MESA VERDE Archeological Excavations in Mesa Verde National Park, Colorado • 1950

EARLY HISTORY OF THE MESA VERDE

MESA VERDE NATIONAL PARK comprises one-half of a great tableland, or mesa, in the southwestern corner of Colorado, only a few miles from the "Four Corners," where the States of Colorado, Utah, Arizona, and New Mexico meet at a common point. The mesa, measuring 15 by 20 miles, rises from 1,000 to 2,000 feet above the surrounding country. Its flat top, which slopes gradually to the south, is cut by a score of rugged canyons, dividing the large mesa into many smaller mesas. Because of its heavy forests of piñon and juniper, the mesa is perpetually green, and at some early date, probably during the 1765-1848 period, when there was much Spanish activity in the area, it was given the name Mesa Verde, or "green table."

First knowledge of the archeological treasures of the Mesa Verde came in 1874, when W. H. Jackson, the famous "Pioneer Photographer," discovered small cliff dwellings in the Mancos River Canyon which borders the Mesa Verde on the east and south (Jackson, 1876, pp. 367-381). In 1888, the major cliff dwellings were discovered, and this date marks the beginning of the tragic period in the history of the Mesa Verde. Shortly after the discovery of the large cliff dwellings, it was learned that there was a ready market for the artifacts which they contained. The cliff dwellings were so thoroughly ransacked in the following 18 years that, so far as is known today, little material of scientific value remains in them. Only one archeologist worked in the Mesa Verde during this period. Baron Gustav Nordenskiöld excavated in a number of cliff dwellings in 1891 and, considering the time, published a most excellent report on his work (Nordenskiöld, 1893).

ARCHEOLOGICAL RESEARCH

Early research, 1908-22. In 1906, a portion of the great mesa was set aside as Mesa Verde National Park. The first research was done 2 years later when Dr. Jesse Walter Fewkes, of Smithsonian Institution, Bureau of American Ethnology, excavated Spruce Tree House, one of the larger cliff dwellings (Fewkes, 1909). During the years between 1908 and 1922, Fewkes excavated a number of cliff dwellings and mesa-top ruins but modern scientific methods were not used and the results leave much to be desired.

More recent research, 1923-38. Following the Fewkes period, research was, of necessity, neglected. The park was created to preserve the ruins and make them accessible to the public. From the time the first roads were built, travel to the area increased year after year. The number of visitors increased more rapidly than the size of the local staff, and consequently research was delayed.

During the summer of 1923, the First National Geographic Beam Expedition, under A. E. Douglass, collected tree-ring specimens in the park (Douglass, 1929, p. 750), and in 1932-33, H. T. Getty, of the Tree-Ring Laboratory, continued this work (Getty, 1935, pp. 21-23). In 1926, Superintendent Jesse L. Nusbaum excavated three early seventh century pithouses in Step House Cave and, during the winters of 1926-29, did a small amount of salvage excavation in the previously disturbed refuse of several cliff dwellings. In 1929, H. S. Gladwin, of Gila Pueblo, surveyed 103 mesa-top and canyon-head sites. Sporadic testing and probing during stabilization of cliff and mesa-top ruins and removal of burials encountered, comprise the only other research for this period.

Research outside the park. Although research lagged in the Mesa Verde itself, the general archeology was comparatively well understood because of extensive work which was done in the surrounding area. Over the past 50 years the reports of many investigators have given us a general overall picture of the archeology of the Mesa Verde region. Outstanding in this regard is the work of the following: E. H. Morris in the Durango, La Plata, Red Rock, and Canyon de Chelly areas to the east, south, and southwest (Morris, 1919b, 1939, 1941); J. A. Jeancon and F. H. H. Roberts, Jr., in the Piedra district to the east (Jeancon, 1922; Jeancon and Roberts, 1923-24; Roberts, 1922, 1925, 1930); E. K. Reed in the Mancos Canyon to the south (Reed, 1943, 1944); S. G. Morley and A. V. Kidder in the McElmo drainage to the west (Morley, 1908; Morley and Kidder, 1917); T. M. Prudden, P. S. Martin, and J. B. Rinaldo in the Montezuma Valley to the northwest (Prudden, 1905, 1914, 1918; Martin, 1929, 1930, 1936, 1938; Martin and Rinaldo, 1939; Rinaldo, 1950); B. Cummings and his students, A. V. Kidder, N. M. Judd, and J. L. Nusbaum in southeastern Utah to the west (Kidder, 1910); and later, J. O. Brew's excavations in the same area (Brew, 1946).

