50 THE NATIVE MANUSCRIPTS iv ba. cxxxiii.

called "el Himyari" because his mother was of Himyar. Dhu el Kila'a was a well-known Himyaritic name (see Chaps. XLiii, LXXV and xciv of

Mas'udi, and Wustenfeld, 3), as were also Kuda'a and 'Adi. There was

also a Himyarite king "Masruk'' son of Abraha (see Mas'odi, Ch. XLiii); and "el Yemani " and " Yemen " suggest Himyaritic origins. In fact, until one reaches the immediate descendants of el 'Abbas the names given in this

paragraph strongly suggest Himyaritic rather than Isma'ilitic affinities.

So, too, the account given by el Mas'udi (Ch. xxxi) as that of an old Copt living in southern Egypt, who was interrogated about 260 a.h. by Ahmad ibn Tulun, would strongly corroborate a theory of Himyaritic affinities

for the Ga'aliin. The old Copt said that the Nuba [i.e. the inhabitants of Dongola and thereabouts] used bows of which the pattern had been borrowed from them by the tribes of the Hegaz and Yemen, and that

" their kings boast that they are Himyarites " {jt<ft>- u-^-^v^'vo^J-^vo-v^^^i) • see V( 1. I, pp. 8 and 168. Of course a late genealogist would not hesitate

to graft an Himyaritic branch on to an Isma'ilitic stem for the sake of exalting the tribe's lineage, and it would appear that this has actually been done. The early (Isma'ilitic) generations are given correctly in para, cxxxiv.

cxxxiv MS. 3 omits this and the following paragraph altogether.

The pedigree given in cxxxiii is more common than the syncopated one favoured here and by the author of AB, on account of its not containing so many non-Arab names. For Serrar see AB, paras. CLXX and ccxii, Bir Serrar, the old rock-hewn wells at the foot of the hill of the same name near Bara in Kordofan, are said to be named after him.

cxxxv Cp. AB, XXIX ; A i, lii; A2, in; A 6, in; A8, ix; B i, xxviii.

For this injunction not to trace predigrees beyond 'Adnan (or his son Ma'ad) see Mas'udi, Chaps, lxix and lxx, where the following occur:

^,^^•^'1^ aJUJI jLaf^)! ,j^ ^__5-^ '•■oJ A.o-ljil «^3 (J^^ 15JI s.,,.— Jb J jU*Jj ^j\

O^ j^^-^!^- O' **vJ i>« '-'J^i ^ w — *- (^ >y*^ j^' ^ y^3 (2)

^j. — Jl J-Al ^aA;^.! jk5^ Jaii j^ji^ ^^J^ ^ Jl ^s J^^ ^t siuj jas jju*

>^y> ^i cr^^l ^j"^ ^o^ «^>JI '->* 0-* ^! — ^^ 'j^ ^^^i U^'j (3)

Caussin de Perceval (Vol. i, p. 183) admits the unimpeachableness of

the pedigrees between Muhammad and 'Adnan and concludes that the birth of '.Adnan cannot be earlier than 130 B.C.

.Muhammad's objection to pedigrees extending beyond 'Adnan was simply that any such must necessarily be guesswork. As he laconically

said ^j^l — Jl vJ^ ^^ O-*) "Beyond this point the genealogists lie" (quoted by ibn Khaldfin, Ar. ed. Vol. li, p. 3).