elements; but if the same should be done in the Akém dialect, it would not sound well." As to the position which the Chwee language holds among the African languages in general, there is no doubt that it belongs to that stock of languages which have been termed "Prefix-pronominal," the principal characteristic of which is the employment of prefixes in the formation, classification, and pluralisation of nouns, and in representing them (before other parts of speech).

For further information on these points let me refer those interested to A Grammar of the Asanté and Fanté Language, called Tshi (Chwee), based on the Aquapem Dialect, with reference to the other (Akán and Fanté) Dialects.' By the Rev. J. G. Christaller, of the Basle German Evangelical Mission on the Gold Coast, published at Basle last year (1875).

XI.-Notes of a Journey from the River St. Francisco to the River Tocantins and to the City of Maranhão. By JAMES W. WELLS, C.E.

On the 15th of February, 1873, I started from Rio de Janeiro, in company with other members of a staff of engineers, for the purpose of surveying the ground for the prolongation of the Dom Pedro Segundo Railway by the rivers Parãopeba and St. Francisco as far as the Cachoeira (cataract) of Pirapora, on the last-named river. From this point the navigation of the river St. Francisco is open and uninterrupted as far as the noble waterfall of Paulo Affonso.

On the 5th of January, 1875, the surveys being completed, I bade adieu to my companions at Pirapora, and started on my journey to Cidade da Barra, 740 miles by river from Pirapora. I went by land at first; passing the mouth of the Rio das Velhas, through the villages of Coração de Jesus and Contendas, on the east side of the river, to the city of Januaria, where I crossed the river and proceeded by the river-side road, on the west bank, to the town of Carunhanha. Here, finding the roads very much flooded, I sold the mules and embarked in an ajoujo (i.e., two canoes lashed together and planked over): in this crazy concern I proceeded to the Cidade da Barra do Rio Grande, situated on the mouth of the river of that name. suffered intensely on this river-journey at night from the attacks of mosquitoes. As all the country to this point has been travelled over and described by various travellers, I shall avoid any description, and proceed at once with the sketch of the

I

journey to the Tocantins and thence to the sea. On the 26th of February, 1875, after a 10-days' stay with the hospitable people of Cidade da Barra, I engaged a passage in a barca returning up the Rio Grande. At the same time I engaged a troop of mules to meet me at the mouth of the Rio Preto, 56 miles up the Rio Grande.

The Rio Grande is one of the most important affluents of the St. Francisco-important not only for its size, but for the combination of navigable streams that flow into it. It is navigable for vessels of 3 to 4 feet draught, as far as the town of Campo Largo, 180 miles from its mouth. Its average width is 318 feet, depth 10 feet, and its velocity 2 feet 4 inches per second. All this length is perfectly free from obstacles, and for yet 80 miles between Campo Largo to Limoeiro; but this last length is troublesome on account of the rapids at some bends of the river. Its chief affluents are Rio Preto, Rio Branco, and Rio das Ondas, all more or less navigable.

I entered the barca (a sort of barge with a commodious cabin built on the deck at the stern) and the polemen, four in number, pushed off.

As the river was very full, our progress was slow. At times the polemen could not reach the bottom with their poles, and in an instant we drifted down stream, soon losing an hour's hard work.

Soon after leaving the town the pilot directed the barca into a lagoa, or overflow from the river, in the adjoining low-lands. Here the progress was better, but many and many times a passage had to be cut through the intervening bushes. It was a slow and wearisome progress, but, nevertheless, better than navigating against the strong current and deep water of the main river.

I found the greater part of the adjoining lands to be very low and flat, and covered with a short, arid grass and a scrubby vegetation. In some places the inundations extend for many miles inland. In these lakes are numerous water-fowl, spoonbills (Culhereiros); and a magnificent white crane, 4 feet high, with a white body and black head and neck, locally termed Jaburu moleque. There were also many other varieties of the crane family, with two species of wild duck.

At sunset, the barca was hauled alongside the sandy bank and secured for the night. Fires then lighted, and dinner prepared and eaten, in a few moments the tired men were asleep, to be away again with the first light of dawn. The nights, though warm, were not disagreeably so, the thermometer registering 76° to 78°, and 82° to 84° at midday in the shade. The remarkable absence of mosquitoes is a notable fact in this river.

Feb. 28th. The morning broke bright and clear; our progress was miserably slow, the stream being very deep, and nothing was to be seen of the banks but dense bush. At 11 A.M. we reached Estreita da Serra, a narrow part of the river where it passes through a gap in a range of low hills that intersect the line of the river at this point. Looking out of the cabin window, my eye rested first on the long belt of mauvecoloured water-lilies (golfoes) skirting the margin of the stream, then rose to the thick, scrubby bushes of the banks; above which, a little inland, rear up, in graceful prominence, a few Carnahuba palms, their spiral stems and feathery foliage standing out in relief against the sombre forest-clad point of the serra behind. The Carnahuba palm (Copernicia cerifera) is one of the most useful trees in Brazil. In these districts it exists only in small numbers, but in the provinces of Ceará and Rio Grande do Norte it grows uncultivated in great luxuriance. Perhaps there is no region of the globe where a tree can be found of such varied uses and so serviceable as this palm. It resists the most severe and longest droughts, keeping always green and flourishing. The roots possess the same medicinal properties as sarsaparilla. From the trunk are extracted strong, light fibres capable of receiving a high polish. The wood is used for props, joists, and other building purposes; as also for stakes and fences, and for musical instruments, tubes and pumps. The inner rind of the young leaf, when fresh, is used as a highly-esteemed and most nutritive food. The tree also affords wine, vinegar, a saccharine substance, and a great quantity of gum like sago, and possessing the same properties and taste. It has often been the only food of the inhabitants of Ceará and Rio Grande do Norte in times of severe drought.

A species of flour, like maizena, is also extracted from it, and a whitish liquor like the liquid contained in coco-nuts. The soft, fibrous substance in the interior of the stalk, and that of the leaves, is a perfect substitute for cork. The pulp of the fruit is agreeable to taste; and the kernel, which is very oily and emulsive, after being roasted and pounded, is used as an excellent substitute for coffee. Of the dried leaves are made mats, hats, baskets, and brooms; and the straw is already sent to Europe in great quantity to be made into fine hats, some of which are sent back to Brazil. From the leaves is extracted a kind of wax, much used for making candles, which are extensively consumed in the northern provinces.

March 3rd. We arrived at Boquerão, a small collection of houses at the foot of the Serra do Boquerão, near the mouth. of the Rio Preto. Here I left the barca, and travelled by land up the Rio Preto (Black River).