At this point excavations have repeatedly been made by the archæologists, and as often filled up again by the workmen in the summer months, until 1876, when an arrangement was made with the proprietor to make a more complete excavation, and put up a door to protect it. In this part, which is just before the _specus_ crosses the last road on its way to the Tiber, seven different branches of other aqueducts cast their surplus water into this earliest and lowest; in many parts this was only a tunnel cut in the bed of tufa, and generally half full of a deposit of clay, left by the water which comes from the clayey fields of Lucullus, but in one part it is built of large blocks of tufa, like the walls of the Kings; it is here six feet high and two feet wide; on each side of it is a terra-cotta water-pipe of very early character, and bringing different water in the same tunnel[212]. Near this point it crossed _over_ the road on a bridge (?), or perhaps from the low level _under_ it, at the site of a gate where the four roads meet, to another old stone-quarry of the same kind nearly under Santa Prisca, where the _specus_ was visible in 1874, but is now concealed by some of the vault having fallen in; and from that it has been traced through another series of old stone-quarries to the mouth on the bank of the Tiber, in the very curious cave called that of Picus and Faunus, and Hercules—having previously passed through another reservoir, with water still in it knee-deep, remaining at a considerable distance under the hill, with a _specus_ from thence to the cave at the mouth, partly a natural channel for the water to escape, and partly cut when the aqueduct was made.