The surveying operations within the oasis were based on a rapid plane-table triangulation from a measured base line within it, details being simultaneously sketched in, on a scale of ¹⁄₅₀₀₀₀. The site chosen for the base was a level stretch of ground extending between the camp at Zubbo and an isolated clump of date-palms to the north-east; the length of this line was found by repeated wheel-measurement to be 3·88 kilometres. The two parties ran off their triangulations from this base to the surrounding hills, and carried on the mapping southward by plane-table, taking stations chiefly on the hills and prominent points of the scarps, and not meeting again till Ain el Haiss, in the southern part of the oasis, was reached. Here a test was made as to agreement of the two sets of maps, only a small difference being found. The position of Ain el Haiss, as found by taking the mean of the two determinations, is latitude 28° 2′ 11″ N., longitude 28° 39′ 19″ E. of Greenwich; this places the spring about 4′ 18″ east of Jordan’s determination (lat. 28° 1′ 55″ N., long. 28° 13′ 47″ E. of Green.). The two parties made a third connection at the extreme south end of the oasis-depression. Our observations for this point give its latitude as 27° 48′ 13″ N., and its longitude as 28° 32′ 19″ E. of Greenwich, placing it very near the position shown on Jordan’s map. The plane-table method making use of the magnetic meridian, it was imperative to determine the amount of declination of the compass. This was done at one point only, viz., at the south end of the oasis, the value found (by observation of the transit of Polaris) being 4° 50′ W. The declination is fairly constant over the entire area, except near the eruptive dolerite masses, the magnetite in which causes a very sensible deflection of the needle; in the neighbourhood of these, however, the surveying was carried on independently of the compass. With regard to the yearly change of declination, we have as data the previous observations of Cailliaud, who found the declination at Zubbo in January, 1820, to be 12° 13′ W.; of Jordan, who obtained the value 6° 56′ W. in March, 1874; and of Capt. Lyons,[12] whose observations with a Bamberg declinatorium at Mandisha in April, 1894, gave the value 5° 8·9′ W. Tabulating these:— Observer. Date. Observed Yearly change. Declination W. Cailliaud January 1820 12° 13′ Jordan March 1874 6° 33′·6 6′·3 Lyons April 1894 5° 8′·9 4′·2 Geological Survey December 1897 4° 50′ 5′·3