On the sides of these inaccessible cliffs we noticed several cliff houses, but so high were they perched above us that they were almost invisible. To reach them at their dizzy altitude was impossible, but we were able to enter some caves a few hundred feet above our camp, finding in them nothing but charred mescal and other evidences of Apache camps. Their walls and entrances are blackened with smoke, but no sign of masonry was detected.

We moved our camp westward from this canyon (which, from a great cliff resembling the Parthenon, I called Temple canyon), following the base of the precipitous mountains to a second canyon, equally beautiful but not so grand, and built our fire in a small grove of scrub oak and cottonwood. In this lonely place Lloyd had lived over a winter, watching his stock, and had dug a well and erected a corral. We adopted his name for this camp and called it Lloyd canyon. There was no water in the well, but a few rods beyond it there was a pool, from which we watered our horses. On the first evening at this camp we sighted a bear, which gave the name Honanki, "Bear-house," to the adjacent ruined dwellings.

The enormous precipice of red rock west of our camp at Lloyd's corral hid Honanki from view at first, but we soon found a trail leading directly to it, and during our short stay in this neighborhood we remained camped near the cottonwoods at the entrance to the canyon, not far from the abandoned corral. Our studies of Honanki led to the discovery of Palatki (figure 247), which we investigated on our return to Temple canyon. I will, therefore, begin my description of the Red-rock cliff houses with those last discovered, which, up to the visit which I made, had never been studied by archeologists.

Palatki

There are two neighboring ruins which I shall include in my consideration of Palatki, and these for convenience may be known as Ruin i and Ruin ii, the former situated a little eastward from the latter. They are but a short distance apart, and are in the same box canyon. Ruin i (plate xcix) is the better preserved, and is a fine type of the compact form of cliff dwellings in the Red-rock country.

This ruin is perched on the top of a talus which has fallen from the cliff above, and is visible for some distance above the trees, as one penetrates the canyon. It is built to the side of a perpendicular wall of rock which, high above its tallest walls, arches over it, sheltering the walls from rain or eroding influences. From the dry character of the earth on the floors I suspect that for years not a drop of water has penetrated the inclosures, although they are now roofless.

A highly characteristic feature of Ruin i is the repetition of rounded or bow-shape front walls, occurring several times in their length, and arranged in such a way as to correspond roughly to the inclosures behind them. By this arrangement the size of the rooms was increased and possibly additional solidity given to the wall itself. This departure from a straight wall implies a degree of architectural skill, which, while not peculiar to the cliff dwellings of the Red-rocks, is rarely found in southern cliff houses. The total length of the front wall of the ruin, including the part which has fallen, is approximately 120 feet, and the altitude of the highest wall is not far from 30 feet.

Fig. 247—Ground plan of Palatki (Ruins i and ii) Fig. 247—Ground plan of Palatki (Ruins I and II)

From the arrangement of openings in the front wall at the highest part there is good evidence of the former existence of two stories. At several points the foundation of the wall is laid on massive bowlders, which contribute to the height of the wall itself. The masonry is made up of irregular or roughly squared blocks of red stone laid in red clay, both evidently gathered in the immediate neighborhood of the ruin. The building stones vary in size, but are as a rule flat, and show well directed fractures as if dressed by hammering. In several places there still remains a superficial plastering, which almost conceals the masonry. The blocks of stone in the lower courses are generally more massive than those higher up; this feature, however, whether considered as occurring here or in the cliff houses of Mesa Verde, as pointed out by Nordenskiöld, seems to me not to indicate different builders, but is due simply to convenience. There appears to be no regularity in the courses of component blocks of stone, and when necessity compelled, as in the courses laid on bowlders, which serve as a foundation, thin wedges of stone, or spalls, were inserted in the crevices. The walls are vertical, but the corners are sometimes far from perpendicular.

PL. XCVIII— THE RED ROCKS; TEMPLE CANYON BUREAU OF AMERICAN ETHNOLOGY—— SEVENTEENTH ANNUAL REPORT PL. XCVIII

THE RED ROCKS; TEMPLE CANYON

The front wall of inclosure B is pierced by three openings, the largest of which is a square passageway into the adjoining room, and is situated in the middle of the curved wall. A wooden lintel, which had been well hewn with stone implements, still remains in place above this passageway, and under it the visitor passes through a low opening which has the appearance of having been once a doorway. Above this entrance, on each side, in the wall, is a square hole, which originally may have been the points of support of floor beams. Formerly, likewise, there was a large square opening above the middle passageway, but this has been closed with masonry, leaving in place the wooden beam which once supported the wall above. The upper edge of the front wall of inclosure B is level, and is but little broken except in two places, where there are notches, one above each of the square holes already mentioned. It is probable that these depressions were intended for the ends of the beams which once supported a combined roof and floor.

On the perpendicular wall which forms the rear of inclosure B, many feet above the top of the standing front walls, there are several pictographs of Apache origin. The height of these above the level of the former roof would appear to indicate the existence of a third story, for the hands which drew them must have been at least 15 feet above the present top of the standing wall.

The front of C is curved like that of inclosure B, and is much broken near the foundations, where there is a passageway. There is a small hole on each side of a middle line, as in B, situated at about the same level as the floor, indicating the former position of a beam. Within the ruin there is a well-made partition separating inclosures B and C.

The size of room D is much less than that of B or C, but, with the exception of a section at the left, the front wall has fallen. The part which remains upright, however, stands like a pinnacle, unconnected with the face of the cliff or with the second-story wall of inclosure C. It is about 20 feet in height, and possibly its altitude appears greater than it really is from the fact that its foundations rest upon a bowlder nearly six feet high (plate cx).

