I was once gravely informed by an Abessinian slave-merchant of the market of Baso in Godjam, that beyond Kaffa there is a country, the male inhabitants of which are all dogs, and the females are women; and that the dogs go out to tend the cattle, while the women occupy themselves with domestic affairs. It was of little avail to inquire how it came to pass that the progeny of this strange union should be canine on the male side, and human on the female. That my informant did not know the other he did know; though he honestly admitted that he had not been so far as to have seen it himself. This story I consider to have originated in the fact, that beyond Kaffa there is truly a "Dog" country, just as, adjoining to Djimma there is a "Monkey" country; that is to say, in Worátta there is a place or district named Ushá,* which word in Amharic means dog," in the same way that Zendjero means "monkey." As is usual in such cases, the story was afterwards invented to account for the name.† *See my map in Journ. Roy. Geogr. Soc., vol. xiii. According to M. Werne (Expedition zur Entdeckung der Quellen des Weissen Nil, p. 325), a fable prevails among the natives of the valley of the White River, respecting a race of cannibals, having heads like dogs and going on all-fours, who are said to inhabit the mountains of Logayà, to the east of Bari. The following reasonable explanation of this monstrous story was, however, given to that traveller by Lákono, the intelligent king of the giant race of Bari. He stated that, "in reality, these wicked people have heads like those of other human beings; only they keep in all their teeth. [It has been remarked (p. 273, note) that the negro inhabitants of the valley of the Nile extract the incisors, in order that they may not resemble wild beasts]; and when they come to eat up others, they creep in on all-fours." On this M. Werne himself remarks, that “ most likely the simple meaning of this is, that these alleged cannibals do not engage in open war with their neighbours, but sneak in among them like dogs, and carry away individuals, whom perhaps they may devour." For myself, I question much the existence of cannibalism among these mountaineers, whom the mere fact of their not extracting their teeth proves to be of less barbarous habits than the natives of the low country. From the earliest times cannibalism has been said to prevail among the inhabitants of Africa. It is only necessary to allude to the A'Díoπes àvdgwτopάyos of Ptolemy, and to the Nyam-Nyam, Lem-Lem, Dum-a-Dum, &c., of the Arabian geographers and of the modern Arabs. But stories of this kind require indisputable evidence to establish their truth; and there is no doubt that they often originate in ignorance, if not in interested motives, on the part of their promulgators. Of the language of these Dokos M. d'Abbadie states that he possesses only twenty-nine words; which is rather surprising, when we consider that he says he had one of these people by him as an interpreter for nearly two years. Its affinity with the language of Woratta is, however, asserted by that traveller.* This will place it in the Gonga class, which is quite in accordance with its geographical position. The only word of the language cited by M. d'Abbadie is elmos, signifying "bread;" which expression, however, I cannot connect with any word in my vocabularies. Dr Krapf states,† that "the language of the Dokos is a kind of murmuring, which is understood by no one but themselves and their hunters." But he also mentions, that Mr Robertson, in his Notes on Africa (London, 1819), pp. 353–354, when treating of the natives of the south bank of the river Congo, says: "The opinion that these, or any other people on this [the west] coast, are Anthropophagi, is ridiculous. One of the traders at Bonny did, indeed, once tell me so plausible a story concerning the Quas eating those who fell into their power, that I was half inclined to believe him. But subsequent inquiry convinced me that there was no truth whatever in it. The Portuguese having taken so much trouble to impress other nations with the dreadful effects of man-eating, they, of course, consider this country of some value; but from their not having been devoured themselves, it seems other animal food is plentiful; or, perhaps, the aboriginal inhabitants think the Portuguese rather coarse meat, and like cattle or fish better." reason. A similar tale of cannibalism is related in Shoa, and for a somewhat similar The wilderness of Gédem, a province in the north of that kingdom, which I visited in April 1841, is "the place of refuge for all the fugitives and bad characters of Shoa" (Journ. Roy. Geogr. Soc., vol. xii. p. 92). The Dankáli slave-merchants trading between Shoa and the coast take care to impress on the minds of their slaves that the people of Gédem are cannibals, who will be sure to eat them up if they run away. Hence, the poor children are content to remain with their (often cruel) masters, rather than run the chance of being devoured by the wicked man-eaters of Gédem! To shew the little value, generally, of these stories of the existence of cannibalism among unknown nations, I may cite, further, the following passage from the interesting Notes sur les Maurs, Coutumes et Traditions des Amakoua of M. Eugène de Froberville, published in the Bulletin de la Société de Géographie, 3d Ser., vol. viii. p. 324 :-" Les traditions les plus effrayantes sont répandues par toute l'Afrique orientale touchant le sort des esclaves transportés dans le pays des blancs. Mes informateurs m'ont unanimement déclaré que l'on croyait généralement que les blancs mangeaient les esclaves qu'ils viennent chercher en Afrique."-Mutato nomine, de te fabula narratur !-22d November 1848. * Bulletin, 2d Ser., vol. xix. p. 439. ↑ Prichard, Natural History of Man, p. 556. 