as may prove to be suited to an African country, a coast belt inhabited by negroes professing Christianity and wearing clothes of European cut, and a hinterland of Mohammedans dressed in the picturesque and wholly suitable costume worn at the present day by the Mandingos and by most Mohammedan negroes between Senegal and the White Nile.

The native races of Liberia, the languages they speak, and the religions they profess may, to a certain extent, be grouped under two classes the Mandingo on the one hand and the Kru negro on the other. I am aware, of course, that the Mandingo type is a very variable one physically, according to the less or greater degree of Caucasian blood which permeates its negro stock, and also that the Kru man proper is confined in his distribution to a small portion of the southern coast of Liberia. But each of these types is sufficiently representative to serve as a general illustration of the two classes of Liberian peoples. Associated with the Mandingos, to a great extent in language, in Mohammedan religion, and in the adoption of the Arab dress, are the Vai and the Gora of western Liberia, and to some extent the Buzi or Kimbuza. All the remaining tribes are more or less related to the Kru stock in language, appearance, physique, customs, and the profession of a pagan and fetishistic religion, similar in general features to the fetish religions of all western and west central Africa, with some points of resemblance to the Bantu beliefs in the southern half of Africa. The Gora language of western Liberia is rather a puzzle in classification, and it must be admitted that it only offers the slightest affinity to the Mandingo group and an equally slender connection with the Kru family. In a still more generalized way it may be said that there are distant resemblances between the languages of the Kru and Mandingo stocks; nor can these slight resemblances be altogether explained by the mere imposition of linguistic influences. The Mandingos, who are destined to play a most important part in the development of Liberia and of much of West Africa, are nothing but a varying degree of cross between the Fula race of the West African park lands and the ordinary West African negro. This crossing and the founding of this group of people-the correct pronunciation of whose name seems to be Mading'a, or Manding'a-may have been a relatively ancient one. The Mandingos, in fact, are the Swahilis of West Africa and offer a striking resemblance in face to the average type of Swahili porter that one meets with on the Zanzibar coast. They also often resemble natives of Uganda, and for the same reason that a good deal of the population of Uganda is infused with a slight Hamitic element derived from the Hima aristocracy. Several Mandingos whom I met at Monrovia were able to speak Arabic in a halting fashion. The Arab words they understood best were those pronounced in the

North African dialect. It seems to me that in some cases the Caucasian element in the Mandingos was derived from direct intermixture of Berbers and Arabs from North Africa with the negroes of the Upper Niger. I doubt if any pure-blooded Fula people extend their range into the northern limits of Liberia; but they have had an undoubted influence in times past over the development of the park land which lies beyond the forest. By their minglings with the indigenous negroes of the Sudanese and West African type they have created the Mandingo peoples and have also carried Mohammedan civilization and tenets into that part of Africa, as well as, no doubt, the Sudanese breeds of cattle and sheep. The domestic sheep of all the forest region of Liberia is that common to the other forested parts of West Africa-the sheep with erect ears, fairly well-developed horns, small size, black and white coloring, a tail without any fatty development, and a long throat mane in the male. Far back in the interior of Liberia I am informed that the Mandingo sheep are similar to those of the Sudan, with fat tails and without the throat mane.

It should hardly come within the limits of the present paper to discuss one of the most interesting problems in Africa-the origin of the Fula race. Personally I am still disposed toward the old theory that the Fulas were an early cross between the Libyans of North Africa and the negroes of Senegal, a cross in which the Caucasian element predominated considerably. They certainly offer marked resemblances, however, to the Hamitic aristocracy of the Upper Nile and the lake regions. Their language is a complete puzzle. At present it can not be said to offer affinities of a marked kind to any group of negro speech, but it is emphatically a negro tongue (with a faint suggestion here and there of the Bantu family), and not in any way influenced by Hamitic, Libyan, or Semitic characteristics. As to the Libyan affinities of Hausa there can be no doubt, but nothing of the kind has as yet been discerned in the structure or vocabulary of the Fulfulde. It even seems to offer less resemblances in structure to the Hamitic language family, for example, than can be discerned in the Bantu.