The foundations of rooms E and F (plate c) are built on a lower level than those of B and C or D, and their front walls, which are really low, are helped out by similar bowlders, which serve as foundations. The indications are that both these inclosures were originally one story in height, forming a wing to the central section of the ruin, which had an additional tier of rooms. There is an entrance to F at the extreme left, and the whole room was lower than the floor of the lower stories of B, C, and D.

The most conspicuous pictograph on the cliff above Ruin i of Palatki, is a circular white figure, seen in the accompanying illustration. This pictograph is situated directly above the first room on the right, A, and was apparently made with chalk, so elevated that at present it is far above the reach of a person standing on any of the walls. From its general character I am led to believe that it was made by the Apache and not by the builders of the pueblo.

There were no names of white visitors anywhere on the walls of Palatki, which, so far as it goes, affords substantial support of my belief that we were the first white men to visit this ruin. While it can not be positively asserted that we were the original discoverers of this interesting building, there is no doubt that I was the first to describe it and to call attention to its highly characteristic architectural plan.

The walls of Palatki are not so massive as those of the neighboring Honanki, and the number of rooms in both ruins which form Palatki is much smaller. Each of these components probably housed not more than a few families, while several phratries could readily be accommodated in Honanki.

PL. XCIX— PALATKI (RUIN I) BUREAU OF AMERICAN ETHNOLOGY—— SEVENTEENTH ANNUAL REPORT PL. XCIX

PALATKI (RUIN I)

Ruin ii lies in a deep recess or cave, the roof of which forms a perfect arch above the walls. It is situated a few hundred feet to the west, and is easily approached by following the fallen débris at the foot of a perpendicular cliff. The front walls have all fallen, exposing the rear wall of what was formerly a row of rooms, as shown in the accompanying illustration (plate ci). There are evidences that this row of rooms was but a single story in height, while those behind it have indications of three stories. Ruin ii is more hidden by the trees and by its obscure position in a cavern than the former, but the masonry in both is of the same general character.

On approaching Ruin ii from Ruin i there is first observed a well-made though rough wall, as a rule intact, along which the line of roof and flooring can readily be traced (plate ci). In front of this upright wall are fragments of other walls, some standing in unconnected sections, others fallen, their fragments extending down the sides of the talus among the bushes. It was observed that this wall is broken by an entrance which passes into a chamber, which may be called A, and two square holes are visible, one on each side, above it. These holes were formerly filled by two logs, which once supported the floor of a second chamber, the line of which still remains on the upright wall. The small square orifice directly above the entrance is a peephole.

In examining the character of the wall it will be noticed that its masonry is in places rough cast, and that there was little attempt at regularity in the courses of the component stones, which are neither dressed nor aligned, although the wall is practically vertical.

At one point, in full view of the observer, a log is apparently inserted in the wall, and if the surrounding masonry be examined it will be found that an opening below it had been filled in after the wall was erected. It is evident, from its position relatively to the line indicating the roof, that this opening was originally a passageway from one room to another. Passing back of the standing wall an inclosure (room A) is entered, one side of which is the rock of the cliff, while the other three bounding walls are built of masonry, 20 feet high. This inclosure was formerly divided into an upper and a lower room by a partition, which served as the roof of the lower and the floor of the upper chambers. Two beams stretched across this inclosure about six feet above the débris of the present floor, and the openings in the walls, where these beams formerly rested, are readily observed. In the same way the beam-holes of the upper story may also be easily seen on the top of the wall. Between the rear wall of this inclosure and the perpendicular cliff there was a recess which appears to have been a dark chamber, probably designed for use as a storage room or granary. The configuration of the cliff, which forms the major part of the inclosing wall of this chamber, imparts to it an irregular or roughly triangular form.

The entire central portion of the ruin is very much broken down, and the floor is strewn to a considerable depth with the débris of fallen walls. On both sides there are nicely aligned, smoothly finished walls, with traces of beams on the level of former floors. Some of these bounding walls are curved; others are straight, and in places they rise 20 feet. Marks of fire are visible everywhere; most of the beams have been wrenched from their places, as a result of which the walls have been much mutilated, badly cracked, or thrown down.

There are no pictographs near this ruin, and no signs of former visits by white men.

Midway between Honanki and the second Palatki ruin a small ancient house of the same character as the latter was discovered. This ruin is very much exposed, and therefore the walls are considerably worn, but six well-marked inclosures, indicative of former rooms, were readily made out. No overarching rock shielded this ruin from the elements, and rubble from fallen walls covers the talus upon which it stands. The adobe mortar between the stones is much worn, and no fragment of plastering is traceable within or without. This evidence of the great weathering of the walls of the ruin is not considered indicative of greater age than the better preserved ruins in the neighborhood, but rather of exposure to the action of the elements. Not only are the walls in a very poor condition, but also the floors show, from the absence of dry soil upon them, that the whole ruin has suffered greatly from the same denudation. There are no fragments of pottery about it, and small objects indicating former habitation are also wanting. A cedar had taken root where the floor once was, and its present great size shows considerable age. If any pictographs formerly existed in the adjacent cliff they have disappeared. There is likewise no evidence that the Apache had ever sought it for shelter, or if they had, their occupancy occurred so long ago that time has effaced all evidence of their presence.

Honanki

The largest ruin visited in the Red-rock country was called, following Hopi etymology, Honanki; but the nomenclature was adopted not because it was so called by the Hopi, but following the rule elsewhere suggested.