66 they address the Supreme Being by crying, "Yer! Yer!" Now Yero, in the Kaffa language, means God;" so that we have here a further proof that the language of these Dokos belongs to the Gonga class. In commenting, on a former occasion, on a few words given by Dr Krapf, apparently from the mouth of Dilbo, as specimens of the language of Kaffa, I remarked,* that "they do not altogether agree with my Kaffa words, which I obtained from persons who were most assuredly natives of Bonga, the capital of that country. From Dilbo's physical appearance and other circumstances, I have reason to believe that he was a native, not of Kaffa itself, but of some neighbouring country, which will account for the difference of language." I may now add, that the description given by M. d'Abbadie of the physical character of the Dokos corresponds so closely with that of Dilbo, that I am inclined to think he was himself one of those savages. Should this really be the case, the words inserted between brackets in my Kaffa vocabulary will most probably belong to some Doko dialect. The next of M. d'Abbadie's unplaced languages is that of Yumbo, which is the name given by that traveller to the negro inhabitants of certain islands formed by the Bahr el Abyad, or the direct stream of the Nile. These islands are placed by him as high up the stream as about 6° N. lat. ; but, as has been shown by me in a communication recently made to the Geographical Society of Paris,† their true position is in about 9° N. lat., below the confluence of the Sobát or River of Habesh. Thus, these Yambos appear to be Denka negroes, and their language will consequently belong to the Nubian class (XIV.). The Bárea is said by M. d'Abbadie to be "spoken by the semi-negroes near the Takkazie ;" that is to say, the “Shánkalas" of that river; so that this language is identical with either the Bárea of Salt (XXIII.), or with the Dalla (XXII.) And lastly, we have the Konfal, who are stated to "live between Kwara and the Awawa"-that is to say, the Agha *Proceedings of the Philological Society, vol. ii. p. 94, n. ghá or Agaus of Agaumider. Of this language M. d'Abbadie says, that he has "no sample beyond the first ten numbers, which are partly Giis ;" and he adds that "the all-but-unknown Kónfal tribes are the most perfect medium between the straight-nosed Ethiopian and the grovelling negro." But the position thus attributed to the Kónfal corresponds precisely with that of the black inhabitants of Gindjar already mentioned; and when M. d'Abbadie asserts that their numerals are partly Geez, i. e., Ethiopic, he should rather have said Arabic; since the fact is beyond dispute that the language of Gindjar is, in great part, a corrupt Arabic, and it is not less a fact that the Ethiopic and Arabic numerals are almost identical. The conclusion to be drawn from this investigation is, that Dr Latham's list, whatever modifications increased information may eventually give occasion to introduce into its arrangement, is, in fact, exhaustive of the languages of Abessinia and the countries immediately adjoining. The map in which the results thus arrived at have been embodied, makes no pretensions to strict accuracy in the limits that are assigned in it to the several classes of languages. Our materials are still too imperfect to admit of precision in this respect. The only merit that this map can lay claim to, is that of affording a general coup-d'œil of the geographical distribution of the various languages spoken in that portion of Africa which has more immediately fallen within my cognizance; and thus of obviating, so far as Abessinia and the neighbouring countries are concerned, the difficulty which, as the author of the Report justly complains, so frequently arises from the absence of any geographical notice respecting the districts within which a particular language is spoken. LONDON, 31st May 1848. Instructions for Collecting and Preserving Invertebrate Animals. By RICHARD OWEN, F.R.S., Hunterian Professor to the Royal College of Surgeons of England. As water is the element in which the greater number of the classes of animals exist, and as the sea is the scene of such existence, and the field of research which will be most commonly presented to those for whom the following instructions for collecting and preserving animals have been drawn up, they will commence with the marine species, and the lowest forms of animal life. Alga, Sponges, Corallines, and Corals. The line of demarcation between the vegetable and animal kingdoms is so obscurely marked in the lowly organised marine species, and the modes of collecting and preserving these are so similar, that the kindred groups above named are associated together as the subjects of the following remarks. Algæ, commonly called sea-weeds, may be divided, for the convenience of the collector, into three kinds, according to their colour : 1. Olive-coloured (Fuci), generally of large size and leathery texture, rarely gelatinous; usually laminate or leafy, rarely filamentous or thready. 2. Red-coloured (Floridea), firm, fleshy, or gelatinous; usually filamentous, sometimes membranaceous. 3. Green (Chlorosperms), membranaceous or filamentous; rarely horny. Sponges are bodies usually adherent in irregular or amorphous masses, rarely in the form of hollow reticulate cones; composed of a soft, jelly-like tissue, supported by siliceous or calcareous spiculæ, or by horny filaments. They are divided, accordingly, into horny, or keratose," siliceous and "calcareous" sponges. Their soft, organic substance is commonly diffluent, and drops from the firmer basis, when removed from the water, or it is easily washed away. It exhibits no signs of sensibility; no contraction or retraction when touched or otherwise stimulated. The evidence of life is afforded, as in the corallines and algæ, by the flow of currents of water through canals, entering by pores, and in the sponges escaping by larger orifices; and an appearance of animal life is given to both algae and sponges by the locomotion of the sporules or gemmules. Corallines are plants coated with a calcareous covering, either red or green when fresh, becoming white and brittle on exposure to the air. Corals, though called "zoophytes," are true animals; the cur |