I 607 iDRYEY OF THE ETHNOGRAPHY OF AFRICA : iND THE FORMER RACIAL AND TRIBAL MIGRATIONS IN THAT CONTINENT. SIR H. H. JOHNSTON, G.C.M.G., K.C.B., 1* PUBLISHED BY THE $nf#ropofogicaf 3n0ftfufe of <&reaf Contain and Jrefanfc, 50, GREAT RUSSELL STREET, LONDON", W TEL. AND CABLE ADDRESS, "U LANTHES, STOCK, LONDON? TELEPHONE 239 LONDON WALL. LfclH!AM) All BUSINESS communicarions ro be addressed lo THE MANAGER. All EDITORIAL THE EDITOR. E.G. 15, Lawford Kent is Dear Or. Gaster, Knowing your interest in anthropology you would like to have the accompanying publicat: not be returned. With kind remembrance to yourself anc and hoping you are both quite well. Yours sincei oad , Town, N.W I thought It need rs . (raster, )F THE ETHNOGRAPHY OF AFRICA: AND IER RACIAL AND TRIBAL MIGRATIONS IN THAT CONTINENT. 2 J 1708 5 TEL. AND CABLE ADDRESS, "U LANTHES, STOCK, LONDON? TELEPHONE 239 LONDON WALL. THE 1 All BUSINESS communicanons \o be addressed \o THE MANAGER. All EDITORIAL THE EDITOR. Dear Br. (raster, Knowing your inter you would like to have the ace not be returned. With kind remembra and hoping you are both quite A SUEVEY OF THE ETHNOGRAPHY OF AFRICA: AND THE FORMER RACIAL AND TRIBAL MIGRATIONS IN THAT CONTINENT. 21 1708 5 375 A SURVEY OF THE ETHNOGRAPHY OF AFRICA: AND THE FORMER RACIAL AND TRIBAL MIGRATIONS IN THAT CONTINENT. BY SIB H. H. JOHNSTON, G.C.M.G., K.C.B., D.Sc. THE paper which I am about to read contains no element of deep research and perhaps no complete solution of any ethnological puzzle. It is rather a statement of the problems and probabilities and the few known facts connected with the archaeology and history of Man in Africa. It has been compiled in the hope that by calling attention to these theories and questions, workers in the field of research may be led to fill in the outlines and to supply further and more detailed informa- tion as to the evolution of races and the gradual human conquest of this still mysterious part of the earth. There is less written history in connection with Africa (outside Egypt) than with Asia or Europe ; though of course in this respect Africa ranks before America and Australasia. In the lower valley of the Nile, history as inscribed by the brush, the reed, of 1 the graving-tool goes back earlier than in any other part of the Old World, except Mesopotamia Mesopotamia which may have passed on from Western Asia the Neolithic civilization that was to cross Arabia and Syria and find a wonderful sphere of development in North-east Africa. But for the deductions that follow, we depend less on written history than on the discoveries of palaeontology and archaeology, on the study of philology, and the consideration of oral tradition. Discoveries of human implements show the Nile Valley to have been inhabited by man of Palaeolithic and Neoli thic culture at a time considerably remote from the present day perhaps 20,000 to 30,000 years ago ; but so far there are no bones, rock pictures or sculptures to show what were the most ancient types of human being using those implements. In the prehistoric graves of Lower Egypt small figures of carved stone have been found which show some resemblance to the Bushman type in their marked steatopygy, 1 but this resemblance is not reinforced by the face, which so far is missing or too rudely liinued to serve as evidence. The earliest pictures given to us by the dynastic Egyptians of the wild aborigines of the Nile Delta are engraved on slate palettes, and depict a dwarfish Negro-like race, not unlike the Congo pygmies of to-day differing from them only in possess- ing rather bigger though flattish " Papuan "-noses with bushy heads of closely curled hair. The males are circumcized (after the Masai fashion) ; they are bearded 1 The same may be said about the little statuettes or figurines found in South-west France. a 2 376 SIR H. H. JOHNSTON. A Survey of the Ethnography of Africa, etc. like the modern pygmies, and in some of the other representations the women (like the Congo pygmies and Tasmanians) grow slight whiskers. I do not know if there are in Egypt or Nubia rock pictures or engravings of as remote a date as those to be seen in Tunis, Algeria, Tripoli, and above all in Southern Morocco and across the Sahara Desert. In the Sahara remarkable discoveries may yet await us, and too little publicity has been given to those already made, owing to the jealousies between ethnologists. Some of these engraved rocks, or photographs of them, can be seen in the Museum or the University of Algiers ; and certain publications have been issued at Lyons dealing with the rock drawings of Northern Africa. Here we see depicted I have had the experience myself, in South-western Algeria an extinct type of buffalo, elephants, and (I am told) giraffes, which have long since vanished from the North African fauna. In the few examples which show man associated with these beasts, the human type is rather that of the Caucasian than the Negro. 1 Yet the skulls of greatest antiquity judged from the depth at which they were embedded which have so far been obtained in North Africa (and stored in the University of Algiers), indicate a Negroid type as being the most primitive Algerian people. Some French anthropologists, however, advance the idea that Neander- thaloid man reached North Africa, especially Tunisia, and that he has even left his descendants there in tribes like the Mogods of the Bizerta district. These Mogods, whom I have seen myself, have a very savage look, with their strongly developed brow-ridges and tendency to prognathism, but their divergence from the average type of Mediterranean man is merely Australoid. The stone implements of Somaliland seem to show the existence of an ancient human population in the eastern horn of Africa. And the same may be said of Senegambia, Inner Nigeria (especially of the northernmost part of the Niger Valley), of the Gold Coast, of West Congoland, and of South Africa from the Zambezi to the Cape of Good Hope. It is, however, remarkable, that so far stone implements have not been discovered in Central Congoland, in the very heart of the Congo Forest. They have only been found along the fringe of the Congo basin ; in the far north, in the open country of the Mubangi, in the region between Southern Tanganyika and Angola, and again in the Cataract region of Western Congo. This and other negative evidence points to the supposition that the most densely forested regions of Central Africa represent areas which have only been penetrated by man within a comparatively recent period. There are traditions at the present day amongst the Bantu tribes of Central Congoland that their ancestors only arrived in this region some few centuries ago ; or at most 1,000 to 1,200 years back (as far as such terms of years can be calculated by the reigns of successive chiefs and dynasties) ; and further it is generally related, both here and in the forested parts of West Africa, that when these metal-using, big-bodied peoples arrived, the forest 1 See Bulletin Trimestriel de la Soctite de Geographic et d? Archeologie d^Oran, December, 1908. Also the writings of L. Pomel, J. B. Flamand, and E. F. Gautier easily found in British Museum Library. Sm H. H. JOHNSTON. A Survey of the Ethnography of Africa, etc. 377 country was only occupied by a dwarfish race, sometimes referred to as " red men," in fact, similar to the Congo pygmies of to-day. These traditions of a pre-existing dwarfish (occasionally red-skinned), negroid race are existent in parts of Inner Nigeria and in the southern oases of the Sahara. It is in South Africa that the most interesting and the most puzzling of discoveries have recently been made regarding the antiquity of prehistoric Man. When the Hottentots of South Africa were questioned by scientific men a hundred years ago and more regarding their traditions, they were wont to refer to their predecessors on the coast of South Africa as a savage race living on the seashore and subsisting on shellfish and the bodies of stranded whales. From their habits these were styled in Dutch the Strandloopers or " Shore-runners." About tw r enty years ago discoveries began to be made of skulls and bones of the Strandloopers in caves on or near the sea-coast, and the first of these that were brought to light, and described by Dr. F. C. Shrubsall and others, exhibited (I am told) some degree of aflinity with the Bushman, but appeared to be more primitive and perhaps more typically Negro in the prognathism they exhibited and the proportionate length of the skull. There would seem to be survivors of this prognathous Strandlooper type amongst the Bushmen and Hottentot peoples of South-west Africa at the present day, for I throw on the screen slide-photographs of a Bushman and of a so-called Hottentot woman (actually, we know, of Bushman race) which exhibit a degree of prognathism exceeding that of any other race known in Africa or elsewhere. They thus differ sharply from the conventional Bushman head, which is round and with a comparatively high facial angle, and little or no protrusion of the jaws. [These photographs are not isolated examples, but only selected instances. See Note at end of paper.] But other specimens of Strandlooper skulls, such as those described recently by Dr. Peringuey, were neither Bushman nor particularly Negro in their affinities. They exhibit considerable brain capacity, and recall much more in general outline the skull form of generalized Negroid-Caucasian types, say, for example, the Hamitic peoples of North-east Africa ; and they have even been compared to the G-alley Hill skull of East Kent a primitive Caucasian type that existed in England some hundred thousand years ago. With this large-brained skull may be apparently associated in South Africa those stone implements which are pre- Bushman, in some respects superior to Bushman arts and inventions. 1 1 Our chief source of information on the subject of the Strandloopers is vol. viii of The Annals of the South African Miweum, written in the main by the Director of that Museum, Dr. L. Peringuey. In Chapter 22 of that work, Dr. Peringuey discusses the evidence of the Strandlooper skulls as analysed up to about 1911 by Dr. F. C. Shrubsall (F.B.A.I., of London). Dr. Shrubsall has studied the Strandlooper question for something like fifteen years, and has written on it in various works, iucluding the present writer's " Uganda Protectorate." It is understood he has much unpublished material on this important question to give to the world, so that when his final verdict is pronounced our conclusions may be strengthened or changed. But as they stand at present the results of his investigations would seem to show That the Strandlooper of the Cape Colony caves (as distinct, perhaps, from the miserable race of shore dwellers exhibiting the prognathism above referred to) preceded the Bushman in 378 SIR H. H. JOHNSTON. A Survey of the Ethnography of Africa, etc. The next oldest type in South Africa was probably the Bushman, and the Bushman certainly seems to have reached Southern Africa from the north-east. We can trace his physical peculiarities up through Eastern Africa into Galaland, and some say even into Egypt. It may be as well at this juncture to consider what are the leading Bush physical peculiarities, apart from skull formation. They consist of the steatopygia already alluded to, and small, folded-over ears, of the hypertrophy of the labia, vulvce minor a in the woman, and, in the men, of the peculiar angle at which the penis is set in relation to the pubic region. The " pepper-corn " aspect of the tightly curled head-hair which seems only to be planted in distinct patches together with the light-yellow, wrinkled skin and the small hands and feet, are features shared by other African or even Oceanic groups of Negroes. This pepper- corn arrangement of the head-hair tends to disappear in the north Kalahari Bush- men, in whom also the skin colour is darker and the wrinkles are usually absent. These North Kalahari or Ovampoland Bushmen usually exhibit some degree of brow-ridge development in the male, their cheek-bones are very prominent, the base of the nose is much sunk and the eyes are very deep set. South Africa ; that the Strandloopers were a race of short but not dwarfish men (the. earlier examples were taller than the later) with a much higher skull capacity than that of the average Bush race 1,500 cubic centimetres as against 1,300 for Bush men and 1,350 as against 1,200 for Bush females. The extreme of cranial capacity in the Strandloopers was a maximum of over 1,600 cubic centimetres, while the extreme minimum among the Bush people descends as low as 955 cubic centimetres ! The frontal region of the skull is much better developed than in the Bush race, and in that respect is more like the Negro. There is little or no brow promi- nence, and one at least of the skulls is as orthognathous in facial angle as that of a European ; others of the skulls are slightly more prognathous, but the Strandlooper type was as orthognathous as the average European. The teeth are well developed and liberally spaced, but are not proportionately as large as those of the true Negro. Dr. Pe'ringuey denies there being any definite resemblance between the Strandlooper cave-dweller and the " Grimaldi " skulls of Monaco. He considers that these South African cave-dwellers were the most primitive known race of South Africa, the oldest examples being the least negroid in features. Yet the affinities between them and the Bushmen were more evident than between Cave Strandlooper and Negro. Professor Henry Balfour (who will shortly publish an essay on the subject in the Journal of the African Society) seems to be of opinion that prior to the occupation of South Africa by Bush, Hottentot, and Bantu, there was a race of superior PahBolithic, almost Neolithic culture, judging from the oldest stone implements and other relics discovered south of the Zambezi. This same race apparently colonized even more markedly the Central Zambezian regions, and to it may be perhaps accredited the ancient pottery discovered at considerable depths below the surface of Nyasaland. Dr. Peringuey, on p. 215 of the work quoted above, sums up his conclu sions regarding the Cave Man of Cape Colony : He was less dolichocephalic than the Bush- men and Hottentots, under 80 in cephalic index. " He was artistically gifted, like the race which occupied and decorated the Altamira . . . and other caves of Spain and France. He painted ; he possibly carved on rocks ; he used bone tools ; he made pottery ; he perforated stones for either heading clubs or to be used as make-weights for digging tools (I have obtained the.e perforated stones from South Tanganyika and Nyasaland H. H. J.) ; his ornaments consisted of sea-shells ; and the ostrich egg-shell discs, which he made, may be said to be a typical product of his industry. And this culture is retained in South Africa by a kindred race, but more dolichocephalic the Bushmen-Hottentots. . . . Analogous are most of his tools and his expression of culture to those of Aurignacian man." SIR H. H. JOHNSTON. A Survey of the Ethnography of Africa, etc. 379 In the ordinarily small stubby feet of the Bushmen the great toe is usually straight, or curves, if at all, inward and away from contact with the other toes, a con- dition which is seen in an exaggerated form in the Oceanic Negroes and the New Guinea Negritos. This " open-toed " aspect of the short foot is one of the few traits which the Bushman shares with the Congo pygmy, though the work of Dr. F. Fiilleborn on the natives of East Africa shows that the Bushman type of foot oblong, with inverted big toe is sporadic in Nyasaland and Mozambique. As to steatopygia, it occurs in a slight degree among the Congo pygmies, but is usually absent from West African Negroes. 1 It is nevertheless observable some- times among the American Negroes (mostly recruited from West Africa), as may be noted in one amongst a number of photographs taken for me at the Tuskegee Institute to illustrate points in negro physiology. In Nileland as on the Kru Coast there is sometimes much steatopygy among the women, but it is less a great backward projection of the nates than a lateral accumulation of fat along the hinder aspect of the thighs. But the Bushman type of steatopygy reappears sporadically among the Somali and Gala women (negroid Caucasians), among Egyptian women of to-day, and even in Syria, Southern Europe, and it is said as far north as Poland, carried thither by the Jews. In France fashion has in the latter part of the 19th century deliberately cultivated the steatopygous outline in woman, and this pre- dilection was based as Dr. Pe"ringuey suggests on a racial admiration for " Venus callipyge," due, possibly, to the deep-down stratum of protonegro race which permeates the Mediterranean peoples. Steatopygia recurs in Arabia and among the Negroid people of South-west Persia (the old Elamites), and according to several German writers is occasionally observable among Oceanic Negroes. As regards the exaggerated growth of the Idbia minor a, the " longinymph " condition which produces what was called a hundred years ago the " Hottentot apron " (illustrated by photographs in Dr. Pe'ringuey's work), this likewise is not a feature exclusively confined to the Bush-Hottentot race, but, as Dr. Karl Weule points out (Native Life in East Africa : Sir Isaac Pitman and Sons, 1909), is of constant occurrence among the Negro women of East Africa, while others have earlier stated that it is a character met with among Egyptian, Abyssinian, Gala, and Somali women. Lastly, the peculiar set of the penis in Bushmen 2 occurred obviously among the Mukenaian and Ionian people of the Eastern Mediterranean, as may be noted in the early Greek designs on painted pottery. Like the steatopygy, the crumpled ear, and the longinymph development, this peculiarity of the Bushman male organ can be matched here and there among the nomad heterogeneous, broken, 1 In all Negro and Bushman children and in the men and women of many Central and East African tribes the development of the nates is actually less than in Europeans. 