siderably less than 4,000 feet, and here its banks begin to be fringed by a tropical vegetation. On descending from the surrounding mountains, which are swept by cold winds, the sensation is like that felt on entering a hothouse.
The Atbara.
After its escape from the region of the plateaux, the Takkazé resumes its westerly course, and at last reaches the plain through a series of rugged gorges. Here it takes the name of the Setit, and is joined by the Atbara, which is much less in volume and hardly half its length; but the mean direction of its valley, beginning immediately west of Lake Tsana, is the same as that of the united streams. The Atbara, like the Mississippi on joining the yellow and turbid waters of the powerful Missouri, gives its name to the hydrographic system; the Goang, one of the tributaries of the Takkaze, rises in the north in the depression of Lake Tsana, from which it is separated only by a ridge 165 feet high. Below the confluence the Atbara, which retains the ancient name given it by Ptolemy under the form of Astaboras, gradually diminishes in volume, as does also its former affluent, the Mareb, which in its upper course describes one of those large semi-circular curves so characteristic of the Abyssinian rivers. In fact, the Mareb, or the "River of the West," as it is called by the Abyssinians from the direction of its course, may be said to have ceased to be an affluent of the Atbara. Called the "Sona" in its middle and "Gash" in its lower course, where it is only an intermittent stream, it flows northwards parallel with the Atbara, and runs out in the alluvial lands before reaching its former outl(;t, called by the Hadendoa nomads "Gash-da," i.e. "Mouth of the Gash." On visiting the country in 1864, Munzinger found that its bed had not been once flooded for twenty years. This change in the local hydrography doubtless arises from the irrigation works constructed on the left bank of the Gash. Embanked on this side, the river flows to the right, eating away its eastern and highest cliffs. Its course, formerly at right angles, now becomes parallel to the Atbara; but as it flows northwards it finally runs dry in the sands. In 1840, Ahmed Pasha, the Egyptian conqueror, tried again to divert the Gash westwards into the Atbara, but his embankment was undermined by the riverain population of the lower plain. Till recently the river Barka, or Baraka, flowing into the swamps on the Red Sea coast not far from Suakin, was also supposed to belong to the Nile basin through a branch of the Mareb. This tradition differs little from that related by Strabo, according to which a branch of the Astaboras flowed to the Red Sea. The hypothesis may perhaps be partly due to a confusion of names, for the plain stretching east of the Mareb towards the Atbara is called Barka, or Baraka, a term also applied to the channel flowing eastwards. However this be, the Axumite Ethiopians, and after them the Abyssinians, who long identified the true Nile with their Takkaze, fancied for centuries that it would be easy to divert their river into the sea and thus deprive Egypt of the water required for its crops. This illusion, however, was also entertained by foreigners, and is referred to by Ariosto in his "Orlando Furioso." Repeating the threat of
