The Project Gutenberg eBook, The Cliff Ruins of Canyon de Chelly, Arizona, by Cosmos Mindeleff This eBook is for the use of anyone anywhere at no cost and with almost no restrictions whatsoever. You may copy it, give it away or re-use it under the terms of the Project Gutenberg License included with this eBook or online at www.gutenberg.org Title: The Cliff Ruins of Canyon de Chelly, Arizona Sixteenth Annual Report of the Bureau of Ethnology to the Secretary of the Smithsonian Institution, 1894-95, Government Printing Office, Washington, 1897, pages 73-198 Author: Cosmos Mindeleff Release Date: November 6, 2006 [eBook #19723] Language: English Character set encoding: ISO-8859-1 ***START OF THE PROJECT GUTENBERG EBOOK THE CLIFF RUINS OF CANYON DE CHELLY, ARIZONA*** E-text prepared by Louise Hope, Carlo Traverso, and the Project Gutenberg Online Distributed Proofreading Team (http://www.pgdp.net/) from psge images generously made available by the Bibliothèque nationale de France (BnF/Gallica) (http://gallica.bnf.fr/) Note: This document is taken from the Sixteenth Annual Report of the Bureau of Ethnology to the Secretary of the Smithsonian Institution, 1894-95, Government Printing Office, Washington, 1897, pages 73-198. Images of the original pages are available through the Bibliothèque nationale de France (BnF/Gallica) http://gallica.bnf.fr/. Transcriber's note: Several Figures, with their captions, were printed sideways in the original book. Their position has been retained for this e-text, rotated 180° where necessary to keep all oriented in the same direction (top of Figure on the left). Each of these Figures links to a separate illustration at the same scale but horizontally positioned. A few typographical errors have been corrected. They have been marked in the text with mouse- hover popups. The spelling "bowlder" is standard for this publication. Brackets within quotations are in the original. THE CLIFF RUINS OF CANYON DE CHELLY, ARIZONA BY COSMOS MINDELEFF CONTENTS 73 75 Page Introduction 79 History and literature 79 Geography 82 Classification and descriptions 89 Ruins of the pueblo region 89 I—Old villages on open sites 93 II—Home villages on bottom lands 94 III—Home villages located for defense 111 IV—Cliff outlooks or farming shelters 142 Details 153 Sites 153 Masonry 159 Openings 164 Roofs, floors, and timber work 165 Storage and burial cists (Navaho) 166 Defensive and constructive expedients 170 Kivas or sacred chambers 174 Chimney-like structures 182 Traditions 190 Conclusions 191 Index Footnotes Numbered sites (list added by transcriber). Text references to "the map" are to Plate XLIII. Site 1 a, b, Plate LXI, Figure 77 Site 2, Plate LV Site 3 a, b, c, Figure 10, Plate LVIII Site 4 a, b, Plate XLVI, Figure 73 Site 6 a, b, Plate LIII, Plate LXII Site 8, Plate LVI; Site 10 a, b, c, d, e, Figure 23, Figure 24, Figure 25, Figure 71, Figure 72, Plate LIX, Plate LXIII; Site 12, Plate XLIV Site 14, Figure 57 Site 15 a, b, Figure 19, Figure 78, Figure 79; Site 16 a, b, c, d, e, f, Figure 27, Figure 80, Plate LI, Plate LVII, Plate LX; Site 17 Site 28, Figure 3 Site 31 a, b, c, d, Figure 20, Figure 70 Site 32 a, b, Figure 21, Figure 69 Site 34, Figure 45 Site 35, Figure 46 Site 36, Figure 35 Site 37 a, b, Figure 47, Figure 74 Site 43, Figure 13 Site 44, Figure 36 Site 45, Figure 63 Site 46, Figure 48 Site 47, Figure 40 Site 48, Figure 17, Figure 18 Site 49, Figure 29 Site 52 Site 58 Site 60 Site 62, Figure 30 Site 64, Figure 62 Site 69, Figure 41 ILLUSTRATIONS Illustrations have been placed as close as practicable to their discussion in the text. Page numbers refer to their original location in the printed book: Figures on text pages, Plates facing text pages. Page Plate XLI. Map of the ancient pueblo region, showing location of Canyon de Chelly 79 XLII. Map of Canyon de Chelly and its branches 85 XLIII. Detailed map of part of Canyon de Chelly, showing areas of cultivable land 93 XLIV. Section of old walls, Canyon de Chelly 95 XLV. General view of ruin on bottom land, Canyon del Muerto 97 XLVI. Village ruin in Canyon de Chelly 103 77 XLVII. Casa Blanca ruin, Canyon de Chelly 105 XLVIII. Mummy cave, central and eastern part 112 XLIX. Eastern cove of Mummy cave 115 L. Reservoir in ruin No. 10 127 LI. Small village, ruin No. 16, Canyon de Chelly 129 LII. Walls resting on refuse in ruin No. 16 131 LIII. Cliff outlook in lower Canyon de Chelly 149 LIV. Cliff ruin No. 14 151 LV. Site marked by pictographs 153 LVI. Site difficult of approach 159 LVII. Masonry in Canyon de Chelly 161 LVIII. Chinked walls in Canyon de Chelly 163 LIX. A partly plastered wall 165 LX. Plastered wall in Canyon de Chelly 167 LXI. Storage cist in Canyon de Chelly 169 LXII. Navaho burial cists 171 LXIII. Kivas in ruin No. 10, showing second-story walls 173 Figure 1. Ground plan of an old ruin in Canyon del Muerto 95 2. Ground plan of a ruin on bottom land in Canyon del Muerto 96 3. Ground plan of small ruin in Canyon de Chelly 96 4. Granary in the rocks, connected with a ruin 97 5. Ground plan of a ruin in a cave 98 6. Ground plan of Pakashi-izini ruin, Canyon del Muerto 99 7. Ground plan of a ruin in Canyon del Muerto 100 8. Ground plan of a ruin in Tseonitsosi canyon 100 9. Ground plan of a much obliterated ruin 101 10. Ground plan of a ruin in Canyon de Chelly 101 11. Ground plan of a village ruin 103 12. Ground plan of kivas in Canyon de Chelly 103 13. Ground plan of a small ruin on bottom land 104 14. Ground plan of the upper part of Casa Blanca ruin 105 15. Ground plan of the lower part of Casa Blanca ruin 106 16. Ground plan of Mummy Cave ruin 113 17. Ruin in a rock cove 117 18. Ground plan of a ruin in a rock cove 117 19. Ground plan of a ruin on a ledge 118 20. Ground plan of ruin No. 31, Canyon de Chelly 119 21. Ground plan of ruin No. 32, Canyon de Chelly 120 22. Section of a kiva wall 122 23. Ruin No. 10 on a ledge in a cove 123 24. Ground plan of ruin No. 10 124 25. Oven-like structure in ruin No. 10 127 26. Plan of oven-like structure 128 27. Ground plan of a small village, ruin No. 16 129 28. Ruins on a large rock 130 29. Ground plan of ruins No. 49 131 30. Ruins on an almost inaccessible site 133 31. Ground plan