Page:Africa by Élisée Reclus, Volume 4.djvu/301

 THE KU-BANGO BASIN. 287 sources within 20O miles of the seuport of Benguella, wheroas the distance in a straight lino thence to the shores of the Indian Ocean is no less than l.'OOO miles. The Eu-Bango, rising in the Bihe district, on the southern slope of the mountains which separate its busin from that of the Cuanza, flows at first in a southerly direction parallel with the Cunene and with the axis of the Angolan coast ranges. Owing to this circumstance many explorers, and amongst others Ijadislus Magyar, accepted the native re|K>rt8 that the Ku-Bango drained through the Cunene to the Atlantic Ocean. Not far from its source the river flows for some distunce in an underground channel, and then reappears here and there for short intervals, until it again becomes a surface stream some G or 7 miles below the point where it first plunged into its rocky subterranean bed. Farther on the Ku-Bango winds through a narrow glen between grassy or forest-clud hills, then trending gradually round to the south-east receives the contributions of the Ku-Kyo, the Ku-Atir, and the Lua-Tuta, all flowing in parallel valleys in the direction from north to south. At the point where it was cro8v<ed by Cupello and Ivens on July 10th, that is some six weeks after the dry season had set in, the current had still a width of 1-30 feet with a mean depth of 10 feet and a velocity of neirly two miles an hour. What becomes of this considerable volume of water, which is greatly increased during the rainy season, and lower down more than doubled by the Ku-Ito affluent, which rises on the transverse continental waterparling south of the Kwango and Kassai, and has a total length of no less than 480 miles ? The two Portuguese explorers advance the hypothesis that the united Ku-Bango and Ku-Ito flow directly eastwards, discharging a considerable portion of their waters through the Kwa-Ndo or Chobe into the Zambese. Elsewhere they remark that " they appa- rently go to feed the southern lakes, or, in some unknown way connect themselves with the Zambese."* This is also the opinion of other travellers, and is confirmed by the reports of many native tribes. Andrew Anderson, who has crossed the district in every direction, merely indicates at this point a marshy tract, where, at least during the rainy season, there is a slow onward movement of the fluvial waters, if not a current in the strict sense of the term. But however this be, the almost perfectly level disposition of the plains traversed by the Ku-Bango, below where it begins to converge towards the Kwa- Ndo, gives rise to some remarkable hydrographic phenomena in this region of uncertain drainage. The Cunene itself would even appear occasionally to com- municate with the Zambese through the irnrmnhas of Lake Etosha, and an intri- cate system of channels spreading eastwards. Thanks to the temporary inunda- tions to which the whole region is subject, the hippopotamus has been able gradually to migrate from lagoon to lagoon as far west as the eastern foot of the Herero highlands. Several streams descend from these highlands, one of which, according to Andorsson, is perennial, never completely drying up even in the heat of summer. At the foot of the Waterberg, an extensive sandstone plateau which absorbs much rainwater, springs one of the largest of these streams, although at first scarcely perceived through the densely matted overhanging foliage. • From Baiguella to the Territory of Yaeea, vol. i, p. 93.