i88i] ANIMAL WORSHIP AND ANIMAL TRIBES 461 



logical system of the Arabs to an ancestor who bore the 

 tribal or gentile name. Thus the Kalb or dog- tribe 

 consists of the Beni Kalb sons of Kalb (the dog), who is 

 in turn son of Wabra (the female rock-badger), son of 

 Thalaba (the she-fox), great-great-grandson of Qoda a, 

 grandson of Saba , the Sheba of Scripture. A single 

 member of the tribe is Kalbi a Kalbite Caninus. 



Such is the system. But can we assign to it historical 

 value ? Is the ancestral dog a real personage or a mere 

 personification of a dog ancestor, the eponym of a tribe 

 which at one time really thought, like the North American 

 Indians, that it was sprung of an animal stock ? That 

 the genealogies of the Arabs, which exhibit the relations 

 of the various tribes and trace them all back to Adam, 

 have been artificially systematised and completed by 

 borrowing from Hebrew and other sources, no one doubts. 

 The shortness of the historical memory of the Arabs has 

 been clearly proved by Noldeke (Uber die Amalekiter, 

 p. 25 seq.), who shows that in Mohammed s time they no 

 longer had any trustworthy traditions of great nations 

 who flourished after the time of Christ. That in many 

 cases gentile unity is ascribed to mere confederations is 

 shown by Sprenger in his Geographic Arabiens. And a 

 conclusive argument against the genealogical system is 

 that it is built on the patriarchal theory. Every nation 

 and every tribe must have an ancestor of the same name 

 from whom kinship is reckoned exclusively in the male 

 line. We know that this system of kinship is not primi 

 tive. According to Strabo (xvi. 4) the Arabs practised 

 Tibetan polyandry (the brothers having one wife in 

 common), of which the levirate customs alluded to in the 

 Qor an (iv. 23) are a relic. 1 The succession from brother 

 to brother, which Strabo mentions as part of the system 

 of marriage and kinship, has left traces even in the Arab 



1 The connection of the levirate with polyandry of the Tibetan type 

 has been shown by Mr. M Lennan, Primitive Marriage, chap. viii. 

 Some of the details of Strabo s account will be noticed below. 