of a large ruin in Canyon del Muerto 134 32. Ground plan of a small ruin in Canyon del Muerto 135 33. Ground plan of a small ruin 135 34. Plan of a ruin of three rooms 136 35. Ground plan of a small ruin, with two kivas 136 78 36. Ground plan of a small ruin, No. 44 137 37. Ground plan of a ruin on a rocky site 137 38. Rock with cups and petroglyphs 138 39. Ground plan of a ruin in Canyon de Chelly 139 40. Site showing recent fall of rock 140 41. Ruin No. 69 in a branch canyon 140 42. Ground plan of a small ruin in Canyon del Muerto 140 43. Ground plan of a small ruin 141 44. Plan of a ruin with curved inclosing wall 141 45. Ground plan of ruin No. 34 142 46. Ground plan of cliff outlook No. 35 143 47. Plan of a cliff outlook 143 48. Plan of cliff ruin No. 46 144 49. Plan of cliff room with partitions 145 50. Plan of a large cliff outlook in Canyon del Muerto 145 51. Plan of a cluster of rooms in Canyon del Muerto 146 52. White House ruin in Tseonitsosi canyon 146 53. Ground plan of a ruin in Tseonitsosi canyon 147 54. Plan of rooms against a convex cliff 147 55. Small ruin with curved wall 147 56. Ground plan of a cliff outlook 148 57. Plan of cliff outlook No. 14, in Canyon de Chelly 148 58. Ground plan of outlooks in a cleft 149 59. Plan of a single-room outlook 149 60. Three-room outlook in Canyon del Muerto 150 61. Plan of a two-room outlook 150 62. Plan of outlook and burial cists, No. 64 150 63. Plan of rectangular room, No. 45 151 64. Rectangular single room 151 65. Single-room remains 152 66. Site apparently very difficult of access 158 67. Notched doorway in Canyon de Chelly 164 68. Cist composed of upright slabs 169 69. Retaining walls in Canyon de Chelly 172 70. Part of a kiva in ruin No. 31 175 71. Plan of part of a kiva in ruin No. 10 176 72. Kiva decoration in white 177 73. Pictograph in white 178 74. Markings on cliff wall, ruin No. 37 178 75. Decorative band in kiva in Mummy Cave ruin 179 76. Design employed in decorative band 180 77. Pictographs in Canyon de Chelly 181 78. Plan of chimney-like structure in ruin No. 15 182 79. Section of chimney-like structure in ruin No. 15 183 80. Plan of chimney-like structure in ruin No. 16 184 81. Section of chimney-like structure in ruin No. 16 185 82. Plan of the principal kiva in Mummy Cave ruin 186 83. Chimney-like structure in Mummy Cave ruin 187 large map Plate XLI. Ancient Pueblo Region Showing Location of Canyon De Chelly Larger View THE CLIFF RUINS OF CANYON DE CHELLY, ARIZONA By Cosmos Mindeleff INTRODUCTION HISTORY AND LITERATURE Although Canyon de Chelly is one of the best cliff-ruin regions of the United States, it is not easily accessible and is practically unknown. At the time of the conquest of this country by the "Army of the West" in 1846, and of the rush to California in 1849, vague rumors were current of wonderful "cities" built in the cliffs, but the position of the canyon in the heart of the Navaho country apparently prevented exploration. In 1849 it was found necessary to make a demonstration against these Indians, and an expedition was sent out under the command of Colonel Washington, then governor of New Mexico. A detachment of troops set out from Santa Fé, and was accompanied by Lieutenant (afterward General) J. H. Simpson, of the topographical engineers, to whose indefatigable zeal for investigation and carefulness of observation much credit is due. He was much interested in the archeology of the country passed over and his descriptions are remarkable for their freedom from the exaggerations and erroneous observations which characterize many of the publications of that period. His journal was published by Congress the next year1 and was also printed privately. The expedition camped in the Chin Lee valley outside of Canyon de Chelly, and Lieutenant Simpson made a side trip into the canyon itself. He mentions ruins noticed by him at 4½, 5, and 7 miles from the mouth; the latter, the ruin subsequently known as Casa Blanca, he describes at some length. He also gives an illustration drawn by R. H. Kern, which is very bad, and pictures some pottery fragments found near or in the ruin. The name De Chelly was apparently used before this time. Simpson obtained its orthography from Vigil, secretary of the province (of New Mexico), who told him it was of Indian origin and was pronounced chay-e. Possibly it was derived from the Navaho name of the place, Tsé-gi. Simpson's description, although very brief, formed the basis of all the succeeding accounts for the next thirty years. The Pacific railroad surveys, which added so much to our knowledge of the Southwest, did not touch this field. In 1860 the Abbé Domenech published his "Deserts of North America," which contains a reference to Casa Blanca ruin, but his knowledge was apparently derived wholly from Simpson. None of the assistants of the Hayden Survey actually penetrated the canyon, but one of them, W. H. Jackson, examined and described some ruins on the Rio de Chelly, in the lower Chin Lee valley. But in an article in Scribner's Magazine for December, 1878, Emma C. Hardacre published a number of descriptions and illustrations derived from the Hayden corps, among others figures one entitled "Ruins in Cañon de Chelly," from a drawing by Thomas Moran. The ruin can not be identified from the drawing. This article is worth more than a passing notice, as it not only illustrates the extent of knowledge of the ruins at that time (1878), but probably had much to do with disseminating and making current erroneous inferences which survive to this day. In an introductory paragraph the author says: Of late, blown over the plains, come stories of strange newly discovered cities of the far south-west; picturesque piles of masonry, of an age unknown to tradition. These ruins mark an era among antiquarians. The mysterious mound-builders fade into comparative insignificance before the grander and more ancient cliff-dwellers, whose castles lift their towers amid the sands of Arizona and crown the terraced slopes of the Rio Mancos and the Hovenweap. 