It seemed to me in making a plan for archeological field work in 1895, that the prehistoric cliff houses, cave dwellings, and ruined pueblos of Arizona afforded valuable opportunities for research, and past experience induced me to turn my steps more especially to the northern and northeastern parts of the territory.[1] The ruins of ancient habitations in these regions had been partially, and, I believe, unsatisfactorily explored, especially those in a limited area called Tusayan, now inhabited by the Moki or Hopi Indians. These agricultural people claim to be descendants of those who once lived in the now deserted villages of that province.

I had some knowledge of the ethnology of the Hopi, derived from several summers' field work among them, and I believed this information could be successfully utilized in an attempt to solve certain archeological questions which presented themselves.[2] I desired, among other things, to obtain new information on the former extension, in one direction, of the ancestral abodes of certain clans of the sedentary people of Tusayan which are now limited to six pueblos in the northeastern part of the territory. In carrying out this general plan I made an examination of cliff dwellings and other ruins in Verde valley, and undertook an exploration of two old pueblos near the Hopi villages. The reason which determined my choice of the former as a field for investigation was a wish to obtain archeological data bearing on certain Tusayan traditions. It is claimed by the traditionists of Walpi, especially those of the Patki[3] or Water-house phratry, that their ancestors came from a land far to the south of Tusayan, to which they give the name Palatkwabi. The situation of this mythic place is a matter of considerable conjecture, but it was thought that an archeological examination of the country at or near the headwaters of the Rio Verde and its tributaries might shed light on this tradition.

It is not claimed, however, that all the ancestors of the Tusayan people migrated from the south, nor do I believe that those who came from that direction necessarily passed through Verde valley. Some, no doubt, came from Tonto Basin, but I believe it can be shown that a continuous line of ruins, similar in details of architecture, extend along this river from its junction with Salt river to well-established prehistoric dwelling places of the Hopi people. Similar lines may likewise be traced along other northern tributaries of the Salt or the Gila, which may be found to indicate early migration stages.

The ruins of Verde valley were discovered in 1854 by Antoine Leroux, a celebrated guide and trapper of his time, and were thus described by Whipple, Ewbank, and Turner in the following year:

A fragment of folklore, which is widely distributed among both the aboriginal peoples of Gila valley and the modern Tusayan Indians, recounts how the latter were at one time in communication with the people of the south, and traditions of both distinctly connect the sedentary people of Tusayan with those who formerly inhabited the great pueblos, now in ruins, dotting the plain in the delta between Gila and Salt rivers. That archeology might give valuable information on this question had long been my conviction, and was the main influence which led me to the studies recorded in the following pages.

An examination of a map of Arizona will show that one of the pathways or feasible routes of travel possible to have been used in any connection between the pueblos of the Gila and those of northern Arizona would naturally be along Rio Verde valley. Its tributaries rise at the foot of San Francisco mountains, and the main river empties into the Salt, traversing from north to south a comparatively fertile valley, in the main advantageous for the subsistence of semisedentary bands in their migrations. Here was a natural highway leading from the Gila pueblos, now in ruins, to the former villages in the north.

The study of the archeology of Verde valley had gone far enough to show that the banks of the river were formerly the sites of many and populous pueblos, while the neighboring mesas from one end to another are riddled with cavate dwellings or crowned with stone buildings. Northward from that famous crater-like depression in the Verde region, the so-called Montezuma Well on Beaver creek, one of the affluents of the Rio Verde, little archeological exploration had been attempted. There was, in other words, a break in the almost continuous series of ruins from Tusayan as far south as the Gila. Ruined towns had been reported as existing not far southward from San Francisco mountains,[5] and from there by easy stages the abodes of a former race had been detected at intervals all the way to the Tusayan pueblos. At either end the chain of ruins between the Tusayan towns and the Gila ruins was unbroken, but middle links were wanting. All conditions imply former habitations in this untrodden hiatus, the region between the Verde and the Tusayan series, ending near the present town of Flagstaff, Arizona; but southward from that town the country was broken and impassable, a land where the foot of the archeologist had not trodden. Remains of human habitations had, however, been reported by ranchmen, but these reports were vague and unsatisfactory. So far as they went they confirmed my suspicions, and there were other significant facts looking the same way. The color of the red cliffs fulfilled the Tusayan tradition of Palatkwabi, or their former home in the far south. Led by all these considerations, before I took to the field I had long been convinced that this must have been one of the homes of certain Hopi clans, and when the occasion presented itself I determined to follow the northward extension of the ancient people of the Verde into these rugged rocks. By my discoveries in this region of ruins indicative of dwellings of great size in ancient times I have supplied the missing links in the chain of ancient dwellings extending from the great towns of the Gila to the ruins west of the modern Tusayan towns. If this line of ruins, continuous from Gila valley to Tusayan and beyond, be taken in connection with legends ascribing Casa Grande to the Hopi and those of certain Tusayan clans which tell of the homes of their ancestors in the south, a plausible explanation is offered for the many similarities between two apparently widely different peoples, and the theory of a kinship between southern and northern sedentary tribes of Arizona does not seem as unlikely as it might otherwise appear.