Recent research in the park, 1939-53. While the general archeological story of the region was relatively well known, it was to be expected that the Mesa Verde would present certain variations. Since 1939, a program of organized research has been carried out in the park in an effort to determine these local manifestations.

(a) Excavation by park staff, 1939-41. In 1939, T. L. Smiley excavated an A. D. 700 pithouse (Smiley, 1949 pp. 167-171), and, in 1941, J. A. Lancaster excavated two pithouses dating about A. D. 600 (Lancaster and Watson, 1943, pp. 190-198) and tested other pithouses and early pueblo structures.

(b) Gila Pueblo Tree-Ring Expedition, 1941. In the fall of 1941, Gila Pueblo collected tree-ring specimens in the park. Dates obtained are included in O'Bryan's recent publication on the Mesa Verde (O'Bryan, 1950, Appendix A, pp. 112-115).

(c) Excavations by Gila Pueblo, 1947-48. During these years, D. O'Bryan excavated an extensive series of ruins in the park (O'Bryan, 1950):

1 shallow pithouse, dating about A. D. 572.
1 deep pithouse, dating about A. D. 664.
2 slabhouse villages, dating about A. D. 840.
1 small pueblo, dating about A. D. 950.
1 small pueblo, dating about A. D. 1024.
1 large canyon-head pueblo, dating about A. D. 1025 to 1200.

(d) Recent excavations by park staff, 1950. In the summer of 1950, members of the local staff excavated six ruins:

2 pithouses, dating about A. D. 700.
1 post and adobe village, dating about A. D. 900.
1 masonry pueblo, dating about A. D. 1000.
1 masonry pueblo, dating about A. D. 1074.
1 masonry pueblo, dating about A. D. 1200.

Reports on these excavations are included in this volume.

(e) Archeological survey by local staff, 1951. An intensive archeological survey of the Mesa Verde was started in the fall of 1951 and will continue for many years. When completed, this survey should provide an overall picture of the prehistoric occupation of the Mesa Verde. All archeological features are included: cliff dwellings, mesa-top ruins, canyon-head ruins, dams, reservoirs, shrines, pictographs, etc. Sherd collections are made, all pertinent data recorded, ruins marked, and, whenever possible, individual ruins are mapped. To date, this survey has been confined to Chapin Mesa south of park headquarters (pl. 1), and in an area of approximately 4 square miles, 472 sites have been surveyed. The only cliff dwellings so far surveyed are those in the head of Cliff Canyon, north of Cliff Palace. In one-half mile of this small canyon, 15 cliff dwellings have been located.

PLATE 1—Map of Chapin Mesa. (click on image for a PDF version)

SUMMARY OF THE ARCHEOLOGY OF THE MESA VERDE

As a result of the excavations and survey within the park and the extensive work which has been done in surrounding areas, the archeology of the Mesa Verde itself is becoming somewhat clearer, although there are still gaps in the story which must be filled through continuing research. It should be mentioned also that, except for some of the early work, all excavation to date has been in the Chapin Mesa area and the remainder of the Mesa Verde is practically untouched. Chapin Mesa is in the center of the Mesa Verde, in an east-west line, and extends from the north rim to the Mancos River Canyon, on the south. It is the largest mesa and, while it appears to have the greatest number of ruins of all known local types, there is reason to believe that archeological variations may occur in other parts of the Mesa Verde.

Occupation of the Mesa Verde seems to have extended from the early part of the Christian Era to almost A. D. 1300. In summarizing this long occupation, the Pecos Classification will be followed as it is the best known and most widely used classification of culture periods (Kidder, 1927, pp. 554-561). The Roberts' Classification is employed in dealing with park visitors as its use of descriptive names, and the fact that it has one less period makes it easier for the uninitiated visitor to whom such classifications are, to say the least, baffling (Roberts, 1935, p. 32 and this volume, table 1).

