to suit the changed environment, yet any general conclusion based on the study of a single ruin would be unsafe. In the present state of knowledge of this field it is not advisable to attempt more than a detailed description, embodying, however, a few inferences, applicable to this ruin only, which seem well supported by the evidence obtained.

The Casa Grande ruin is located near the southwestern corner of the group, and the ground surface for miles about it in every direction is so flat that from the summit of the walls an immense stretch of country is brought under view. On the east is the broad valley of Gila river rising in a great plain to a distant range of mountains. About a mile and a half toward the north a fringe of cottonwood trees marks the course of the river, beyond which the plain continues, broken somewhat by hills and buttes, until the view is closed by the Superstition mountains. On the northwest the valley of Gila river runs into the horizon, with a few buttes here and there. On the west lies a range of mountains closing the valley in that direction, while toward the southwest and south it extends until in places it meets the horizon, while in other places it is closed by ranges of mountain blue and misty in the distance. In an experience of some years among northern ruins, many of them located with special reference to outlook over tillable lands, the writer has found no other ruin so well situated as this.

The character of the site occupied by the ruin indicates that it belongs to a late date if not to the final period in the occupancy of this region, a period when by reason of natural increase of numbers, or perhaps aggregation of related gentes, the defense motive no longer dominated the selection of a village site, but reliance was placed on numbers and character of structures, and the builders felt free to select a site with reference only to their wants as a horticultural people. This period or stage has been reached by many of the Pueblo tribes, although mostly within the historical period; but some of them, the Tusayan for example, are still in a prior stage.

PLATE LIV

STANDING WALL NEAR CASA GRANDE

A ground plan of the ruin is shown in plate LII, and a general view in plate LIII. The area covered and inclosed by standing walls is about 43 feet by 59 feet, but the building is not exactly rectangular, and the common statement that it faces the cardinal points is erroneous. The variation from the magnetic north is shown on the ground plan, which was made in December, 1890. The building comprised three central rooms, each approximately 10 by 24 feet, arranged side by side with the longer axes north and south, and two other rooms, each about 9 by 35 feet, occupying respectively the northern and southern ends of the building, and arranged transversely across the ends of the central rooms, with the longer axes running east and west. Except the central room, which was three stories in height, all the rooms were two stories above the ground. The northeastern and southeastern corners of the structure have fallen, and large blocks of the material of which they were composed are strewn upon the ground in the vicinity. It is probable that the destruction of these corners prior to that of the rest of the building was due to the disintegration of minor walls connected with them and extending, as shown by the ridges on the ground plan, northward from the northeastern corner and eastward from the southeastern corner. These walls doubtless formed part of the original structure and were probably erected with it; otherwise the corners of the main structure would not have been torn out or strained enough to fall before the rest of the building was affected.

It is not likely that the main building originally stood alone as at present. On the contrary there is every reason to suppose that it was connected with other buildings about 75 feet east of it, now marked by a bit of standing wall shown on the map (plate LI), and probably also with a small structure about 170 feet south of it, shown in plate LIV. These connections seem to have been by open courts inclosed by walls and not by continuous buildings. The court east of the ruin is well marked by the contours and seems to have been entered by a gateway or opening at its southeastern corner.

DIMENSIONS.

It is probable that the area immediately adjacent to the ruin, and now covered by mounds, carried buildings of the same time with the main structure and was occupied contemporaneously with it or nearly so. This area, well marked on the map, measures about 400 feet north and south, and 240 feet east and west. It is not rectangular, although the eastern and western sides, now marked by long ridges, are roughly parallel. The northeastern corner does not conform to a rectangular plan, and the southern side is not more than half closed by the low ridge which extends partly across it. This area is doubtless the one measured in 1776, by Padre Font, whose description, was copied by later writers, and whose measurements were applied by Humboldt and others to the ruin itself. Font gave his measurements as those of a circumscribing wall, and his inference has been adopted by many, in fact most, later writers. A circumscribing wall is an anomalous feature, in the experience of the writer, and a close inspection of the general map will show that Font's inference is hardly justified by the condition of the remains today. It seems more likely that the area in question was covered by groups of buildings and rows of rooms, connected by open courts, and forming an outline sometimes regular for a considerable distance, but more often irregular, after the manner of pueblo structures today. The long north and south ridge which forms the southeastern corner of the area, with other ridges extending westward, is quite wide on top, wide enough to accommodate a single row of rooms of the same width as those of the ruin, and it is hardly reasonable to suppose that a wall would be built 10 or 12 feet wide when one of 4 feet would serve every purpose to which it could possibly be put. Furthermore, the supposition of an inclosing wall does not leave any reasonable explanation of the transverse ridges above mentioned, nor of the long ridge which runs southward from the southeastern corner of the ruin.

The exterior walls rise to a height of from 20 to 25 feet above the ground. This height accommodated two stories, but the top of the wall is now 1 to 2 feet higher than the roof level of the second story. The middle room or space was built up three stories high and the walls are now 28 to 30 feet above the ground level. The tops of the walls, while rough and much eroded, are approximately level. The exterior surface of the walls is rough, as shown in the illustrations, but the interior walls of the rooms are finished with a remarkable degree of smoothness, so much so as to attract the attention of everyone who has visited the ruin. Mange, who saw the ruin with Padre Font in 1697, says the walls shine like Puebla pottery, and they still retain this finish wherever the surface has not cracked off. This fine finish is shown in a number of illustrations herewith. The walls are not of even thickness. At the ground level the exterior wall is from 3½ to 4½ feet thick, and in one place at the southern end of the eastern wall, is a trifle over 5 feet thick. The interior walls are from 3 to 4 feet thick at base. At the top the walls are reduced to about 2 feet thick, partly by setbacks or