that Tibesti, naturally another part of Tripoli's Hinterland, was assigned to France. The work contains some useful maps, by Bartholomew, of the Nile basin. In that showing political data, the whole British and Egyptian territory from Alexandria to Mombasa is divided into four areas, according to the four grades of British domination spoken of in the text. The reason for the severance from the Sudan by the parallel passing through Fashoda, of the Bahr-el-Ghazal, etc., as a "British sphere of influence," is not evident, as this is in accordance neither with the terms of the Anglo-Egyptian agreement of January, 1899, nor with the fact that Egyptian stations have already been re-established south of that line. The map showing zones of vegetation is misleading in one or two points. Thus Uganda, with its abundant vegetation, is shown of the same tint as the arid plains of Somaliland, while the coast lands of British East Africa are supposed to be covered with dense forests comparable to those of the Congo basin.

Reorganization of French West Africa.-By a decree of the French Colonial Minister, dated October 17, 1899, the French possessions in West Africa have been placed under a new organization, which considerably modifies the distribution of territory between the various provinces. The reason given is the difficulty which has been experienced in controlling the operations in the far interior-on the Middle Niger-from so distant a centre as Kayes, the capital of Senegal. The Middle Niger region, including the greater part of the area within the bend of the river, is therefore placed under a separate military administration, while the western portion of the old Sudan province is united with that of Senegal to form a single administrative division, which reaches eastward to Lake Debu. French Guinea, the Ivory Coast, and Dahome remain, broadly speaking, within the same limits as formerly, the Ivory Coast extending roughly to 10° N., and Dahome a little north of the latitude of Say. The approximate boundaries of the new divisions are shown on the opposite map, which also illustrates the territorial arrangements between Great Britain and Germany in the same region. They are taken from a map published in the Bulletin du Comité de l'Afrique Française, and reproduced in the Politique Coloniale for November 14.

Baron von Grünau's Visit to Siwa.-During the winter of 1898-99 a visit to the Siwa oasis was made by Baron von Grünau, who gives a short account of his journey in the Zeitschrift of the Berlin Geographical Society (1899, No. 3). Between the Natron lakes and Moghara the German traveller seems to have taken a route not often followed across the plateau, by which the length of the journey was reduced by a day. The ground here entailed a succession of ascents and descents over a rocky surface, but the differences in elevation were but slight. Baron von Grünau made a compass survey of the route, and was able to correct some slight errors in the maps. He also carried out a regular series of meteorological observations, as well as determinations of the temperature of the water of wells. In Siwa he took measurements of some of the ruins, but does not seem to have added much to the information collected by Mr. Silva White. He seems to have been incorrectly informed that Mr. White did not visit Aghormi, of which a photograph is given in that traveller's book. Possibly, however, the particular ruins which he claims to have discovered were not seen by Mr. White. The temperature of the famous "fountain of the sun," near the temple of Jupiter Ammon, was repeatedly taken, at various hours of the day, and the same result (31° C. or 87.8° Fahr.) obtained in each case. The popular idea that it is cooler at midday and warmer in the evening, is thus due only to the divergence of the temperature from that of the outer air. Baron von Grünau met with a friendly reception from the sheiks, even from the representative of the Senussi, but any

attempt to proceed to Jarabub would have been strenuously opposed. He accomplished the return march in eighteen days without serious mishaps, but considers the journey to Siwa the most difficult of the nine caravan journeys he has hitherto undertaken.