The following brief summary is an effort to present the archeology of the Mesa Verde as it is known today and to point out some of the problems yet to be solved. Few comparisons will be made with findings in nearby areas, for this has been done in past publications and in the three reports included in this volume. Dates vary somewhat from those which have been given for other areas. It must be borne in mind that there was constant progress throughout the entire occupation, and it is impossible to draw sharp lines and set exact dates for the various periods. Many different dates have been used by archeologists, but those used below seem best for the Mesa Verde as the archeology is known at present.

The presentation of each of the following periods is divided into two parts. First, a general summary is given of the period, as known throughout the Mesa Verde region. Following this is a summarization of the known aspects of the period in the Mesa Verde itself.

Basketmaker I

This is a postulated, preagricultural stage set up at the first Pecos Conference, since it was obvious that a hunting-gathering culture must have preceded the later farming cultures.

Basketmaker II

Dates: From about the beginning of the Christian Era to A. D. 450. Dates obtained indicate that by the first century A. D., agricultural people were well established in the Four Corners region. Corn and squash were cultivated and there was also great use of wild plant foods and game. Pottery was absent and excellent baskets and bags were made in profusion. Slab-lined storage cists were characteristic, but houses have been found only in the Durango area, where there is evidence of crude, early structures. The atlatl and dart, and curved sticks served as weapons for the bow and arrow were not known. Other traits were woven bands; square-toed sandals; string aprons; fur-string blankets; soft, padded cradles; troughed metates; jewelry of stone, bone, seeds and shells; and wide use of animal skins. Dogs were present. This widespread culture served as the base out of which the later, more highly developed cultures grew.

No remains dating from this period have been found in the Mesa Verde, but there is reason to believe they will be found when the necessary, difficult excavation can be performed. Material dating from this period has, except in one area, been found in caves. Since Mesa Verde caves contain cliff dwellings it will be necessary to excavate under these structures in order to establish this early occupation. Basketmaker II remains have been found to the east of the park in the Durango area, to the west in southeastern Utah and to the southwest in Arizona, and it would be surprising if the Mesa Verde were not occupied during this period. Investigation of lower cave levels, deep under cliff dwellings, is high on the priority list of future research projects.

Basketmaker III

Dates: A. D. 450 to 750. This period saw marked development and is characterized by the advent of pottery, widespread use of pithouses of a standardized nature, appearance of the bow and arrow and the first use of hafted axes and mauls. Beans were first grown, turkeys apparently were domesticated, and turquoise came into use. Except for these new items, other material traits remained much the same as listed for Basketmaker II. However, fur-string blankets declined in favor of feather-string blankets, and sandals with notched or scalloped toes replaced the square-toed variety.

The Mesa Verde contains abundant remains dating from the latter part of this period. Twelve pithouses dating from A. D. 572 to 700 have been excavated in the park but no work has been done, as yet, in the A. D. 450-572 period. Because of the lack of work in the earlier part of this period the dates and types of the earliest houses and pottery are unknown. Seven houses dating at about A. D. 600 have been excavated. The common structure at this time appears to have been a large, shallow pithouse with a comparatively large, connected antechamber (Lancaster and Watson, 1943, pp. 190-198: O'Bryan, 1950, pp. 55-58). The three pithouses excavated in Step House Cave were somewhat different. Instead of having antechambers they had ventilators which were much too small to have served as entrances. This variation may have resulted from the cave location.

Five pithouses dating from the late seventh century show definite changes. These pithouses were deeper, averaging 4 to 5 feet in depth. The antechambers were smaller than at an earlier date, and certain features of the later ceremonial rooms, or kivas, such as bench, sipapu, deflector, and ventilator, were well established (Lancaster and Watson, "Excavation of Two Late Basketmaker III Pithouses" in this volume; Smiley, 1949, pp. 167-171; O'Bryan, 1950, pp. 58-61).