The reader will notice that I accept without question the belief that the so-called cliff dwellers were not a distinct people, but a specially adaptive condition of life of a race whose place of habitation was determined by its environment. We are considering a people who sometimes built dwellings in caverns and sometimes in the plains, but often in both places at the same epoch. Moreover, as long ago pointed out by other students, the existing Pueblo Indians are descendants of a people who at times lived in cliffs, and some of the Tusayan clans have inhabited true cliff houses in the historic period. By intermarriage with nomadic races and from other causes the character of Pueblo consanguinity is no doubt somewhat different from that of their ancient kin, but the character of the culture, as shown by a comparison of cliff-house and modern objects, has not greatly changed.

While recognizing the kinship of the Pueblos and the Cliff villagers, this resemblance is not restricted to any one pueblo or group of modern pueblos to the exclusion of others. Of all modern differentiations of this ancient substratum of culture of which cliff villages are one adaptive expression, the Tusayan Indians are the nearest of all existing people of the Southwest[6] to the ancient people of Arizona.

The more southerly ruins of Tusayan, which I have been able satisfactorily to identify and to designate by a Hopi name, are those called Homolobi, situated not far from Winslow, Arizona, near where the railroad crosses the Little Colorado. These ruins are claimed by the Hopi as the former residences of their ancestors, and were halting places in the migration of certain clans from the south. They were examined by Mr Cosmos Mindeleff, of the Bureau of American Ethnology, in 1893,[7] but no report on them has yet been published.

While, however, the Homolobi group of ruins is the most southerly to which I have been able to affix a Hopi name, others still more to the southward are claimed by certain of their traditions.[8] The Hopi likewise regard as homes of their ancestors certain habitations, now in ruins, near San Francisco mountains. In a report on his exploration of ZuÑi and Little Colorado rivers in 1852, Captain L. Sitgreaves called attention to several interesting ruins, one of which was not far from the "cascades" of the latter river. After ascending the plateau, which he found covered with volcanic detritus, he discovered that "all the prominent points" were "occupied by the ruins of stone houses, which were in some instances three stories in height. They are evidently," he says, "the remains of a large town, as they occurred at intervals for an extent of eight or nine miles, and the ground was thickly strewn with fragments of pottery in all directions."

In 1884 a portion of Colonel James Stevenson's expedition, under F. D. Bickford, examined the cliff houses in Walnut canyon, and in 1886 Major J. W. Powell and Colonel Stevenson found scattered ruins north of San Francisco mountains having one, two, or three rooms, each "built of basaltic cinders and blocks of lava." These explorers likewise reported ruins of extensive dwellings in the same region made of sandstone and limestone. At about 25 miles north of the mountains mentioned they discovered a small volcanic cone of cinders and basalt, which was formerly the site of a village or pueblo built around a crater, and estimated that this little pueblo contained 60 or 70 rooms, with a plaza occupying one-third of an acre of surface.[9]

Twelve miles eastward from San Francisco mountains they found another cinder cone resembling a dome, and on its southern slope, in a coherent cinder mass, were many chambers, of which one hundred and fifty are said to have been excavated. They mention the existence on the summit of this cone of a plaza inclosed by a rude wall of volcanic cinders, with a carefully leveled floor. The former inhabitants of these rooms apparently lived in underground chambers hewn from the volcanic formation. Eighteen miles farther eastward was another ruined village built about the crater of a volcanic cone. Several villages were discovered in this locality and many natural caves which had been utilized as dwellings by inclosing them in front with walls of volcanic rocks and cinders. These cavate rooms were arranged tier above tier in a very irregular way.

At this place three distinct kinds of ruins were found—cliff villages, cave dwellings, and pueblos. Eight miles southeastward from Flagstaff, in Oak creek canyon, a cliff house of several hundred rooms was discovered. It was concluded that all these ruins were abandoned at a comparatively recent date, or not more than three or four centuries ago, and the Havasupai Indians of Cataract canyon were regarded as descendants of the former inhabitants of these villages. The situation of some of these ruins and the published descriptions would indicate that some of them were similar to those described and figured by Sitgreaves,[10] to which reference has already been made.

