are, in fact, mere collections of undigested notes—adds to the difficulty of grasping the author's reasoning. Indeed, I am by no means sure that I have grasped it; but his reasoning is to be carefully distinguished from his facts, and these—so far as apparent from his by no means lucid presentation—are certainly valuable. His main contention, we gather, is that "concurrent with fetishism or Jujuism, there is in Africa a religion giving us a much higher conception of God than is generally acknowledged by writers on African modes of thought." This religion, to which Mr. Dennett gives the name of Nkicism, has, he thinks, been overlaid by Ndongoism (equivalent to what is usually understood by witchcraft and fetishism), and in great part forgotten. It was handed down in connection with certain formulae embodying what may be called a system of philosophy, theology, and ethics, which were taught to the people by their kings (Maloango) in the sacred groves. Much of this traditional law has evidently been lost, and the kingly office itself has fallen from its ancient estate. Maloango, the paramount chief of the Bavili tribes in Loango, was once tributary, along with Kakongo, to the Ntotela, or King of Kongo (Sao Salvador), but has been virtually independent for the last three hundred years—at least till the French took over the country in 1883. The present chief, however, was never officially crowned,—a fact which, for various reasons, is to be deplored. But the bearing of these matters on native administration—not the least important of the issues raised by Mr. Dennett—does not come within the scope of this notice.
As we shall have occasion to point out later on, Mr. Dennett has failed to employ his excellent linguistic knowledge to the best advantage, for want of such acquaintance with other Bantu languages as would have enabled him to employ the comparative method. In the same way, one fancies, he does not perceive the real bearing and connection of many of his facts; which, indeed, I must confess, left me in a state of helpless bewilderment, till, fortunately, I lit on the clue supplied by M. Van Gennep, viz. MM. Durkheim and Mauss' article in L'Anne Sociologique for 1903. This places the Bavili "Categories" (p. 108) in their true light as elementary attempts at classifying