Present evidence indicates that during this period, at least after A. D. 600, the bulk of the population lived on the mesa-tops. The three structures found in Step House Cave show there was some cave occupation, but surface evidence of innumerable pithouses points to a preference for the open mesas. These mesa-top pithouses were grouped into villages but, to date, only individual pithouses have been excavated and the plan of the village is unknown. Testing has located as many as 9 pithouses in an area less than 300 feet in diameter.

The earliest Mesa Verde pottery, to date, came from a pithouse which gave a bark date of A. D. 572 (O'Bryan, 1950, pp. 55-58). The dominant pottery was Lino Gray, with a small amount of Lino Black-on-gray. At this site, and in an A. D. 664 pithouse, O'Bryan reported considerable Lino-like pottery which shows some degree of polishing. To this he gave the name Twin Trees Plain, and for a similar but decorated type, the name Twin Trees Black-on-white (O'Bryan, 1950, p. 91). In two late seventh century pithouses which are reported in this volume, Lino Gray was still the dominant type, while the decorated ware was La Plata Black-on-white. This pottery is basically like Lino Black-on-gray except that the paint is inorganic, rather than organic. There is strong support for the belief that La Plata Black-on-white was far more common in the Mesa Verde at this time than Lino Black-on-gray, but some earlier reports have lumped the two together under the more widely used latter name. The difficulty of drawing a line of demarcation between polished and unpolished pottery and the presence of crushed rock temper in the five above mentioned types, rather than sand temper as specified in published descriptions, indicate the need for an intensive study of the pottery of this period.

The minor traits of the period appear to be the same in the Mesa Verde as in the surrounding area.

Pueblo I

Dates: A. D. 750 to 900. It is difficult to draw a definite line between this and the preceding period as the differences, for the most part, are a matter of continuing development rather than abrupt change. One radical change did occur, however, for at about the beginning of this period the soft, padded cradle was discarded and a rigid, wooden cradle was adopted. This was responsible, in part, for the cranial deformation which was prevalent after this time.

This period is characterized by important developments in architecture and ceramics. Surface living rooms of stone-slab, post and adobe construction developed and were joined together in long, curving rows. In front of the rows of living rooms were pitrooms which grew very deep and began to lose their domiciliary functions as they became more kiva-like. Villages were sometimes very large. By the end of the period experiments with stone masonry had started. Ceramic improvement came with the introduction of the slip and decorated types were polished. Banded-neck vessels are diagnostic of the period. Red wares were made in some areas. Minor arts and crafts continued much as in the preceding period but sandals changed from a scalloped to a rounded toe. Cotton was in use by the end of the period.

In the Mesa Verde this period as well as the end of the preceding period is not well understood for there has been no excavation of ruins dating between A. D. 700 and 825. O'Bryan excavated 2 slabhouse ruins, 1 with deep pitrooms, which produced bark dates ranging from A. D. 829 to 845. (O'Bryan, 1950, pp. 37-43 and 51-53). Neither was completely excavated but they were, in general, typical of the slabhouse villages with associated pitrooms, mentioned above. The date for the advent of the first surface living rooms in the Mesa Verde is unknown. Slabhouses are believed to have developed from slab-lined storage cists which, in some areas, have been found associated with Basketmaker III pithouses. No slab-lined storage cists were found associated with any of the five late Basketmaker III pithouses which have been excavated in the Mesa Verde, however, so the date and manner of development of surface structures awaits further research.

Pottery of the preceding period continued in use in the Mesa Verde with Lino Gray common. Kana-a Gray, essentially the same as Lino Gray but with banded necks, appeared and is a period diagnostic. The dominant Pueblo I decorated pottery of the Mesa Verde is undetermined. Lino Black-on-gray practically disappeared but La Plata Black-on-white continued in use and may, eventually, prove to be the common decorated type. La Plata Black-on-red accounted for 2.0 percent of the sherds in the two Pueblo I sites excavated by O'Bryan (O'Bryan, 1950, p. 92). There has been some question as to whether this pottery was actually made in the Mesa Verde, but its prevalence on the surface in some parts of the park indicates local manufacture.

