branches to enable the philologist to satisfy himself that he is justified in regarding them as forming homogeneous groups. But the various members of any one of these families often differ widely in their Vocabularies, and sometimes are unlike in many

indicates the range of the Bantu family of languages.

indicates the position of languages possibly allied.

Map No. 5.-Map of Africa, showing range of Bantu Languages.

grammatical points. For instance, Galla and Šilha (Shlu) (the language of the Moroccan Berbers) are both members of the Hamitic group, yet, beyond their structural resemblance and their agreement in certain main features of grammar, they scarcely possess a root or test-word in common.1 I need not multiply instances

1 Compare with this Lu-ganda and Oci-herrero, languages spoken at

of this; the fact is already recognized by students of African languages, that in the northern half of the continent there are bewildering multitudes of diverse tongues belonging to many independent families, and apparently irreducible to a common origin. Yet cross the irregular boundary-line which runs over the continent from 6° N. on the west coast to the Equator on the east coast-viz. the northern limit of the Bantu speech-and what do we find? Why that the whole of the southern half of Africa, with the exception of the Masai and Galla intrusion in the north-east and the Hottentot enclave in the south-west, is the domain of a single homogeneous family of languages, the Bantu differing perhaps less among themselves than do the many offshoots of the Aryan stock. The only other African tongues which in any way approach the Bantu (as regards similarity of structure) occupy little isolated patches of country in Central and Western Africa, surrounded often by totally dissimilar forms of speech. Originally, there is little doubt, the primal Bantu language was as one of these, a member of a little group of prefix-governed tongues developed somewhere in the heart of Africa. Peculiar circumstances gave the people who spoke it the opportunity of playing a great rôle in unwritten African history; and the Bantu negroes, at one time very likely an obscure and unimportant tribe like the Temne and Bulom of Sierra Leone, the Efik of Old Calabar, or the Tumale of Kordofan, became the ruling and almost

either extremity of the Bantu language-field, separated as widely as are Galla and Silha, yet differing from each other only as Latin from Greek.

2 I hardly think it worth while to except the languages of the Pigmy races (Obongo and others) from this general statement, as we know too little of them to pronounce as to their affinities.

the exclusive race of Southern Tropical Africa, swallowing up, obliterating, absorbing the previous inhabitants of the land, and carrying their own form of language triumphantly from the Upper Nile to Natal, and from the River Tana on the east coast to Fernando Po on the west. In the spread of the Masai race and language we may see, on a much smaller scale, a parallel to this extension of the Bantu. It will be remembered that I already have pointed out the relations which the Masai language bears to the Šiluk and other tongues spoken in the basin of the White Nile. Some of these forms of speech resemble Masai about as much as the Non-Bantu prefix-languages resemble Bantu-proper. Most of them are spoken over small and confined areas. Yet the range of the Masai tongue and its various dialects extends almost uninterruptedly from 3° N. to 5° S. of the Equator, wherever the Masai race has spread in its warlike raids. Now we know that in many parts of Eastern Equatorial Africa the Masai only arrived, as it were, the other day. In many districts, too, in the centre of their present home the remnant of aboriginal natives still retain the oral tradition that several generations ago the Masai were unknown in the land. Yet, with the exception of a few isolated mountain-ranges where the Bantuspeaking inhabitants still linger, the Masai language is dominant throughout the whole length and breadth of the huge country known as Masai-land. Moreover, it is probable that had this warlike race not encountered natives armed with firearms and dominated by Arabs on the south, they would have overrun a considerably larger district than the one they at present occupy, and their language would have acquired an even greater range, and in course of time have lost its uniformity,

and become split up into many separate dialects. Yet it is probable that some centuries ago the Masai were only a small section of a Nilotic race which adverse circumstances drove from its original home, and forced farther and farther south, the Masai all the time developing their fighting qualities, first to save themselves from extinction, and then in turn to pursue a successful policy of rapine and dominion. Such, in all likelihood, were the origin and history of the spread of the Bantu race and language. It is quite possible that before their advent the southern half of Africa was sparsely populated, perhaps by a low type of negro in the north and west, and Hottentots, Bushmen, and Pigmies in the south and centre. The invading Bantu carried all before him in the early days of his invasion, and was probably little but a fightingman, as is the Masai at the present day. In the west, however, he was stopped-choked, one might almost say-by the dense negro population, and in the southwest by the Hottentots, who were probably driven to bay in this corner of the continent. In the centre of Africa the race and language of the Bantu people are purest to this day, probably from the much less intermixture which took place between them and any previous inhabitants. Judged by his purest types-the inhabitants of the great lakes and the Congo basinthe Bantu was probably a red race in its origin. Intermixture with Hottentots has given the Zulus a yellowish tinge, and clicks in their language, and the fusing of Bantu and Negro on the west coast has debased the character of the speech and deepened the colour of the skin.

Taking into consideration the fact that the unwritten languages of a savage people vary

vary with

extraordinary rapidity, and noting at the same time how great is the uniformity in grammar, construction, and vocabulary between all the Bantu languages, we can hardly believe that they are a very ancient offshoot of the negro race, or that the date of their original dispersion from their centre of development is very far removed. I should think, as far as we may judge, that it is about 2000 to 2500 years since the Bantu people commenced their invasion of Southern Africa; and that before this date-before the dispersal of what was evidently a compact and homogeneous tribe -one common language, the Bantu mother-speech, was spoken. One of the principal reasons which induces me to hazard this guess as to the recent period at which this family of languages split up into separate branches and spread out over half the continent of Africa, is the interesting fact that in nearly every form of Bantu speech the domestic fowl is indicated by words of one type, all pointing to a common origin. This in its most archaic form is nguku. This is the term used in Ki-nyoro, in Lu-ganda (the language of Buganda, on the north-west shore of the Victoria Nyanza), in the tongues of the Upper Congo, of Lake Tanganyika, of Kilima-njaro, at the present day, while variants of the same word are found in Kafir (inkuku), Zulu, Mpongwe, and, in short, almost all the known members of the Bantu group. Now this leads us to believe that the domestic fowl must have been known to the original Bantu race before it commenced its migrations from the centre of its development, from

3 Itself, no doubt, in those early days but a sister-tongue or a lateral branch of many cognate forms of prefix-governed speech descended from an earlier stock.

4

Ki-nyoro, one of the most northerly forms of Bantu language, is spoken in Bu-nyoro on the Victoria Nile, in lat. 2° N.