Of all the peoples in Liberia affiliated with the Kru stock perhaps the most numerous group is that of the Kru, which occupies the coast of Liberia between the French frontier at the Kavalli River and the river Sestos. With the Kru I associate the Grebo, as the two peoples differ but little in language and scarcely at all in physical type. The Grebos are inclining strongly toward Christianity, but very few, if any, converts to that religion have been made among the Kru people proper who inhabit the coast between Greenville and Garraway. Krus and Grebos together number something like 375,000. The next most important group of people, as regards numbers, are the Mandingos, of whom there are perhaps 300,000 within the limits of Libe

ria. After them may be ranked the Kpwesi people, a general term for a congeries of tribes speaking dialects of a common language. These Kpwesi (familiarly known by the Americo-Liberians as Pessi, or Pessa) may be as many as 250,000. In an appendix to this paper the rest of the tribes and their approximate numbers are enumerated. In the central parts of Liberia, within the limits of the forest, there is no doubt that cannibalism prevails. This is a very marked feature in the life of the Beila, or Bele. These people are said to relish most keenly the hands and feet, and this very dainty dish is usually set before a king or chief alone. Nowhere in Liberia have I noticed-nor has any explorer encountered or reported—any race of negroes wholly naked, either among men or women, such as are so commonly met with in eastern Equatorial Africa, or until a few years ago in parts of South Central Africa. A certain degree of complete nudity in unmarried women was at one time quite a common feature of the natives of the Niger delta, the Cross River, and the Kamerun, while on the upper Cross River complete nudity among the men was just beginning to disappear twenty years ago. Throughout Liberia no one has ever observed complete nudity among either men or women. Though there are a few rare exceptions to this rule, it may generally be observed that the marked feature of male nudity so characteristic of the Upper Nile, the eastern equatorial regions, and originally of the north end of Lake Nyassa and central Zambezia, is never met with in the forested regions of Africa, except possibly here and there among the Pigmies. Throughout the Kongo basin and countries as completely savage as the innermost parts of Liberia, the men wear a minimum of clothing, which is a concession to ideas of decency, and which, when the race is quite out of touch with the trade of the outer world, is generally a strip of bast (bark cloth) from a fig tree. I have not observed any of the savages from the interior of Liberia wearing dressed skins. I am told that so greedy are they after food that when any beast is killed the hide is roasted and eaten. On the other hand, the civilized Mandingos of the north have learned from the Fulas or from the Moors, or possibly from both, the most beautiful work in leather.

Nowhere along the coast of Liberia is there a harbor in the sense of the bay at Sierra Leone. All the anchorages, in fact, are open roadsteads. But on the other hand, this is not a particularly dangerous condition for ships, as the south wind never blows strongly enough to raise a big sea, while the north wind, coming from off the land, can only affect the Atlantic at some distance from the shore. But of course this portless condition adds very much to the discomfort of dealings with Liberia. Although the swell from the choppy surface raised by the wind may not be sufficiently serious to affect big vessels lying at anchor, it is not at all nice for small boats or steam launches, and generally during the rainy season of the year transfer

ence from the big steamer to the shore-going boat has to be effected by means of a crane and a cradle. Still more to be dreaded on some points on the coast is the landing or the going off on account of the surf. Perhaps the best approach to a sheltered harbor which exists is at the capital, Monrovia. Here there is a bar to a small river or creek which communicates with the St. Pauls River. The bar is very seldom rough, being to a certain extent sheltered by a promontory, and once across the bar you can land quite comfortably on the beach in perfectly smooth water. But at places like Grand Basá, Sinó, and Cape Palmas, the landing can be extremely dangerous and disagreeable. One seldom arrives at or departs from these places without, at any rate, a wetting from the rollers that break over the stern or bow of the boat. However, real accidents to persons or property are, it must be confessed, of rare occurrence, and the whole question depends very much on the good or ill will of the Kru boys who direct the steering.

