to diminish the numbers of Imajeghan as compared with the Imghad. The hard-and-fast rule among all the Tuareg, that nobles can only be born of a noble mother irrespective of the caste of the father, has done much to preserve the type and characteristics of the race. In recent years the custom has tended to break down, for where a noble father, who has taken unto himself a servile wife, is sufficiently powerful to assert himself he will often succeed in passing off his sons and daughters as Imajeghan. Ahodu has done so with his boy; but had this been impossible the child would have been accounted of the Irejanaten or mixed people. The old laws of succession are said by von Bary to have become especially slack among the Kel Owi, but even here the status of noble women has remained so unassailable that it would still be impossible to-day for them to marry outside their own class.

The laws of inheritance and succession also show the strength of the matriarchal tradition. Although hereditary office is rare among the Tuareg nowadays, it seems to have been more frequent in the past.[139] Ibn Batutah states that the heir of the Sultan of Tekadda was the son of the ruler’s sister.[140] Similarly of the Mesufa who were Tuareg, he records that descent is traced through the maternal uncle, while inherited property passes from a deceased man to the children of his sister to the exclusion of his own family.[141] The traveller adds that nowhere except among the infidel Indians of Malabar did he observe a similar state of things.[142] Bates thinks that Egyptian records tend to show that the succession of the chieftainship of the Meshwesh Libyans passed in the female line. The genealogy of many of the kings of Agades is recorded by their female parentage. The Tuareg of Ghat not only treat their women-folk in much the same way as their brethren further south, but Richardson specifically states that the succession of the chiefs and Sultans of those parts is similar to the practice of the Tekadda house and at Agades. It is the son of the sister of the Sultan who succeeds.[143] It seems clear that before the advent of Islam, which has tended to modify the system, the Tuareg had a completely matriarchal organisation. In this earlier state of society may perhaps be found the explanation of the reputed Amazons of the west of North Africa, recorded by Diodorus Siculus in a grossly exaggerated version of some story which he had probably heard concerning the status of certain Libyan women.[144]

I know of no reason to suppose that these matriarchal customs were derived from association with the negro people; the reverse is quite as likely to have occurred, as the culture contacts of North Africa, following the trend of migration, seem to have taken a course from north to south and not the opposite direction.[145] The matter is one of great interest,[146] for the matriarchate is found in a highly developed state in Ashanti, and it would be of interest in connection with the origin of this people to learn if the system can be traced to a common origin.[147] I cannot agree with Barth’s[148] conclusion that the descent of the Sultan of Tekadda “is certain proof that it was not a pure Berber State, but rather a Berber dominion ingrafted upon a negro population, exactly as was the case in Walata,” where he cites the case of the Mesufa. Moreover, this remark is in contradiction with his previous assumption,[149] to wit: “With respect to the custom that the hereditary power does not descend from the father to the son but to the sister’s son . . . it may be supposed to have belonged originally to the Berber race; for the Askar (Azger), who have preserved their original manners tolerably pure, have the same custom. . . . It may therefore seem doubtful whether . . . this custom belonged to the black native,” with which statement I am decidedly inclined to agree. The problem, however, is one which I prefer on the whole to leave to qualified anthropologists.

[Footnote 107: Not to be confused with Tanut in Damergu. The word “tanut” means a shallow well; there are consequently many places of this name.]

[Footnote 108: Just north of Auderas.]

[Footnote 109: Barth, _op. cit._, Vol. I. p. 385.]

[Footnote 110: Jean, _op. cit._, pp. 148-9.]

[Footnote 111: Von Bary’s Diary (French edition), p. 183, etc.]

[Footnote 112: The available data are in the hands of the author, if some more fortunate traveller can check and examine the place.]

[Footnote 113: The “El Hakhsas,” Barth: _op. cit._, Vol. I. p. 416.]

[Footnote 114: The extremes in variation, for the first rains of sufficient volume to fill stream beds of a certain size with flood water, are recorded by von Bary east of Bagezan on 3rd June, 1877, and by Barth in Northern Air on 1st September, 1850. Both these dates seem to be exceptional.]

[Footnote 115: This, and not T’efira, is presumably the point south of Auderas where Barth saw “natron” encrustations on the ground (see Vol. I. p. 389). Salt or “ara” is collected at T’efira further east, but Barth would not have described “entering” the Buddei valley after seeing the “natron,” for the road past Auderas to T’efira winds down the Buddei valley.]

[Footnote 116: This is the vowel which in English words “oft_e_n,” “_a_non,” “_u_ntil,” may be written as _o_, _e_, _a_, or _u_.]

[Footnote 117: Cf. Barth, Vol. I. p. 350, and von Bary, p. 169, on the Kel Ataram of Auderas. The people of this village were simply “People of the West” for the inhabitants of Ajiru in Eastern Air, where von Bary was living.]

[Footnote 118: As Barth would have it: _op. cit._, Vol. I. p. 339.]

[Footnote 119: Cf. Barth, _op. cit._, Vol. I. pp. 339 and 347.]

