When eventually we reached our destination, some 15 to 20 metres beyond the last man-hole, it was only to find that the tunnel just stopped. There was no more water emerging from the rock at the end than I had seen entering at a dozen small fissures along the course we had traversed; there was no vertical excavation downwards; there were no traces of inscriptions; nothing, in fact. After all our trouble, we had drawn a blank. Although I positively dreaded the long crawl back, the very fact that every step took us towards the fresh air made me feel comparatively cheerful. After making my examination of the end of the tunnel, however, I found the space was insufficient to allow of my turning round, so that I was compelled to proceed backwards for some distance before I found room in which to turn. I must confess to never having experienced such a feeling of relief as when we eventually arrived at the bottom of the open shaft, and were able to fill our lungs with copious draughts of the air descending from above. We could see the tiny pin-hole of light far above us, and when sufficiently rested we swung the rope as a signal to be wound up. Although I got back to the surface none the worse for my adventure, with the exception of skin bruises, I would not recommend anyone to attempt the exploration of other similar tunnels unless there were open shafts on either side to insure some circulation of air. It is now quite certain that the aqueducts derived their water from the numerous small fissures traversed, so that the yield of any single tunnel must have depended to a great extent on its total length. The general formation of the country determined the general direction of the tunnels, which, in order to keep within reasonable distance of the surface, had to follow the valleys. The latter, without exception in this district, run in a north and south direction, having formed at one time drainage-lines from the high plateau to the north. Many of the short branch tunnels are, however, driven along east and west fissures, though these could not be followed to any great distance, owing to the high ridges separating the valleys. It is also quite certain that the work of excavation was eventually stopped by the ever-increasing labour involved in the construction of the vertical shafts, the depth of which necessarily increased at a rapid rate as the surface of the ground rose when nearing the actual escarpment. For instance, in the space of only 230 metres, between the ninth and the first shafts, the depth of the man-holes increases from 40 to 54 metres, and to the north of this the increase would have been still more rapid, until the amount of vertical excavation would have become altogether disproportionate to the horizontal distance gained by the aqueduct. CHAPTER XIII BORING METHODS: ANCIENT AND MODERN