NAVAJO The Excavation and Repair of Betatakin

As Plate 10 of his "Preliminary report on a visit to the Navaho National Monument," Doctor Fewkes (1911) publishes a ground plan of Betatakin prepared by W. B. Douglass. This plan I naturally assumed to be correct; it forms the basis of that which appears herein as Plate 3.¹³ House numbers previously assigned are retained in each instance, but since these end with room 96 those additional dwellings disclosed in 1917 have been designated rooms 100-135. Perhaps a dozen others, evidenced by characteristic wall seatings, might have been included had we found opportunity carefully to examine and delineate their respective sites.

---

¹³Although without instruments of precision, we observed certain minor discrepancies in the Douglass survey. This, it should he borne in mind, was made under pressure of time and before all the house walls were exposed. The principal differences between his plan and ours will be manifested by direct comparison of the two and by our description of the individual rooms.

---

The notes which now follow were made during the course of our excavations. Room measurements taken at the same time will be found in the table beginning on page 72.

Room 1, above and west of room 2 in the extreme western end of the cave, was obviously designed for storage. Its floor is the rough, unworked cliff, slightly filled in at the front. Three ceiling beams, extending from the front wall to a masonry fill at the rear, support short cross poles and a thick layer of willows, covered directly by adobe. Two poles overlain with split cedar so parallel the front wall as to suggest a former hatchway in the north corner. Three pegs for hanging articles protrude from the inner front wall.

A 16 by 25 inch (0.40 by 0.64 m.) door opens through the northeast wall. Its sill, widened outside by a sandstone slab on edge, formerly supported a stone door which fitted into grooves on the outer jambs and was held firmly in place by wooden wedges inserted through feather-cord loops. One such loop occupies a hole at each side of the door; that on the north is held in place by mud. At the outer southeast side a bench 5 by 12 inches (0.13 by 0.31 m.) by 11 inches (0.27 m.) high fills a narrow crack.

Through the lower front wall four roof poles protrude from room 2. This latter structure was obviously built first, for there is a distinct line of separation between the large blocks forming its rear wall and the smaller stones in the front wall of room 1.

We attempted no repairs in rooms 1-8; nor on the masonry of court 10.

Room 2 was probably a storeroom; no smoke stains appear on its walls. It floor is mostly of fairly level, native rock, but a shallow fill in front concealed several east-west poles horizontally embedded about 6 inches (0.15 m.) above the base of the east wall. These were undoubtedly deemed necessary as supports or anchors, since the masonry stands flush with the cliff edge. Two pegs protrude from the southeast wall; two beams extend through from room 3. Six beams support the ceiling of willows, cedar bark, and adobe mud; in the northeast corner of this is a hatchway, the only entrance.

Five feet six inches above the outer southeast corner of the roof are two holes, drilled through a cliff angle; near by is a similar, single eyelet. From these, various light objects were doubtless suspended by the ancient inhabitants.

Room 3, a dwelling, exhibits the smoke stains of long occupancy. The lower half of its three masonry walls was repeatedly plastered, but the west, or cliff, side was not similarly surfaced. Two small, parallel timbers next the cliff and a larger, central beam support 10 east-west ceiling poles carrying layers of willows, reeds, and mud. A small smoke vent through the middle roof, next the north wall, had been closed with cedar bark and plastered over. In the southeast corner the solid sandstone surface is about a foot lower than the remaining floor, which lies some 8 inches (0.20 m.) higher than that of room 4.

At the floor level an opening 11 inches (0.27 m.) square connects with room 4. Directly in front of this opening, embedded in the floor of room 3, is a willow loop—one of two loom anchors. The second is wanting, but directly above, two loops still hang from a ceiling cross pole. Four wall pegs remain in position; holes for four others may be noted.

A door opens through the north half of the east wall. Just within this door, and formerly protected by a now missing screen, is the fireplace. The retaining wall continuing from court 10 ends against the outer east wall, south of the door. Above its south jamb six shallow holes had been drilled.

On the cliff face above the roof a sandal figure had been pecked; pointed and flat implements had been sharpened and white paint crudely spread in several places. Above the southeast roof corner one may note portions of a pair of drilled holes on a broken cliff slab.

