^{page 861 note 1} The Shilluk People, p. 35Google Scholar

^{page 862 note 1} I am deeply indebted to Professor Westermann for going over the MS. of this article with me, and also for the loan of the MS. of his own article Charakter und Einteilung der Sudansprachen (Africa, 1935), which goes into these points in greater detail and to which the reader is here referred. As further source I have used Dr. Alice Werner's treatise Structure and Belationship of African Languages.Google Scholar

^{page 863 note 1} For the range of ideas covered by these two terms, see Meinhof, , Die Sprachen der Hamiten, p. 22 et seq.Google Scholar Detailed discussion on this language family will also be found in De Lacy Evans O'Leary's Characteristics of the Hamitic Languages. See also Vycichl, Werner, “ Was sind Hamitensprachen ?” (Africa, 1935)Google Scholar; Brockelmann, C., “ Gibt es einen hamitischen Sprachstamm ? ” (Anthropos, 1932)Google Scholar; Cohen, Marcel, “ Les Langues dites Chamitiques ” (Congr. de l'Inst. des Lang, et Civ. afr., 1933).Google Scholar

^{page 864 note 1} Under “ group ” is to be understood here a number of languages and dialects with close grammatical and vocabulary affinities, to the point of complete or partial mutual intelligibility. This loose definition embraces also, of course, the idea of a common original language, whether actual or hypothetical.

^{page 865 note 1} Under a combined grant from the International Institute of African Languages and Cultures and the School of Oriental Studies.

^{page 867 note 1} “ Avukaya ” or “ Abukaya ” is said to be a Zande nickname for the Ojila and Ojigo. The name “ Keliko ” is of obscure origin.

^{page 868 note 1} My questionings of a Moru speaker, whom I had with me while on tour, elicited the fact that, of all the Moru-Madi languages which were foreign to him, he got on best with Abukaya and Madi, but found Logo the most difficult to understand. He himself w as a Moroondri.

^{page 869 note 1} Not to be confused with the Ma'di of Opari district, whose language belongs to the Moru-Madi group.

^{page 870 note 1} Not to be confused with the Shilluk-speaking “ Jur ” (Luo) near Wau. The word “ Jur ” means “ stranger ” in Dinka, and the Dinka apply the term to all foreigners except Europeans and Arabs.

^{page 870 note 2} Of all these tribes the Beli are the most numerous. Seligman mentions another tribe, GBERI, living west of Mvolo, and speaking a dialect akin to Mittu (Pagan Tribes of the Nilotic Sudan, p. 474Google Scholar). I was unable to locate this tribe.

^{page 870 note 3} From information supplied me by Fathers Simoni and Santandrea.

^{page 870 note 4} From information supplied me by P. B. Broadbent and A. J. Arkell, of the Sudan Political Service.

^{page 871 note 1} Documents sur les Langues de l'Oubangui-Chari. I have been able to compare aspects of Bongo grammar with notes taken on Sara by Professor Westermann and Dr. H. J. Melzian, and justify at least the inclusion of that language in the Bongo group.

^{page 871 note 2} Essai de Grammaire de la Langue Baguirmienne. Gaden was able to note only the vocabulary similarity, but comparison of his grammar with my field notes has established the connection beyond all dispute.- It is interesting to note here that, of all the Bongo-Baka languages in the Sudan, Bagirmi most approaches the Morokodo sub-section in grammatical structure; this is strange, seeing that these languages have themselves been largely influenced (in other respects) by Moru-Madi.

^{page 871 note 3} Sammlung und Bearbeitung Central-Afrikanischer Vocabularien. Dr. J. Lukas, who read this article in MS., informs me from personal experience that BULALA, KUKA, and MUDOGO (not mentioned above), are all practically identical and all closely related to Bagirmi.

