1II.3.V. THE GUHAYNA GROUP 275
numbers of Arabs did push southwards from Tunis Algiers and Morocco to Central Africa in the centuries following the Hilalian
invasion of North Africa^, and though one may admit that the pre-
valence of the Abu Zayd el Hilali tradition among the Bakkara is a
little suggestive, we have the definite statement of Ibn Khaldun that
in the first half of the fourteenth century the Guhayna swarmed over Nubia and rapidly pushed farther afield "following the rainfall^," and modern expert opinion has heavily preponderated in favour of
the view that the Bakkara came from the east. For instance, Barth
says^ of the Shoa (Shawi'a)—the name given locally to the semi-
sedentary Bakkara Arabs of Bornu Bakirmi and Chad, and par-
ticularly to the Salamat :
Of the migration of these Arabs from the east there cannot be the
least doubt. They have advanced gradually through the eastern part of Negroland.... Their dialect is quite different from the Maghrebi, while in many respects it still preserves the purity and eloquence of the language of
Hijaz.... These Shuwa are divided into many distinct families or clans, and
altogether may form in Bornu a population of from 200,000 to 250,000 souls. He adds that they appear to have immigrated gradually from the
east from very early times, "although at present we have no direct
historical proofs of the presence of these Arabs in Bornu before the time of Edris Alawoma" (1571-1603); and he mentions the systems
of blood-money (" dhia"—which by the way maintains among all
the nomad Arabs of the Sudan) and infibulation of females as con-
necting the Salamat-Shoa with the east^. Similarly M. Carbou, who
divides the Shoa into two groups, one from the north and the other,
the "Guhayna" group, from the east, also remarks^ that the use of
the word "Nuba" for "all indigenous non-Arab Muhammadans"
lends weight to the current tradition of an early sojourn in what
is now the Anglo-Egyptian Sudan^. In the same way the habit
in use among the western Arabs of denoting the Kanembu as "Hamag'''" points as clearly to a connection with the "Tribes of
^ Q.v. in Part II. The Arabs of Bornu, Bakirmi and Chad district—chiefly semi- sedentar^' Bakkara—are called Shoa, or Shawia (Barth, "Shuwa" and "Shlwa") by the indigenous tribes.
^ See p. 139, above.
* Vol. n, Ch. XXXII, pp. 355-356. * Vol. in, Ch. lii, p. 465. * Vol. II, pp. 4, 8, 9, 20, 28, 48.
** Carbou, 11, 47.
' Ibid. I, 36. Carbou is apparently unaware of the existence of a tribe called Hamag in the eastern Sudan and therefore fails to explain the term. So, too, Nachtigal (Voy. au Ouada'i, p. 74) says " Les Hammedj's sont les derniers auto- chtones du Kanem. lis sont de la famille des Boulala's "—also, it seems, in ignorance of the Hamag of Sennar. There is no evidence of a Hamag movement from east to west or vice versa at any time and the term Hamag was probably no more than an Arab importation, used to denote any uncivilized people. An exactly similar use of the word occurs in MS. C 3 11.