3i8 A TROPICAL DEPENDENCY and the coast, extended at one period to the north of Songhay and Haussaland. It has been related in the early myths of Gao, of Kano, of Daura, that the killing of the fetish and the substitution of a higher form of religion was the beginning of their recorded history. This higher form of paganism, which would appear to have been derived from sources similar to those that furnished the religions of Phoenicia, Egypt, and Arabia, seems gradually, at some very early period, to have pushed fetishism southward before it, and to have held the ground to the north until, in the four hundred years between the tenth and the fourteenth centuries, Mohammedanism was generally accepted along the northern border. The higher form of paganism then suffered the fate which it had itself inflicted upon fetishism. It was driven south, and in turn drove fetishism farther before it, until, as in the present day, the religions of West Africa could almost be defined by latitudinal lines. If, following the opinion quoted in the earlier portion of this book from a distinguished French authority, we take io° as the most southerly limit of Mohammedanism, and give 7°, as I think we may be justified in doing, as the farthest extension northwards of fetishism, we get three degrees, from 7° to 10°, in which the higher forms of paganism may be held still to prevail. As a matter of fact, these latitudinal divisions will be proved to be too arbitrary. M. de Lauture wrote upwards of half a century ago, and Mohammedan influence has since his day extended farther south. On the other hand, the higher paganism also extends in places farther north than the limit which he assigned. In the Nigerian Provinces of Zaria, Bautchi, and Yola, through which the tenth parallel extremity passes, but which extend to and even beyond 11°, many tribes of pagans still exist. And I have little doubt that the same observation would hold good in French territory farther west, with which I am unacquainted.

The general drift, however, of the observation is, I think, sound, and it is with this that we are for the