79 80 Of the Chaco ruins it is said: In size and grandeur of conception, they equal any of the present buildings of the United States, if we except the Capitol at Washington, and may without discredit be compared to the Pantheon and the Colosseum of the Old World. In the same year Mr J. H. Beadle gave an account2 of a visit he made to the canyon. He entered it over the Bat trail, near the junction of Monument canyon, and saw several ruins in the upper part. His descriptions are hardly more than a mention. Much archeologic data were secured by the assistants of the Wheeler Survey, but it does not appear that any of them, except the photographer, visited Canyon de Chelly. In the final reports of the Survey there is an illustration of the ruin visited by Lieutenant Simpson about thirty years before.3 The illustration is a beautiful heliotype from a fine photograph made by T. H. O'Sullivan, but one serious defect renders it useless; through some blunder of the photographer or the engraver, the picture is reversed, the right and left sides being interchanged, so that to see it properly it must be looked at in a mirror. The illustration is accompanied by a short text, apparently prepared by Prof. F. W. Putnam, who edited the volume. The account by Simpson is quoted and some additional data are given, derived from notes accompanying the photograph. The ruin is said to have "now received the name of the Casa Blanca, or White House," but the derivation of the name is not stated. In 1882 Bancroft could find no better or fuller description than Simpson's, which he uses fully, and reproduces also Simpson's (Kern's) illustration. In the same year investigation by the assistants of the Bureau of Ethnology was commenced. Colonel James Stevenson and a party visited the canyon, and a considerable amount of data was obtained. In all, 46 ruins were visited, 17 of which were in Del Muerto; and sketches, ground plans, and photographs were obtained. The report of the Bureau for that year contains an account of this expedition, including a short description of a large ruin in Del Muerto, subsequently known as Mummy Cave. A brief account of the trip was also published elsewhere.4 The next year a map of the canyon was made by the writer and many new ruins were discovered, making the total number in the canyon and its branches about 140. Since 1883 two short visits have been made to the place, the last late in 1893, and on each trip additional material was obtained. In 1890 Mr F. T. Bickford5 published an account of a visit to the canyon, illustrated with a series of woodcuts made from the photographs of the Bureau. The illustrations are excellent and the text is pleasantly written, but the descriptions of ruins are too general to be of much value to the student. In recent years several publications have appeared which, while not bearing directly on the De Chelly ruins, are of great interest, as they treat of analogous remains—the cliff ruins of the Mancos canyon and the Mesa Verde. These ruins were discovered in 1874 by W. H. Jackson and were visited and described in 1875 by W. H. Holmes,6 both of the Hayden Survey. This region was roamed over by bands of renegade Ute and Navaho, who were constantly making trouble, and for fifteen years was apparently not visited by whites. Recent exploration appears to have been inaugurated by Mr F. H. Chapin, who spent two summers in the Mesa Verde country. Subsequently he published the results of some of his observations in a handsome little volume.7 In 1891 Dr W. R. Birdsall made a flying trip to this region and published an account8 of the ruins he saw the same year. At the time of this visit a more elaborate exploration was being carried on by the late G. Nordenskiöld, who made some excavations and obtained much valuable data which formed the basis of a book published in 1893.9 This is the most important treatise on the cliff ruins that has ever been published, and the illustrations can only be characterized as magnificent. All of these works, and especially the last named, are of great value to the student of the cliff ruins wherever located, or of pueblo architecture. GEOGRAPHY The ancient pueblo culture was so intimately connected with and dependent on the character of the country where its remains are found that some idea of this country is necessary to understand it. The limits of the region are closely coincident with the boundaries of the plateau country except on the south, so much so that a map of the latter,10 slightly extended around its margin, will serve to show the former. The area of the ancient pueblo region may be 150,000 square miles; that of the plateau country, approximately, 130,000. The plateau country is not a smooth and level region, as its name might imply; it is extremely rugged, and the topographic obstacles to travel are greater than in many wild mountain regions. It is a country of cliffs and canyons, often of considerable magnitude and forming a bar to extended progress in any direction. The surface is generally smooth or slightly undulating and apparently level, but it is composed of a series of platforms or mesas, which are seldom of great extent and generally terminate at the brink of a wall, often of huge dimensions. There are mesas everywhere; it is the mesa country. Although the strata appear to be horizontal, they are slightly tilted. The inclination, although slight, is remarkably persistent, and the thickness of the strata remains almost constant. The beds, therefore, extend from very high altitudes to very low ones, and often the formation which is exposed to view at the summit of an incline is lost to view after a few miles, being covered by some later formation, which in turn is covered by a still later one. Each formation thus appears as a terrace, bounded on one side by a descending cliff carved out of the edges of its own strata and on the other by an ascending cliff carved out of the strata which overlie it. This is the more common form, although isolated mesas, bits of tableland completely engirdled by cliffs, are but little less common. The courses of the margins of the mesas are not regular. The cliffs sometimes maintain an average trend through great distances, but in detail their courses are extremely crooked; they wind in and out, forming alternate alcoves and promontories in