Pueblo II

Dates: A. D. 900 to 1100. This period is characterized by wide experimentation and rapid improvement in architecture, and marked changes in pottery. At the beginning of the period post and abode villages were still being built, and crude stone masonry had also come into use in some areas. Once stone masonry was accepted it quickly supplanted all other types of construction and this period saw it develop from its crude beginning of rough stones laid in excessive amounts of mud mortar to good, double-coursed, horizontal masonry. The villages usually were small, consisting of a few rooms, in front of which was an isolated kiva. The kiva saw its real development in this period, advancing from a deep pitroom almost to its final Classic stage (Lancaster and Pinkley, Steps in the Development of the Mesa Verde Kiva, in "Excavation at Site 16", in this volume). The dominant decorated pottery of the area was a black-on-white, iron paint type. Corrugated ware, which grew out of the earlier banded-neck variety, came into use and was widely accepted. Red ware declined rapidly in popularity and disappeared. Flat metates came into use as well as large stone blades, or tchamabias, and the so-called sandal-lasts. Other minor traits changed little.

Three Pueblo II ruins have been excavated in the Mesa Verde by members of the park staff and are reported in this volume. Two more were excavated by O'Bryan and portions of a third ruin which he excavated, Site 34, may also date from this period (O'Bryan, 1950, pp. 32-36; 44-51; 79-80). The findings indicate that Pueblo II in the Mesa Verde followed closely the pattern for Pueblo II in the general region, as described above.

The three ruins excavated at Site 16 illustrate clearly the architectural changes (Lancaster and Pinkley, "Excavation at Site 16," in this volume). The first consists of post and adobe living rooms and a crude, four-post kiva with earthen walls. Sitting directly on this ruin is a small pueblo of single-coursed masonry, accompanied by a six-pilastered kiva with masonry below the bench. On top of this ruin is a pueblo built of double-coursed masonry with an eight pilastered kiva which has masonry lining to the top of the pilasters. This kiva has a southern recess which marks the advent of this feature in the Mesa Verde. In this one site are illustrated the important stages in Pueblo II architecture. One new feature, the circular tower, appeared during the last occupation of Site 16, which gave bark dates of A. D. 1074.

Although small quantities of Lino Gray pottery were still made, Mesa Verde pottery of the Pueblo II period consisted almost entirely of two types: Mancos Black-on-white and corrugated. The former was the only decorated pottery associated with the three ruins at Site 16, which covered the entire span of A. D. 900 to 1100. Minor traits show little change except that the flat-slab metate appeared during this period.

Pueblo III

Dates: A. D. 1100 to 1300. The beginning date assigned to this period in the Mesa Verde is somewhat later than that usually given for the region in general. Some archeologists have placed it at 1050, others at 1000, and even A. D. 950 has been suggested as a beginning date. The matter of dividing the period is merely a matter of definition, and it is possible to use any of these dates and develop a perfectly logical classification. As far as the Mesa Verde itself is concerned, the late date seems best from present knowledge. This climax stage is often called the "Great" or "Classic" Pueblo Period. In interpretive work with park visitors, these terms are used and for that reason there is a tendency to lean rather heavily on the term "classic." If the "classic" stages of architecture, ceramics, village layout, and many of the minor crafts serve as the criteria, A. D. 1100 is not too late for the beginning of this period in the Mesa Verde.

This was the climax period, marked by large communities, extensive local specialization and high development of arts and crafts. Multistoried pueblos were built, masonry was superior and kivas were numerous and standardized, although there were variations. Structures with unusual ground plans were also built. Pottery of two types, black-on-white and corrugated, characterize the period. The iron paint of earlier types lost favor and was supplanted by carbon paint. The decorated pottery was of excellent quality with high polish and with skillfully executed black designs on a white background. Excellent cotton cloth was produced and most of the minor traits exhibit superior workmanship. Sandals continued to change and the jog-toe shape appeared.

During the earlier part of the period, most of the villages were small and the population was widely dispersed. Later there were a shifting and, apparently, a decline of the population. Toward the end, the people concentrated in certain areas and large communities were built, many with an obvious defensive intent. Small villages were still present but this later period is characterized by the development of large, compact pueblos, often with defensive aspects.