In 1896 two amateur explorers, George Campbell and Everett Howell, of Flagstaff, reported that they had found, about eighteen miles from that place, several well-preserved cliff towns and a remarkable tunnel excavation. The whole region in the immediate neighborhood of San Francisco mountains appears, therefore, to have been populated in ancient times by an agricultural people, and legends ascribe some of these ruins to ancestors of the Hopi Indians.

There are several ruins due south of Tusayan which have not been investigated, but which would furnish important contributions to a study of Hopi migrations. Near Saint Johns, Arizona, likewise, there are ruins of considerable size, possibly referable to the Cibolan series; and south of Holbrook, which lies about due south of Walpi, there are ruins, the pottery from which I have examined and found to be of the black-and-white ware typical of the Cliff people. Perhaps, however, no ruined pueblo presents more interesting problems than the magnificent Pueblo Grande or Kintiel, about 20 miles north of Navaho Springs. This large ruin, lying between the Cibolan and Tusayan groups, has been referred to both of these provinces, and would, if properly excavated, shed much light on the archeology of the two provinces.[11] Kinnazinde lies not far from Kintiel.

The ruins reported from Tonto Basin, of which little is known, may later be found to be connected with early migrations of those Hopi clans which claim southern origin. From what I can judge by the present appearance of ruins just north of the Mogollon mountains, in a direct line between Tonto Basin and the present Tusayan towns, there is nothing to show the age of these ruined villages, and it is quite likely that they may have been inhabited in the middle of the sixteenth century. While it is commonly agreed that the province of "Totonteac," which figures extensively in certain early Spanish narratives, was the same as Tusayan, the linguistic similarity of the word to "tonto" has been suggested by others. In the troublesome years between 1860 and 1870 the Hopi, decimated by disease and harried by nomads, sent delegates to Prescott asking to be removed to Tonto Basin, and it is not improbable that in making this reasonable request they simply wished to return to a place which they associated with their ancestors, who had been driven out by the Apache. Totonteac[12] is ordinarily thought to be the same as Tusayan, but it may have included some of the southern pueblos now in ruins west of ZuÑi.

Having determined that the line of Verde ruins was continued into the Red-rock country, it was desirable to see how the latter compared with those nearer Tusayan. This necessitated reexamination of many ruins in Verde valley, which was my aim during the most of June. I followed this valley from the cavate dwellings near Squaw mountain past the great ruin in the neighborhood of Old Camp Verde, the unique Montezuma Well, to the base of the Red-rocks. Throughout this region I saw, as had been expected, no change in the character of the ruins great enough to indicate that they originally were inhabited by peoples racially different. Stopped from further advance by a barrier of rugged cliffs, I turned westward along their base until I found similar ruins, which were named Palatki and Honanki. Having satisfied myself that there was good evidence that the numbers of ancient people were as great here as at any point in the Verde valley and that their culture was similar, I continued the work with an examination of the ruins north of the Red-rocks, where there is substantial evidence that these were likewise of the same general character.

The last two months of the summer, July and August, 1895, were devoted to explorations of two Tusayan ruins, called Awatobi and Sikyatki. In this work, apparently unconnected with that already outlined, I still had in mind the light to be shed on the problem of Tusayan origin. The question which presented itself was: How are these ruins related to the modern pueblos? Awatobi was a historic ruin, destroyed in 1700, and therefore somewhat influenced by the Spaniards. Many of the survivors became amalgamated with pueblos still inhabited. Its kinship with the surviving villagers was clear. Sikyatki, however, was overthrown in prehistoric times, and at its destruction part of its people went to Awatobi. Its culture was prehistoric. The discovery of what these two ruins teach, by bringing prehistoric Tusayan culture down to the present time and comparing them with the ruins of Verde valley and southern Arizona, is of great archeological interest.

While engaged in preparing this report, having in fact written most of it, I received Mr Cosmos Mindeleff's valuable article on the Verde ruins,[13] in which special attention is given to the cavate lodges and villages of this interesting valley. This contribution anticipates many of my observations on these two groups of aboriginal habitations, and renders it unnecessary to describe them in the detailed manner I had planned. I shall therefore touch but briefly on these ruins, paying special attention to the cliff houses of Verde valley, situated in the Red-rock country. This variety of dwelling was overlooked in both Mearns' and Mindeleff's classifications, from the fact that it seems to be confined to the region of the valley characterized by the red-rock formation, which appears not to have been explored by them. The close resemblance of these cliff houses to those of the region north of Tusayan is instructive, in view of the ground, well taken, I believe, by Mr Mindeleff, that there is a close likeness between the Verde ruins and those farther north, especially in Tusayan.