Room 4 adjoins room 3 on the north and is apparently of later construction. Its east wall is missing; of the north wall, a section at the northwest corner stands 3 feet (0.91 m.) high against the cliff. Above this fragment and formerly concealed by the wall are several holes, pecked through the cliff corner; two pairs of similar holes on the cliff face had been plastered over. Through the west end of the south wall, 8 inches (0.20 m.) above the floor, a 10 by 18 inch (0.25 by 0.46 m.) opening marks the position of a former room 3 door, the lintels of which still show through three coats of plaster.

In the middle floor is a slab-lined fireplace measuring 16 by 18 inches (0.40 by 0.46 m.); its deflector slab, standing on the east, suggests a former door on that side.

Room 5, beneath room 6, was used for storage. At least the smoke stains which ordinarily identify a dwelling are lacking. Vertical cliff forms its south and west walls, while the other two are of masonry, freely chinked. The floor, of native rock, is slightly higher than the terrace outside.

A single beam, paralleling the cliff, supports the ends of six cross poles; upon and at right angles to these are layers of willows and reeds. On the floor we found a cedar pole, slightly shorter than the room length, which may formerly have been an additional ceiling support, since its butt end fitted in a hole at the north side when the opposite end rested on a protruding rock between the main beam and the cliff. Through the middle east wall is a 15 by 24 inch (0.38 by 0.61 m.) door whose outer jambs are grooved for the usual slab. A loop fastener appears on the north side only. Four sticks support the sandstone lintel; in its stone sill are two grooves where axes were sharpened.

Fire stains on the cliff at the outer southeast corner mark the position of an open fireplace.

Room 6, above room 5, probably served also for storage. Its walls are unsmoked; its, floor, smooth and hard. The cliff forms its south and west sides, while the other two are composed of small, irregular stones laid in quantities of adobe mud, chinked with sandstone spalls. The outer east wall, smooth and regular, is superior to most Betatakin masonry; the outer north wall is plastered. (Pl. 7, A.)

Paralleling the south cliff wall is a single large ceiling beam, the west end of which rests on a narrow wedge of masonry, built in a crevice. This south beam supports one end of a timber which parallels the west cliff, and the latter beam, in turn, supports the ends of 14 cross poles. Above these is a thick layer of willows with adobe mud spread directly upon it to complete the roof. Six central cross poles have been broken by rock falling from the overhanging cliff.

A door through the middle north wall measures 18 by 24 inches (0.46 by 0.61 m.). Its lintel includes four sticks and a large sandstone block; its sill is a well-worn slab showing two shallow grooves where axes were sharpened. This sill and both outer jambs are deeply grooved for the door slab characteristic of storerooms; on each side is a willow loop for wooden fasteners. An inset step, 7-1/2 inches (0.19 m.) wide by 3 inches (0.076 m.) high by 2-1/2 inches (0.063 m.) deep, lies in the outer north wall 19 inches (0.48 m.) below the door sill and 17 inches (0.43 m.) above the roof, of room 7. A similar step, 4 by 1-3/4 by 2 inches, lies upon the lintel slab, 24 inches below the roof of room 6. The protruding end of the west beam doubtless served as an additional step.

Room 7, adjoining room 5 on the north, is a dwelling whose west wall is formed, by the cliff. Its floor is mostly solid rock, but the east third was filled in to a depth of from 6 to 18 inches. Two north-south beams support a ceiling of 18 cross poles overlain with willows and cedar bark. One of the cross poles was spliced, its two portions being tied with yucca strips; three cross poles extend only from the central beam to the east wall. Of equal interest is the fact that the west beam, hewn off on top, was braced in the middle by a post resting on a sandstone block. A former hatchway, 3 feet 3 inches (0.99 m.) from the north wall, is evidenced by the cut ends of several cross poles showing in the broken roof of the southeast quarter.

Abutting the outer northeast corner of room 5 is a section of masonry, 20 inches (0.51 m.) wide, the perpendicular east end of which is slightly convex, carefully chinked with numerous spalls, and plastered over as though to form a door jamb. Similar treatment is noted for the upper south end of the broken east wall. Together, these finished wall ends suggest for room 7 a southeast corner door extending from floor to ceiling. (Pl. 8, B.) Such an entrance would be most unusual; none like it is observed elsewhere in Betatakin nor, indeed, can I recall having seen one in any other prehistoric Pueblo village.