2 "The penis is normally carried horizontally and in some } r oung subjects in a semi-vertical or even vertical position. In paintings it is the same ; man's genitalia are never hanging when Bush people are delineated." P6ringuey. The practice of pushing up the testicles when the child was young, so that the growth of a pendant scrotum might be avoided as much as possible, lest it impeded flight, was not peculiar to the Bushmen and Hottentots, but was recorded by Greek geographers as among the practices of the Hamitic tribes of the Red Sea littoral. 380 SIR H. H. JOHNSTON. A Survey of the Ethnography of Africa, etc. hunting tribes of Equatorial East Africa, who evidently contain elements of a now absorbed Bushman race. As regards linguistic evidence of Bushman distribution and migrations, I might point out that Ludovico di Varthema, an Italian adventurer who visited Mozambique about 1505, relates how in a mountain of caves on the mainland of Mozambique, was a people of dwarfish stature and yellow skin, whose speech was full of clicks. He compares these clicks to the noises made by Sicilian mule- drivers in urging on their beasts. I have collected traditions amongst the Bantu peoples of Nyasaland as to the existence down to quite recent times of a dwarfish, yellow-skinned, click-using, stone-throwing people on the tops of certain high mountains. At the present day there is a click-using language of undecided affinities the Sandawi, spoken by a cattle-keeping, semi-nomad tribe to the south of the Victoria Nyanza in German East Africa. The existence of clicks in this language is undoubted, but I have not been able to trace much affinity in word roots between this language and either Bushman or Hottentot, though it is noteworthy that the word for four, ha^a, is almost identical with the word for four in all the Hottentot dialects, while the phonology of the language is reminiscent of Bushman in its nasals and gutturals. 1 Various travellers have thought that they saw a physical resemblance to the Bushman in some of the Andorobo hunting people of Eastern Equatorial Africa. The Andorobo are mixed in their racial elements like the Sandawi. Some examples are almost Hamitic in face, others are dwarfish and offer a slight resemblance to the Bushman type, while others again are very prognathous and present a facial resemblance to certain Melanesian types of Oceania. 2 The language spoken by the Andorobo no doubt imposed on them by conquerors is an outlying member of the Nandi-Nilotic group. Where did the true Negro type, with more or less black skin, tall or normal stature, large feet, big teeth, thickly-growing woolly hair, originate ? This is a problem that perhaps lies outside Africa and the scope of my paper, and it is one as to which we can only hazard vague guesses. We find a markedly Negro type in Papuasia, more especially in the islands of the Bismarck Archipelago to the east of New Guinea. The distribution of the Negro throughout Southern Asia occurs sporadically, but scarcely anywhere in a pure form. The type of Asiatic Negro existing at the present day which most nearly resembles the Negro races of Africa, is that which is to be found in the northern Solomon Islands and the Bismarck Archipelago. The Grimaldi skulls of Monaco were Negroid in some features, but they show a brain capacity much exceeding that of African or Asiatic 1 In the Pygmies of the north-eastern corner of the Congo basin and amongst the Bantu tribes of the Equatorial East .African coast there is a tendency to faucal gasps or explosive consonants which suggests the vanishing influence of clicks. 2 The Nandi people on the western and the Kikuyu on the eastern side of the Rift Valley in Equatorial East Africa offer similar divergent types of prognathous, small-featured, low- statured Negroids and comely, big-nosed, orthognathous Caucasoids. SIR H. H. JOHNSTON. A Survey of the, Ethnography of Africa, etc. 381 Negroes. Their age has been guessed at some 30,000 years ago, but this may be an under estimate. 1 It is conceivable that the Negroid sub-species originated in Western Europe and thence spread southwards into Africa and across Western and Southern Asia into Australasia, reaching in its most generalized and primitive type the far distant Island of Tasmania. There is some evidence to show that Algeria (and North Africa generally) was inhabited by a Negro race before these regions were invaded by the Caucasian. The Congo Pygmy seems to be little else than a primitive and dwarfed form of the Forest Negro, perhaps representing one of the earliest types of Negro that invaded Africa. The Negro specialized in the swamps of Nileland into what we know as the Nilotic Negro : tall, even to gigantic, long-legged, but as a rule with comelier features and a better development of brain than in the Forest Negro. There is perhaps a third variety to be distinguished : the Sudan Negro : the type so often met with in Bornu. in the Bahr-al-Ghazal, Kordofan and the western flanks of Abyssinia ; also in French Nigeria. This Sudanese Negro shares the tall stature and thin shanks of the Nilotic, but sometimes exhibits a good deal of prognathism and has the triangular face and the aggressively projecting cheek- bones of the Bushman-Hottentot and the exaggerated eversion of the lips, which is also a feature in the West African Forest Negro. The fusion of these three varieties, dashed here and there with Pygmy and with Bushman blood, gives us the " Bantu " Negro, a linguistic term which merely connotes a vast congeries of Negro tribes united by the common bond of language. Except as indicating an average Negro, and a fusion of Negro types, the term Bantu has no value in physical classification. Looking back into the past by the very dim light furnished by the discovery of crania and bones, a few objects of human art, and the indications of language affinity, we seem to see an Africa 15,000 years ago which was mainly Negro or Negroid in population. But already at this period the Mediterranean man a Caucasoid had entered Egypt and occupied the Nile Delta, driving away or absorbing its dwarfish Negro population. He might even have penetrated up the Nile into the highlands of Abyssinia. He had certainly much earlier begun to colonize Tunisia, Algeria, and Morocco, and was pursuing and destroying the remarkable mammalian fauna of North Africa, which even at as recent a date as about 15,000 years ago had a decidedly African fades, as already indicated in my opening remarks. What was the type of language spoken by the earliest white colonists of North Africa ? Some French ethnologists have suggested that it may have been the ancestor of the Fula, Wolof, Temne, Bantu, and Kordofan groups, a type of language offering faint resemblances in structure with the Lesghian speech of the Caucasus and the Dravidian tongues of Baluchistan and India ; a speech in which 1 See the monograph of Dr. Verneaux on the Grottes de Grimaldi. Dr. Verueaux has also made an exceedingly interesting anthropological survey of Abyssinia, published (1911) by the French Government, which bears on some of the theories advanced in this paper. 382 SIR H. H. JOHNSTON. A Survey of the Ethnography of Africa, etc. nouns were divided into more or less numerous classes with distinctions not based on sexual gender, and in which as a rule the root was unchanging, while much use was made of detachable prefixes or suffixes, linked up throughout the speech by concordant adjectival and pronominal particles. According to such a theory, therefore, the first Caucasoid invaders of North Africa would have used tongues akin to that of the Fula, and when they were forced by the impact of other immigrants from the north or east to wander into Negro Africa they became the ancestors of the Tula and carried their language family with them far into the Sudan and West Africa ; perhaps in the direction of Kordofan or the Equatorial Nile developing the Bantu family. In the far west of Africa, by intermixture with the Negroes, they possibly originated the Wolof, the Temne, Mosi, and other groups, and even influenced the tongues of the Niger Delta. The same theorists would contend that the Fula type of " white man " and of Caucasian (i.e., Lesghian, Georgian) language was displaced in North Africa by the invasion of the Hamites and the Libyans or Berbers. The Haniites were no doubt of common origin, linguistically and racially, with the Semites, and perhaps originated in that great breeding-ground of conquering peoples, South-west Asia. They preceded the Semites, and (we may suppose) after a long stay and concen- tration in Mesopotamia invaded and colonized Arabia, Southern Palestine, Egypt, Abyssinia, Somaliland, and North Africa to its Atlantic shores. The Dynastic Egyptians were also Hamites in a sense, both linguistically and physically ; but they seem to have attained to a high civilization in Western Arabia, to have crossed the Eed Sea in vessels, and to have made their first base on the Egyptian coast near Berenice in the natural harbour formed by Eas Benas. From here along, broad wadi or valley then no doubt fertile led them to the Nile in the Thebaid, the first seat of their kingly power. The ancestors of the Dynastic Egyptians may have originated the great dams and irrigation works in Western Arabia ; and such long struggles with increasing drought may have first broken them in to the arts of quarry- ing stone blocks and building with stone. Over population and increasing drought may have caused them to migrate across the Eed Sea in search of another home ; or their migration may have been partly impelled by the Semitic hordes from the north, whom we can imagine at this period some 9,000 to 10,000 years ago pressing southwards into Arabia and conquering or fusing with the preceding Hamites ; just as these latter, no doubt, at an earlier day, had wrested Arabia from the domain of the Negroid and Dravidian. The Dynastic Egyptians were not far distant in physical type from the Gala of to-day, but they had perhaps some element of the proto-Semite ; and their language, which is still rather a puzzle to classifiers, though mainly Kushite in its features, exhibited early in its history the influence of Semitic speech, and no doubt absorbed into itself elements of its Libyan sister, which it perhaps found already extending to the valley of the Nile. The Dynastic Egyptians evidently concentred themselves in the narrow strip of fertility along the banks of the Nile, not colonizing very markedly the Eed Sea coast-lands. By about 8,000 years ago they SIR H. H. JOHNSTON. A Survey of the Ethnography of Africa, etc. 383 had become the conquerors and rulers of Lower and Upper Egypt. The inhabitants of Egypt were thenceforth a people in which Hamitic (Libyan-Kushite), Semitic, Nilotic and even Sudanese-Negro elements were fused ; but the Caucasian type predominated. In the early days of their colonization, the Dynastic Egyptians domesticated only the indigenous African wild ass, the Nubian ibex, the audad or African wild sheep, the Leucoryx and Addax antelopes, and the striped hyena. 1 They probably brought with them from Arabia various types of domesticated dog. As they increased their trade relations with Syria and Mesopotamia they received from the north-east first the long-horned " Gala " ox (which originated seemingly in Western Asia), next the hairy, long-tailed, long-legged, inaned, domesticated mouflon of Europe and Western Asia, a little later the fat-tailed, dewlapped variety of this sheep, and, in succession, the " Wallachian " type of domesticated sheep (in which the spiral horns grew out at right angles to the line of the head), and the humped cattle which we know as the Indian breed, but which were near akin in origin to the early long-horned, straight- backed Egyptian ox. The first true goat which they received from Syria in a domestic form, and which rapidly displaced the ibex in their estimation, was a dwarf, plump, compact breed, with short hair and rather reduced horns the Guinea goat of to-day. 2 This was succeeded at a later date by the Roman-nosed Syrian goat with long hair, pendent ears, and a tendency to become hornless (a breed which has never reached real Negroland beyond northern Nigeria). It was not until about 500 B.C. that the Egyptians received the domestic fowl from Syria and Persia. The horse was brought by the Hyksos from Syria about 2000 B.C., but was not fully established in Egypt till about 1500 B.C., though it may have been separately imported into North Africa 3 at an earlier date. As early, perhaps, as the fifth millennium before Christ, the Pharaohs of Egypt turned their attention to the Sudan, the land of black men, which began in those days somewhere about the Second Cataract. Between the First and Second Cataract of the Nile the population was of a mixed type, Negro mingled with Kushite in various forms and degrees ; but south of the Second Cataract it was almost purely Nubian, that is to say, Nuba or Sudanese Negro. Higher up still, beyond the junction of the Blue and White Niles, the population was Nilotic 1 See Claude Gaillard's " Les Tatonneraents des Egyptians de 1'Ancien Empire a la recherche des animaux a domes tiquer " (Revue cP Ethnographic, etc., Paris, December, 1912). 2 This obviously ancient breed the only oue known to true Negro Africa penetrated through India to Malaysia and Borneo, where it still exists, displaying much likeness to the domestic goat of West and Central Africa. 3 It is very doubtful whether North Africa any more than Egypt or Arabia possessed a wild horse of the Caballus sub-genus. The bones of extinct equines indigenous to Algeria rather suggest a type of zebra than a horse. (See the writings of L. Pomel.) But the ancestors of the Barb or Numidian horse may have been brought by the Berbers from Spain at a comparatively early date. After 1000 B.C., the Phoenicians no doubt imported the Arab type of horse into Mauretauia and crossed it with the Spanish horse. So far, we have discovered no traces of any true indigenous wild horse in Syria or Arabia, only asses. It is probable that the true horse was confined in its area of evolution to Europe, Central Asia and North India. 384 SIR H. H. JOHNSTON. A Survey of the Ethnography of Africa, etc. Negro, though at a later date this region was colonized by Kushites of the Beja, Hadendawa, and Gala types. A brisk commerce in slaves, ivory, gold, leopard-skins, gums, and strange beasts and birds began between Egypt and the Nilotic Sudan, and the coming and going of caravans, probably composed of Nuba Negroes, with here and there an adventurous Egyptian, must have opened up communications between Kordofan, Darfur, and the regions round Lake Chad, as well as the purely Negro countries of the Bahr-al-Ghazal and the Mountain Nile. Thus as early as 4,000 years before the time of Christ, goats of the plump Guinea breed, hairy sheep of the long-legged or the fat-tailed varieties, and straight-backed, high-shouldered cattle of the long-horned Egyptian variety, had begun to penetrate Negro Africa across Kordofan to the regions-of Lake Chad, the basins of the Niger and the Benue and the Northern Kamerun, and all parts of West Africa. Together with these domesticated animals there came all musical instruments superior to the musical bow and the drum, several types of games played with hollowed or divided boards and counters, and a good many Egyptian notions about religion. Moreover, these emissaries of trade and culture may have brought with them to the Neolithic and Paleolithic Negroes and Negroids the earliest idea of using metals, copper in the first place, though perhaps not iron. 1 Athwart this fan-shaped spread of Egyptian influence over Negro Africa the apex of the fan being at the Second Cataract of the Nile and the two outermost spokes of its partially-opened spread lying (one) on the east, up the Nile Valley to the Great Lakes, (two) on the west through Kordofan, Darfur, and Lake Chad to the Niger Valley we can trace strands of other Caucasian migrations and influences coming due south from North Africa : possibly " Pelasgian " or " Mukenaian." But though here and there they leave traces of Pelasgian or Minoan civilization in the Western Sudan and West Africa, it is clear that Negroland, anterior to the Christian era,, owed less of its awakening to conditions of civilized human existence to the white men of North Africa proper (Mauretania and the Tripolitaine) than to those of Egypt. For example, all the domestic animals of Negroland (except such as have been quite recently derived from America or Europe) are of Egyptian origin or type and not North African. The early peoples of North Africa had a well- marked form of domestic dog, which exists down to this day amongst the Berber nomad tribes. This was of the long-haired, prick-eared, broad-headed Chow type, with a plumed tail tending to curl over the back a type of dog common in Southern Europe from the Bronze period onwards and far spread in its range over all Northern and Eastern Asia and Northernmost America. This Chow-like dog, though apparently it reached the Canary Islands, never penetrated (till modern tunes) across the Sahara Desert to Negro Africa, neither did the taurine (as contrasted with the zebu) type of ox, the woolly sheep, the long-horned European goat, or any other domestic animal or cultivated plant of European origin. It was Egypt that furnished Negroland not only with all its earliest domestic animals, but 1 See postscript at the end of this essay. Sm H. H. JOHNSTON. A Survey of the Ethnography of Africa, etc. 385 with all its cultivated plants, and Egypt derived these for -the most part from Syria and Arabia. The Dynastic Egyptians of later times having recreated their kingdom on the Lower Nile, after the invasion of the Hyksos, seemed to have lost interest in Negro Africa and thought mainly of extending their conquests across Syria to Asia Minor ; but they retained dim traditions of the land of their origin, South-west Arabia. Moreover, a considerable boat trade had sprung up on the Red Sea, and civilized South-west Arabia was beginning to colonize with its present Semitic population the highlands of Abyssinia, 1 and possibly a good deal of Somaliland. Thus the Arabs had become acquainted (in Somaliland) with trees yielding aromatic gums, present, 110 doubt, also in Arabia, but not in such considerable quantities. It is possible, there- fore, that Arabia first drew the attention of Egypt to the western land of Punt the eastern horn of Africa whither the Egyptians themselves sent ships about 1500 B.C. to open up a trade in incense and other products. Did the Dynastic Egyptians colonize Negro Africa as well as influence it ? Perhaps relatively little, at any rate until during the thousand years whiclr preceded the Christian era. We then seem to see traces of enforced migrations of rebels or malcontents, probably not of very pure Egyptian stock. Bands of such a kind seem to have entered the northern part of Hausaland and to have assisted in moulding the Hausa language as a medium of trade intercourse. This speech, Sudanese Negro in phonology and perhaps in most of its word roots, is emphatically Hamitic in its grammatical features and pronouns. But the Hamitic element is thought by experts to be as much Kushite, or even Koptic, as Libyan. On the whole, it seems probable that the Hausa speech was shaped by a double influence : from Egypt, and Hamiticized Nubia, as well as by Libyan immigrants from across the Sahara. 