As far as the Mesa Verde itself is concerned, this is the most confusing and confused period. Two different shifts of the population occurred and much work must be done in 12th century sites before they can be understood.

At the beginning of the period the villages were small and widely dispersed over the mesa tops. Single and double-coursed walls were built and pecked-faced building stones were just coming into use as well as the use of small spalls in the adobe mortar. Round towers had appeared, as evidenced by the finding of three at Site 16 (Lancaster and Pinkley, "Excavation at Site 16," in this volume). In plan the villages consisted of a few rooms in a compact group, to the south of which was an isolated kiva. The kivas themselves had practically reached their classic form and usually contained their standard features: ventilator, deflector, firepit, sipapu, wall niches, bench, southern recess, and six pilasters.

As the period progressed the plan of the villages changed. The kiva was placed inside the house block so that it was surrounded by the houses. Often a round tower was constructed beside the kiva and connected to it by a tunnel, as at Sun Point Pueblo (Lancaster and Van Cleave, "Excavation of Sun Point Pueblo," in this volume). Still farther into the period, and the date for this is unknown, many of the villages grew larger and the population appears to have concentrated in certain areas near the north rim of the mesa. An excellent example of this is the Far View House group where 1 large, three-storied pueblo was closely surrounded by at least 15 smaller ones. Several canyon heads also contain groups of large and small pueblos.

About A. D. 1200, a movement to the caves began. In the many canyons of the Mesa Verde are hundreds of caves, and during the 13th century cliff dwellings were built in almost every cave. The majority of tree-ring dates obtained from cliff dwellings fall in the 1230-60 period, indicating that this was a time of great building activity. In size the cliff dwellings range from 1 room to the largest, Cliff Palace, which has about 200 rooms and 23 kivas.

Cliff dwellings, in general, show no definite plan, for the builders were forced to fit the structures to the available cave space. Architecturally there were no radical changes. Few double-coursed walls were built and single-coursed walls average somewhat thinner than in the earlier mesa-top pueblos. This may have been an effort to save space, or possibly the thick walls were not needed since the caves provided shelter against the destructive forces of the elements. Masonry varied greatly with rough and superior types side-by-side. The finest examples of walls contain well shaped, evenly sized stones which have smoothed or pecked faces, the latter being more common. Many rooms were plastered and bore painted designs. Wide use of small chinking stones was characteristic.

At the beginning of the period Mancos Black-on-white was the common decorated pottery in the Mesa Verde. A radical change came at about this time when iron paint was supplanted by carbon paint. The earliest type of carbon paint pottery has been called McElmo Black-on-white, but Brew stated it mildly when he said, ". . . the definitions of McElmo and the illustrated specimens labelled McElmo vary so that the safest procedure at present seems to be to call it early Mesa Verde" (Brew, 1946, p. 285). The origin of this carbon paint pottery is not clear but further study probably will show wider use of carbon paint during Pueblo II than has been suspected. It seems not to appear in the Chapin Mesa area until after A. D. 1100, although O'Bryan dated it much earlier in other parts of the Mesa Verde (O'Bryan, 1950, p. 26). This early carbon paint type is simply the beginning of Classic Mesa Verde Black-on-white which, apparently, was well developed by A. D. 1200 (Lancaster and Van Cleave, "Excavation of Sun Point Pueblo," in this volume). Typical forms of this finest Mesa Verde pottery are bowls, ladles, water jars, kiva jars, and mugs. Corrugated pottery continued in wide use throughout the period.

Since artifacts have been preserved in the caves, the minor crafts of the 13th century are well represented in collections. In general they are much like those of earlier periods but exhibit superior workmanship. Baskets and sandals, however, had declined somewhat in quality since Basketmaker times. The jog-toe sandal was typical of this period.

The events of Pueblo III in the Mesa Verde are puzzling and lack of excavation of 12th century mesa-top ruins makes this a difficult period to understand fully. However, it is possible to point out certain trends and emphasize the problems yet to be solved.

