El Masoudi, who has been called the Herodotus of Arabia, gives us still further details about the race, speaking of Sofala as a place to which the Arabs of his time went habitually to obtain gold and precious stones from the natives. He is more explicit about the descent from the north of the Zindj tribes, which took place not long before his day; and unless there was a previous wave of barbarians, concerning whom we have no account, it may be supposed that it was owing to their advent that the gold settlements up country were finally abandoned, and the Arab traders restricted to the coast. Describing the natives of the land behind Sofala, he speaks of them as negroes naked except for panther skins; they filed their teeth and were cannibals; they fought with long lances, and had ambuscades for game. They hunted for elephants, but never used for their own purposes the ivory or gold in which their country abounded. From this picture it is easy to see that in those days the inhabitants were just as they are now, an uncultured wild race of savages. We get another testimony to this in the voyage of two Arabs who went to China in 851 A.D., and returned by the east coast of Africa. M. Renaudot has translated their experiences, in which they describe the Zindj as follows: ‘Among them are preachers who harangue them, clad in a leopard skin. One of these men, with a staff in his hand, shall present himself before them, and having gathered a multitude of people about him, preach all the day to them. He speaks of God and recites the actions of their countrymen who are gone before them.’ In this account we easily recognise the witch-doctor and ancestor worship, the Mozimos and Muali of the present race. Abou Zeyd’s evidence is also to the same effect. He thus speaks of the Zindj: ‘Religious discourses are pronounced before this people, and one never finds elsewhere such constant preachers. There are men devoted to this life who cover themselves with panther and monkey skins. They have a staff in their hands, and go from place to place.’ Quite an accurate description of the South African witch-doctor. Consequently, from this mass of evidence we may affirm with absolute certainty that for a thousand years at least there has been no change in the condition of this country and its inhabitants. Further testimony to the same effect is given us by Edrisi in his geography, who alludes to the Zendj tribes as inhabiting this country, and occupying the coast towns Dendema and Siorma, ‘which latter is situated on a gulf where foreign vessels come to anchor.’ He speaks, too, of the iron trade which the Zendj carried on with the Indians, and of the abundance of gold in the mountains behind Sofala, adding, ‘nevertheless, the inhabitants prefer brass, making their ornaments of the latter metal.’