We have now enough data to calculate the amount of material excavated from these underground works, and this gives us a good idea of the magnitude of the task. The average depth of the man-holes (from the ground-surface to the roof of the horizontal tunnel) works out at 22 metres— that is to say, they represent a total length of vertical excavation amounting to 3·3 kilometres. Adding 10 per cent. to the length of the aqueduct, to allow for the numerous bends and the side branches, we get 3·2 kilometres of horizontal excavation, or a grand total of 6·5 kilometres. From a number of measurements the cross-section of both shafts and tunnels works out at an almost identical figure—_i.e._, three-quarters of a square metre. Thus the amount of rock excavated in this system alone is about 4,875 cubic metres, and we may safely say that the construction of the four subterranean aqueducts and the 600 or 700 vertical shafts meant the excavation and removal of over 20,000 cubic metres of solid rock. In contemplating the time and labour involved in the excavation and removal of this great mass of material, it must not be forgotten that the latter represents the construction of underground passages over twenty miles in length, and of such restricted size that only one man could have found room to work on the face of the tunnel or shaft at a time. And although the sinking of more than one vertical shaft may have been carried on simultaneously, the cutting of the long aqueducts had almost certainly to be conducted from one side only—_i.e._, from their exits. The accumulation of water would probably have prevented the cutting out of different sections at one and the same time by separate gangs of men. One day, in the summer of 1908, I was foolhardy enough to follow the aqueduct to its termination, from the bottom of the ninth shaft—the one I had previously descended—and in case some future visitor to Um el Dabâdib should feel inclined to examine for himself these wonderful subterranean works, it may not be out of place to give a brief account of my examination of the upper portion and termination of the aqueduct in question (the second, counting from the most westerly), so that he may benefit by my experience. My particular objects were to ascertain if the bulk of the water came from one or more large fissures, to determine whether any special characteristics marked the termination of the tunnel, and to discover whether there were inscriptions which would yield valuable information regarding the making of the aqueduct. My intention was to descend one of the uppermost shafts by means of the windlass with which, remembering my previous experience with nothing better than a very elastic native-made rope, I had been careful to provide myself. I reckoned that the distance thence to the point of origin of the aqueduct could not be so far as to entail any great difficulty or danger from insufficiency of ventilation.