PL. C— PALATKI (RUIN I) BUREAU OF AMERICAN ETHNOLOGY—— SEVENTEENTH ANNUAL REPORT PL. C

PALATKI (RUIN I)

This ruin lies under a lofty buttress of rock westward from Lloyd's canyon, which presented the only available camping place in its neighborhood. At the time of my visit there was but scanty water in the canyon and that not potable except for stock. We carried with us all the water we used, and when this was exhausted were obliged to retrace our steps to Oak creek. There are groves of trees in the canyon and evidences that at some seasons there is an abundant water supply. A corral had been made and a well dug near its mouth, but with these exceptions there were no evidences of previous occupancy by white men. We had hardly pitched our camp before tracks of large game were noticed, and before we left we sighted a bear which had come down to the water to drink, but which beat a hasty retreat at our approach. As previously stated, the knowledge of this ruin was communicated to me by Mr Schürmann.

Fig. 248—Ground plan of Honanki Fig. 248—Ground plan of Honanki

The Honanki ruin (figure 248) extends along the base of the cliff for a considerable distance, and may for convenience of description be divided into two sections, which, although generally similar, differ somewhat in structural features. The former is lineal in its arrangement, and consists of a fringe of houses extending along the base of the cliff at a somewhat lower level than the other. The walls of this section were for the greater part broken, and at no place could anything more than the foundation of the front wall be detected, although fragments of masonry strewed the sides of the declivity near its base. The house walls which remain are well-built parallel spurs constructed at right angles to the cliff, which served as the rear of all the chambers. At the extreme right end of this row of rooms, situated deep in a large cavern with overhanging roof, portions of a rear wall of masonry are well preserved, and the lateral walls of one or two chambers in this portion of the ruin are still intact. Straggling along from that point, following the contour of the base of the cliff under which it lies, there extends a long row of rooms, all destitute of a front wall.

The first division (plate cii), beginning with the most easterly of the series, is quite hidden at one end in a deep cavern. At this point the builders, in order to obtain a good rear wall to their rooms, constructed a line of masonry parallel with the face of the cliff. At right angles to this construction, at the eastern extremity, there are remnants of a lateral wall, but the remainder had tumbled to the ground. The standing wall of z is not continuous with that of the next room, y, and apparently was simply the rear of a large room with the remains of a lateral wall at right angles to it. The other walls of this chamber had tumbled into a deep gorge, overgrown with bushes which conceal the fragments. This building is set back deeply in the cave, and is isolated from the remaining parts of the ruin, although at the level which may have been its roof there runs a kind of gallery formed by a ledge of rock, plastered with adobe, which formerly connected the roof with the rest of the pueblo. This ledge was a means of intercommunication, and a continuation of the same ledge, in rooms s, t, and u, supported the rafters of these chambers. At u there are evidences of two stories or two tiers of rooms, but those in front have fallen to the ground.

The standing wall at u is about five feet high, connected with the face of the cliff by masonry. The space between it and the cliff was not large enough for a habitable chamber, and was used probably as a storage place. In front of the standing wall of room u there was another chamber, the walls of which now strew the talus of the cliff.

The highest and best preserved room of the second series of chambers at Honanki is that designated p, at a point where the ruin reached an elevation of 20 feet. Here we have good evidence of rooms of two stories, as indicated by the points of insertion of the beams of a floor, at the usual levels above the ground. In fact, it is probable that the whole section of the ruin was two stories high throughout, the front walls having fallen along the entire length. From the last room on the left to the eastern extremity of the line of houses which leads to the main ruin of Honanki, no ground plans were detected at the base of the cliffs, but fallen rocks and scattered débris are strewn over the whole interval.

The eastern part of the main ruin of Honanki, however, lies but a short distance west of that described, and consists of many similar chambers, arranged side by side. These are lettered in the diagram h to u, beginning with h, which is irregularly circular in form, and ends with a high wall, the first to be seen as one approaches the ruin from Lloyd canyon. This range of houses is situated on a lower foundation and at a lower level than that of the main quarter of Honanki, and a trail runs along so close to the rooms that the whole series is easily visited without much climbing. No woodwork remains in any of these rooms, and the masonry is badly broken in places either by natural agencies or through vandalism.

Beginning with h, the round room, which adjoins the main quarter of Honanki, we find much in its shape to remind us of a kiva. The walls are in part built on foundations of large bowlders, one of which formed the greater part of the front wall. This circular room was found to be full of fallen débris, and could not be examined without considerable excavation. If it were a kiva, which I very much doubt, it is an exception among the Verde valley ruins, where no true kiva has yet been detected.[27]

PL. CI— FRONT WALL OF PALATKI (RUIN II) BUREAU OF AMERICAN ETHNOLOGY—— SEVENTEENTH ANNUAL REPORT PL. CI

FRONT WALL OF PALATKI (RUIN II)

Room r (plate ciii) is a finely preserved chamber, with lateral walls 20 feet high, of well-constructed masonry, that in the rear, through which there is an opening leading into a dark chamber, occupying the space between it and the cliff. It is braced by connecting walls at right angles to the face of the solid rock.

At s, the face of the cliff forms a rear wall of the room, and one of the side walls is fully 20 feet high. The points of insertion of the flooring are well shown, about 10 feet from the ground, proving that the ruin at this point was at least two stories high.