2 Egyptian adventurers no doubt from time to time found their way up the Nile through the countries of the Nilotic Negroes, and there became chiefs and almost demigods, so great being the Impression made by their lighter skins, their beautiful features, and possibly their domestic animals or simple arts. Sometimes one is disposed to think that those remarkable cattle-keeping aristocracies of the heart of Central Africa the Bahima, the Batusi, the Makarka, 3 and Mangbettu are descended from Egyptian colonists of 2,000 and 3,000 years ago. I have certainly seen individuals in Western Uganda and Unyoro who were so remarkably Egyptian (rather than Gala) in features that I took them actually for Egyptians left behind 1 The Himyaritic Arabs who Semiticized the Abyssinian highlands and incidentally Northern Somaliland, probably found Abyssinia populated by three principal racial elements : the Negro, the dark-skinned Kushite-Hamitic, and the white-skinned sometimes blue or grey-eyed and brown-haired Libyan. This, at least, is Dr. Verneaux's assumption. 2 See the article in the Mittnilungen des Seminars fiir Orientalische Sprache, Berlin, 1906, by Julius Lippert (" Uber die Stellung der Hausasprache," etc.). 3 A common name for the Makarka, Azande, and allied tribes ranging over the northern Bahr-al-Ghazal and the Northern Mubangi basin is " Nyam-Nyam " a cant term meaning " Meat ! meat ! " or " cannibals." Some of the Nyam-Nyam are broad-headed and brachycephalous. 386 SIR H. H. JOHNSTON. A Survey of the Ethnography of Africa, etc. by Emin Pasha's expeditions. But they turned out to be local aristocrats who knew absolutely nothing of Egypt and could enumerate their ancestors for centuries as residents in Equatorial Africa. Others again of this type were so strikingly like Galas and Somali that the Somalis of my party declared them to be of their own race. The pure-bred Gala resembles physically the Dynastic Egyptian. It is possible these aristocracies of East-central Africa and of the Central Sudan do not owe their origin to Egypt, but to former attempts of the Galas and Abyssinians to penetrate Negro Africa. Personally I am inclined to invoke both influences. It has indeed been suggested by Father Torrend in his Grammar of the South African languages, that the Zimbabwe civilization (which I would assert with emphasis is not Negro) is due to some unrecorded Abyssinian penetration of South- east Africa along the East African coastline. This theory is not quite so wild as it might appear, seeing that more or less degenerate Hamitic (Kushite) nomad tribes have penetrated, as it is, a considerable distance into German East Africa, as far south as the 4th or 5th parallel of S. latitude. Considered linguistically the Hamites of Africa became at an early date divided into two main sections (to which later was added a third, the Dynastic Egyptian). These two categories might be distinguished as Libyan and Kushite. The Kushite group passed down to the south along the west coast of Arabia and left a considerable element behind in South-west Arabia and in Sokotra, which afterwards became Semiticized. Then (we may suppose) they crossed the lied Sea 12,000 years ago or more, colonized North-east Africa and Somaliland, became the ancestors of the Bishari and Hadendawa, the Saho, Agau, Bogos, Kaffa, Afar, Gala, and Somali groups. 1 Entering also the valley of the Nile in the tropical and equatorial regions, they mingled with the Nilotic Negroes and formed subsidiary groups of peoples, whose languages to this day, though they are truly Nilotic and Negro, yet bear evidence of Hamitic influence. Of such, for example, are the Masai and Lotuka. Meantime the Libyan section of the Hamitic peoples, reinforced by the Berbers (Iberians) from Spain, much whiter in skin-colour, colonized all North Africa to the limits of the Sahara and the Libyan Deserts. They in their turn mingled with the pre-existing Tula type and the Negroids, and from some such mixtures, as well as the far earlier intercourse between the Fula and the Negro, were formed the Songhai, or " Habe " (Habe, plural of Kado, is only " slaves " or serfs in the Fula tongue), the Tibus or Teda, the Kanem and Kanuri peoples ; who, as lithe, handsome, and intelligent Negroids, ranged across the Sahara from Fezzan to Agades and Bornu and the oases of the Libyan Desert. These hybrids assisted to introduce into Negro Africa the white man's weapons and early domestic animals. 1 It is remarkable that the Bishari (Baja) and Hadendawa peoples (which formerly extended their rauge across Senaar to the Nile and Dongola ; Dongola = Dauagla, Dauakil) should have preceded and outlived the Dynastic Egyptians ; and equally remarkable that the British should have occupied Egypt and the Eastern Sudan since 1882 and have made no efforts to study and illustrate their interesting speech. SIR H. H. JOHNSTON. A Survey of the Ethnography of Africa, etc. 387 But if French aud German theorists are right. Northern Africa, after beinetween 100 B.C. and A.D. 500, bringing Roman beads to Hausaland and the Gold Coast, the pattern of Roman oil-lamps in metal as far south as the Baya country of the Nortli Cameroons, and creating the wonderful art and the strange religions, idols, and metal-work of Borgu, Yoruba, Benin, Iboland, and Calabar. The second and the principal line of Bantu migration (No. 2) at first skirted the dense Congo Forest on the west, penetrated round the Victoria Nyanza and down both coasts of Tanganyika. From the region between the Victoria Nyanza and Tanganyika a great stream (No. 3) of conquering peoples must have swept eastwards to the Zanzibar coast, . while a thinner migration (No. 4) came from Kavirondo eastwards and occupied the Kikuyu highlands and the slopes of Kiliman- jaro and Kenya. In the actual coast belt, north of Mombasa, the Bantu invaders have had to fight their way northwards being often exterminated and pushed back against the pre-existing Kushite, Nilotic and even Sudanese, Bushman, and Hottentot peoples. Dotted here and there over Eastern Africa, south of the Equator, are mysterious little groups of Negroes speaking languages that offer no marked relationship with any other. Such tongues as the Bakumu of the North-east Congo bend, the Sandawi, Mbulungu, and the Mbugu of German East Africa may be confined nowadays to a few thousand or a few hundred people in a little group of villages. They are evidently the relics of many unrecorded Negro colonizations of East Africa. All Bantu Africa of to-day, except the heart of the Congo Forest and the regions south of the Zambezi, must have been more or less thickly populated b 2 392 Sm H. H. JOHNSTON. A Survey of the Ethnography of Africa, etc. before the Bantu impressed with extraordinary rapidity and completeness their own type of language on the tribes they conquered. There are faint indications of people of pre-Bantu speech having existed in South-east Africa, south of the Zambezi ; pre-Bantu, but not Hottentot or Bushman. There are small enclaves of non-Bantu people on the Northern and North-eastern Congo. These latter languages are distantly connected with isolated families of speech in the Southern Bahr-al- Ghazal and the Egyptian Sudan, and may be classified with vagueness as " Sudanese," and distantly affiliated with the Nyam-Nyam and Mangbettu groups. From the west coast of Tanganyika the Bantu migrants are divisible into three main groups (Nos. 5, 6, and 7), which may have set out almost concurrently. The most northern (No. 5) pushed its way along the main stream of the Lualaba- Congo and crossed over from the Western Congo to the Caineroons coastline, where it fused in speech and culture with the descendants of the first great east-to-west Mubangi-Cameroons invasion. The second section of the Tanganyika Bantu (No. 6) conquered and colonized Southern Congoland the Luba and Lunda territories, Angola, and what came to be known as the kingdom of Kongo, expending the last of its force in Loango, to the north of the Congo mouth, where it encountered the southernmost range of Yoruba-Benin culture. The third and most southerly section of the West Tanganyika groups (No. 7) may have received some great reinforcement from the east or north, a reinforcement accompanied by the straight-backed, long-horned cattle and the fat-tailed sheep. This section originated the remarkable Herero people of Damaraland and the tribes of Northern and Central Zarnbezia. But the Herero section early specialized itself when it got to the west of the Upper Zambezi. It colonized Southern Angola (leaving here and there an unassimilated element) but developed its most marked characteristics of speech and culture in South-west Africa, where it was obliged for centuries to fight against the Hottentots, Bushmen, 1 and a pre-existing Negro type which had already accepted a Bushman form of speech. This was the Berg-Damara or Ova-Herero, a people that in physique are typically Bantu (generalized Negro with a suggestion here and there of the Forest Negro), but in speech are Bushman or Hottentot. It has been suggested by some that the Ova-Herero (called Dama by the Hottentots) may be the Ma-tama described by the Portuguese as existing in Southern Angola in the 16th century. Another great Bantu migration (No. 8) descended from the south of Tangan- yika and occupied Eastern Zambezia, Nyasaland, and Mozambique. These were the ancestors of the great Nyanja, Yao, and Makua groups, and of the Karafia peoples of Southern Zambezia. A considerable stream of tribes (No. 9), speaking primitive Bantu languages, made its way in successive waves down the east coast of Tanganyika, reached to the north coast of Lake Nyasa, and penetrated finally 1 Or the mysterious, little-known, Bushman-like tribes of the westernmost limits of the Zambezi basin : the Kankala, Kasekere, Ma-kwengo, and Ma-sarwa. SIR H. H. JOHNSTON. A Sin-vei/ <>f the Ethnography of Africa, etc. 393 along the IJufiji and Kuvuma valleys to the shores of the Indian Ocean, where it has left its traces in the Makonde people. From out of tribes of the 3rd group the Arabs of South Arabia, who had certainly settled sporadically along the east coast of Africa before the Christian era, built up the Swahili people and the Swahili language, which bids fair to become the Linyua Franca of all Bantu Africa. But the most mysterious of all the Bantu migrations (No. 10) is that which created firstly the Bechuana people, and secondly the Zulu-Kafir tribes. Although the Zulu-Kafir group is the most isolated of all divisions of the Bantu speech family, it possesses marked affinities with the Sechuana group and less marked relationships with the Ochi-Herero (No. 7). On the whole its relationships beyond those tongues might be described as East African and Central African. It would seem as though the ancestors of the Bechuana and the Zulu rushed down from East- central Africa in one of those great raids which they have repeated and made so familiar to us in the course of the 1 9th century, and passed over the heads (so to speak) of the earlier Nyanja and Karafia tribes of Zambezia and Nyasaland. They first settled in South-central Zambezia, and then penetrated down the Limpopo River into South- east Africa, passing up the coast as far north as Inyambane, and conflicting here with the pre-existing Karafia invasion, and with still earlier settlers who may not have been of Bantu speech. At all events a considerable body of these Bechuana-like people passed south through Swaziland across the Drakensberg Mountains into Zululand, Natal, and what is now the eastern part of Cape Colony. Shut in from the north by the great ranges of lofty mountains, and thus cut off from intercourse with their Bechuana relations, yet having somewhat more intercourse with the tribes of Nyanja-Karana stock on the north-east they developed a type of language which in a large proportion of its vocabulary is absolutely original and self-developed. At the same time they found in the lands along the coast of the southern Indian Ocean a considerable Bushman population. 1 They probably killed the Bush men and espoused the Bush women ; and the Bush mothers imparted to the children of their Bantu husbands three or four of their ugly clicks, 1 It has been questioned by some authorities whether there were ever any Bushmen in Zululand and Amatongaland, because no Bushman paintings have yet come to light in easternmost South Africa ; and yet there is an obvious strain of Bushman blood in the serf (not the high class) population of Zululand and South-east Africa. May it not, however, be a misleading idea on the part of some ethnologists that wherever cavern paintings are found in Africa and South-west Europe they are to be attributed to the presence of a former Bushman people in those regions, and conversely that where in South Africa no cavern paintings exist, in such regions there have been no Bushmen ? So far as I am aware, cavern paintings are only met with in the centre of southernmost Africa and in great Namakwaland, and some are associated with the pre- Bushman, Strandlooper race. No paintings or rock engravings have as yet been discovered due north of the Kalahari Desert, in South-west Africa, where, nevertheless, Bush tribes exist to this day. The Strandlooper may have introduced this imitative art into central South Africa and have taught it to the Bushmen. In all probability the ancient graphic arts of Northern Africa and Western Europe were the work of Palaeolithic proto-Caucasians or Neolithic Caucasians. Somewhat similar work was done by the Australoids of Australia and the vanished, blond Neolithic race of Western Siberia. 394 Sffi H. H. JOHNSTON. A Survey of the Ethnography of Africa, etc. disfigure the Zulu-Kafir speech to this day. The Bechuana peoples between the Orange Kiver and the Drakensberg on the south and the Limpopo and Ngami j basiu on the north, also seem to have been rather isolated for a time and thus to have developed marked linguistic peculiarities. According to their traditions their ancestors crossed the Zambezi from the north about the 14th to 15th centuries A.D. They did not hesitate to mingle with the pre-existing Bushmen, but were little Influenced in their speech by the click languages. This has only occurred in the southernmost part of their range Basutoland. Before passing to the consideration of the recent migrations of Caucasian peoples in Africa, we might touch on an invasion of Central Congoland which had a profound influence on Bantu Africa : I refer to the " Bushongo " ("or more correctly, Bashi-bushongo) civilization of Central Congoland made known to us and made famous by the splendid work of Mr. Emil Torday and his collaborator, Mr. Athol Joyce. On such evidence as Mr. Torday has placed before us, we may hazard the guess that the ancestors of the Bashi-bushongo (these words seem to mean in Bantu the "fathers" bashi "of the metal- spear-blade " bushongo) came from the region of the Shari Eiver about A.D. 600, say 1,300 years ago. They were a type of Sudanese Negro with a considerable infusion of Hamitic blood in their veins, and spoke a language apparently belonging to the vaguely-defined Central Sudanese group represented by some of the tongues still extant on the Shari Eiver and in Baghirmi. In fact, they may have derived some of their inherent culture from the relatively old civilization of Baghirmi. No doubt much Egyptian influence of an indirect kind passed from Lake Chad up the Shari and Logun Eivers, as it similarly reached the Upper Benue. The ancestors of the Bushongo according to their traditions would seem to have crossed the Mubangi Kiver and the main Congo, and to have made their way to the Sankuru and the eastern basin of the Kasai. Here they met with an aboriginal race of pygmies and with the Bantu invaders of Nos. 5 and G (No. 6 is that of the Luba-Tabwa tribes). The Bushongo finally abandoned their own Sudanese language and assisted to create a very corrupt form of Bantu speech (quite different from the beautiful Luba tongue) which they use at the present day. But they perfected in the very heart of Congoland the arts of working metals, plaiting and weaving, and developed, in conjunction with the Baluba (whose inspiration came from the highlands north of Tanganyika), a culture which is most remarkable in comparison with the savagery round the innermost circle of the Congo basin. In their magnificent physique the Bushongo aristocrats bear to this day traces of the Caucasian element which must have reached their far-back Negro ancestors from North-east Africa. Once well established in the heart of Congoland their influence, conjoined with that of the Baluba, radiated over their Bantu neighbours to the west and south, so that finally through the Awemba people it penetrated to the regions north-west of Lake Nyasa, or even farther south still to the Central Zambezi. 