the wall, and frequently they are cut through by valleys, which may be either narrow canyons or interspaces 10 or even 20 miles wide. The whole region has been subjected to many displacements, both flexures of the monoclinal type and faults. Some of 81 82 these flexures attain a length of over 80 miles and a displacement of 3,000 feet, and the faults reach even a greater magnitude. There is also an abundance of volcanic rocks and extinct volcanoes, and while the principal eruptions have occurred about the borders of the region, extending but slightly into it, traces of lesser disturbances can be found throughout the country. It has been said that if a geologist should actually make the circuit of the plateau country, he could so conduct his route that for three-fourths of the time he would be treading upon volcanic materials and could pitch his camp upon them every night. The oldest eruptions do not go back of Tertiary time, while some are so recent as probably to come within the historic period—within three or four centuries. The strata of the plateau country are remarkable for their homogeneity, when considered with reference to their horizontal extensions; hardly less so for their diversity when considered in their vertical relation. Although the groups differ radically from each other, still each preserves its characteristics with singularly slight degrees of variation from place to place. Hence we have a certain amount of similarity and monotony in the landscape which is aided rather than diminished by the vegetation; for the vegetation, like the human occupants of this country, has come under its overpowering influence. The characteristic landscape consists of a wide expanse of featureless plains, bounded by far- off cliffs in gorgeous colors; in the foreground a soil of bright yellow or ashy gray; over all the most brilliant sunlight, while the distant features are softened by a blue haze. The most conspicuous formation of the whole region is a massive bright-red sandstone out of which have been carved "the most striking and typical features of those marvelous plateau landscapes which will be subjects of wonder and delight to all coming generations of men. The most superb canyons of the neighboring region, the Canyon de Chelly and the Del Muerto, the lofty pinnacles and towers of the San Juan country, the finest walls in the great upper chasms of the Colorado, are the vertical edges of this red sandstone." Of the climate of the plateau country it has been said that in the large valleys it is "temperate in winter and insufferable in summer; higher up the summers are temperate and the winters barely sufferable." It is as though there were two distinct regions covering the same area, for there are marked differences throughout, except in topographic configuration, between the lowlands and the uplands or high plateaus. The lowlands present an appearance which is barren and desolate in the extreme, although the soil is fertile and under irrigation yields good crops. Vegetation is limited to a scanty growth of grass during a small part of the year, with small areas here and there scantily covered by the prickly greasewood and at intervals by clumps of sagebrush; but even these prefer a higher level, and develop better on the neighboring mesas than in the valleys proper. The arborescent growth consists of sparsely distributed cottonwoods and willows, closely confined to the river bottoms. On intermediate higher levels junipers and cedars appear, often standing so closely together as to seriously impede travel, but they are confined to the tops of mesas and other high ground, the valleys being generally clear or covered with sagebrush. Still higher up yellow pines become abundant and in places spread out into magnificent forests, while in some mountain regions scrub oak, quaking asp, and even spruce trees are abundant. In the mountain regions there is often a reasonable amount of moisture, and some crops, potatoes for example, are grown there without irrigation; but the season is short. In the Tunicha mountains the Navaho raise corn at an altitude of nearly 8,000 feet, but they often lose the crop from drought or from frost. On the intermediate levels and in the lowlands cultivation by modern methods is practically impossible without irrigation, except in a few favored localities, where a crop can be obtained perhaps two years or three years in five. But with a minute knowledge of the climatic conditions, and with methods adapted to meet these conditions, scanty crops can be and are raised by the Indians without irrigation throughout the whole region; but everywhere that water can be applied the product of the soil is increased many fold. Near the center of the plateau country, in the northeastern corner of Arizona, a range of mountains crosses diagonally from northwest to southeast, extending into New Mexico. In the north an irregular cluster of considerable size, separated from the remainder of the range, is called the Carrizo; and the range proper has no less than three names applied to different parts of it. The northern end is known as the Lukachukai, the central part as the Tunicha, and the southern part as the Chuska or Choiskai mountains, all Navaho names. The two former clusters attain an altitude of 9,500 feet; the Tunicha and the Chuska are about 9,000 feet high, the latter having a flat top of considerable area. On the east these mountains break down rather abruptly into the broad valley of the Chaco river, or the Chaco wash, as it is more commonly designated; on the west they break down gradually, through a series of slopes and mesas, into the Chin Lee valley. Canyon de Chelly has been cut in the western slope by a series of small streams, which, rising near the crest of the mountain, combine near its head and flow in a general westerly direction. The mouth of the canyon is on the eastern border of the Chin Lee valley. It is 60 miles south of the Utah boundary and 25 miles west of that of New Mexico; hence it is 60 miles east and a little north from the old province of Tusayan, the modern Moki, and 85 miles northwest from the old province of Cibola, the modern Zuñi. Its position is almost