^{page 871 note 4} The name “ Belanda ” is the cause of much confusion, as it is used to denote two tribes, speaking totally unrelated languages: the Bvrai (also called “ Mvegumba ”) speaking a Ndogo dialect, and the BOB (also called “ Mverodi”) speaking a Shilluk dialect. These two tribes are neighbours, intermarry, and mostly know each other's languages, however. See my article, “ The Tribal Confusion around Wau ” (Sudan Notes and Records, vol.xiv, pt. 1, 1931). The name “Belanda” is of Bongo origin.Google Scholar

^{page 873 note 1} See Lagae “La Langue des Azande” for a full list of Zande sub-sections.

^{page 873 note 2} Note that Poutrin, in Principales Populations de VAfrique Equitoriale Française, considers Kreish a sub-section of the Banda group. I can find no correspondences in the two languages to justify this. Kreish has, if anything, more in common with Bongo-Baka.

^{page 873 note 3} Burssens's estimate of Bangba tribal strength is 50, 000, and he would place them between Dungu and Niangara.

^{page 874 note 1} Kongo-Overzee, Oct. 1934.Google Scholar

^{page 874 note 2} Schebesta, in Les pygmées du Congo belge, remarks considerable vocabulary correspondence between Efe and the Lendu-Logo languages, but considers this due to mutual borrowing.

^{page 874 note 3} Mention is made here of a tribe “ Ndo ” (13, 947); it is grouped with the Keliko and Lendu.

^{page 875 note 1} Ndogo intonation, from a lexical point of view, has been fairly well recorded in Father Ribero's Elementi di Lingua Ndogo, but the grammar rules are yet to be worked out.

^{page 875 note 2} By “ characteristic ” I mean a vowel which harmonizes, according to distinct phonetic rules, with the vowel in the verb stem.

^{page 876 note 1} It is impossible to give here the names of all the sub-sections and clans within the Dinka tribes and sub-tribes, so, as far as possible, I shall confine myself to those sections which are known to speak definite dialects and which give their names to these dialects.

^{page 877 note 1} For those concerned in orthographical problems, it is probable that the Bororthography will cover the Eastern and North-Eastern dialects, and the Rek orthography the Western and Central dialects. Owing to different processes in wordformation, the same orthography will not do for both sections.

^{page 878 note 1} Here again it will be impossible to give the names of all the Nuer sub-tribes and clans.

^{page 878 note 2} The spellings here are taken from the introduction to Father Crazzolara's Nuer Grammar, as the official spellings of these sub-tribes are not to hand.Seligman's arrangement of the most important Nuer tribes, based on Dr. Evans Pritchard's ethnographical field work, is as follows:-Western, Bui, Jagei, Lek, Nuong, Dok; Eastern (i.e. central), Thiang, Lak, Gaweir, Lau; Far Eastern, Gaajok, Gaajak. (See Pagan Tribes of the Nilotic Sudan, p. 207.)Google Scholar

^{page 879 note 1} I am told by Dr. Evans-Pritchard that most of these so-called “ Anuak ”, lying between the Ngok Dinka and Nasser, are really BALAK Dinka.

^{page 879 note 2} Not to be confused with the “ Rumbek Jur ”. See footnote above.

^{page 879 note 3} See footnote on Bviri above.

^{page 880 note 1} This contention is vigorously opposed by Seligman, , Pagan Tribes of the Nilotic Sudan, pp. 415–422Google Scholar, who relies largely on cephalic index, both of the present inhabitants and of the early inhabitants (700 B.C.). Both are found to be mesaticephalic with typical prognathous negroid faces, whereas the Nilotes are dolichocephalic, and in many cases, especially among the Shilluk, “ with long shapely faces, thin lips, noses anything but coarse, and well-modelled foreheads.”

^{page 881 note 1} According to Driberg (“ Hill, Lafon,” Sudan Notes and Records, 1925) the Pari are divided into three sections; the Pugéri (of Shilluk origin), and the Boi and the Kor (of Anuak origin).Google Scholar

^{page 882 note 1} It should be added here that Nuer uses a few suffixes as well in its inflexions.