1 1 Mr. Emil Torday thinks that the Bushongo came only as warriors to take possession of a portion of civilized Lubaland, and that the Baluba brought from the east or north-east the arts SIR H. H. JOHNSTON. A Survey of the Ethnography of Africa, etc. 395 There has evidently been an old civilization in what may be termed the Butonga country of Barotseland, and perhaps also in Katanga and on the coasts and islands of the northern half of Tanganyika. In both these regions there are or were remarkable breeds of domestic fowl those of Butonga (the so-called " Makololo " fowls) being of small size, but of extraordinary prolificness in laying, while the fowls of North-west Tanganyika were a giant breed more like the Cochin China. In Butonga also there was a remarkable breed of pygmy cattle, only 3 feet high. The influence of the Bushongo-Luba led, no doubt, to the creation of great Bantu empires such as that of Lunda, which spread over the Luba countries towards Tanganyika and Lake Mweru. But the empire of this Mwata Yanvo only arose as far as I can ascertain about the 16th century. Its growth seems to have followed on one of those racial eruptions in Negroland which baffle explanation and seem incredible had we not witnessed similar ones in our own day in the northward raids of the Bechuana 'and Zulu peoples. Early in the 16th century a cannibal people of Bantu speech, but with little or none of the Bushongo or Luba civilization about them, suddenly boiled over, and devastated much of Western Congoland and Angola, of Central Zambezia, and of East Africa. These were the Jagga and the Ba-zimba of the Portuguese. Their descendants still a somewhat truculent folk are the Imbangala of the Upper Kwango, and the Ba-jok of Mr. Torday, the Va-kioko or Va-chibokwe of Livingstone, Cameron, and Swan. Their devastations on the north almost overwhelmed and exterminated the semi-civilized kingdom of the Congo, which had already come under Portuguese influence. They occupied much of Angola, and after swarming across the Zambezi-Congo water-parting, they surged down the valley of the Luangwa until they came to the northern bank of the Zambezi opposite Tete. Here they fought for years with the Portuguese, who eventually got the better of them, though they practically wrecked Portuguese Zambezia for half a century. But one body of these dreaded Ba-zimba 1 fought their way northwards through South Nyasaland and the Yao countries till they crossed the Euvuma and laid siege to the Portuguese-Arab town of Kilwa. This they finally took and sacked, and ate thousands of their prisoners. Then they continued their ruthless course till they similarly encamped on the mainland west of Mombasa. Only by desperate efforts did the Portuguese, allied with the local Negroid tribes of Gala affinities, defeat them here. Their defeat led no doubt to their practical annihilation, for they seem to have left no traces behind in this region. Let us retrace our steps in the prehistory which is founded on archaeological research and in the history which is derived from written records, and consider the of metal working, weaving, etc. There certainly was, many centuries ago, a remarkable Bantu civilization, emanating from the Ruanda and Tabwa countries (north and west of Tanganyika), and extending south to metalliferous Katanga. 1 It is said by French missionaries such as Arbousset that another section of the Ba-zimba wandered across Matebeleland to the Limpopo and became the ancestors of the Bavenda tribe, and introduced the custom of cannibalism amongst the south-eastern Bechuana peoples. The Chi-venda language, the most interesting and purely Bantu in South Africa, is, how ver, of East African affinities. 396 SIK H. H. JOHNSTON. A Survey of the Ethnography of Africa, etc. effect on Africa of the Semitic invasions and settlements. Here again, as in the term " Hamite," designation is of linguistic rather than racial import. Several distinct types of Mediterranean white man the Aramaean (Jew, Assyrian, Phoeni- cian), the Armenian, and the Arab ; the Kurd ; besides the negroid Elamite and the brown-skinned Hamite and Mongoloid Sumerian : have fused partially in the com- position of the white men of Semitic speech. These composite types of Semitic race more European in physique, for the most part, than the Kushite Hamites, less so than the true Berber have colonized Africa at intervals from about 6,000 to 7,000 years ago. At some such distant period as about 4500 B.C., Semitic settlements were apparently being formed in the Isthmus of Suez and the Nile Delta. About 4,000 years ago began the great Hyksos or Haqshu invasions of Egypt, which for more than 500 years paralysed the power of the Dynastic Egyptians. These " shepherd kings " may have been akin to the Arabs in their pastoral stage, though some Egyptologists think there was an admixture of Turk or Turkoman or even of the Medes (Skythian Aryan). Their use of the horse suggests a Central or West Asian influence, but their speech seems to have been Semitic. Their migration from Syria took place before the domestication of the Arabian camel, but they brought the horse with them as a draught animal. When Pharaonic dynasties had regained power and overcome these barbarians, some of the survivors of the Hyksos probably retreated into Midian and Arabia and fused with the Arabs ; others must have contributed to the complex elements of the residential Egyptian population. There is no need to suppose that any large body of them went westward to form the Fula or the Songhai peoples. At least there is not the slightest evidence as yet to support this conjecture. We may hazard the guess that an important Neolithic civilization had remained behind in South-west Arabia even after the departure of the Dynastic Egyptians ; that this peninsula became gradually Semitic in religion and language ; and that having sent one of its peoples the Phoenicians the "Pun," perhaps of the Egyptian Pun-t or Puanit from the Persian Gulf to the Syrian coast, and hordes of Himyarites to colonize Abyssinia, it still possessed sufficient energy in its west Arabian states of Saba and Mina to explore the east coast of Africa, perchance in boats with mat sails like the existing "mtepe" of Zanzibar. I see nothing inherently improbable in the finding of gold by proto-Arabs in the south-eastern part of Zambezia ; nor in the pre-Islamic, Arab origin of Zimbabwe. In any case we know from the Periplus of the Erythrasan Sea a compilation dating back to the 1st century of the Christian era that at that time the Arabs of the south-west corner of Arabia traded with the Zanzibar coast. Probably by this time, or soon afterwards, their ships had found the north end and east coast of Madagascar. Here they introduced, before the Islamic period, the use of their Himyaritic dialect, and thither they conveyed (from Bantu Africa opposite) the long-horned cattle and the domestic fowl. From various indications one is entitled to believe that the East African coast was brought by Arab sea-trade into touch with India, and even China and Malaysia, as much as 2,000 years ago, if not SIR H. H. JOHNSTON. A Survey of the Etknmj niphy of Africa, etc. 397 earlier. I think Arab ships introduced the cultivated banana 1 (whose nearest wild relation grows in South-eastern Asia, not in Africa) into East Africa, whence it spread from tribe to tribe right across Equatorial Africa to the West Coast. They also introduced hemp from Asia ; possibly also the cotton plant and the sugar-cane, rice and wheat. The Arabs may also have brought the domestic fowl from Persia or India to East Africa and Madagascar. They conveyed cattle of the long-horned ancient Egyptian type into Madagascar, and certainly brought thither, early in the Christian era, numbers of Negro slaves from the Zangian and Motjambique coasts. The " Carthaginians " of course originated in the Phoenician settlements of Sidon and Tyre on the coast of Syria. They established themselves on the coasts of Tunis and northern Morocco about 1100 B.C. But the real growth of their power and colonization only dates (it may be) from the foundation of Carthage in 822 B.C. A good deal of the coastlands of Tunis and western Tripoli was actually colonized by the Phoenicians, who left many monuments behind which testify to their former presence there. They undoubtedly opened up trading stations along the Atlantic coast of Morocco, and in one or more voyages penetrated to the mouth of the Senegal, and even to the southern limits of Sierra Leone. 2 They seem to have got in this way into relations with the Fula and Wolof peoples, and to have used Fula interpreters 3 to communicate with the shy, black Negroes. Their farthest-south permanent trading station on the west coast was the little island of Kerne (still called Herne) in the narrow gulf of the Rio de Oro. This region has for some years been in the possession of the Spaniards, and it is surprising that no excavation or research work has been made in Herne Island. The archaeological results might be of the highest interest. No doubt, even as far back as 1000 B.C., a certain amount of overland trade went on between North Africa and the Western Sudan, but such trade probably only followed the Atlantic coastline. Several of the more important oases of the Sahara seem to have become uninhabited. In fact, the Neolithic population of the Sahara would appear to have died out, no doubt owing to the increasing drought. But there is nothing in the culture or the languages of West Africa to suggest any great amount of intercourse between the Carthaginians and Negroland, after Hanno's celebrated voyage to Sierra Leone in the 6th century B.C. Things became a little more active when the Romans displaced the Carthaginians in North Africa. In the early part of the Christian era they the Romans had crossed the Atlas Mountains of Morocco, and had sent expeditions far down into the Sahara Desert from Tripoli, probably to Bilma. They were 1 This may even have been brought by Malay canoes across the Indian Ocean to Madagascar and the Comoro Islands. The earlier Arab writers on East Africa between the 10th and 13th centuries mention both Larnu on the North Zangian Coast and the Comoro Islands as special centres of banana cultivation. 2 Mr. Frederick Migeod, in his Languages of West Africa, suggests from certain evidence that the Carthaginians were in touch by sea with the Gold Coast. 3 A deduction made by the French colonial administrator and historian, M. Louis Binger, on fairly good grounds. 398 SIR H. H. JOHNSTON. A Survey of the Ethnography of Africa, etc. assisted in this last direction by the Garamantes, who were perhaps Tibus. But it would seem as though the Eoman beads which turn up all over West Africa and the Sudan had entered Negroland through the Egyptian trade across Nubia, Darf ur, and the Lake Chad region. The Jews both before and after the opening of the Christian era took up the work of their kindred, the Phoenicians. They established themselves in such numbers in Cyrenaica that they practically ousted the Greeks ; indeed, in a Jewish uprising of A.D. 117 the Greeks of this old Greek colony (now Italian territory) were almost exterminated. 1 The Romans retaliated by killing the Jews, but the Jewish settlements at Carthage and elsewhere in Tunisia and in Morocco prospered mightily under Roman rule ; or rather outside the limits of Roman rule amongst the Berber princes, many of whom became converts to the Jewish faith. The Jews played a great part in the history and development of North Africa in those days. They instigated the Berbers to oppose the Arab invasion and to assert their independence against the Vandals. The Arabs and Berber historians of North Africa even state that it was the Jews who, in their trading enterprise, first occupied or re-occupied important oases in the Northern Sahara such as Twat. They may possibly have found still living at Twat a Negro or a Negroid race, and unquestionably they were not the first inhabitants of any Sahara oases. But they seem in many directions to have preceded the Berbers in their south-west migra- tions. Some of these Jews are thought even to have crossed the Sahara, carrying with them a garbled blending of the Jewish and Christian faiths, and to have reached the country of Borgu to the south of the Central Niger, establishing there a form of religious faith which has only recently given way to Islam. Jews, how- ever, early in the Christian era (according to traditions gathered up by Arab historians) made their appearance in the Mandingo kingdoms of Ghana ta and Melle and at the capitals of the Songhai empire which followed. On the eastern side of Africa, Jews or Idumoaans had evidently settled in Abyssinia several centuries before the Christian era, and many of the Jews of Arabia migrated to Abyssinia after the establishment of Islam. The other Arab invasions of Africa before the Christian era were restricted to occasional settlement on the west coast of the Red Sea and in Somaliland, and to the trading colonies on the Zangian coast already described. But after the first Islamic convulsions in Arabia, Arabs began to pour into Lower Egypt from the direction of Syria. Thence they passed at the head of conquering armies through the Tripolitaine into Roman Africa and Morocco. Allied with the Berbers they descended in thin streams the Atlantic coast of the Sahara till they reached the Senegal River in the early part of the 10th century. Other bands of them found 1 There would seem to have been a Pelasgian, Minoan, or early Greek colonization of Eastern Tunisia, and the Greeks, as we know historically, founded a prosperous colony in Cyrenaica as early as 631 B.C. It is possible that through the intervening Berbers and Tibus, Greek influence to a slight degree penetrated (in objects of trade) across the Sahara Desert into the Sudan. SIR H. H. JOHNSTON. A Survey of the Ethnography of Africa, etc. 399 their way from Arab Egypt along the old trade route to Darfur, and commenced the Islamizing of the Central Sudan, from Bornu to the Niger. But the great invasion of Northern Africa and the Sudan by Islamic Arabs did not take place till the llth century that invasion which is named the Hilalian by French writers. This was the result of the driving out from Arabia of two turbulent tribes, the Beni Soleim and the Beni Hilal, who crossed the Red Sea, forced their way through the Hamitic peoples of the Nubian Desert, and settled on the Nile in the Cataract region. From this direction many of them made their way in succeeding centuries across Darfur to Wadai, Bornu, and Baghirmi, where they are represented at the present day by the Shawia. Others, again, mingled with Hamitic and Negro elements and founded the powerful Funj dynasty of Senaar. A large proportion, however, of this Hilalian invasion in the llth century was directed by the decadent power of Arabized Egypt (under the Fatimite puppet- Caliphs) towards Mauretania. Northern Africa, when it ceased to be Graeco-Ronian, was becoming practically a series of Berber principalities which retained Arabic as their official language and remained Muhammadan in religion, but were inclined to enter into friendly relations with the European States across the Mediterranean. But the Arab character of North Africa and the Muhammadan fanaticism were sharpened by this Hilalian invasion, which planted in Tunis, Algeria, Morocco, and the Western Sahara hundreds of thousands of Arabs, who have retained to this day their racial characteristics, except where they have mingled with the Negroes. Thus there are Arabic-speaking tribes along the northern Senegal and within the Western Sahara. The Islamic Arabs had resumed intercourse with the Somali and Zangian coasts as early as the 8th century A.D. According to their historians, ships from India or Malaysia were trading soon afterwards with the natives of (what we now call) the Angoshe district of MoQambique for iron ; and making use of the Comoro Islands as trading depots. By the llth century of this era the Arabs were established at Kilwa and as far south as Sofala. Soon afterwards they formed a settlement at Sena on the Lower Zambezi, and their influence and culture probably went far towards the creation of the powerful Negro empire of the Monomotapa. The arrival of the Portuguese, however, arrested for a time the Arabizing of East Africa, though the Arabs had already got into touch with Unyamwezi and Lake Nyasa. They were reinforced at different periods after the llth century by colonies of Persians from Shiraz, and these Persians introduced much culture into the East Africa coastlands between Somaliland and Kilwa. Although the Islamized Berbers of the Sahara Desert the so-called Tuaregs (this nickname is itself of Arab origin, from Tarqi, a raider) had after the first Arab invasions of Mauretania pushed farther and farther into the Sahara Desert (often, in the oases, preceded by Jews), and had in the 12th century founded or refounded Timbuktu and given a Caucasian reinforcement to most of the Negroid tribes and dynasties in the Central and Western Sudan, their civilizing work was hardly to be distinguished from that of the Arabs who accompanied them, or who 400 SIR H. H. JOHNSTON. A Survey of the Ethnography of Africa, etc. supplemented their efforts in introducing everywhere throughout the Sudan (north of the forest belt) the religion, the language and culture of Arabia. On the east coast of Africa a great revival of Arab emigration took place after the decline of the Portuguese power in the 17th century. The Arab cause was taken up by the bold seafaring traders of Oman in South-east Arabia. The Portuguese were driven from all their positions north of the Ruvuma Ptiver by the 18th century. In the 19th century, from the Arab metropolis of Zanzibar, East Africa, as far inland as Tanganyika and Nyasa, was considerably Arabized. Arab adventurers reached even to the extreme Upper Zambezi, concurrently with Livingstone, besides discovering the Upper