The first problem is the change in the layout of the villages which has already been mentioned. At the beginning of the period the plan was similar to that of earlier periods, with the kiva separated from the houses. Early in the period the plan changed and the kiva was drawn into the village and surrounded by the houses. Very often a round tower was built beside the kiva and the two were connected by a tunnel. It is difficult not to interpret this change as resulting from a defensive need. The kiva was used chiefly by the men, and this isolated, underground room would have been a death-trap in case of a quick raid. Placing the kiva inside the house structure and connecting it with a tall watchtower certainly hints at a need for defense.

The next problem is an apparent shift of the population. Prior to the early 12th century there was a dense population in a wide belt running east and west across the mesa. On Chapin Mesa, this area of dense population ranges in altitude from about 6,700 to 7,200 feet. Hundreds of ruins ranging from Basketmaker III through Pueblo II times are in this area.

Sometime during the 12th century the population seems to have shifted to the north and there are few Pueblo III mesa-top ruins at the 6,700-7,200 foot elevation. Sun Point Pueblo is one of the few examples on Chapin Mesa (Lancaster and Van Cleave, "Excavation of Sun Point Pueblo" in this volume). The movement evidently was up the mesa for the bulk of the late surface ruins is in the northern part of the Mesa Verde. On Chapin Mesa this is above the 7,500 foot level. Large groups of late pueblos are found on the mesa top and in the broad canyon heads near the north rim. This shift of population is difficult to explain and only one suggestion will be made. The higher portions of the mesa receive more rain and much more snow than the lower elevations. Many summer rains miss the lower sections entirely. Tree-ring records show that from A. D. 1090 through 1101 there were 12 consecutive years during which precipitation was below normal (Schulman, 1947, p. 6). This drought was more severe than in any period of the same length during the great drought of A. D. 1276-99. One can only wonder whether this period of drought may have caused a significant portion of the population to shift to the higher elevations where there was more rainfall.

The second population shift, a movement to the caves, is even more puzzling. Lack of tree-ring dates for mesa-top pueblos makes the beginning date for this shift uncertain but it appears to have been under way by A. D. 1200. As has been mentioned, cliff dwellings were built in practically all Mesa Verde caves during the 13th century. Some people may have remained in surface pueblos, but certainly it was a small percent of the total population.

This move to the caves must have resulted from a need for security. If one accepts it as an effort to provide for easier defense of the villages, it is possible to see the beginning of this defensive trend late in Pueblo II times. Circular towers appeared before A. D. 1100. If these were watchtowers, it means there was some threat to the security of the people. A little later, the kiva was placed within the village walls and was connected to the tower by a tunnel. Still later the population concentrated in certain areas and many large, multistoried pueblos were built. The final step was to move to the caves and certainly this must have resulted from a need for defense. Many of the cliff dwellings were located high on the cliff faces and often they were additionally fortified with defensive walls.

To explode this entire theory one might suggest that the towers were not watchtowers and that the kiva-tower combination was merely a "psychological unit." If the latter is true, it is surprising that the feature was not continued in the cliff dwellings. Connected kivas and towers have not been found in any cliff dwelling in the Meas Verde.

If the "defense" theory is accepted, another question rises immediately. Against whom were the people defending their homes? It has been suggested that in Pueblo III times there was strife within the pueblo group and that the people were warring among themselves. A more widely held theory, however, is that at this time nomadic Indians entered the area and the people were forced to defend themselves against an outside enemy. Only further research will solve the problem.

It must again be pointed out that the 12th century is a confused period in the Mesa Verde. There has been almost no excavation of ruins dating from this century and solutions for the problems will come only through intensive excavation.

Occupation of the Mesa Verde ended just before A. D. 1300. This was the period of the great drought of 1276-99. Since abandonment seems to have occurred at the time of the drought, it is considered the chief cause for the desertion of the Mesa Verde by the agricultural Indians. Internal strife, enemy trouble and other unknown factors may have been contributing causes for the abandonment of the area. As the people left the Mesa Verde they seem to have moved southeast to the Rio Grande. There they merged with other Pueblo groups and soon lost their Mesa Verde identity.

Table 1.—Southwest Classificatory Systems as They Relate to the Mesa Verde Region