areas seen in the illustrations are composed of this disintegrated limestone. The general appearance of the ruin at the present time must not be accepted as its normal condition. It is probable that the débris has undergone a process of artificial selection, the flat slabs and most available stones for building probably having been removed by neighboring settlers and employed in the construction of stone fences, which are much used in this region. Even with a fair allowance for such removal, however, there is no evidence that the rooms were higher than one story. The quantity of potsherds scattered about the ruins is noticeably small.

Plate XXII. General View of Ruin 8 Miles North of Fossil Creek.

Fig. 286.—Ground plan of ruin
8 miles north of Fossil creek.

About 8 miles north of the mouth of Fossil creek, on the eastern side of the Verde, there is a ruin which, though very small, is interesting. At this point there is a long narrow mass of rock, the remains

of a volcanic dike, some 80 or 90 feet long, which at the southern end overhangs the stream, while the other end is merged into the ground level. At its southern end the rock is some 50 feet above the water, but 150 feet northward the dike is no longer traceable. A general view of this dike is given in plate XXII, while the ground plan, figure 286, shows the character of the site. There were rooms on all that portion of the dike that stands out prominently from the ground level, and traces of other rooms can be seen on the ground level adjoining on the north and in the causeway resulting from the breaking down and disintegration of the dike. Remains of eight rooms in all can be traced, five of which were on the summit of the rock. The wall lines on the summit are still quite distinct and in places fragments of the original walls remain, as shown on the ground plan. The plan shows typical pueblo rooms of average size, and the masonry, though rough, is of the same character as that of other ruins in the vicinity.

Facility of defense undoubtedly had something to do with the choice of this location, but that it was not the only desideratum consulted is evident from the occurrence of a large area of fertile bottom land or flat river terrace immediately adjoining the ruin on the east and overlooked by it; in fact, the volcanic dike on which the ruin occurs occupies the western end of a large semicircular area of tillable land, such as already described. Viewed, however, as a village located with reference to defense it is the most perfect example—facility of obtaining water being considered—in this region. It may be used, therefore, to illustrate an important principle governing the location of villages of this type.

A study of the ground plan ([figure 286]) and the general view ([plate XXII]) will readily show that while the site and character of this village are admirably adapted for defense, so well adapted, in fact, as to suggest that we have here a fortress or purely defensive structure, still this adaptation arises solely from the selection of a site fitted by nature for the purpose, or, in other words, from an accident of environment. There has not been the slightest artificial addition to the natural advantages of the site.

The statement may seem broad, but it is none the less true, that, so far as our knowledge extends at the present time, fortresses or other purely defensive structures form a type which is entirely unknown in the pueblo region. The reason is simple; military art, as a distinct art, was developed in a stage of culture higher than that attained by the ancient pueblo builders. It is true that within the limits of the pueblo region structures are found which, from their character and the character of their sites, have been loosely described as fortresses, their describers losing sight of the fact that the adaptability of these structures to defense is the result of nature and not of art. Numerous examples are found where the building of a single short wall would double the defensive value of a site, but in the experience of the writer the