Congo. Besides this, a pre-Portuguese element of Arab settlement continued to exist, and exists still, in the southernmost part of the Mozambique region, chiefly the district of Angoshe. Arabs from Syria and North Africa, together with large contingents of Berbers from Algeria, Morocco, and the Northern Niger, had colonized Spain between the 8th and the llth centuries. Here they had mingled with Jewish and Christian elements, and the amalgam of all these strains produced a magnificent human type of great physical beauty and a tendency towards whiteness of skin ; a most talented people, artistic to their finger-tips, passionately fond of music, clever at mechanics, industrious in agriculture. Chased out of Spain because of their imperfect acceptance of Christianity and their inherent dislike of Castilian rule, these Pamias as they were styled by the people of North Africa or " Andaluzi " became to Muhammadan Africa what the French Huguenots were to England, Ireland, and Holland. They brought into Africa the bravery, the ideas of discipline, and the cultivated mind and artistic instincts of the European. They became the sharpshooters, the artillerymen of the Moorish Empire, and were the main agents in carrying the rule of Morocco to Western Nigeria. As a civilizing element on the Upper Niger, and even as far south as Senegambia, they cannot be overrated. They also figured considerably in the founding of the Fula kingdoms, and the eventual Fula empire, even as far east as Bornu, and as far south as Baghirmi. A colony of them also settled at or near Lamu on the equatorial East coast of Africa, and their descendants assisted the Maskat Arabs to expel the Portuguese. Amongst the many race movements due to the influence of the Muhammadan religion were those of the Hausa and the Fula. The Hausa-speaking Negroes of the Central Sudan, by their industry and intelligence (helped, however, largely by the unconscious gospel of art and agriculture, preached by the dark-skinned Caucasian hybrids who have permeated this region for about two thousand years, and perhaps long before), had founded important states between the Benue basin and the Sahara Desert; and their trading language and influence by the 16th century had crossed the Niger and travelled westward over the semi- civilized kingdoms of the Mossi group of Negroes, until they were stayed from any further western progress by the equally potent Mandingos. The Hausification, so to speak, of all these regions of the Central Sudan prepared the way for the establishment of the three remarkable Fula SIR H. H. JOHNSTON. A Survey of the Ethnography of Africa, etc. 401 empires over West, Central, and Eastern Nigeria which have only l>een replaced of late by French or British rule. The ancient history of the Fulas has already been alluded to. Their original home, according to Arab-Berber tradition, was in the western oases of the Sahara, the valley of the Draa, the hill country of Adrar, and the Atlantic coast of the Sahara Desert. They may have been the " Pharusii " of Roman geographers. The Berber emigrations which resulted from the wars of conquest of the Romans, followed by the convulsions due to the invasion of Mauretania by the Islamic Arabs, are believed to have driven the Fulas south towards Senegambia, their place being taken by Berbers and Arabs. They were at first quiet people, herdsmen and shepherds, with a high and intricate type of pagan religion which still survives in parts of Nigeria. But large numbers of them became converted to Islam from the 12th century onwards and gained some knowledge of the world outside Africa by their pilgrimages to Mekka. In Senegambia they came gradually to form proud little principalities of Muhammadan people ; which, often assisted by Arab and Moorish adventurers or fakirs, finally dominated the basin of the Upper Senegal at the close of the 18th century. Their colonies and wandering gypsy tribes of cattle- keepers, however, had long before this penetrated to Borgu (at the back of Yoruba), and to most parts of the Sudan within the Niger basin. They had even penetrated as far to the east as the Shari River and Darfur. At the end of the 18th and the beginning of the 19th centuries an uprise of Muhammadan fanaticism and a proud consciousness of their racial superiority to the mere Negro (a principle they had learnt by their pilgrimages to Mekka) armed them as an aristocracy to wrest the political control of all Nigeria (between the Futa Jallon highlands and the Upper Benue) from the hands of Negro rulers or the decaying power of the Tuareg and Songhai. The Fula conquests effected little or nothing against the eastern Mandingos or the powerful and warlike Mossi group of Muhammadan Negro states ; but made them completely dominant in the basin of the Middle Niger and the Upper Benue. The pure-bred Fulas were at least semi-Caucasian in physique and brain ; and this race was all unconsciously carrying on the Caucasian invasion and penetration of 'Africa. 1 1 Since the discussion on my paper (verbal, and in Mr. Morel's paper, the African Mail) I will sum up my views on the Fula as follows : Very little material is still available for the definition of the precise physical type of their pui'est breeds. Those I have seen myself in Portuguese Guinea, the Gambia and Sierra Leone, and in French West Africa struck me forcibly with their likeness to very good-looking Arabs, the Arabs of Western Arabia. Their complexions were no darker and their hair was in long ringlets. Though betraying the same ancient " negrification " that is to be seen in so much of the North African population and in that of Egypt, they were emphatically a three-quarter Caucasian race, the one undoubted touch of the Negro being the kinky hair. It is strange that so handsome and remarkable a race should be almost unrepresented in the books and museums of Europe by any good series of photographs. According to the best authorities of the 19th century the purest-blooded, most Caucasian-looking Fulas are or were to be found in what is now French West Africa, between the Upper Senegal and Upper Gambia and the " lacustrine " Niger (Jenn6 also in Borgu 402 SIR H. H. JOHNSTON. A Survey of the Ethnography of Africa, etc. There remain five other noteworthy racial migrations in Africa and its great adjacent island of Madagascar which should be described before we close this survey at the last phase : the recent invasion of the dark continent by the white peoples of Europe. These five would be (1) the Malay colonization of Madagascar ; (2) the southern and south-eastern migrations of the Nilotic Negroes ; (3) the northward movement of the Hottentots in South-west Africa ; (4) the northward raids of the Zulus and Bechuana in South-central and East Africa, and (5) the descent of the Fang tribes on the coast lands of the Gaboon and South Cameroons. (1) The colonization of Madagascar by a race or races which came thither in prehistoric times right across some 3,500 miles of open Indian Ocean from Sumatra or Java, in apparently nothing better than outrigger canoes, must remain one of the most astounding adventures in the history of man. Yet there are the Malagasy tribes of Madagascar at the present day, speaking one single language which has no deep-seated differences between its dialects, and which bears the unmistakable signs of its kinship with the Malayo-Polynesian tongues. It has been evolved much earlier (between Gao and Yoruba) ; and among the semi-nomad herdsmen or " Cow" Fulani of British Nigeria from Hausaland to Lake Chad. According to Denham, in 1825 there were communities of very pure-blood pagan Fulas in the Mandara country, south of Bornu. Most of the conquering and ruling Fulani of the 19th century, more especially in British Nigeria, are so mixed with Hausa blood that they are more Negro in physique than anything else. Although the religious beliefs of the pagan Fula are described as of a lofty mysticism, very little exact information has been published about them. The Western Fulas became Tslamized comparatively early in the history of the Arabization of West Africa say in the 10th to 14th centuries of this era. The West African Fulas have always shown a great sympathy for and interest in the Arab, and have furnished a large proportion of the " Takruri " pilgrims to Mekka. On the other hand, the original eastern Fulani of Bornu and Baghirmi (apart from the Fula conquerors who came from the Northern Niger under Sheikh Omar in the early 19th century) have shown themselves hostile to Islam and.to the Arab. The trend of migration at any rate during the Christian era has been from west to east, if one may go by the evidence of the existing Fula dialects, that of the regions between the Senegal and Niger being apparently the mother-speech. The physical type of the pure-bred Fula would seem to be definable as follows : Tall of stature (but not gigantic, like the Nilote and South-east Sudanese), olive-skinned or even a pale yellow ; well-proportioned, with delicate hands and feet, without steatopygy, with long, oval face, big nose (in men), straight nose in women (nose finely cut, like that of the Caucasian), eyes large and " melting," with an Egyptian look about them, head-hair long, black, kinky or ringlety, never quite straight, hair of body almost absent except at pubes and axillae (that is to say, the body is not hairy as in most Berbers and in the Forest Negroes), but a fairly abundant beard and moustache in the men. No Negro skin odour (in the pure-blood Fula). Their original culture was mainly pastoral. They seem to have originated no architecture or industries. Their domestic animals and cultivated plants were just those of the Northern Sudan and of indirect Egyptian origin. They did not have the plough or the Libyan chow- like dog and ignored the use of stone for building. Their language most highly developed betrays absolutely no structural affinity with the Semitic or Libyo-Hamitic groups, or with any other speech families outside Africa, except in the principle of " class " distinctions (which it shares with many Negro speech groups). The Semitic and Berber vocabulary elements in Fula are all borrowed words of comparatively modern use, due to the introduction of Muhammadanism. Such affinities as Fula may have with other speech families are confined to the class languages of West and Central Africa, including Bantu. SIR H. H. JOHNSTON. A Survey of the Ethnography of Africa, etc. 403 than the actual Malay language, which is a comparatively recent birth of time ; but the general affinities of the Malagasy speech with the Malay o-Polynesian group are undoubted. In physique the people of Madagascar vary from the almost Negro Sakalavas on the west and the Betsileo, Tanala, and Ibara on the south, to the straight-haired, short-statured Javanese-like Hovas in the centre and east. Along the north coast there is clear evidence of ancient intermixture with Arabs and Hindus, while, of course, there is much Negro blood in the west. On the whole, however, the people of Madagascar bear a remarkable resemblance to the different types of Mongoloid and Melanesian inhabitants of the Malay Archipelago and Papuasia. We may surmise that as soon as the enterprising Indonesian immigrants into Malaysia, three to four thousand years ago (or more), began to affect the Mon- golian-Melanesian-Negrito savages with their Caucasian energy and daring spirit of adventure, Malay voyages in outrigger canoes with mat sails (like those of Polynesia) may have extended across the Bay of Bengal from Sumatra to Ceylon and Southern India, and later to the Maldiv and Chagos Archipelagoes. But there is as yet no historical or archaeological evidence of such Malay voyages. There are indications of an ancient human habitation of the Sechelles Archipelago, but not to any great extent. These equatorial islets were without inhabitants or traces of inhabitants when first discovered by the Portuguese and French. There is absolutely no evidence of any ancient human settlement on Rodriguez, Mauritius or Bourbon (the Mascarenes), or on the Aldabra group, or even the Komoro Islands. This last string of volcanic islands which lies between the Zangian coast and Madagascar and which may be the submerged remains of a great peninsula that nearly connected Africa with Madagascar in the Miocene and Pliocene nearly, but not quite does not seem to have received a Malagasy population but to have been uninhabited until it was discovered by Arab ships about 2,000 years ago or more, discovered and partially settled by Arabs and by the Negro slaves they brought thither from the mainland. These slaves must have been of fairly recent importation perhaps since the Islamic period began, for the dialect they speak is nearly related to the Swahili of Zanzibar, though to an archaic form of Swahili. The Malay colonization of Madagascar therefore, though very thorough and several times repeated, was singularly limited in its scope, since the Malayo- Polynesian adventurers apparently never reached the Mascarene Islands or the Komoros, 1 or Zanzibar or the Zangian or Mozambique coasts. Once they had landed on Madagascar they appear to have abandoned sea-faring habits ; and though they retained the outrigger canoes they never ventured in them far from shore. The outrigger canoes were afterwards introduced by the Arabs into the Komoros and to the Island of Zanzibar, but were never spontaneously adopted by the Bantu 1 There is a good deal of Malagasy blood in the Komoros to-day, but it is due to compara- tively recent immigration and the slave trade of the 18th and 19th centuries. 404 SIR H. H. JOHNSTOX. A Survey of the Ethnography of Africa, etc. Negroes of the East African coast. According to tradition there have been repeated Malay invasions of Madagascar, the earliest perhaps about 3,000 years ago : the latest that of the Hovas in (some suggest) the 7th century of the present era. 1 The Hovas instinctively fought their way through the antecedent peoples of the east coast-belt till they reached the healthy plateaux of the interior. Here in the course of some hundreds of years they became a warlike, vigorous nation which in the early 18th century set out to conquer all Madagascar and would have succeeded, but for French interference. The Hovas have kept their Malay racial characteristics much more than the Betsileo, Sakalavas, and other tribes of earlier arrival. The Sakalavas have evidently mixed much with the Bantu Negroes of the East African coast. A number of words of Bantu origin have been imported into the Malagasy dialects, but more into those of the west coast than into the speech of the Hovas. Still the dialectal differences in Malagasy are slight and point to a certain racial homogeneity. There is, per contra, not the slightest trace of Malagasy influence or blood among the peoples of East Africa. There is a tradition in East and South Madagascar that the earliest Malay immigrants found a preceding Negro or Negroid race still existing in the land. These are generally called the " Va- zimba," and in many districts mounds of seemingly artificial origin are called " Vazimba graves." But although both French and English explorers have made such remarkable discoveries of a fossil and sub-fossil fauna in Madagascar, they have either never troubled to investigate these alleged sepulture places of an ancient and vanished race of Negro type which may have preceded the Malagasy ; or they have found nothing to support the theory of the existence of a pre-Malagasy population. The word " Vazimba " may have been introduced by the Arabs at the time the cannibal Ba-zimba hordes were ravaging the opposite African coast. But it is difficult to believe that the actual, historical Ba-zimba, who had no canoes with which to ford the Zambezi at first, and were thwarted in their attack on Mombasa because it was an islet separated from the mainland by a narrow tidal creek, could have found the means of crossing the 300-miles-wide Mozambique Channel and invading Madagascar at the time when the Portuguese held command of these seas. Considering, moreover, that neither the Bantu Negroes of East Africa nor the Bushmen who preceded them were possessors of any better means of navigation than dug-out canoes or rafts, it is difficult to believe in an ancient Negro peopling of Madagascar before the Arabs could serve as transport agents. Of course it is just possible that the mysterious race that built the Zimbabwe 2 walls and temples and commenced mining for gold in South-east 1 In the opinion of the late Professor A. H. Keane, who has given an excellent resume of the question in his Man: Past and Present (1899), the Hova departure from Java or Sumatra must have been much earlier than only twelve hundred years ago. The reader is also referred to Madagascar, etc., by E. F. Gautier, Paris, 1902. 