Major Gibbons' Zambezi Expedition.-Writing to us from Lialui on August 31, 1899, Major Gibbons sketches the results of the explorations of himself and his colleagues since March, in which month his last letter was despatched (Journal, vol. xiv. p. 93). In accordance with the plans previously announced, the field first chosen for exploration was that to the west of the upper Zambezi, on the upper courses of the Okavango, Kwando and other streams of the western Zambezi basin. Captain Quicke, after reaching, from Lialui, the confluence of the Kubangui and Kwando, followed the latter to its source, and then struck north-east to the Lungwebungu, which, where first seen, was a strong deep stream about 200 yards wide. Its valley, like that of other streams of the region, is bounded by undulations of white sand, which diminish in height in the direction of the Zambezi. The lower part of the Kwando was explored by Captain Hamilton, who found it a much smaller stream than either the Kubangui or the Kwito, both of which are, or could easily be made, navigable for stern-wheelers. The Kwito, according to Major Gibbons' survey, has a very different course to that hitherto shown on our maps, on which it is placed some 25 miles too far east. After crossing 19° E. in about 15° 5′ S., it makes a wide curve to the west, again crossing 19°, and finally passing east of that line only in about 16° 65' south.* It joins the Okavango in 20° 27′ E. and 17° 58′ 30′′ S., beyond which point the latter river crosses and recrosses 18° south, and then, flowing a little south of east, enters a wide swampplain without any of the windings shown on many maps.† Major Gibbons, like Mr. Reid (see below), speaks of a channel connecting the Okavango with the Kwando (Linyanti). It is known as the Mag'wekwana. Following its course from near the Kwando, he was struck with the magnitude of its bed, which inclined him to the idea that it had once formed the main channel of the Okavango, which would then have belonged to the Zambezi system. The channel, however, subsequently became less definite, and he found that it merely forms an overflow from the Okavango during two months of the year, becoming waterless in the dry season. After its bed had become dry, Major Gibbons still found water in pans well above its level, though these in turn are dry in July. Within the bend of the Kwito above described, a large number of tributaries of the Kwando take their rise, very little water flowing to the Kwito from this side. The ground falls considerably towards the Kwando both from the vicinity of the Kwito and of the Lungwebungu, enclosing a generally level triangular plain which extends to the Zambezi, and may be called the great Zambezi plain. On the east of the Zambezi the ground again rises rapidly. Major Gibbons' plans for further exploration were as follows: Captain Hamilton was to descend the Kafukwe, joining his surveys with those of Major Gibbons in 1895-96. Captain Quicke was to ascend the upper Kafukwe and meet his chief at the Kabompo source, afterwards proceeding west along the watershed en route for Loanda. Major Gibbons himself hoped to ascend the Zambezi to its source, then proceeding east along the Kafukwe-Congo watershed, and

*The Kwito was crossed by Capello and Ivens in about 15° 45′ S., its longitude being supposed by them to be then about 19° 35' E. They showed its course without any decided bend to the west, though they were correctly informed as to its navigability down to the Okavango.

As to the lower course of the Okavango, Major Gibbons agrees with Dr. Passage (Journal, vol. xiv. p. 311).

making his way to Tanganyika. Thence he will, if possible, make for the Nile and Khartum. His proceedings have been somewhat hampered by the impossibility of obtaining supplies from the lower river, owing to the unfortunate death of Mr. Muller and the retirement through ill health of Mr. Weller.

Survey of the Linyanti River by Mr. Percy Reid. Mr. Percy Reid, whose explorations in the Barotse country in 1895-96 are well known to our readers, writes to us from Kazungula, under date September 17, announcing his return to that place from a trip round the marshes of the Linyanti river (Chobe of Livingstone), as far as Maini's, placed in our maps in about 23° E. Mr. Reid had been able to make a fair number of astronomical observations, and had fixed the position of Maini's and of the Sunta river (the arm which diverges to the south) by occultations. He hoped also to fix the position of Kazungula (opposite the mouth of the Linyanti) in the same way. An interesting discovery is that of the waterway connecting the Okavang› marshes with those of the Chobe, which has long been reported to exist, though the fact has been doubted.* Mr. Reid has paid much attention to the mysterious subject of the rise of the Linyanti at the height of the dry season, to which attention was first, we believe, called by Mr. Selous (Proc. R.G.S., N.S., vol. iii. p. 171), and we may hope that his observations will throw light on the problem. Mr. Reid says that the river rises twice in the year. Mr. Selous found the flood-water at its highest in September, while Dr. Bradshaw (ibid., p. 212), woo seems to have been in the country at all times of the year, says that the rise commenced in January, and was at its full height in March, falling till January again. It would almost seem as if the régime of the floods varied from year to year, and this is in some meas ire borne out by Mr. Selous' statement that between 1874 and 1879 the overflow seemed to be growing less year by year. Mr. Arnot, who crossed the headstreams of the river in 1884, accounted for its rise in the dry season by the porous nature of the ground near its sources, which was said to absorb the rainfall until the close of the wet season (Proceedings, N.S., vol. xi. p. 69).