[Footnote 120: The Cortier 1/500,000 map shows a large affluent to the right bank joining the Auderas valley below the village. This is incorrect: a small affluent called the Mafinet joins at the point shown, but the valley purporting to be the upper part of the Mafinet valley is the Tagharit valley, which falls into the Ben Guten, and not into the Auderas basin. The Cortier map is generally somewhat incorrect in this area, especially in regard to the position of Mount Dogam.]

[Footnote 121: I am indebted to Sir J. Currie of the Empire Cotton-growing Corporation for these reports.]

[Footnote 122: For fear of appearing to misinform people who are always ready to mind other people’s business before looking after their own, I hasten to add that the legal practice of slavery has, of course, been abolished in Air since the advent of the French. The psychology and habit of slavery, nevertheless, still remain as strong as ever, and master and slave continue to regard each other by _mutual consent_ in the light of their former relationship. I therefore propose to refer to slaves and the custom of slavery as if they were still sanctioned by law.]

[Footnote 123: Respectively “Akel” and “Irawel” in the singular.]

[Footnote 124: Barth, _op. cit._, Vol. I. p. 344 _sq._]

[Footnote 125: Cf. _infra_, Chaps. XI. and XII.]

[Footnote 126: Bates, _op. cit._, p. 115.]

[Footnote 127: _Vide infra_, Chap. XI., _et apud_ Barth, _op. cit._, Vol. I. pp. 235 and 239.]

[Footnote 128: When von Bary (_op. cit._, p. 184) says that Imajeghan were never enslaved, he is wrong. Although the Air Tuareg, when they raided the Aulimmiden, often used to lift their cattle but spare the men because they were of the same race, some of the latter division nevertheless, became Imghad of the Air Kel Ferwan, for instance, in the course of these raids.]

[Footnote 129: This is, of course, not an absolute rule, for the “I name” might have been forgotten, as previously explained. The supposition that “Kel names” represent Imghad and the “I names” Imajeghan is, of course, quite untenable.]

[Footnote 130: The singular form of Imghad.]

[Footnote 131: There are several instances of this among the Northern Tuareg, as will be seen from the data contained in Chap. XI.]

[Footnote 132: Cf. Schirmer’s note in von Bary, _op. cit._, p. 184.]

[Footnote 133: Barth’s statement, _op. cit._, Vol. I. p. 237, that the Imghad are not allowed to carry arms is not substantiated: he seems at this point to have confused the Imghad with slaves.]

[Footnote 134: Cf. _supra_, p. 134. Von Bary, _op. cit._, p. 181, notes that the distinction between Imghad and Imajeghan among the Kel Owi seemed to have broken down. This is perhaps exaggerated, but interesting, as this division in a sense is the most modern in development in Air.]

[Footnote 135: Barth erroneously calls him the Astafidet.]

[Footnote 136: Cf. Badges of Office among Libyan rulers given by Bates, _op. cit._, p. 116.]

[Footnote 137: Von Bary, _op. cit._, pp. 172 and 188-9.]

[Footnote 138: By Jean, _op. cit._, p. 106.]

[Footnote 139: Cf. Bates, _op. cit._, pp. 112, 114-15.]

[Footnote 140: Ibn Batutah (ed. Soc. Asiatique), Vol. IV. pp. 388 and 443. Cf. also Appendix IV.]

[Footnote 141: The Mesufa are a surviving section of the Sanhaja, and are specifically described by Ibn Batutah and Ibn Khaldun as a part of the People of the Veil, _i.e._ not negroes or negroids (_vide infra_, Chap. XI.).]

[Footnote 142: This statement is made in spite of the reference a little later to the succession of the Sultan of Tekadda, who, though a Tuareg, does not seem to have been of the Mesufa. This little inaccuracy is, however, of no importance.]

[Footnote 143: Richardson: _Travels_, etc., Vol. II. pp. 65-6.]

[Footnote 144: Diod. Sic., iii. 53 _sq._ See also Silius Italicus, ii. 80. Bates, _op. cit._, pp. 112-13 and 148, agrees that the existence of matriarchal society would be a reasonable explanation of the Amazon story.]

[Footnote 145: Nevertheless the matriarchate is known to have existed in classical times as far south as Æthiopia, in the Meroitic kingdom as well as in early Egypt.]

[Footnote 146: Perry (_The Children of the Sun_) would doubtless suggest that it came from Egypt.]

[Footnote 147: See Rattray, _Ashanti_, 1924. This authority thinks that the Ashanti people themselves came from the north. Many of the details of their matriarchal system accord closely with that of the Tuareg.]

[Footnote 148: Barth, Vol. I. p. 388.]

[Footnote 149: _Ibid._, p. 341. On page 342 he says the Aulimmiden, who have the same custom, consider the practice shameful, “as exhibiting only the man’s distrust of his wife’s fidelity; for such is certainly its foundation.” I don’t agree with this conclusion; the origins of matriarchy are certainly not as simple as this.]