in the heart of the ancient pueblo region; the Chaco ruins lie about 80 miles east, and the ruins of the San Juan from 60 to 80 miles north and northeast. The geographic position of Canyon de Chelly has had an important effect on its history, forming as it does an available resting place in any migratory movement either on the north and south line or east and west. The Tunicha mountains are a serious obstacle to north and south movement at the present day, but less so than the arid valleys which border them. Except at one place, and that place is difficult, it is almost impossible to cross the mountains with a wheeled vehicle, but there are innumerable trails running in all directions, and these trails are in constant use by the Navaho, except in the depths of winter. The mountain route is preferable, however, to the valley roads, where the traveler for several days is without wood, with very little water and forage, and his movements are impeded by deep sand. To the traveler on foot, or even on horseback, Canyon de Chelly is easily accessible from almost any direction. Good trails run northward to the San Juan and northeastward over the Tunicha mountains to the upper part of that river; Fort Defiance is but half a day's journey to the southeast; Tusayan and Zuñi are but three days distant to the traveler on foot; the Navaho often ride the distance in a day or a day and a half. The canyon is accessible to wagons, however, only at 83 84 85 its mouth. map Plate XLII. Map of Canyon De Chelly and Its Branches Surveyed by Cosmos Mindeleff Larger View The main canyon, shown on the map (plate XLII) as Canyon de Chelly and known to the Navaho as Tsé-gi, is about 20 miles long. It heads near Washington pass, within a few miles of the crest of the mountain, and extends almost due west to the Chin Lee valley. The country descends by a regular slope from an altitude of about 7,500 feet at the foot of the main crest to about 5,200 feet in the Chin Lee valley, 25 miles west, and is so much cut up locally by ravines and washes that it is impassable to wagons, but it preserves throughout its mesa-like character. About 3 miles from its mouth De Chelly is joined by another canyon almost as long, which, heading also in the Tunicha mountains, comes in from the northeast. It is over 15 miles long, and is called on the map Canyon del Muerto; the Navaho know it as Én-a-tsé-gi. About 13 miles above the mouth of the main canyon a small branch comes in from the southeast. It is about 10 miles long, and has been called Monument canyon, on account of the number of upright natural pinnacles of rock in it. In addition to those named there are innumerable small branches, ranging in size from deep coves to real canyons a mile or two long. Outside of De Chelly, and independent of it, there is a little canyon about 4 miles long, called Tse-on-i-tso-si by the Navaho. At one point near its head it approaches so near to De Chelly that but a few feet of rock separate them. On the western side of the mountains there are a number of small perennial streams fed by springs on the upper slopes. Several of these meet in the upper part of De Chelly, others in Del Muerto, and in the upper parts of these canyons there is generally water. But, except at the time of the autumn and winter rains and in the spring when the mountain snows are melting, the streams are not powerful enough to carry the water to the mouth of the canyon. The flow is absorbed by the deep sand which forms the stream bed. Ordinarily it is difficult to procure enough water to drink less than 8 or 10 miles from the mouth of De Chelly, but occasionally the whole stream bed, at places over a quarter of a mile wide, is occupied by a raging torrent impassable to man or beast. Such ebullitions, however, seldom last more than a few hours. Usually water can be obtained anywhere in the bottom by sinking a shallow well in the sand, and it is by this method that the Navaho, the present occupants of the canyon, obtain their supply. The walls of the canyon are composed of brilliant red sandstone, discolored everywhere by long streaks of black and gray coming from above. At its mouth it is about 500 feet wide. Higher up the walls sometimes approach to 300 feet of each other, elsewhere broadening out to half a mile or more; but everywhere the wall line is tortuous and crooked in the extreme, and, while the general direction of De Chelly is east and west, the traveler on the trail which runs through it is as often headed north or south. Del Muerto is even more tortuous than De Chelly, and in places it is so narrow that one could almost throw a stone across it. At its mouth the walls of Canyon de Chelly are but 20 to 30 feet high, descending vertically to a wide bed of loose white sand, and absolutely free from talus or débris. Three miles above Del Muerto comes in, but its mouth is so narrow it appears like an alcove and might easily be overlooked. Here the walls are over 200 feet high, but the rise is so gradual that it is impossible to appreciate its amount. At the point where Monument canyon comes in, 13 miles above the mouth of De Chelly, the walls reach a height of over 800 feet, about one-third of which consists of talus. The rise in the height of the walls is so gradual that when the canyon is entered at its mouth the mental scale by which we estimate distances and magnitudes is lost and the wildest conjectures result. We fail at first to realize the stupendous scale on which the work was done, and when we do finally realize it we swing to the opposite side and exaggerate. At the junction of Monument canyon there is a beautiful rock pinnacle or needle standing out clear from the cliff and not more than 165 feet on the ground. It has been named, in conjunction with a somewhat similar pinnacle on the other side of the canyon, "The Captains," and its height has been variously estimated at from 1,200 to 2,500 feet. It is less than 800. A curious illustration of the effects of the scenery in connection with this pinnacle may not be amiss. The author of Western Wilds (Cincinnati, 1878) thus describes it: 86 But the most remarkable and unaccountable feature of the locality is where the