Two walled inclosures, one within the other, characterize room u. On the cliff above it there is a series of simple pictographs, consisting of short parallel lines pecked into the rock, and are probably of Apache origin. This room closes the second series, along the whole length of which, in front of the lateral walls which mark different chambers, there are, at intervals, piles of débris, which enabled an approximate determination of the situation of the former front wall, fragments of the foundations of which are traceable in situ in several places.

The hand of man and the erosion of the elements have dealt harshly with this portion of Honanki, for not a fragment of timber now remains in its walls. This destruction, so far as human agency is concerned, could not have been due to white men, but probably to the Apache, or possibly to the cliff villagers themselves at the time of or shortly after the abandonment of the settlement.

From the second section of Honanki we pass to the third and best-preserved portion of the ruins (figure 249), indicated in the diagram from a to g. To this section I have referred as the "main ruin," for it was evidently the most populous quarter of the ancient cliff dwelling. It is better preserved than the remainder of Honanki, and is the only part in which all four walls of the chambers still remain erect. Built at a higher level than the series of rooms already considered, it must have towered above them, and possibly served as a place of retreat when danger beset the more exposed quarters of the village.

Fig. 249—The main ruin of Honanki Fig. 249—The main ruin of Honanki

Approaching the main ruin of Honanki (plate civ) from the east, or the parts already described, one passes between the buttress on which the front wall of the rounded room h is built and a fragment of masonry on the left, by a natural gateway through which the trail is very steep. On the right there towers above the visitor a well-preserved wall of masonry, the front of room a, and he soon passes abreast of the main portion of the ruin of Honanki. This section is built in a huge cavern, the overhanging roof of which, is formed by natural rock, arching far above the tops of the highest walls of the pueblo and suggesting the surroundings of the "Cliff Palace" of Mesa Verde, so well described by the late Baron G. Nordenskiöld in his valuable monograph on the ruins of that section of southern Colorado. The main ruin of Honanki is one of the largest and best preserved architectural monuments of the former people of Verde valley that has yet been described. Although somewhat resembling its rival, the well-known "Casa Montezuma" of Beaver creek, its architecture is dissimilar on account of the difference in the form of the cavern in which it is built and the geological character of the surrounding cliffs. Other Verde ruins may have accommodated more people, when inhabited, but none of its type south of Canyon de Chelly have yet been described which excel it in size and condition of preservation. I soon found that our party were not the first whites who had seen this lonely village, as the names scribbled on its walls attested; but so far as I know it had not previously been visited by archeologists.

PL. CII— HONANKI (RUIN II) BUREAU OF AMERICAN ETHNOLOGY—— SEVENTEENTH ANNUAL REPORT PL. CII

HONANKI (RUIN II)

In the main portion of Honanki we found that the two ends of the crescentic row of united rooms which compose it are built on rocky elevations, with foundations considerably higher than those of the rooms in the middle portion of the ruins. The line of the front wall is, therefore, not exactly crescentic, but irregularly curved (figure 249), conforming to the rear of the cavern in which the houses are situated. About midway in the curve of the front walls two walls indicative of former rooms extend at an angle of about 25° to the main front wall. All the component rooms of the main part of Honanki can be entered, some by external passageways, others by doorways communicating with adjacent chambers. None of the inclosures have roofs or upper floors, although indications of the former existence of both these structural features may readily be seen in several places. Although wooden beams are invariably wanting, fragments of these still project from the walls, almost always showing on their free ends, inside the rooms, the effect of fire. I succeeded in adding to the collection a portion of one of these beams, the extremity of which had been battered off, evidently with a stone implement. In the alkaline dust which covered the floor several similar specimens were seen.

The stones which form the masonry of the wall (figure 250) were not, as a rule, dressed or squared before they were laid with adobe mortar, but were generally set in place in the rough condition in which they may still be obtained anywhere under the cliff.

All the mortar used was of adobe or the tenacious clay which serves so many purposes among the Pueblos. The walls of the rooms were plastered with a thick layer of the same material. The rear wall of each room is the natural rock of the cliff, which rises vertically and has a very smooth surface. The great natural archway which covers the whole pueblo protects it from wind and rain, and as a consequence, save on the front face, there are few signs of natural erosion. The hand of man, however, has dealt rudely with this venerable building, and many of the walls, especially of rooms which formerly stood before the central portion, lie prone upon the earth; but so securely were the component stones held together by the adobe that even after their fall sections of masonry still remain intact.

Fig. 250—Structure of wall of Honanki Fig. 250—Structure of wall of Honanki

There are seven walled inclosures in the main part of Honanki, and as each of these was formerly at least two stories high there is substantial evidence of the former existence of fourteen rooms in this part of the ruin. There can be little doubt that there were other rooms along the front of the central portion, and the fallen walls show them to have been of large size. It would likewise appear that the middle part was higher than the two wings, which would increase the number of chambers, so that with these additions it may safely be said that this part of Honanki alone contained not far from twenty rooms.

PL. CIII— WALLS OF HONANKI BUREAU OF AMERICAN ETHNOLOGY—— SEVENTEENTH ANNUAL REPORT PL. CIII

WALLS OF HONANKI

Room c is the first of a series of chambers, with external passageways, but its walls are very much broken down, and the openings thereby enlarged. The front wall is almost straight and in one place stands 30 feet, the maximum height of the standing wall of the ruins. In one corner a considerable quantity of ashes and many evidences of fire, some of which may be ascribed to Apache occupants, was detected. A wooden beam, marking the line of the floor of a second story, was seen projecting from the front wall, and there are other evidences of a floor at this level. Large beams apparently extended from the front wall to the rear of the chamber, where they rested on a ledge in the cliff, and over these smaller sticks were laid side by side and at right angles to the beams. These in turn supported either flat stones or a layer of mud or clay. The method of construction of one of these roofs is typical of a Tusayan kiva, where ancient architectural forms are adhered to and best preserved.