2 There is seemingly no etymological connection between -zimba, a Bantu tribal name-root applied amongst others to the cannibal raiders from Southern Congoland in the 16th century), and Zim-babwe, which simply means " stones " in the Karafia dialects, from -babwe, -bawe, a stone, and Zin-, a plural prefix. SIR H. H. JOHNSTON. A Survey of the Ethnography of Africa, etc. 405 Africa may also have invaded Madagascar. But in such case they are more likely to have been pre-Islamic Arabs or Karaites rather than Negroes. It has been stated by Capitaine M. Guillain, Sir Kichard Burton, and other writers on East Africa in the middle 19th century that Malay sailing vessels have been blown by strong easterly winds from Java to the East African coasts once or twice within the last hundred years. But the thorough and repeated colonization of Madagascar by a Malay o-Polynesiau race was scarcely a matter of mere accident, and argues a certain design, though a bold one, at the period of its inception. Another enigma in the history of East Africa is : If the Malay -Polynesians could colonize Madagascar deliberately or even accidentally, why did they not reach East Africa in similar numbers and leave unmistakable marks of their former presence there in the native races, the languages, manners, and customs ? There is no greater puzzle in the history of human migrations. (2) The Nilotic Negroes early came under the influence of that primitive or modified form of white man, the Hamite. This influence is to be seen in not only the physique of some of the Nilotic types, but in several of their language- groups in which there exist numerals obviously borrowed from Gala or Somali, and in which there are other word-roots of apparently ancient incorporation. Whether the discrimination of sex, masculine and feminine, is directly derived from the impact and influence of the Hamite, is open to question. Most, if not all, of the Nilotic tongues, and the mysterious and isolated Bongo of the Bahr-al-Ghazal, distinguish in their pronoun-particles between masculine and feminine, " he " and " she." But this may be only a further development of a principle already at work in the Bantu tongues, in which there has been a tendency from remote times towards the institution of a feminine particle or prefix (without a concord) derived from a word meaning " mother " (na, nya), or to identify the diminutive prefix Jca (which possesses a concord) with a feminine sense, and the root-word for "woman," -ka. At the same time a male prefix, se, sa, or so, exists in some of the Bantu tongues and is related to a root-word for " father." It is, at any rate, noteworthy that the feminine particle or prefix in the Nilotic tongues, which is usually na and which consequently offers some resemblance to the Bantu " mother " prefix, is unlike anything associated with the feminine gender in the Hamitic or Semitic groups, wherein the feminine letter or particle (prefix and suffix) is t or ti in its pristine form. But about the 17th century of the Christian era (as nearly as we can calcu- late by deductions from native traditions and from the prevalence of Bantu place- names) some distance to the north of the present Bantu frontier there began a decided movement southwards up the valley of the Mountain Nile on the part of the Nilotic Negroes, such as are represented at the present day by the Shiluk, Bari, and Lotuka. The Shiluk 1 type sent considerable colonies southwards, overrunning Sudanese 1 This word is an Arabic corruption of a triba name which was probably something like Choli originally, but it is a convenient term for the inclusion of a special group of the Nilotic peoples of which the existing Shiluk are representative. c 406 SIR H. H. JOHNSTON. A Survey of the Ethnography of Africa, etc. tribes like the Madi (of West-central African affinities). Thus there were founded settlements of the Aluru group on the shores of the Albert Nyanza, and of the Acholi and Gang tribes between Karamojo and the Victoria Nile, and the much farther-afield colony of the Ja-luo in the Kavirondo country east of the Victoria Nyanza. In Kavirondo, Bukedi, and Bunyoro, and possibly even the southern part of the Madi country, the Nilotic tribes were certainly preceded by the Bantu, judging from the prevalence of a Bantu type of place-name, especially those with the prefix pa-. Earlier still, perhaps, the Nandi group of Nilotic peoples had migrated southwards from the west coast of Lake Rudolf (perhaps expelled by the Turkana) and had colonized the plateaus and mountains of Eastern Equatorial Africa from about 3 N. latitude to 6 S. latitude. They have imparted their present characteristics in language and customs to the hunting tribe of the Andorobo, which now ranges nearly to the equatorial coast of the Indian Ocean and far south into German East Africa. Such of the broken tribes of northern German East Africa near the Victoria Nyanza as are not a kind of bastard Hamitic (Es-segeju, etc.), or are unclassed like the Mbugu, Mbulungu, or Sandawi, belong to this Nandi group in language. Simultaneously with these convulsions (we may guess), a third group of Nilotic Negroes, the Bari-Lotuka-Turkana, sent out its pastoral herdsmen and its cattle-loving giants. These in their southward march have left the Teso or Elgumi people near Mount Elgon, the Turkana and allied tribes on the west coast of Lake Eudolf, and most noticeably the Masai of British and German East Africa. The Masai became noteworthy as migrants and raiders, and touched the fringe of recorded history no earlier than the first half of the 19th century. At that time, when Krapf and Rebmann and other German missionaries were penetrating from the coast towards the great Snow Mountains, they were made aware of the Masai or Oigob raiding peoples, who were pressing the Bantu in all directions, driving them towards the coast or up into the mountains. They noted that there were two types of Masai, though it might be added, there was practically no linguistic difference between the two. The earlier arrivals were known by the Swahili name of Kuavi or Kuafi, and apparently described themselves as Oigob. They were agricultural as well as cattle-loving, and became much more placable than the prouder Masai, who pressed on them from the west and virtually destroyed them in course of time. Before European influence interposed a check, the Masai in their southward raids had reached as far as 7 S. latitude. (3) In the south-west of Africa far-reaching convulsions were caused by a relatively small band of invading Hottentots. The Dutch settlers in Cape Colony had imparted some knowledge of Christianity and elements of material civilization to their Hottentot serfs and servants, whom they did not always treat very well. One family of three brothers, which had assumed the Dutch nickname of Afrikaner, revolted against the cruel treatment of their master or employer, shot him, and escaped north of the Orange River with all the plunder and arms on which they could lay hands. These Afrikaners were daring men of reckless courage, ; SIR H. H. JOHNSTON. A Survey of the Ethnography of Africa, etc. 407 strong swimmers, ferocious fighters, and ruthless slayers of big game. North of the Orange River they soon collected a band of adherents, mostly Hottentots and Bushmen, but also a few nondescript Negroes, escaped Bantu slaves from Cape Colony. Their numbers increased by birth and immigration ; and at last in the sixties of the 19th century they had become the dominant power in what is now German South-west Africa, and carried on a bloody warfare with the Bantu- speaking Ova-Herero. The Hottentot race is possibly composed of an ancient Negro ? proto-Bantu or even semi-Caucasian (Hamitic) invading element in South-western Africa, mixed with the blood of the indigenous Bushman population. The Bushman element now greatly predominates in the Hottentot physique. When South Africa was first dis- covered by the Portuguese and the Dutch, there seem to have been two very distinct types of Hottentot, one much more civilized than the other. The more savage Hot- tentots were simply larger and stronger Bushmen, using weapons superior to those of the Bushman, but not possessing domestic cattle or sheep. The other tribes found in the more central part of southernmost Africa eastern Cape Colony were much more pleasing in appearance, and not only possessed domestic cattle of an ancient Egyptian breed, but had already trained them as riding animals. Hot- tentots of this type, especially amongst the women, are described by the European pioneers as being good-looking or handsome, the women especially so. Not only in language 1 but in culture and in bodily characteristics, there would seem to have been some slight element of the Hamitic Caucasian in the Hottentot, and this, considerably reinforced by the constant intermarriages between Dutch settlers and Hottentot women, made of the Hottentots in the 19th century a conquering race wherever they came into contact with pure-blood Negroes or Bushmen. It was, of course, mainly the possession of firearms and the knowledge how to use them that gave to the Namakwa Hottentots, under the leadership of that remarkable Afrikaner clan, the position of a conquering race throughout South- west Africa during the second half of the 19th century. The Afrikaner clan and its adherents penetrated northwards during the last hundred years till they actually menaced the Portuguese settlements in Southern Angola. But for the intervention of the Germans they would probably have exterminated or enslaved the tall Bantu Herero and Ovambo. On the other hand, in the more central parts of South Africa, the bastard Hottentots or " Grikwa," working in closer sympathy almost alliance with the missionaries and the British authorities, opened up the way to Bechuanaland and the Western Transvaal. In 1884 Germany intervened in the affairs of South-west Africa (because of the maltreatment of German missionaries by the Namakwa 1 The existence of the click language Sandawi south-east of the Victoria Nyanza which has slight root affinities with Hottentot, yet is spoken by a degraded Hamitic race, is an instance of how the Hottentots of South-west Africa may have arisen. The Hamitic affinities of the Hottentot dialects are confined to sex distinction and sex-indicating particles, a somewhat parallel instance to Hausa. There is apparently no sex discrimination in Sandawi. C 2 408 SIR H. H. JOHNSTON. A Survey of the Ethnography of Africa, etc. Hottentots). Subsequently, between 1890 and 1906 it cost the Germans several millions sterling and the lives of several thousand German soldiers, before these thirty or forty thousand Hottentots and Hottentotized Negroes could be subdued, mainly by extermination. (4) The Zulu conquests of South-east Africa began about the same period, the opening years of the 19th century. Dingiswayo, a Zulu refugee returning after years of exile in Portuguese East Africa, secured the chieftainship of a small clan in Zululand proper, and laid the foundations of the Zulu monarchy. 1 He was succeeded by Chaka, who became a mighty conqueror, but whose conquests were largely futile owing to his wild-beast love of slaughter. But he was the indirect means of sending two of his great captains on subsidiary careers of conquest Mosilikatse on the north-west, and Sochangane on the north-east. In addition to this, other Zulu bands, known generally as the Bangoni, wandered northwards on their own account, raiding and ravaging, until at last from sheer weariness of travel they settled down as overlords amongst submissive populations. Mosilikatse carried the Zulu language and the Zulu blood across the Transvaal to what is now known as Matebeleland and the western Karafia or Mashona country. His rule, however, and his tribal influence stopped short at the Zambezi. His efforts were countered in that direction by the equally warlike Makololo, the Basuto followers of Sebituane. The general upset caused by these Zulu convulsions in South-east Africa had sent streaming northwards hundreds of thousands of Bechuana, 2 first as refugees and then, as they grew bold, as raiders and con- querors. A hundred years ago or more one band of these the Ba-hurutse or Bafurutse had turned northwards and partly settled the Ngami Lake region and partly the banks of the Upper Zambezi, where they left behind them the tribal name " Barotse." But Sebituane's followers, the Makololo, belonged rather more to the Basuto or southernmost Bechuana stock. They had conquered nearly all Upper Zambezia by the middle of the 19th century, and although the indigenous tribes and dynasties returned to power after 1864, the ugly Makololo dialect of Sechuana remains (unfortunately) the official language of Northern Zambezia at the present day, where it does not give way to the still uglier " Kitchen Kafir." The descendants and slaves of Mosilikatse's Zulus have settled down as the 1 Zulu power arose from one of the easternmost of the many Kafir clans of South Africa a tall, good-looking Bantu people, with a suspicion of a Caucasian strain in their blood, derived, it may be, from those unknown Caucasian builders of Zimbabwe, or even from that remote Egyptian strain which is so prominent to the eye in the more aristocratic Bantu tribes of Central and Eastern Africa. The Zulus were armed for these far-reaching conquests by the ideas of discipline and organization which they obtained from the example of the white man in the late 18th and early 19th centuries. 2 The Bechuana, anciently allied to the Zulu-Kafir group, had reached Central Zambezia about the 14th to 15th centuries. Some connect them in origin with the Ba-zimba cannibals of South Congoland. They forced their way in the 16th to 17th centuries through the Karafia people of Monomotapa and reached the Transvaal, which they colonized, after expelling or absorbing the Bushmen. SIR H. H. JOHNSTON. A Survey of the Ethnography of Africa, etc. 409 " Matebele," a name given to them by the Bechuana. The followers of Sochangane mostly became known as the Aba-gaza, and founded, so to speak, the Zulu State of Gazaland in South-east Africa. They were first made famous by their ravages on the Portuguese Zambezi settlements in the early 19th century. In this direction they were called the " Landins," no doubt a corruption of the Zulu proper name, Umlandine. But the Landins may also have belonged to the bands ordinarily termed Abangoni. The Abangoni under various leaders crossed the Zambezi in the twenties and thirties of the 19th century and founded powerful kingdoms in West and South-west Nyasaland. Others of them passed to the east side of the lake or round its northern extremity, and became the dreaded Ma-gwangara of German East Africa, also known as the Maviti, Mazitu, Machonde, or Manindi. Bands of these Abangoni founded powerful kingdoms in South-west and West Nyasa- land and in the far north became known as the Batuta, and reached to the east coast of Tanganyika and even to within sight of the waters of the Victoria Nyanza, before their efforts died out and they became merged in the resident population. These facts have to be borne in mind by the ethnologist and philologist ; otherwise, finding Zulu clicks and Zulu word-roots, head-dresses, shields, and customs on the verge of Equatorial Africa, he may arrive at wrong conclusions regarding the development of Bantu speech and history generally, unless he realizes that the presence of the Zulu speech and phonology and Zulu customs and names in Eastern Africa and Nyasaland can be explained by the tribal convulsions which occurred in South Africa a hundred years ago and led to those extraordinary Zulu raids which reached from the 30th northward to the 3rd degree of south latitude. (5) Traders and missionaries who had begun to settle on the Gaboon coast in the first half of the 19th century, recorded the turmoil and loss of life which was taking place due to the invasion of these equatorial coastlands in the basin of theOgowe by a much dreaded race of cannibal Negroes of splendid physique. These were the Ba-fang (as they sometimes call themselves generically), the Pahouin of the French and Fan of the earlier British explorers. The Fang were lighter in skin colour and better looking in facial features than the Negroes of the Gaboon-Cameroons coastline. They used cross-bows and were clever iron-smiths. The men went nude before they came under coast influence. We now know that the Fang migration was due west from the Western Mubangi and Sanga rivers ; also from the Upper Sanga, down the Nyong river to South Cameroons. They are to be divided at the present day into three different groups: the Bulu (farthest east), the Yaunde (north), and the Osieba or Fang (south). The Fang would appear to have migrated from the Baya country, north-west of the Mubangi country, some five or more centuries ago. Here they were impressed with the southernmost influence of the Sudan, and tinctured with a superior strain of northern blood. They issued from this region speaking a semi-Bantu tongue, but then for some centuries dwelt in close contact with West Congo Bantu, from whom they borrowed numerals, many word-roots, and some prefixes. 