canyons meet. There stands out 100 feet from the point, entirely isolated, a vast leaning rock tower at least 1,200 feet high and not over 200 thick at the base, as if it had originally been the sharp termination of the cliff and been broken off and shoved farther out. It almost seems that one must be mistaken; that it must have some connection with the cliff, until one goes around it and finds it 100 feet or more from the former. It leans at an angle from the perpendicular of at least 15 degrees; and lying down at the base on the under side, by the best sighting I could make, it seemed to me that the opposite upper edge was directly over me—that is to say, mechanically speaking, its center of gravity barely falls with the base, and a heave of only a yard or two more would cause it to topple over. (Page 257.) The dimensions have already been given. The pinnacle is perfectly plumb. The rock of which the canyon walls are formed is a massive sandstone in which the lines of bedding are almost completely obliterated. It is rather soft in texture, and has been carved by atmospheric erosion into grotesque and sometimes beautiful forms. In places great blocks have fallen off, leaving smooth vertical surfaces, extending sometimes from the top nearly to the stream bed, 400 feet or more in height and as much in breadth. In the lower parts of the canyons the walls, sometimes of the character described, sometimes with the surfaces and angles smoothed by the flying sand, are generally vertical and often overhang, descending sheer to the canyon bottom without talus or intervening slopes of débris. The talus, where there is any, is slight and consists of massive sandstone of the same character as the walls, but much rounded by atmospheric erosion. The enlarged map (plate XLIII) shows something of this character. map Plate XLIII. Detailed Map of Part of Canyon De Chelly Showing Areas of Cultivatable Land Larger View The varying distances between vertical lines in this thumbnail view are a byproduct of scanning. The larger version has been artificially rescaled to approximate the correct dimensions. Numbers and words in red were added by the transcriber because the original text was unreadable at any practical magnification. Original Plate (very large, unedited) Near its mouth the whole bottom of the canyon consists of an even stretch of white sand extending from cliff to cliff. A little higher up there are small areas of alluvium, or bottom land, in recesses and coves in the walls and generally only a foot or two above the stream bed. Still higher up these areas become more abundant and of greater extent, forming regular benches or terraces, generally well raised above the stream bed. At the Casa Blanca ruin, 7 miles up the canyon, the bench is 8 or 10 feet above the stream. Each little branch canyon and deep cove in the cliffs is fronted by a more or less extended area of this cultivable bottom land. Ten miles up the talus has become a prominent feature. It consists of broken rock, sand, and soil, generally overlying a slope of massive sandstone, such as has been described, and which occasionally crops out on the surface. With the development of the talus the area of bottom land dwindles, and the former encroaches more and more until a little above the junction of Monument canyon the bottom land is limited to narrow strips and small patches here and there. These bottom lands are the cultivable areas of the canyon bottom, and their occurrence and distribution have dictated the location of the villages now in ruins. They are also the sites of all the Navaho settlements in the canyon. The Navaho hogans are generally placed directly on the bottoms; the ruins are always so located as to overlook them. Only a very small proportion of the available land is utilized by the Navaho, and not all of it was used by the old village builders. The Navaho sites, as a whole, are far superior to the village sites. The horticultural conditions here, while essentially the same as those of the whole pueblo region, present some peculiar features. Except for a few modern examples there are no traces of irrigating works, and the Navaho work can not be regarded as a success. The village builders probably did not require irrigation for the successful cultivation of their crops, and under the ordinary Indian methods of planting and cultivation a failure to harvest a good crop was probably rare. After the Harvest season it is the practice of the Navaho to abandon the canyon for the winter, driving their flocks and carrying the season's produce to more open localities in the neighboring valleys. The canyon is not a desirable place of residence in the winter to a people who live in the saddle and have large flocks of sheep and goats, but there is no evidence that the old inhabitants followed the Navaho practice. 87 88 During most of the year there is no water in the lower 10 miles of the canyons, where most of the cultivable land is situated. The autumn rains in the mountains, which occur late in July or early in August, sometimes send down a little stream, which, however, generally lasts but a few days and fails to reach the mouth of the canyon. Late in October, or early in November, a small amount comes down and is fairly permanent through the winter and spring. The stream bed is even more tortuous than the canyon it occupies, often washing the cliffs on one side, then passing directly across the bottom and returning again to the same side, the stream bed being many times wider than the stream, which constantly shifts its channel. In December it becomes very cold and so much of the stream is in shade during a large part of the day that much of the water becomes frozen and, as it were, held in place. In the warm parts of the day, and in the sunshine, the ice is melted, the stream resumes its flow, and so gradually pushes its way farther and farther down the canyon. But some sections, less exposed to warmth than others, retain their ice during the day. These points are flooded by the water from above, which is again frozen during the night and again flooded the next day, and so on. In a short time great fields of smooth ice are formed, which render