The Kavalli River, though probably less in volume than the St. Pauls, is the most navigable as a means of access to the interior. Unfortunately there is a very bad bar at the mouth of the Kavalli or it would have played a very different part in the history of West Africa. Once across the bar, a steam launch or a rowing boat can ascend the river for about 80 miles, when the first rapids are reached. The upper Kavalli is imperfectly known, and, indeed, its extreme upper course is still a matter of conjecture. On the definition of its course depends to a great extent the laying down of the northeastern frontier of Liberia according to the French treaty of 1892.

The St. Pauls River is only navigable as far as a place called White Plains, about 20 miles from the mouth. There is no difficulty about the bar at the mouth of the St. Pauls, for this reason, that one can land with little or no risk at Monrovia; and once inside the Mesurado River (which is merely a creek), one can embark on a small river steamer, of which there are two or three in existence, enter the St. Pauls River by the Stockton Creek, and thence ascend the St. Pauls to the first rapids near the settlement known as White Plains. All the lower part of the St. Pauls River is pretty thick with Americo-Liberian settlements, some of them of a distinctly prosperous and prepossessing appearance. Several of the officials at Monrovia have country houses on the banks of the St. Pauls amid charming surroundings. This great virtue may be attributed to the AmericoLiberians, that they certainly know how to build houses, comely in appearance, sanitary, and lasting, or as reasonably durable as one can expect in a land of heat and moisture. In fact, the leading characteristics of the Americo-Liberians are their love of building and their remarkable politeness.

For some distance above its first rapids the St. Pauls River is scarcely navigable for canoes; but in the little-known region north of Dobli Zulu Island, it is said to be navigable for canoes up to its junction with the Tuma, which is also stated to be an important

stream.

Eastward of Monrovia, the Mesurado River or Creek reaches, as a navigable piece of water, to within a very short distance of the Junk Creek, which is a branch of the Dukwia River; so that, except for a little isthmus of sandy soil, Monrovia might be situated at the western extremity of a long island. With developments that might come later on, it would be very easy to cut a canal to join the Mesurado and the Junk rivers. This would then give access for vessels of light draft to the Dukwia and Farmington rivers. These streams debouch at a settlement of some importance named Marshall; but the bar at Marshall (mouth of the river Junk) is extremely bad and dangerous. As a matter of fact, with a very little expenditure of money Monrovia might be made a good port.

The population of Monrovia, so far as Americo-Liberians and foreigners are concerned, is about 2,500. It has a system of telephones which connects it with the settlements on the St. Pauls River. This has been set up and is well worked by a native Liberian. It is, perhaps, hardly necessary to remind you that Monrovia was named after the celebrated President of the United States, Monroe, who is responsible for that doctrine which inhibits any European nation from further conquests in the New World. The capital of Liberia is divided into two parts, the low and shoreward section being given over to large settlements of Kru boys and indigenous negroes, while the upper part of the town is inhabited by Americo-Liberians and European consuls, traders, etc. This civilized part of the town is composed of broad, grass-grown streets, and substantial, well-built, comely looking houses, churches, offices, and public buildings. The smart appearance of the houses, in fact, is in somewhat striking contrast to the neglected condition of the roads. These have never been made, and are simply the unleveled rock of more or less flat surface. Consequently at the present time they are absolutely unsuited to any vehicle, though I have seen an enterprising Liberian negotiate them with a bicycle. They are generally covered with a very short, close turf of thickly growing plants, which is kept in the condition of turf by the constant nibbling of the pretty little cattle that frequent the streets of Monrovia. A less agreeable feature are the pigs, which exist in great numbers and perform the office of scavengers. The appearance of most of the houses is, as I have said, either very striking or comely both in form and color, and this appearance is enhanced by the beautiful clumps of trees and the gardens which surround most of the houses. The Americo-Liberians seem to be