G. W. Grahham— Wells of North-Eastern Sudan. 317 



of nitre, or nearly an ounce to the gallon.^ The reduction of these 

 nitrates, for instance by iron pyrites, would account for the large 

 amount of niti'ogen gas found in the artesian waters of Kharga. 



The origin of the water in the sandstone is too extensive a subject 

 to be entered into here. The bulk of the water, at any rate witliin 

 the area dealt with here, is probably supplied by seepage from the 

 river. This takes place, for the most part, during the flood when 

 the Nile is high and the rapidly flowing current is keeping the pores 

 of the bed freely open. It does not appear that the Nubian Sandstone 

 extends much further south than the limits shown in Rolland's map.^ 

 The small amount of rainfall occurs in violent storms of short duration, 

 and provided the water forms rapidly flowing streams, the proportion 

 soaking into the ground must be very large. Often it remains in 

 pools, or is retained by the surface soil so that it evaporates away 

 without adding to the underground water. 



Concluding Eemaeks. 



There are some special cases of well supplies that do not fall among 

 the classes already described, but call for some notice. 



For some time after the rains, water can be obtained from shallow 

 holes made in the sandy beds of khors, but later it either evaporates 

 or percolates away, and seldom lasts until the following rainy season. 

 The occurrence of water in this way presents a curious problem, for 

 in some cases at least, if not in all, there is no impermeable substratum 

 to hold it up; yet there it remains, while the permanent ground water- 

 level may be a hundred feet below. This phenomenon is not limited 

 to sandstone regions, but may occur on the surface of the Valley Fill. 

 As an example the wells in Wadi Mugaddam may be cited. Here 

 the natives find water close to the surface immediatelj' after the rainy 

 season. They deepen the wells as the water falls, but finally it 

 disappears and they have recourse to deep wells in the sandstone. 

 I saw another instance of a similar thing in a small vallej^ half-way 

 between Khartoum and Wadi Mugaddam. The water had been 

 obtained from shallow holes in the stream-bed for some time after the 

 rains, but when I was there a few months later there was no water 

 in a well over a hundred feet deep in the sandstone close by. 



Another special case of well supplies is found connected with Xhor 

 Gash. In summer this stream comes down in flood from the mountains 

 of Eritrea and loses itself on the plains north of Kassala, and during 

 the rest of the year good supplies of water are obtained in the stream- 

 bed, but the permeable material is not extensive alongside. The case 

 is really analogous to the Valley Fill, with impervious soil taking the 

 place of crystalline rock. The ground around is either alluvial or 

 Cotton Soil, and wells at a distance from the stream-bed have pene- 

 trated to a considerable depth without obtaining any water. I am 

 told that the natives have a fairly shrewd idea of the limits within 

 which water is to be found, and these, no doubt, depend on the extent 

 the stream-bed has wandered before assuming its present course. 



Large supplies of water are derived from wells in the alluvial 

 deposits alongside the Nile, but these call for no special notice here. 



1 Wellcome Research Lcaboratories, Khartoum, Third Eeport, 190S, ji. 397. 

 ^ Reproduced in E. Suess, La Face de la Terre, vol. i, p. 259. 