travel on horseback very difficult and even dangerous. This, and the scant grazing afforded by the bottom lands in winter, doubtless is the cause of the annual migration of the Navaho; but these conditions would not materially affect a people living in the canyon who did not possess or were but scantily supplied with horses and sheep. The stream when it is flowing is seldom more than a foot deep, generally only a few inches, except in times of flood, when it becomes a raging torrent, carrying everything before it. Hence irrigation would be impracticable, even if its principles were known, nor is it essential here to successful horticulture. One of the characteristic features of the canyons at the present day is the immense number of peach trees within them. Wherever there is a favorable site, in some sheltered cove or little branch canyon, there is a clump of peach trees, in some instances perhaps as many as 1,000 in one "orchard." When the peaches ripen, hundreds and even thousands of Navaho flock to the place, coming from all over the reservation, like an immense flock of vultures, and with disastrous results to the food supply. A few months after it is difficult to procure even a handful of dried fruit. The peach trees are, of course, modern. They were introduced into this country originally by the Spanish monks, but in De Chelly there are not more than two or three trees which are older than the last Navaho war. At that time, it is said, the soldiers cut down every peach tree they could find. But, aside from the peaches, De Chelly was until recently the great agricultural center of the Navaho tribe, and large quantities of corn, melons, pumpkins, beans, etc, were and are raised there every year. Under modern conditions many other localities now vie with it, and some surpass it in output of agricultural products, but not many years ago De Chelly was regarded as the place par excellence. It will be clear, therefore, that prior to very recent times De Chelly would be selected by almost any tribe moving across the country, and, barring a hostile prior occupancy, would be the most desirable place for the pursuit of horticultural operations for many miles in any direction. The vicinity of the Tunicha mountains, which could be reached in half a day from any part of the canyons, and which must have abounded in game, for even now some is found there, would be a material advantage. The position of the canyon in the heart of the plateau country and of the ancient pueblo region would make it a natural stopping place during any migratory movement either north and south or east and west, and its settlement was doubtless due to this favorable position and to the natural advantages it offered. This settlement was effected probably not by one band or tribe, nor at one time, but by many bands at many times. Probably the first settlements were very old; certainly the last were very recent. CLASSIFICATION AND DESCRIPTIONS RUINS OF THE PUEBLO REGION No satisfactory general classification of the ruins of the ancient pueblo region has yet been made; possibly because the material in hand is not sufficiently abundant. There are thousands of ruins scattered over the southwest, of many different types which merge more or less into each other. In 1884 Mr A. F. Bandelier, whose knowledge of the archeology of the southwest is very extensive, formulated a classification, and in 1892, in his final report,11 he announces that he has nothing to change in it. The classification is as follows: I. Large communal houses several stories high. (a) Composed of one or two, seldom three, extensive buildings, generally so disposed as to surround an interior court. (b) Polygonal pueblos. (c) Scattered pueblos, composed of a number of large many-storied houses, disposed in a more or less irregular manner; sometimes in irregular squares or on a line. (d) Artificial caves, resembling in number, size, and disposition of the cells the many-storied communal dwelling. (e) Many-storied dwellings, with artificial walls, erected inside of natural caves of great size. II. Detached family dwellings, either isolated or in groups forming villages. Many hundreds of ruins have been examined by Mr Bandelier, and doubtless the classification above afforded a convenient working basis for the region with which he is most familiar, the basin of the Rio Grande and its tributaries. It does not apply very well to the western part of the pueblo region. The distinguishing characteristics of the first group (of five classes)—houses several stories high—are as follows: Each building consisted of an agglomeration of a great number of small cells, without any larger halls of particularly striking dimensions. All the buildings, except outhouses or additions, were at least two stories high, and the lower story was entered only from the roof. The various stories receded from the bottom to the top. The prevalence of the estufa (kiva) 89 90 generally, or often, circular in form. Ruins of class II—detached family dwellings—consist sometimes of a single room; more often of several rooms. The rooms are generally built of stone, although examples constructed of mud and adobe are also found in certain regions. The average size of the room is larger than in the communal building, and there is a gradual increase in size of rooms from north to south. There are front doorways and light and air holes are larger than in the communal houses. Mr Bandolier suggests that the detached family dwelling was the early type, and that only when enemies began to threaten were the communal houses resorted to for purposes of defense. This classification is apparently based on external form alone, without taking into account the numerous influences which modify or produce form; and while no doubt it was sufficient for field use, it is not likely to be permanently adopted; for there does not appear to be any essential or radical difference between the various classes. Moreover, there does not appear to be any place in the scheme for the cliff ruins of the variety especially abundant in De Chelly and found in many other localities, unless indeed such ruins come under class II—detached family dwellings; yet