The entrance to room d is very much enlarged by the disintegration of the wall, and apparently there was at this point a difference in level of the front wall, for there is evidence of rooms in advance of those connected with the chambers described, as shown by a line of masonry, still standing, parallel to the front face of inclosures c and d.

Room e communicates by a doorway with the chamber marked f, and there is a small window in the same partition. This room had a raised banquette on the side toward the cliff, recalling an arrangement of the floor similar to that in the cavate dwellings opposite Squaw mountain which I have described. This platform is raised about three feet above the remainder of the floor of f, and, like it, is strewn with large slabs of stone, which have fallen from the overhanging roof. In the main floor, at one corner, near the platform, there is a rectangular box-like structure made of thin slabs of stone set on edge, suggesting the grinding bins of the Pueblos. Room f communicates with g by a passageway which has a stone lintel. The holes in the walls, in which beams were once inserted, are seen in several places at different levels above the floor. The ends of several beams, one extremity of which is invariably charred, were found set in the masonry, and others were dug from the débris in the floor.

As a result of the curve in the front wall of the ruin at that point, the shape of room f is roughly quadrate, with banquettes on two sides. There are six large beam holes in the walls, and the position of the first floor is well shown on the face of the partition, separating f from g. The passageway from one of these rooms to the other is slightly arched.

Room g is elongated, without an external entrance, and communicates with f by a small opening, through which it is very difficult to crawl. Its longest dimension is almost at right angles to the front face of the remaining rooms, and it is raised above them by its foundation on an elevated rock like that of a, b, and c. There is a small, square, external opening which may have served as the position of a former beam or log. The upper level of the front wall is more or less broken down in places, and formerly may have been much higher. Beyond g a spur of masonry is built at right angles to the cliff, inclosing a rectangular chamber at the end of the ruin which could not be entered. Possibly in former times it was accessible by means of a ladder from the roof, whence communication with other portions of the structure was also had.

PL. CIV— APPROACH TO MAIN PART OF HONANKI BUREAU OF AMERICAN ETHNOLOGY—— SEVENTEENTH ANNUAL REPORT PL. CIV

APPROACH TO MAIN PART OF HONANKI

A circle painted green and inclosed in a border of yellow is undoubtedly of Apache origin. There is at one point a row of small pits, arranged in line, suggesting a score or enumeration of some kind, and a series of short parallel lines of similar import was found not far away. This latter method of recording accounts is commonly used at the present time in Tusayan, both in houses and on cliffs; and one of the best of these, said to enumerate the number of Apache killed by the Hopi in a raid many years ago, may be seen above the trail by which the visitor enters the pueblo of Hano on the East Mesa. The names of several persons scratched on the face of the cliff indicate that Americans had visited Honanki before me.

The majority of the paleoglyphs at both Palatki and Honanki are of Apache origin, and are of comparatively modern date, as would naturally be expected. In some instances their colors are as fresh as if made a few years ago, and there is no doubt that they were drawn after the building was deserted by its original occupants. The positions of the pictographs on the cliffs imply that they were drawn before the roofs and flooring had been destroyed, thus showing how lately the ruin preserved its ancient form. In their sheltered position there seems to be no reason why the ancient pictographs should not have been preserved, and the fact that so few of the figures pecked in the cliff now remain is therefore instructive.

One of the first tendencies of man in visiting a ruin is to inscribe his name on its walls or on neighboring cliffs. This is shared by both Indians and whites, and the former generally makes his totem on the rock surface, or adds that of his gods, the sun, rain-cloud, or katcinas. Inscriptions recording events are less common, as they are more difficult to indicate with exactitude in this system of pictography. The majority of ancient pictographs in the Red-rock country, like those I have considered in other parts of Verde valley, are identical with picture writings now made in Tusayan, and are recognized and interpreted without hesitation by the Hopi Indians. In their legends, in which the migrations of their ancestors are recounted, the traditionists often mention the fact that their ancestors left their totem signatures at certain points in their wanderings. The Patki people say that you will find on the rocks of Palatkwabi, the "Red Land of the South" from which they came, totems of the rain-cloud, sun, crane, parrot, etc. If we find these markings in the direction which they are thus definitely declared to exist, and the Hopi say similar pictures were made by their ancestors, there seems no reason to question such circumstantial evidence that some of the Hopi clans once came from this region.[28]

One of the most interesting of the pictographs pecked in the rock is a figure which, variously modified, is a common decoration on cliff-dweller pottery from the Verde valley region to the ruins of the San Juan and its tributaries. This figure has the form of two concentric spirals, the ends of which do not join. As this design assumes many modifications, it may be well to consider a few forms which it assumes on the pottery of the cliff people and on that of their descendants, the Pueblos.

The so-called black-and-white ware, or white pottery decorated with black lines, which is so characteristic of the ceramics of the cliff-dwellers, is sometimes, as we shall see, found in ruins like Awatobi and Sikyatki; but it is so rare, as compared with other varieties, that it may be regarded as intrusive.

One of the simplest forms of the broken-line motive is a Greek fret, in which there is a break in the component square figures or where the line is noncontinuous. In the simplest form, which appears prominently on modern pottery, but which is rare or wanting on true black-and-white ware, we have two crescentic figures, the concavities of which face in different directions, but the horns overlap. This is a symbol which the participants in the dance called the Húmiskatcina still paint with pigments on their breasts, and which is used on shields and various religious paraphernalia.