^{page 882 note 2} Acholi and Lango make no distinction between dental and alveolar consonants while the vowel system in most of the Southern languages is much simpler than in Shilluk and Anuak.

^{page 883 note 1} I am interested to hear that Dr. Ida Ward, in her researches in Nigeria, came lately across a tribe which also distinguishes words by means of voice quality: The ABUA in the Niger Delta. This phenomenon is also suspected in Kalahari. Neither language shows any vocabulary or grammar affinities with the Nilotic languages.

^{page 883 note 2} This phenomenon is also to be found in Hausa, Ful, and Nubian. See also the Nilo-Hamitic languages.

^{page 884 note 1} This, however, would be to ignore the Nubian factor, which probably also plays part. The reader is referred to Murray, 's ” The Nilotic Languages,“ JRAI., 1920.It is interesting to note, in passing, that the undeniable Hamitic element in the Nilotic languages tallies well with Professor and Mrs. Seligman's description of these people as being, both in cranial structure and in culture, strongly hamiticized. (See Pagan Tribes of the Nilotic Sudan, chap, i.)Google Scholar

^{page 885 note 1} The term “ Nilo-Hamitic ” must be taken here in its strictly linguistic sense. Ethnologically and culturally the Shilluk, Nuer, etc., are Nilo-Hamitic, and the Western Bari speakers Sudanic.

^{page 885 note 2} The Rejaf Conference Report mentions 8, 000 Bari living in the Belgian Congo, but I have been unable to find any support for this statement. The Belgian Government statistics, however, give two sections of “ Bari ” (with alternative name “ Bale ”) one, 1, 149, being a sub-section of the Lugwara, and the other, 1, 629, living in Mamvu Monbutu territory. There are also the Bari-Logo, with a similar name.

^{page 886 note 1} Singular: Lotuko.

^{page 887 note 1} Not to be confused with the Shilluk-speaking Lango of Uganda. See above.

^{page 887 note 2} In “ Dialects, Lotuko ”, American Anthropologist, 1932.Google Scholar

^{page 888 note 1} Since there are at least two “ Omo ” rivers in Abyssinia, it should be mentioned here that the river referred to above is that which flows south into the north end of Lake Rudolf.

^{page 889 note 1} Struck, for example, places Suk in the Masai group and Kwafi and Ndorobo in the Nandi-Tatoga group. Other tribes of his which I have been unable to follow up are the Elgumi, speaking Masai, and the Japtuleil, Sabei, and Burkeneji, speaking Nandi dialects. (See Ueber die Sprachen der Tatoga und Irakulente and compare Hollis, , The Masai, Oxford, 1905.Google Scholar)

^{page 890 note 1} The Nandi, their Language and Folklore, Oxford, 1909.Google Scholar

^{page 890 note 2} Probably another name for the Mutei, as the localities of these two alleged tribes seem to coincide.

^{page 890 note 3} The Suk, their Language and Folklore, Oxford, 1911.Google Scholar

^{page 891 note 1} In the Western Bari dialects, especially Pojulu and Kakwa, the kw and gw and nw sounds of Bari have been “ Sudanized ” to kp and gb and nm.

^{page 892 note 1} Note that in Masai and Nandi there is no passive form proper, but a form derived from the third person active, after the manner of Sudanic languages.

^{page 892 note 2} In some locative expressions the old Sudanic construction is to be met. Thus in Bari, i mukōk na mere (at the foot of the mountain) may also be rendered i mere mukōk In Nandi: mi tulwet nony.

^{page 893 note 1} It is possible that these people are Driberg, 's “ Lokathan ” in which case their language belongs to the Lotuko group. See p. 887.Google Scholar

^{page 894 note 1} “ The Didinga Language,” MSOS., 1931.Google Scholar

^{page 89 note 1} This conclusion was arrived at after seeing the MS. notes on this language, compiled by Mr. Shackleton of the Kenya Political Service.

^{page 896 note 2} See Murray, G.W., “ The Nubian and Bari Languages Compared,” Sudan Notes and Records 1920.Google Scholar