this would imply precedence in time, and the ruins themselves will not permit such an inference. The essential uniformity of types which prevails over the immense area covered by the ancient pueblo ruins is a noteworthy feature, and any system of classification which does not take it into account must be considered as only tentative. What elements should be considered and what weight assigned to each in preparing a scheme of classification is yet to be determined, but probably one of the most important elements is the character of the site occupied, with reference to its convenience and defensibility. There are great differences in kind between the great valley pueblos, located without reference to defense and depending for security on their size and the number of their population, of which Zuñi and Taos are examples, and the villages which are located on high mesas and projecting tongues of rock; in other words, on defensive sites where reliance for security was placed on the character of the site occupied, such as the Tusayan villages of today. Within each of these classes there are varieties, and there are also secondary types which pertain sometimes to one, sometimes to the other, and sometimes to both. Such are the cliff ruins, the cavate lodges, and the single house remains. The unit of pueblo architecture is the single cell, and in its development the highest point reached is the aggregation of a great number of such cells into one or more clusters, either connected with or adjacent to each other. These cells were all the same, or essentially so; for while differentiation in use or function had been or was being developed at the time of the Spanish conquest, differentiation in form had not been reached. The kiva, of circular or rectangular shape, is a survival and not a development. Large aggregations of many cells into one cluster are the latest development of pueblo architecture. They were immediately preceded by a type composed of a larger number of smaller villages, located on sites selected with reference to their ease of defense, and apparently the change from the latter to the former type was made at one step, without developing any intermediate forms. The differences between the largest examples of villages on defensive sites and the smallest appear to be only differences of size. Doubtless in the early days of pueblo architecture small settlements were the rule. Probably these settlements were located in the valleys, on sites most convenient for horticulture, each gens occupying its own village. Incursions by neighboring wild tribes, or by hostile neighbors, and constant annoyance and loss at their hands, gradually compelled the removal of these little villages to sites more easily defended, and also forced the aggregation of various related gentes into one group or village. At a still later period the same motive, considerably emphasized perhaps, compelled a further removal to even more difficult sites. The Tusayan villages at the time of the Spanish discovery were located on the foothills of the mesas, and many pueblo villages at that period occupied similar sites. Actuated by fear of the Ute and Comanche, and perhaps of the Spaniards, the inhabitants soon after moved to the top of the mesa, where they now are. Many villages stopped at this stage. Some were in this stage at the time of the discovery—Acoma, for example. Finally, whole villages whose inhabitants spoke the same language combined to form one larger village, which, depending now on size and numbers for defense, was again located on a site convenient for horticulture. The process sketched above was by no means continuous. The population was in slow but practically constant movement, much the same as that now taking place in the Zuñi country; it was a slow migration. Outlying settlements were established at points convenient to cultivable fields, and probably were intended to be occupied only during the summer. Sometimes these temporary sites might be found more convenient than that of the parent village, and it would gradually come about that some of the inhabitants would remain there all the year. Eventually the temporary settlement might outgrow the parent, and would in turn put out other temporary settlements. This process would be possible only during prolonged periods of peace, but it is known to have taken place in several regions. Necessarily hundreds of small settlements, ranging in size from one room to a great many, would be established, and as the population moved onward would be abandoned, without ever developing into regular villages occupied all the year. It is believed that many of the single house remains of Mr Bandelier's classification12 belong to this type, as do also many cavate lodges, and in the present paper it will be shown that some at least of the cliff ruins belong to the same category. The cliff ruins are a striking feature, and the ordinary traveler is apt to overlook the more important ruins which sometimes, if not generally, are associated with them. The study of the ruins in Canyon de Chelly has led to the conclusion that the cliff ruins there are generally subordinate structures, connected with and inhabited at the same time as a number of larger home villages located on the canyon bottom, and occupying much the same relation to the latter that Moen-kapi does to Oraibi, or that Nutria, Pescado, and Ojo Caliente do to Zuñi; and that they are the functional analogues of the "watch towers" of the San Juan and of Zuñi, and the brush shelters or "kisis" of Tusayan: in other words, they were horticultural outlooks occupied only during the farming season. Mr G. Nordenskiöld, who examined a number of cliff and other ruins in the Mancos canyon and the Mesa Verde region, adopts13 a very simple classification, as follows: 91 92 I. Ruins in the valleys, on the plains, or on the plateaus. II. Ruins in caves in the walls of the canyons, subdivided as follows: (a) Cave dwellings, or caves inhabited without the erection of any buildings within them. (b) Cliff dwellings, or buildings erected in caves. From its topographic character it might be expected that the Canyon de Chelly ruins would hardly...