A study of any large collection of decorated Pueblo ware, ancient or modern, will show many modifications of this broken line, a number of which I shall discuss more in detail when pottery ornamentation is considered. A design so distinctive and so widespread as this must certainly have a symbolic interpretation. The concentric spirals with a broken line, the Hopi say, are symbols of the whirlpool, and it is interesting to find in the beautiful plates of Chavero's Antigüedades Mexicanas that the water in the lagoon surrounding the ancient Aztec capital was indicated by the Nahuatl Indians with similar symbols.

Objects Found at Palatki and Honanki

The isolation of these ruins and the impossibility of obtaining workmen, combined with the brief visit which I was able to make to them, rendered it impossible to collect very many specimens of ancient handiwork. The few excavations which were made were limited almost wholly to Honanki, and from their success I can readily predict a rich harvest for anyone who may attempt systematic work in this virgin field. We naturally chose the interior of the rooms for excavation, and I will say limited our work to these places. Every chamber was more or less filled with débris—fragments of overturned walls, detached rock from the cliff above, dry alkaline soil, drifted sand, dust, and animal excreta. In those places where digging was possible we found the dust and guano so dry and alkaline that it was next to impossible to work for any length of time in the rooms, for the air became so impure that the workmen could hardly breathe, especially where the inclosing walls prevented ventilation. Notwithstanding this obstacle, however, we removed the accumulated débris down to the floor in one or two chambers, and examined with care the various objects of aboriginal origin which were revealed.

In studying the specimens found in cliff-houses due attention has not always been given to the fact that occupants have oftentimes camped in them subsequently to their abandonment by the original builders. As a consequence of this temporary habitation objects owned by unrelated Indians have frequently been confused with those of the cliff-dwellers proper. We found evidences that both Honanki and Palatki had been occupied by Apache Mohave people for longer or shorter periods of time, and some of the specimens were probably left there by these inhabitants.

The ancient pottery found in the rooms, although fragmentary, is sufficiently complete to render a comparison with known ceramics from the Verde ruins. Had we discovered the cemeteries, for which we zealously searched in vain, no doubt entire vessels, deposited as mortuary offerings, would have been found; but the kind of ware of which they were made would undoubtedly have been the same as that of the fragments.

No pottery distinctively different from that which has already been reported from the Verde valley ruins was found, and the majority resembled so closely in texture and symbolism that of the cliff houses of the San Juan, in northern New Mexico and southern Utah, that they may be regarded as practically identical.

The following varieties of pottery were found at Honanki:

I. Coiled ware.
II. Indented ware.
III. Smooth ware.
IV. Smooth ware painted white, with black geometric figures.
V. Smooth red ware, with black decoration.

By far the largest number of fragments belong to the first division, and these, as a rule, are blackened by soot, as if used in cooking. The majority are parts of large open-mouth jars with flaring rims, corrugated or often indented with the thumb-nail or some hard substance, the coil becoming obscure on the lower surface. The inside of these jars is smooth, but never polished, and in one instance the potter used the corrugations of the coil as an ornamental motive. The paste of which this coiled ware was composed is coarse, with argillaceous grains scattered through it; but it was well fired and is still hard and durable. When taken in connection with its tenuity, these features show a highly developed potter's technique. A single fragment is ornamented with an S-shape coil of clay fastened to the corrugations in much the same way as in similar ware from the ruins near the Colorado Chiquito.

The fragments of smooth ware show that they, too, had been made originally in the same way as coiled ware, and that their outer as well as their inner surface had been rubbed smooth before firing. As a rule, however, they are coarse in texture and have little symmetry of form. Fragments identified as parts of bowls, vases, jars, and dippers are classed under this variety. As a rule they are badly or unevenly fired, although evidently submitted to great heat. There was seldom an effort made to smooth the outer surface to a polish, and no attempt at pictorial ornamentation was made.

The fragments represented in classes iv and v were made of a much finer clay, and the surface bears a gloss, almost a glaze. The ornamentation on the few fragments which were found is composed of geometric patterns, and is identical with the sherds from other ruins of Verde valley. A fragment each of a dipper and a ladle, portions of a red bowl, and a rim of a large vase of the same color were picked up near the ruin. Most of the fragments, however, belong to the first classes—the coiled and indented wares.

There was no evidence that the former inhabitants of these buildings were acquainted with metals. The ends of the beams had been hacked off evidently with blunt stone axes, aided by fire, and the lintels of the houses were of split logs which showed no evidence that any metal implement was used in fashioning them. We found, however, several stone tools, which exhibit considerable skill in the art of stone working. These include a single ax, blunt at one end, sharpened at the other, and girt by a single groove. The variety of stone from which the ax was made does not occur in the immediate vicinity of the ruin. There were one or two stone hammers, grooved for hafting, like the ax. A third stone maul, being grooveless, was evidently a hand tool for breaking other stones or for grinding pigments.

Fig. 251—Stone implement from Honanki Fig. 251—Stone implement from Honanki

Perhaps the most interesting stone implement which was found was uncovered in the excavation of one of the middle rooms of the western part of the ruin, about three feet below the surface. It consists of a wooden handle rounded at each end and slightly curved, with a sharpened stone inserted midway of its length and cemented to the wood with pitch or asphaltum. The stone of this implement would hardly bear rough usage, or sustain, without fracture, a heavy blow. The edge is tolerably sharp, and it therefore may have been used in skinning animals. Judging from the form of the handle, the implement is better suited for use as a scraper than for any other purpose which has occurred to me (figure 251).

