272 THE GUHAYNA GROUP iii. 3. i.

its high protruding forehead, wide mouth and weak chin at once suggests the Fellata, and it is an estabhshed fact that large numbers of that race have become incorporated with the Bakkara tribes since they first settled in Central Africa. This is of course more especially true of the Salamat and Haymad, the most westernly Bakkara, v/ho live among a population that is largely Fellata and who appear to include sections of that people; but it also applies to the more easternly Bakkara as we shall see in dealing with the Hawazma.

II As a whole the Bakkara are, with the possible exception of the Shaikia, the most warlike Arabs in the Sudan: they are also the most inveterate slave traders and raiders, and living as they do on the northern confines of the negro country they have indulged their predatory propensities ad libitum for so long as they have not been repressed by the firm hand of the Government.

The same qualities that have made them bold fighters and hunters have at all times since their settlement in Africa brought them into collision with the rulers of the more sedentary people who inhabit the zone immediately north of them, with the Sultans and Meks, that is, of Bornu, Wadai, Darfur and Kordofan.

In the dry season of the year the Bakkara move with all their cattle to the rivers of the south and there hunt the elephant and raid the negroes, but when the rains render the southern Bakkara country a swamp of cotton-soil infested by the fly they move northwards to the clean pastures of the higher ground and cultivate or graze their herds. It is then that they have been apt to become involved in quarrels with the sedentary people of the Sultanates.

They have not, however, been invariably successful, except in so far as a perennial evasion of the full tribute demanded may be counted success, and in consequence they have at diff'erent periods migrated eastwards or westwards along the line of least resistance and various sections have been transplanted from place to place and from tribe to tribe until it is impossible to say how they were originally grouped. In the account which follows I have taken as the units the tribes as they appear at present and specified the various sections subject to each, but it will be at once obvious from a comparison of the lists of these sections and from a study of the past history of the Bakkara that no real racial dividing line can be drawn between any one tribe of them and any other.

In Kordofan, where, if one omit the brief orgy of the " Mahdta,'* there has been a settled Government for nearly a century, the Carbou, ii, 51.

See Chevalier, Afrique centrale frangaise, p. 321, quoted by Carbou, 11, 60.