Squier says that this "fortress" was "the southern limit of the dominions of the first Inca." "The fortress reaches from the mountain, on one side, to a high, rocky eminence on the other. It is popularly called 'El Aqueducto,' perhaps from some fancied resemblance to an aqueduct--but the name is evidently misapplied." Yet he admits that the cross-section of the wall, diminishing as it does "by graduations or steps on both sides," "might appear to conflict with the hypothesis of its being a work of defense or fortification" if it occupied "a different position." He noticed that "the top of the wall is throughout of the same level; becomes less in height as it approaches the hills on either hand and diminishes proportionately in thickness" as an aqueduct should do. Yet, so possessed was he by the "fortress" idea that he rejected not only local tradition as expressed in the native name, but even turned his back on the evidence of his own eyes. It seems to me that there is little doubt that instead of the ruins of Rumiccolca representing a fortification, we have here the remains of an ancient azequia, or aqueduct, built by some powerful chieftain to supply the people of Piquillacta with water. A study of the topography of the region shows that the river which rises southwest of the village of Lucre and furnishes water power for its modern textile mills could have been used to supply such an azequia. The water, collected at an elevation of 10,700 feet, could easily have been brought six miles along the southern slopes of the Lucre Basin, around Mt. Rumiccolca and across the old road, on this aqueduct, at an elevation of about 10,600 feet. This would have permitted it to flow through some of the streets of Piquillacta and give the ancient city an adequate supply of water. The slopes of Rumiccolca are marked by many ancient terraces. Their upper limit corresponds roughly with the contour along which such an azequia would have had to pass. There is, in fact, a distinct line on the hillside which looks as though an azequia had once passed that way. In the valley back of Lucre are also faint indications of old azequias. There has been, however, a considerable amount of erosion on the hills, and if, as seems likely, the water-works have been out of order for several centuries, it is not surprising that all traces of them have disappeared in places. I regret very much that circumstances over which I had no control prevented my making a thorough study of the possibilities of such a theory. It remains for some fortunate future investigator to determine who were the inhabitants of Piquillacta, how they secured their water supply, and why the city was abandoned. ------ FIGURE Sacsahuaman: Detail of Lower Terrace Wall ------ ------ FIGURE Ruins of the Aqueduct of Rumiccolca ------