The inhabitants of the two ruins of the Red-rocks used obsidian arrowpoints with shafts of reeds, and evidently highly regarded fragments of the former material for knives, spearheads, and one or two other purposes.

The stone metates from these ruins are in no respect characteristic, and several fine specimens were found in place on the floors of the rooms. One of these was a well-worn specimen of lava, which must have been brought from a considerable distance, since none of that material occurs in the neighborhood. The existence of these grinding stones implies the use of maize as food, and this evidence was much strengthened by the finding of corncobs, kernels of corn, and charred fragments at several points below the surface of the débris in the chambers of Honanki. One of these grinding stones was found set in the floor of one of the rooms in the same way that similar metates may be seen in Walpi today.

Of bone implements, our limited excavations revealed only a few fragments. Leg bones of the turkey were used for awls, bodkins, needles, and similar objects. In general character the implements of this kind which were found are almost identical in form with the bone implements from Awatobi and Sikyatki, which are later figured and described. Although the bone implements unearthed were not numerous, we were well repaid for our excavations by finding an ancient fireboard, identical with those now used at Tusayan in the ceremony of kindling "new fire," and probably universally used for that purpose in former times. The only shell was a fragment of a bracelet made from a Pectunculus, a Pacific coast mollusk highly esteemed in ancient times among prehistoric Pueblos. The majority of the wooden objects found showed marks of fire, which were especially evident on the ends of the roof and floor beams projecting from the walls.

Fig. 252—Tinder tube from Honanki Fig. 252—Tinder tube from Honanki

There were many specimens of basketry found on the surface of the rubbish of the floors which, from the position of their occurrence and from their resemblance to the wickerwork still used by the Apache, seem without doubt to have been left there by temporary occupants of the rooms. There were likewise many wisps of yucca fiber tied in knots which must probably be regarded as of identical origin. The Yucca baccata affords the favorite fiber used by the natives at the present time, and it appears to have been popular for that purpose among the ancients.

Several specimens of sandals, some of which are very much worn on the soles, were found buried at the floor level. These are all of the same kind, and are made of yucca leaves plaited in narrow strips. The mode of attachment to the foot was evidently by a loop passing over the toes. Hide and cloth sandals have as yet not been reported from the Red-rock ruins of Verde valley. These sandals belonged to the original occupants of the cliff houses.

Fabrics made of cotton are common in the ruins of the Red-rocks, and at times this fiber was combined with yucca. Some of the specimens of cotton cloth were finely woven and are still quite strong, although stained dark or almost black. Specimens of netting are also common, and an open-mesh legging, similar to the kind manufactured in ancient times by the Hopi and still worn by certain personators in their sacred dances, were taken from the western room of Honanki. There were also many fragments of rope, string, cord, and loosely twisted bands, resembling head bands for carrying burdens.

A reed (figure 252) in which was inserted a fragment of cotton fiber was unlike anything yet reported from cliff houses, and as the end of the cotton which projected beyond the cavity of the reed was charred, it possibly was used as a slow-match or tinder-box.

Several shell and turquois beads were found, but my limited studies of the cliff-houses revealed only a few other ornaments, among them being beads of turkey-bone and a single wristlet fashioned from a Pectunculus. One or two fragments of prayer-sticks were discovered in a rock inclosure in a cleft to the west of the ruin.

Conclusions Regarding the Verde Valley Ruins

The ruins of the Verde region closely resemble those of Tusayan, and seem to support the claim of the Hopi that some of their ancestors formerly lived in that region. This is true more especially of the villages of the plains and mesa tops, for neither cave-houses nor cavate dwellings are found in the immediate vicinity of the inhabited Tusayan pueblos. The objects taken from the ruins are similar to those found universally over the pueblo area, and from them alone we can not say more than that they probably indicate the same substratum of culture as that from which modern pueblo life with its many modifications has sprung.

The symbolism of the decorations on the fragments of pottery found in the Verde ruins is the same as that of the ancient pueblos of the Colorado Chiquito, and it remains to be shown whether the ancestors of these were Hopi or Zuñi. I believe it will be found that they were both, or that when the villages along the Colorado Chiquito[29] were abandoned part of the inhabitants went to the mesas of Tusayan and others migrated farther up the river to the Zuñi villages.

Two centers of distribution of cliff houses occur in our Southwest: those of the upper tributaries of the Colorado in the north and the cliff houses of the affluents of the Salt and the Gila in the south. The watershed of the Rio Grande is, so far as is known, destitute of this kind of aboriginal dwellings. Between the two centers of distribution lie the pueblos of the Little Colorado and its tributaries, the home of the ancestors of the Hopi and the Zuñi. The many resemblances between the cliff houses of the north and those of the south indicate that the stage of culture of both was uniform, and probably the same conditions of environment led both peoples to build similar dwellings. All those likenesses which can be found between the modern Zuñi and the Hopi to the former cliff peoples of the San Juan region in the north, apply equally to those of the upper Salado and the Gila and their tributaries to the south; and so far as arguments of a northern origin of either, built on architectural or technological resemblances, are concerned, they are not conclusive, since they are also applicable to the cliff peoples of the south. The one important difference between the northern and the southern tier of cliff houses is the occurrence of the circular kiva, which has never been reported south of the divide between the Little Colorado and the Gila-Salado drainage. If a kiva was a feature in southern cliff houses, which I doubt, it appears to have been a rectangular chamber similar to a dwelling room. The circular kiva exists in neither the modern Hopi nor the Zuñi pueblos, and it has not been found in adjacent Tusayan ruins; therefore, if these habitations were profoundly influenced by settlers from the north, it is strange that such a radical change in the form of this room resulted. The arguments advanced that one of the two component stocks of the Zuñi, and that the aboriginal, came from the cliff peoples of the San Juan, are not conclusive, although I have no doubt that the Zuñi may have received increment from that direction.