410 SIR H. H. JOHNSTON. A Survey of the Ethnography of Africa, etc. Finally, there is that question of vast importance, the modern European colonization of Africa. Africa, we have seen so far as indications of archaeology and language go has been anciently and continuously permeated by the white man, who in some shape or form or attenuated intermixture has provoked all race movements of great importance. The definite and the modern European colonization of the Dark Continent began with the establishment of the Koman Empire. After a long break, it was followed by the maritime wars and slave trade of the Saracens and Moors, which imported hundreds of thousands of Greek, Slav, Italian, French, and even British captives into Northern Africa. In the 19th century there followed the commercial Greek and Italian colonization of Lower Egypt, and the Maltese, French, Italian and Spanish colonization of North Africa. The Dutch had begun to colonize South Africa at the close of the 17th century. The Portuguese colonized Angola to a limited extent at the same time, and especially during the 18th and 19th centuries, producing, however, not as the Dutch had done, a solidly established, pure white people on African soil, but a race of half-castes ; who, however, have had a most potent influence on the civilization and the commercial development of West-central Africa. The Dutch were followed in South Africa by the British and the Germans. The French have partially colonized Madagascar, as well as their old plantation colonies of Mauritius and Reunion. The Italians will shortly make a new Italy in the Tripolitaine. Spain undoubtedly will colonize a good deal of Northern Morocco. The British are founding white men's colonies in Nyasaland, Rhodesia, and Equatorial East Africa. Once again, Africa is about to receive a most powerful infusion of Caucasian blood. SUMMARIZED CONCLUSIONS AND SPECULATIONS OF THE AUTHOR. The broad central part of Africa from Senegambia to Somaliland, north of the Mubangi-Wele River, has evidently been densely and continuously populated by Negro and Negroid races for a hundred thousand years or so. The reasons for arriving at this conclusion are the evident ancient destruction of forests and diminution of game by human agency over all but the very humid districts, the apparent ancientness of the stone implements, the extraordinary number and diversity of the speech families and distinct languages, and the isolated character of many of these. Such linguistic developments and singularities would suggest a considerable space of time, an abundant and persistent population. Further, the more primitive Negro peoples of forested (i.e., West and West- central) Africa offer a remarkable degree of affinity in the phonology of their speech, in their customs, weapons, utensils, and body-adornments with the Mela- nesians and Oceanic Negroes of New Guinea, the Bismarck Archipelago, and Solomon Islands, and even Fiji. These affinities and peculiarities, however, are not shared with the Bushman of South Africa in the same degree. Sm H. H. JOHNSTON. A Survey of the Ethnography of Af/ica, etc. 411 South Africa was seemingly the home of man of a generalized (perhaps Negroid) type at a period which may be computed at some 10,000 to 50,000 years distant from the present day j 1 though as yet the approximate remoteness of the earliest human remains and artefacts cannot be guessed with any precision. Apparently the Bushman type was not the earliest human race in South Africa, but was preceded by a race with fairly large brain capacity and slightly Caucasian affinities. As yet there are no definite discoveries of undeniably ancient pure Negro inhabitants in the regions to the south of the Zambezi. The Bushmen are aberrant, specialized, and, perhaps, degenerate Negroids, displaying scarcely any affinity with the true Negro beyond the spiral, tightly curled hair. Amongst themselves they share certain peculiarities of physical con- formation but vary in regard to the shape of the head, the degree of prognathism, development of brow-ridges, and stature. Though they all agree in general features of phonology the excessive use of clicks, of nasal and guttural sounds, and tenuity or paucity of vowels yet judged by the degree of affinities in word- roots their languages differ so widely from one another that a considerable period of time is required to account for their divergence from a common stock. On the other hand, the resemblance between the three or four Hottentot dialects is close and shows that no very long period can have elapsed since the ancestors of the existing Hottentots formed one people living within a restricted area. Though a Bushman-like race extends northwards between the Kunene and Upper Zambezi rivers, so far no specimen of Bushman pictorial art has been found north of the Zambezi or in South-east Africa. Indications of the existence of a race using stone implements like those of the Strandloopers and Bushmen occur not only all over South Africa, but also through Northern Zambezia to the vicinity of Tanganyika and eastward to Mozambique. One or two historical references and a variety of native traditions lead to the conclusion that Bushmen-like, click- language people persisted in Nyasaland and the Mozambique province down to within a few hundred years ago. Other fainter indications chiefly the occurrence of the Sandawi click-language and sporadic instances of Bushman physical peculiarities suggest the former existence of a Bushman-like race over all East Africa. " Strandlooper " (South African cave-men) stone and bone implements and shell ornaments and pictorial designs can be matched in Algeria. The arrival of the Hottentots as a band of conquering, cattle and sheep- keeping people in South-west Africa did not much precede that of the Bantu in time. The cultural and linguistic affinities of the Hottentots suggest their being 1 Very few data exist by which the age of the Cave men of South Africa and their artefacts can be determined. These remains, however (and others in Northern Rhodesia), are associated with the bones of mammals long since extinct, though how long we do not know ; for instance, with the gigantic buffalo (Bos baini), a large horse (Equm maximus, an exaggerated develop- ment of the zebra), a vanished type of rhinoceros, and even in the Transvaal with a mastodon. But such mammalian types may have persisted to a very late date in Southern Africa. Similar archaic mammals have survived all climatic changes and the attacks of savage man, only to be extinguished in a few years by British sportsmen in the 19th century. 412 Sm H. H. JOHNSTON. A Survey of the Ethnography of Africa, etc. influenced by some Hamitic invasion of Eastern Equatorial Africa, say 2,000 to 3,000 years ago. The Congo- Angola coastlands, especially north of the Kwanza Eiver, and the coastlands of Zanzibar, down to the Euvuma Kiver, and the east coast of Lake Nyasa and actual valley of the Zambezi, contain some evidence of comparatively ancient human habitation. But the heart of Central Africa mainly the central basin of the Congo as yet produces no sign of ancient human penetration and settlement. So far no stone implements have been obtained between the Mubangi River on the north and the table-lands and hill country which fringe the Congo basin on the south and west, though they are common in the cataract region of the Western Congo. But the south-eastern fringe of the Congo Basin, and especially the West Tanganyika and Katanga regions, are rich in stone implements and evidences of comparatively ancient metal workings (copper and iron). The reported vast sub- terranean dwellings and caves in Katanga call urgently for scientific investigation. Some of the existing cave-dwellers are said to be Bushman-like in appearance. This region yields " Strandlooper " implements. The Bantu occupation of Trans-Zambezian South Africa does not go back traditionally more than about 1,200 to 1,300 years. The only immediate predecessors of the Bantu in South-central and South Africa seem to have been the Congo Pygmies in the Congo basin and Bushmen and Hottentots south of the Zambezi. There are no traces remaining of pre-Bantu people in East Africa (other than Bushmen) south of the 8th parallel of S. latitude ; though the Zimbabwe ruins and a certain non-Bantu, non-Bushman element in the tongues of the south-east coast belt (Zambezi Delta to Sabi River) suggest the previous existence in that region of peoples who were neither Bushman nor Bantu in language, and far superior to either in culture. The ancient pottery dug up from a depth of 10 feet or so in East Nyasaland is somewhat superior in make to the pottery made there at the present day. In East Africa but not markedly so, south of the 4th degree of S. latitude there is some linguistic, cultural, and physical evidence to show that this region was anciently inhabited by Bushmen, Pygmy, Forest, and Sudanese Negroes. But there is no trace of an aboriginal Negro people in Somaliland proper. Much of Abyssinia, Galaland, Egypt and North Africa (besides the Sahara Desert and the Northern Sudan) had an ancient Negro population. All North Africa and much of the Sahara to the northern limits of the Sudan abounds with ancient rock engravings recalling in style and fidelity of reproduction of animal life the drawings of the cave-men of Western Europe rather more than the style of the Strandlooper and Bushman. Such indications as there are in these records of the rocks of the human figure suggest a man of generalized Caucasian type rather than a Negro. Stone worship and the use of stone in building and sepulture extend from North Africa southwards across the desert region to Sene- gambia (sporadically) and the northern parts of the Sudan, and to Somaliland. Sm H. H. JOHNSTON. A Survey of the Ethnography of Africa, etc. 413 The superstitious use of stone in connection with religion, burial, and after-death memorial reappears again in Yoruba, in the North-west Cameroons and adjoining Calabar region (Ekoi-land). Elsewhere there is scarcely a trace of the structural use of stone (except the piling-up of cairns of small stones as a superstition) over all Negroland until one reaches the region between the Zambezi, the Limpopo, and the Sabi the Zimbabwe country. In the heart of forested Negro Africa, from French Guinea to South-central Congoland, wood everywhere takes the place of stone as material for utensils, as object of worship and as memorials of the dead. The area of wood use and wood worship, moreover, was of much wider extension formerly and covered nearly all East Africa. The Bantu languages were called into existence by the immigration into the Central Sudan of some Fula-like race of semi-Caucasians, speaking "class and concord " language. In word-roots the Bantu tongues have their nearest affinities with the "semi-Bantu" languages of the Benue, Lower Niger, and Calabar- Cameroon s. In syntax they have also slight affinities with the Kordofan and the Nilotic groups. I think we may further surmise that after they were created by the impact of some language influence like that of the Fula type, they were additionally at a later date reacted upon by slight Hamitic influence on the part of some such people as the Gala. This impact probably took place after the ancestors of the Bantu had occupied the central region of the Great Nile Lakes, and after they had sent out their first hordes of invaders into the Northern Congo forest-belt, towards the Atlantic (Cameroons) coast. I believe this Bantu conquest and occupation of the southern third of Africa to have been an event of comparatively recent times, beginning, perhaps, not more than 2,000 years ago. Parts of the Congo basin were perhaps only occupied 400, 600, and 900 years ago. Some of the pre-existing Sudanese Negroes, whose ancestors were overwhelmed and "Bantuized," still remain with their non-Bantu languages along the course of the Northern Congo. Elsewhere, however, in Northernmost and in Central Congoland, there have been subsequent Sudanese invasions which have overridden and forced back the Bantu. The Bantu-speaking Negroes, with their cattle-keeping aristocracies (anciently of Gala origin), have from their earliest days and traditions been associated with the use of metals, first copper and then iron. No doubt it was this knowledge and these superior weapons which enabled them to displace and assimilate so rapidly the pre-existing Negro tribes, still in the Stone age or only using weapons and utensils of wood, horn, or bone. The Malay colonization of Madagascar was an episode which extended perhaps^ $ over a period of about 1,200 years, beginning about 3,000 years ago and terminating X with the arrival of the Hovas about 1,800 years ago. There are vague traditions C , among the Malagasies that they found this great island sparsely populated by ^ ' some pre-existing Negro race which buried its dead in graves surmounted with u stones. Later on, the Himyaritic Arabs, trading to Madagascar about the beginning of the Christian era, transported thither many Negro slaves from the opposite Mozambique and Zanzibar coasts. These Negroes presumably had then, 414 SIR H. H. JOHNSTON. A Survey of the Ethnography of Africa, etc. no more than now, no suitable vessels for sea-faring, therefore they could only have come to the island on board the Arab daus and mitepe. But by some means or other they came in sufficient numbers to colour the skin of the Malay invaders,, crisp their hair and influence their speech with numerous Bantu roots. Lastly, in deciphering the faintly recorded human history of Africa, one seems to see in the White Man, the Caucasian, the primum mobile, the chief causer and inspirer of racial migrations, disturbances, remoulding of peoples, uprise of religious beliefs, creation of new languages, new arts, especially of agriculture and the domestication of animals. The White Man has been the cause of all good progress as well as of all the annectant misery and strife which hang on the flanks of upward evolution. And so potent has been the Caucasian in the history of Africa,, as of Asia and Oceania perhaps even of prehistoric North America that it has needed but the slightest admixture of his blood with that of the Negro to effect these far-reaching results. When and since my paper was read before the Royal Anthropological Institute in June, the following criticisms have been received ; together with others verbal and epistolary which I have attended to in my revised text. But I append the remarks of Mr. Emil Torday, Mr. T. Athol Joyce and Professor C. Gr. Seligmann as they stand : The intimate personal acquaintance Sir Harry Johnston has with the various inhabitants of Africa (including the white men inhabiting the north, among whom he began his brilliant career) makes him a much fitter person to judge their respective merits than most of us who only know one race or the other and conse- quently are prejudiced in its favour. I admit my partiality for the black man and consequently have to distrust my own judgment when I feel convinced by certain arguments that seem to prove to my satisfaction that we are indebted to the Negro- for the very keystone of our modern civilization and that we owe him the discovery of iron. That iron could be discovered by accident in Africa seems beyond doubt : if this is so in other parts of the world, I am not competent to say. I will only remind you that Schweinfurt and Petherick record the fact that in the northern part of East Africa smelting furnaces are worked without artificial air current and, on the other hand, Stuhlmann and Kollmann found near Victoria Nyanza that the natives simply mixed powdered ore with charcoal and by introduction of air currents obtained the metal. These simple processes make it possible that iron should have been discovered in East or Central Africa. No bronze implements have ever been found in black Africa ; had the Africans received iron from the Egyptians, bronze would have preceded this metal and all traces of it would not have disappeared. Black Africa was for a long time an exporter of iron and even in the 12th century exports to India and Java are recorded by Idrisi. It is difficult to imagine that Egypt should have obtained iron from Europe SIR H. H. JOHNSTON. A Survey of tlie Ethnography of Africa, etc. 415 where the oldest find (in Hallstadt) cannot be of an earlier period than 800 B.C., 1 or from Asia, where iron is not known before 1000 B.C., and where, in the times of Ashur Nazir Pal it was still used concurrently with bronze, while iron beads have been only recently discovered by Messrs. G. A. Wainwright and Bushe Fox in a predynastic grave and where a piece of this metal, possibly a tool, was found in the masonry of the great pyramid. We, must, however, not forget that no archaeological work whatever has been carried out in the greatest part of Africa, and future discoveries may disprove all that is said above. The country west from Lake Mweru and of the Lukalaba (Lualaba) with its great number of undisturbed caves ought to attract the archaeologist who wants to contribute to this most fascinating research in the ancient history of Africa. E. TORDAY. It is a matter of great regret to me that I am unable to be present at the reading of Sir Harry Johnston's paper this afternoon ; but through the kindness of the author and the Secretary of the Institute I have had the opportunity of seeing an advance proof. I wish I could have expressed personally my congratulations to the author for the clear-sighted manner in which he has reduced a most bewildering mass of evidence to order ; the material which must have passed under his hands is very vast, and its nature makes the continent of Africa perhaps the most difficult of all to view as an organized whole as far as the tribal movements are concerned. The paper represents, I know well, years of labour, and contains only the results. On the importance of those results it is unnecessary to dwell ; but the ethnologist is always asking for more, and I should like to express the hope that Sir Harry will at some time give us the chief points of evidence concerning his most important conclusions as well. This could not, of course, be done within the limits of a paper ; I am hoping for a book, and I trust that Sir Harry will perform yet another service to ethnology in producing it. Among the various kinds of evidence which the author has considered, I believe that linguistic evidence has taken the principal place. Unfortunately my studies have been limited to the culture and traditions of the various tribes, and it is a great satisfaction to me personally to find that the conclusions which I have been led to form correspond in the main to those which Sir Harry has based on a more extensive knowledge. Nearly all my remarks consist in that request for more which, as I said above, is always on the lips of the insatiable race of ethnologists. First as to the Fula. I cannot quite gather from the paper the exact ethnological position which the author would give them. In one place he seems to regard them as present in Africa before the immigration of the Berbers, and as distinct from the " Libyan " section of peoples of which these Berbers are the nucleus. If so, whence did they come ? [Presumably from western North Africa, 1 German authorities would make this 2000 B.C., or even earlier. H. H. J. 416 SIK H. H. JOHNSTON. A Survey of the Ethnography of Africa, etc. like the Berbers, whom they preceded. They may be the more or less negrified remains of the Neolithic hunters and herdsmen of Mauretania. H. H. J.]