The ruin most easily visited (fig. 2) in this canyon is on a comparatively low shelf in a shallow cave, 40 feet high, a few feet above the top of the talus. Like many other cliff houses it is divided into two parts, called the upper and the lower, according to the level they occupy. The lower is practically buried under rocks fallen from the walls of the upper house. The front wall of the upper part is well preserved and closely follows the contour of the low ridge on which it stands. The masonry is fairly good, but the floors of the rooms are buried under a thick deposit of sheep droppings, solidly packed, showing that the enclosures have been used secondarily as corrals for these domesticated animals. The partition walls of the rooms end on the vertical wall of the precipice, the face of the precipice serving as their rear wall. It thus happens that there is no recess between the back of the rooms and the rear of the cave, as commonly found in cliff dwellings. Circular rooms are absent in the upper part of this ruin, and kivas, if any, must be sought buried under the accumulated débris of the lower part. The front wall of the upper house measures 64 feet, and can be traced throughout its whole extent. At one end of the ruin there are four narrow rooms separated by partitions, each containing a grinding bin, where maize (corn) was reduced to meal. The remaining rooms are roofless, plastered, and evidently used as dwellings. In the lower series of rooms, buried beneath a mass of fallen rocks, are circular depressions, which may be ceremonial rooms; but no excavations were made in these depressions and their significance is unknown. [Illustration: FIG. 2.—Ground plan of cliff ruin in Nashlini Canyon.] Another cliff house, a few miles farther up in the canyon, is almost hidden in an inaccessible recess of the cliff, but so high that it was not visited. On the dizzy top of a cliff overlooking the canyon, near the second ruin, artificial walls were observed but not visited. An Indian guide claimed that they were towers; they are certainly so situated as to permit a wide view up and down the canyon. These walls are mentioned by Dr. Prudden.