Vol. 59.] SEMNA CATAKACT OP THE NILE. 67 



set up. At low Nile the gneiss-band entirely bars the stream, 

 except for a narrow central channel about 40 metres (130 feet) in 

 width. 



The accompanying map and section (PI. Ill), based on a rapid 

 reconnaissance-survey, represent the condition of things at low 

 Nile in the middle of March 1902. The shore-line and the rocky 

 bar were surveyed with a 3-inch tacheometer ; this portion of the 

 map may, I believe, be relied on within the scale-limits. The 

 village of Kumna was measured by pacing, aided by tacheometric 

 fixation of its principal points. The hill-features away from the 

 river, and the brick-ruins of Semna, are only roughly sketched in. 

 Absolute levels were not taken, as this would have involved a long 

 line of spirit-levelling from distant bench-marks. The relative 

 levels shown were found tacheometrically, with untrained fellahin 

 as staff-holders ; they are believed to be accurate to within 0*2 metre 

 (8 inches). The Assuan gauge-reading at the time of the survey 

 was 1 pic 11 kirats, corresponding to the reduced level of 84*93 

 metres (279 feet), and a discharge past Assuan of 390 cubic metres 

 (13,650 cubic feet) per second. There is so little water taken off 

 between Semna and Assuan, that we may very fairly assume the 

 discharge past the Semna barrier at the time of the survey to 

 be about 400 cubic metres (14,000 cubic feet) per second. The 

 whole of this discharge passed through the 40-metre central channel 

 without any foaming, though the throttling of the entrance by rocky 

 islets produced numerous eddy- currents ; and the velocity of the 

 narrow stream, as roughly estimated by the eye, was only about 

 4 kilometres (2^ miles) per hour, so that it was at once evident 

 that the channel was deep. A sounding taken in the centre of the 

 rapid gave the depth of water as 23 metres (754 feet), but probably 

 3 metres (10 feet) or so should be taken off, to compensate for 

 deflection of the plumb-line by the current. No soundings were 

 taken outside the central channel; but there doubtless exists a 

 great deepening down-stream of the barrier, due to the action of 

 the water in falling over it. On both sides of the barrier the water 

 was so placid as to resemble a lake more than a river. How great 

 is the contrast between the flow across the bar and that on either 

 side of it, will be evident from the sketch reproduced in fig. 1 (p. 68), 

 which is taken from a photograph. 



The ruined temples of Kumna and Semna, with their surrounding 

 forts, lie on elevated portions of the same gneiss-band as that which 

 forms the barrier, and the fissile gneiss has been used by the ancients 

 to form embankments around their erections. The temples, as they 

 now stand, date from the 18th dynasty ; but the forts around 

 them are considerably older, and appear to have marked the southern 

 frontier of Egypt proper (not of the Empire) during the 12th and 

 13th dynasties 1 : thus they formed an appropriate place for 

 registering the height of the flood, which was important as infor- 

 mation for the great irrigation-engineers of the time. A large 



1 Maspero's 'Dawn of Civilization' 4th ed. (1901) p. 485 ; also id. 'The 

 Struggle of the Nations ' pp. 89, 230. 



17 9 