. They are first heard of in Senegal, and all their movements of which we have evidence have been from west to east. Their geographical position and their physical appearance make them a very convenient link between the Berber, Tuareg and Tibbu on the one hand, and the Mandingo and Negro on the other ; and I have hitherto regarded them as an intermediate tribe. It is true that the Berbers are in the main agriculturists, the Tula cattle-keepers, but here the Tuareg and Tibbu furnish the link, since these tribes have been forced to abandon agriculture owing to their desert environment. I gather that the author is inclined to regard the Libyans as, in the main, of Iberian stock, and therefore distinct from the Hamite. This point of view is supported by the cultural evidence. Wherever the Hamite is found, pure or diluted with Negroid admixture, the keeping of cattle always assumes paramount importance, while the Berber, except where he has had a pastoral life forced upon him by his environment, is principally an agriculturist. I am inclined to regard the Hamitic element in the Libyan peoples as entirely negligible unless some evidence other than linguistic can be found. We know that linguistic evidence is at times unstable, as in the case of the Barotse country, where the Sesuto speech survives, though the Makololo who introduced it were wiped out to a man in historical times. When the author states with emphasis that the Zimbabwe ruins are not Negro, I am quite prepared to agree with him if he uses the term in the narrow sense. But if he means to exclude the Bantu also, then I find myself, to my regret, on the other side. We are beginning to find that the Bantu is not so primitive as at first thought ; recent investigations have brought to light a sociology and system of government considerably more complicated than at first suspected, especially as far as the greater kingdoms, Bushongo, Lunda, Benin, etc., are con- cerned. The architecture of the ruins is on a very low plane, and we know that the economical activity of the Negroid, as far as work implying communal labour is concerned, depends solely on the personality of the chief ; and Africa has produced many chiefs of great force of character. No remains other than fragments of porcelain of comparatively recent date, implying external connection have been found in connection with the ruins. The ruins themselves do not resemble any buildings elsewhere, and in any case the onus of proof must lie with those who wish to invoke some external influence. If Sir Harry has evidence of such I should like to urge him to publish it. As to Egyptian influence ; I am inclined to think that the author is inclined to overrate its extent. That the tribes in the immediate radius of the Nile valley were affected by its culture it would be hard to disbelieve ; but I would urge that there is no evidence for supposing that this influence reached West Africa. The presence of beads of Mediterranean type is no evidence. Beads are the most exasperating objects with which the museum man has to deal. They are universally SIR H. H. JOHNSTON. A Sumey of the Ethnography of Africa, etc. welcome, and they pass readily from hand to hand. Nor are they alone in thei propensity for travel ; I need only mention Katanga copper, which in early days travelled far to the north and north-west, and formed a regular article of import among tribes who were never near the Katanga. Also we know maize and tobacco in the interior of Africa anticipated by centuries the arrival of the white man who introduced it on the coast. Nor can I admit that the art of the Bushongo or Bini bears any striking resemblance to that of Egypt. In fact I have for a long time held the other view, viz., that many points of Egyptian culture, and especially of religion, are based on practices and beliefs which are Negro in origin. With regard to the specialized culture of the Bushongo, I think it is clear that they brought little of it with them. Tradition is quite categorical at least on the subject of weaving. It is expressly stated that they wore barkcloth until a very recent date, and this material survived until contemporary times as the garb of ceremony. We even know the name of the king in whose reign weaving was adopted from the tribes to the west of their present home. There are two movements of peoples on the importance of which to Africa I should like to lay stress. The first of these is that which caused the Bushongo migration, and which, I believe, gave rise also to the easterly movement of the Azande and the westerly movement of the Fang. These three tribes appear to be connected by many points of similarity, especially the use of the throwing-knife (Shongo means undoubtedly throwing-knife). Moreover their traditional routes of migration seem to coincide in the region of the Shari where this weapon is still in use. It is possible that the Fang migration may be later than that of the Azande and Bushongo, between whom the cultural connection is particularly close. The second movement is that attendant upon the formation of the Lunda Empire, which had such far-reaching effects upon Angola and the Kwango, Kwilu and Loange districts. I have recently worked out the complicated tribal movements to which it gave rise, from a number of sources including certain information brought back by Mr. Torday. This account now awaits publication in Belgium, and will be found to confirm Sir Harry Johnston. Also I have information concerning the westerly spread of tribes from Tanganyika, relating mainly to the great Batetela people, and based almost entirely on Mr. Torday's material. This though worked out two years ago, is still in the Belgian printer's hands. I wish that I had had the time to give a longer study to a communication of this importance ; the foregoing remarks embody my first impressions. In concluding I should like to offer the author my thanks, as a student of African ethnology, for his courage in attacking so complicated a subject, and in presenting a concrete scheme which must be a valuable guide for future research. I hope that he will not stop here, but will eventually give us in book form a fuller and more detailed treatment of the question, in which he will discuss at length the various points of evidence upon which his conclusions are based. T. ATHOL JOYCE. 418 SIR H. H. JOHNSTON. A Survey of the Ethnography of Africa, etc. In answer to the foregoing criticisms I add the following remarks on (a) the originating or the introduction of iron working in Negro Africa, and (6) the " ethnological " character of the Zimbabwe buildings and culture. With regard to iron, I would refer readers to the article by W. Belck on " The Discoveries of the Art of Iron Manufacture," in the Berlin Zeitschrift fur Ethnologic, 1910 (there is an English translation published in the 1911 Annual Eeport of the the Smithsonian Institute), and to the article on " Hallstadt " in the Encyclopaedia Britannica and other writings of Professor W. Eidgeway, based on much German and Austrian research work ; also to my book, George Grrenfell and the Congo. It seems to be increasingly evident that the art of iron manufacture originated first in Europe or "West Asia, and that, like the domestication of the horse, it was one of the important discoveries of the Aryans which made them in so many directions a conquering people, and from them passed to Semite and Mongol. It is quite possible that Negroes or Negroids in the Northern Sudan were manufacturing iron implements before iron became a much used metal in Egypt ; but these Sudanese, Nubians, or Nilotes may have learnt the art from Caucasian emissaries of trade and knowledge who were Minoan or Phoenician adventurers. This much is clear from an examination of African archaeology, legends, folk-lore, and languages : that copper was used before iron, and immediately replaced stone, horn, bone, and wood, and that the use of iron and the knowledge of iron working does not go back very anciently, except in Nubia. In the Congo basin, the introduction of iron instead of copper had apparently much to do with the great Bantu conquests and formations of kingdoms between an approximate 600 A.C. and 1400 A.c. " The hunter with the iron spear " is in many cases the legendary founder of this and that Central African dynasty, I myself require very convincing proof that the pure-blood Negro ever origin- ated anything. This is one reason why I find it distasteful to attribute to a Negro Bantu people, or any other more or less pure Negro race, the Zimbabwe architecture and culture. No doubt the Makarana subsequently copied the art of stone-building and of gold-mining learnt from the Caucasian or semi-Caucasian immigrants into South-east Africa. But I am convinced they did not originate the gold-mining industry or build the old and original Zimbabwe. The remarkable Bantu languages may have been evolved in the Central Sudan from some diluted impetus of white man's intelligence, but the purest " Bantu " and the least Negro in physiognomy are the tribes of the Victoria Nyanza and Tanganyika ; and although they stand high in brain capacity and religious and even social developments, in the structural arts they like the Herero and Zulu have remained low and have never conceived of the use of stone, nor have they taken any interest in gold. The Arab historians quoted by Capt. M. Guillain 1 specially 1 Chiefly Idrisi, who wrote at the court of King Robert II of Sicily in the 12th century. The passages I refer to more especially are on pp. 224-226 of Guillain's Documents sur I'Hiatoire (etc.), de VAfrique Orientale. SIR H. H. JOHNSTON. A Survey of the Ethnography of Africa, etc. 419 mention that although the Sofala countries abounded in iron and gold, the native Negroes preferred copper to other metals, and made all their ornaments of it. It is a remarkable fact that up till now no indigenous Bantu word has been found for " gold " the natives of Central, East, or South Africa employ for the concept, " gold," corruptions of the English, Portuguese, or Arab word. In Zambezia the word for gold is (ordinarily) ndalama, which is simply a corruption of the Arab dirham = money. The conical towers of the Zimbabwe architecture recall irresistibly the early Islamitic minarets on the east coast of Africa, which in their turn may have been developed from antecedent Phoenician (or South-west Arabian) pre-Islamic architectural influence on the east coast of Africa. I still think the balance of probability lies in favour of an Arab-built Zimbabwe of from 2,300 to 2,000 years ago. H. H. JOHNSTON. The paper we have just heard covers so vast a territory and takes account o so long a period of time that any attempt at criticism or discussion must necessarily resolve itself into the consideration of a few sentences which, divorced from their context, may very well appear to bear a different significance to that intended by their author. Bearing this in mind, I am nevertheless tempted to cavil at Sir Harry's suggestion (if I understand him aright) that the Forest Negro arose directly from the Pygmy type, and I should absolutely dissent from the proposition that the Forest Negro specialized in the swamps of Nileland and became the Nilotic Negro. It is clear to my mind that the peculiarities of the Nilote are not due to specializa- tion in situ of a primitive or comparatively primitive Negro race, but must be attributed to the influence of a foreign element, this element being identical with that non-Negro strain which also occurs in the Masai and kindred East African tribes. In other words the Nilote owes his comelier features and better developed brain to invading Hamitic influence, an influence which is perhaps to be found in its purest form in the foreign (Gala) aristocracy of Bunyoro and Ankole. A good deal of interest attaches to the tall stature of this group, of the allied Bahima and of the tribes of the Masai group in East Africa ; indeed the Bahima are probably the tallest men in the world, yet all seem to have sprung from the fusion of the short, slim Hamite and the moderately-grown, stoutly-built Forest Negro. As far as I have been able to ascertain the numerous breeding experiments on Mendelian lines that have been carried on in recent years afforded no analogous instances, indeed nothing seems to be known of the conditions producing variation of size in animals; there may be, however, an interesting parallel to the stature of the Nilotes and related tribes in the result of the union of the dwarf procumbent pea (commonly called the cupid pea) and the bushy pea, the latter being a plant of stiff upright habit of short to medium height : the first generation of hybrids are all giants. Again, the view put forward concerning the Bantu Negro (pp. 390,391^ seq.) quite ignores the comparatively large foreign, and as I should say Hamitic, element in such 420 SIR H. H. JOHNSTON. A Survey of the Ethnography of Africa, etc. tribes as the Zulu Kafirs. This is perfectly logical from the author's standpoint ; if there is no Hamite in the Nilotes then there need be none in the Zulu, yet a consider- able foreign non-Negro element is generally admitted in the Southern Bantu. In the last few months important evidence has come from the linguistic side in favour of the Hamitic element in the Nilotes. This will be found in Wester- mann's recent work The Shilluk People. This authority states that the Nilotic languages, whether they belong to his Niloto-Sudanic group (Dinka, Shilluk, etc.), or to his Niloto-Hamitic group (Bari, Masai, etc.), all show more or less strong Hamitic influences, and as further showing how widely spread is this Hamitic influence, I may quote his conclusion that even such languages as those of the Abukaya, Lendu and Moru are connected with the Niloto-Sudanic group. I may now refer to the suggestion that the cattle-keeping aristocracies of Central Africa, e.g., the Bahima, are descended from Egyptian colonists of two or three thousand years ago. Surely a simpler explanation is that these represent the immigrant Hamitic element in a comparatively pure form. Here is a simple explanation of Sir Harry's observation: "Others again of this type [resembling Egyptians] were so strikingly like Galas and Somali that the Somalis of my party declared them to be of their own race." I feel sure that Sir Harry would not suggest an Egyptian origin for the Somali, though like the proto-Egyptians they are pure or almost pure Hamitic. [I have sometimes fancied, in studying the facial features of the Hima aristocracies in Central Africa, that I could distinguish two non-Negro strains : one akin to the Dynastic Egyptian, and the other to the Gala- Somali. Their long-horned cattle are akin to the breed of Ancient Egypt and modern northern Galaland ; but this type of cattle (existing in the Chad region also) is not found in southern Galaland or in Somaliland. H. H. J.] To sum up ; my impression of one aspect of the paper is that the author lays too much stress on Egypt as the great civilizing factor in Africa and ignores the far- spread Hamitic influence (of which the Egyptian civilization was only a special development) which was leavening dark Africa, perhaps for thousands of years before Egypt herself emerged into the light of history. Before sitting down I should like to be allowed to add my tribute of admiration to that expressed by the President, and I may perhaps be allowed to suggest that the value of the paper, great as it already is, would be enhanced if Sir Harry could see his way to give brief footnotes giving his authorities. I very much doubt if any one in this room except Sir Harry could refer to the original account of Javanese junks which reached the African coast during the last century. C. G. SELIGMANN. The following are my comments on Professor Seligmann's remarks : It matters very little whether the Forest Negro arises from a Pygmy stock or whether the Pygmy is a specialized development of the Forest Negro. The two groups are evidently closely allied in bodily type. But I am inclined to regard II I I II II II II SIR H. H. JOHNSTON. A Survey of the Ethnography of Apia . A 9?? 95 1 dwarfishness (i.e., stature below a 5 ft. 3 in. standard for men) to be an acquired rather than a primitive character in mankind since the formation of the human genus. I think the long-leggedness and tallness of the Nilotic Negroes are physical traits due quite as much to conditions of environment as to racial intermixture. The Bahima are not, as Professor Seligmann suggests, "the tallest men in the world." This statement should be applied to the Turkana of Lake Kudolf ; but gigantic stature occurs in most of the tribes of more or less Nilotic affinities. I cannot find anything in any statement of mine which underrates the importance of the Harnitic blood-element in the higher types of Zulu or of other great Bantu peoples, though this feature is capable of exaggeration at the hands of some writers. It is a great mistake to suppose that the whole mass of the Zulu-Kafirs are a race of tall handsome Negroids : it is only the aristocracies, so to speak, and a tribe or family here and there which answers to a Eider Haggard description. Many of the Pondoland and Natal Kafirs and the Swazis are pronouncedly Negro in build and physiognomy, while even among the Zulus of Zululand the Bushman element is patent to the eye of a trained observer. The linguistic influence of Hamite on Nilote is not denied by me but is, I think, correctly defined and limited. As regards the cattle-keeping aristocracies of the Central Sudan and of Bantu Africa, I still prefer to assign much of the slight Caucasian element in their blood and almost all their culture to infiltration from Ancient Egypt rather than to influences from Galaland and Somaliland for reasons too numerous and wordy to be given here. Most of them have been already presented in my published works on Africa. References to photographic illustrations of this essay appear here and there and are partly allusions to lantern slides used at the time of the lecture and partly to my original intention to illustrate this little treatise by a very large number of photographs and drawings of racial types, skulls (ancient and modern), domestic animals, architecture, and implements. But this collection proved to be far too numerous and expensive for reproduction by the Institute in its Journal. Therefore I decided to take the advice of Mr. Athol Joyce and reserve this material, together with maps, statistics, and bibliography, for production at some future date in a work on the Ethnography of Africa. H. H. JOHNSTON. \_Reprinted from the Journal of the Royal Anthropological Institute, Vol. XLIII, July-December, 1913.] Harrison and Sons, Printers in Ordinary to His Majesty, St. Martin's Lane, London. d REC'S IMJW. APR28IW .r tc if' Universil Southe Libre