The difficulty of getting correct information in such a country as this, as daily experience proved to us, is one of the most serious difficulties of the traveller, and depends quite as much on the incapacity of the natives as on the habitual suspicion with which all strangers are regarded. One day when Kaïd el Hasbi appeared to be in unusual good-humour we were endeavouring to obtain from him information as to some place on our route, and the interpreter being told to ask if he could make the matter clearer by reference to the cardinal points, he answered in a tone of contempt, ‘Does he take me for the captain of a steamer?’ After riding with us for two or three miles the Kaïd of Amsmiz bade us a friendly farewell; and we continued in our course across the plain, with occasional halts, in order more closely to examine the vegetation, which was here less varied and interesting than usual. Most of the surface was under grain crops—chiefly wheat and barley—now ripe, and in great part cut and carried away. In the drier waste tracts we once more came upon the characteristic vegetation of the plains, _Acacia_, _Zizyphus Lotus_, _Rhus pentaphylla_, and _Withania frutescens_ being the prevailing shrubs. Of herbaceous plants _Elæoselinum meoides_, and other large _Umbelliferæ_, with _Compositæ_ of the thistle tribe, were most conspicuous. In Beaudouin’s map the chief branch of the Oued Nfys is shown as flowing parallel to the Atlas range from the south-west of Amsmiz, and receiving as tributaries the Amsmiz torrent and the broad stream that we had crossed near Gurgouri. On this day we satisfied ourselves that this representation is erroneous. The unanimous statements of the natives, confirmed by our own observations, proved that all the waters flowing northward from the mountains between Amsmiz and the borders of Mtouga are united in the stream that we had traversed at Sheshaoua on its way to the main river Tensift. The practice of intercepting the streams from the Atlas, and carrying them across the plain through irrigation channels below the surface, makes it extremely difficult to unravel the hydrography of this region.