Fluctuations in the Level of Tanganyika.-The Mouvement Géographique for October 22 last (No. 43, 1899) prints an extract from a letter from Père Daull, missionary at Karema, relative to recent changes in the level of Lake Tanganyika. Since 1879, the writer says, the lake has fallen at least 26 feet (8 metres), there being now more than a kilometre of cultivated land between the shore-line and the low hills near Karema, which the water reached a few years ago. This does not mean, he says, that the lake is in course of disappearance, but merely that it has retreated within its natural bed, its former high level being due to the blocking of the outlet. It now remains stationary. Père Daull also gives some particulars respecting the present state of Ujiji, or rather of the chief centre of that district. Arab influence has entirely declined, most of the traders having left for the coast. A Greek trader now occupies the house of the former English mission. The Germins, under whose rule there is complete security for Europeans, are introducing the rupee as the official medium of exchange.

Surveys by Captain Ashburnham in Buddu and Ankoli.-Cap'ain Ashburnham, who before starting for Uganda took some lessons in surveying under the Society's instructor, writes to Mr. Coles from Buddu, giving an account of his surveys in that country, with sketch-maps. The chief geographical result has been the first correct delineation of Lake Kachera, which, though long known to

* Dr. Aurel Schulz found that a considerable stream, which he concluded to be a branch of the Okavango, entered the Linyanti from the west. His idea was confirmed by native accounts.

No. I. JANUARY, 1900.]

F

exist, has hitherto been incorrectly shown on the maps. The lake seems to have been first visited by Lieut. Hobart (Journal, vol. xii. p. 189), who, however, saw a part of it only, and considered this to be distinct from the lake previously shown on the maps from native accounts. Captain Ashburnham has traced its shores without losing sight of it for more than an hour at a time, and has found that there is only one lake, split up into a number of narrow arms. The greatest length of the lake about 25 miles according to Captain Ashburnham's sketch-is from west to east, but there is a long narrow arm running north for about 12 miles to the neighbourhood of Kabula (on Lugard's route of 1891-92). The Ruezi river enters the lake at the west end, while the Kivale, known lower down as the Bukora, flows out at the east, entering the Victoria Nyanza a little north of the mouth of the Kagera, with which river it has no connection, though erroneously shown as a tributary on some maps.* The name Kachera is known to all the neighbouring tribes, though each separate inlet (of which there are nearly a dozen) seems to have its own local name. A depression among the hills to the south-west, forming the continuation of one of the arms of the lake, constitutes the boundary between Koki and Ankoli, and another natural trough, continuing the line of the northern arm, separates the latter from Buddu. Neither the Ruezi nor the Bukora are navigable, though the Bukora is deep even in the dry weather, but, like a great part of the lake, is choked with papyrus. There are no crocodiles in either, though they swarm in the Kagera. The district of Sango, south of the Bukora mouth, is densely cultivated, but a forest of large timber stretches to the south and south-west of this. Captain Ashburnham's observations for latitude confirm Lieut. Hobart's statement that the German operations have overstepped the boundary into the British sphere. He has had a good deal of trouble with the rebels, but thinks that their power is now broken.

AMERICA.

Forest and Prairie in Nebraska.-A short paper on recent changes in the extent of forest-land on the Nebraska plains appears in Science for November 24 last, having been read by Mr. C. E. Bessey at the last meeting of the American Association for the Advancement of Science. Two years before, the same author had shown reason for believing that the pines of Western Nebraska are advancing eastward where destruction by fire or cattle is prevented, and where cutting does not proceed too rapidly. He now brings forward evidence from Eastern Nebraska, which, he says, is still more conclusive that tree areas are there advancing with a good deal of rapidity. He has found that, in travelling up-stream and passing out to the side branches, the trees are invariably smaller and younger, it being a very rare occurrence to find large trees near the upper end of a forest belt. This alone would hardly be conclusive, but it is supported by the statements of old settlers, who invariably tell of an advance of trees up the valleys, sometimes a mile or more, while the width of the timber belts on the streams has also greatly increased. The keeping out of fire seems the general cause of the spread of the trees.

The Harriman Expedition to Alaska.-An account of the expedition to Alaska last summer, organized by Mr. E. H. Harriman, is given by Mr. Henry Gannett, of the U.S. Geological Survey, in the Bulletin of the American Geographical Society, vol. xxxi., No. 4, 1899. This expedition was planned for the

The Ruezi ("Rwizi "), the lake, and the Kivali ("Kiware") were shown in correct relation in Stanley's map of 1878, though the latter river was made to join the Kagera. The name of the lake was not then known, but was correctly given in Mr. Ravenstein's map illustrating Captain Lugard's journeys.