Cushing has, I believe, furnished good evidence that some of the ancestors of the Zuñi population came from the south and southwest; and that some of these came from pueblos now in ruins on the Little Colorado is indicated by the great similarity in the antiquities of ancient Zuñi and the Colorado Chiquito ruins. Part of the Patki people of the Hopi went to Zuñi and part to Tusayan, from the same abandoned pueblo, and the descendants of this family in Walpi still recognize this ancient kinship; but I do not know, and so far as can be seen there is no way of determining, the relative antiquity of the pueblos in Zuñi valley and those on the lower Colorado.

The approximate date of the immigration of the Patki people to Tusayan is as yet a matter of conjecture. It may have been in prehistoric times, or more likely at a comparatively late period in the history of the people. It seems well substantiated, however, that when this Water-house people joined the other Hopi, the latter inhabited pueblos and were to all intents a pueblo people. If this hypothesis be a correct one, the Snake, Horn, and Bear peoples, whom the southern colonists found in Tusayan, had a culture of their own similar to that of the people from the south. Whence that culture came must be determined by studies of the component clans of the Hopi before the arrival of the Patki people.[30]

The origin of the round shape of the estufa, according to Nordenskiöld (p. 168), is most easily explained on the hypothesis that it is a reminiscence of the cliff-dwellers' nomadic period. "There must be some very cogent reason for the employment of this shape," he says, "for the construction of a cylindrical chamber within a block of rectangular rooms involves no small amount of labor. We know how obstinately primitive nations cling to everything connected with their religious ideas. Then what is more natural than the retention, for the room where religious ceremonies were performed, of the round shape characteristic of the original dwelling place, the nomadic hut? This assumption is further corroborated by the situation of the hearth and the structure of the roof of the estufa, when we find points of analogy to the method employed by certain nomadic Indians in the erection of their huts." This theory of the origin of the round form of dwelling and its retention in the architecture of the kiva, advanced by Nordenskiöld in 1893, has much in its favor, but the rectangular form, which, so far as known, is the only shape of these sacred rooms in the Tusayan region, is still unexplained. From Castañeda's narrative of the Coronado expedition it appears that in the middle of the sixteenth century the eastern pueblos had both square and round estufas or kivas, and that these kivas belonged to the men while the rooms of the pueblo were in the possession of the women. The apparent reason why we find no round rooms or kivas in the southern cliff houses and in Tusayan may be due to several causes. Local conditions, including the character of the building sites on the Hopi mesa, made square rooms more practical, or the nomadic stage was so far removed that the form of the inclosure in which the ancients held their rites had not been preserved. Moreover, some of the most ancient and secret observances at Walpi, as the Flute ceremony, are not performed in special kivas, but take place in ordinary living rooms.

As in all the other ruins of Verde valley, circular kivas are absent in the Red-rock country, and this fact, which has attracted the attention of several observers, is, I believe, very significant. Although as yet our knowledge of the cliff houses of the upper Gila and Salado and their numerous tributaries is very fragmentary, and generalization on that account unsafe, it may be stated provisionally that no circular kivas have yet been found in any ruins of the Gila-Salado watershed. This form of kiva, however, is an essential feature of the cliff dwellings of Rio Colorado, especially of those along its affluents in southern Colorado and northern New Mexico. Roughly speaking, then, the circular kiva is characteristic of the ruins of this region and of certain others in the valley of the Rio Grande, where they still survive in inhabited pueblos.

Circular ruins likewise are limited in their distribution in the Southwest, and it is an interesting fact that the geographic distribution of ancient pueblos of this form is in a general way the same as that of circular kivas. There are, of course, many exceptions, but so far as I know these can readily be explained. No ruins of circular dwellings occur in the Gila-Salado drainage area, where likewise no circular kivas have been observed. Moreover, the circular form of dwelling and kiva is distinctively characteristic of prehistoric peoples east of Tusayan, and the few instances of their occurrence on its eastern border can readily be explained as extra-Hopi.

The explanation of these circular kivas advanced by Nordenskiöld and the Mindeleffs, that they are survivals of round habitations of nomads, has much to commend it; but whether sufficient or not, the geographic limitation of these structures tells in favor of the absence of any considerable migration of the prehistoric peoples of the upper Colorado and Rio Grande watersheds southward into the drainage area of the Gila-Salado. Had the migration been in that direction it may readily be believed that the round kiva and the circular form of dwelling would have been brought with it.

The round kiva has been regarded as a survival of the form of the original homes of the nomad, when he became a sedentary agriculturist by conquest and marriage.

The presence of rectangular kivas in the same areas in which round kivas occur does not necessarily militate against this theory, nor does it oblige us to offer an explanation of a necessarily radical change in architecture if we would derive it from a circular form. It would indeed be very unusual to find such a change in a structure devoted to religious purposes where conservatism is so strong. The rectangular kiva is the ancient form, or rather the original form; the round kiva is not a development from it, but an introduction from an alien people. It never penetrated southward of the Colorado and upper Rio Grande drainage areas because the element which introduced it in the north was never strong enough to influence the house builders of the Gila-Salado and tributary valleys.