At the outer northwest corner, just below the roof level, a single step had been pecked in the cliff face.

Room 8 is merely a recessed platform, 20 inches wide by 6 feet long (0.51 by 1.82 m.), formed by a convex wall on a ledge above the northwest corner of room 7. (Pl. 7, A, B.) Externally this wall stands 26 inches high to the solid rock floor; there is no trace of upper wall masonry.

Room 9. The remains of a small storeroom opposite the northeast corner of room 7; outside of and below the retaining wall of court 10. It is shown but not numbered on Douglass's ground plan.

Court 10, long and narrow, lies between rooms 7 and 11; it is formed by the irregular retaining wall which extends from the south corner of room 11 to the east side of room 3. This wall appears not to have stood more than 1 foot above the court floor; the latter consists mostly of native rock, with a débris fill along the east side. Shallow, close-lying steps on the uneven rock surface next the cliff were probably pecked by children at play. Opposite the northeast corner of room 7 and fronting rooms 3 and 4 the retaining wall had largely disappeared prior to 1909; elsewhere it is still well preserved. At the northeast corner of room 4 the walk formed by this wall was extremely narrow.

Against the retaining wall and 5 feet 3 inches (1.6 m.) from room 11 is a fireplace measuring 20 by 34 by 12 inches deep (0.51 by 0.86 by 0.31 m.). In the east corner, with its sill at the court level, a 10 by 11 inch (0.25 by 0.27 m.) vent opens downward into room 11. Nearby, the fragment of a small post stands against the inclosing wall.

At the north end of court 10 a slender pine pole, 30-35 feet long, gave access to the gallery ledge above. (Pl. 26, A.) The ladder leading to the roof of room 11 is one we substituted for steps pecked in the cliff at the north corner.

Room 11 is one of the most interesting chambers in Betatakin, since it obviously is an old dwelling, remodeled for ceremonial purposes. Next the cliff wall is a ceiling beam whose west end rests on a shelf of masonry; two other beams, with butt ends opposite, lie side by side across the middle room. On these timbers are 25 cross poles, overlain by layers of willows, cedar bark, and sand. Where the cross poles were too short to reach from the southeast wall to the cliff two were placed together, butt ends opposite. The lower third of the southwest wall and the lower half of the southeast and northeast walls were surfaced; traces of plaster also remain on the northwest, or cliff, side.

Incised drawings appear in the smoke-blackened plaster of all four walls. (Fig. 1.) On the north half of the southwest wall 22 parallel and approximately vertical lines varying in length from 1 to 4 inches have been scratched. Other scratches occur below this group, but no geometric design is distinguishable among them. So-called "pottery" or "basket" designs were incised in the plaster of the southeast wall.

FIGURE 1.—Incised figures on the walls of Room 11. (The bar represents 1 inch)

A south corner roof opening, 17 by 29 inches (0.43 by 0.73 m.), had been cut through after completion of the ceiling and perhaps after conversion of the room. This hatchway, 8 inches from the southwest wall, was rimmed with slabs on all except the southeast side; its inner jambs, above the cross poles, were plastered with adobe and chinked with small spalls. Two feet from the west corner and 13 inches above the floor a former southwest door, 17 by 29 inches (0.43 by 0.73 m.), had been carefully blocked from the outside, leaving its inner sill exposed to form a shelf 5 inches wide.

In the south corner an opening, 12 inches (0.31 m.) high by 17 inches (0.43 m.) wide, formerly pierced the southeast wall at the floor level; its east jamb sloped to the west. Externally this ventilator was a full third less than its inner dimensions. Apparently it did not prove entirely satisfactory, for it had been closed with masonry from the inside and a second vent, 12 inches (0.31 m.) high by 11 inches (0.27 m.) wide, cut through the southwest wall flush with the south corner, and 9 inches above the floor. The slab sill of this second opening slopes upward to the level of court 10. Within the room and directly below this latter vent are the remains of a shallow, basinlike depression, 16 inches (0.41 m.) in diameter by 1-1/2 inches (0.038 m.) deep. Formerly a ladder extended through the hatchway directly above this depression. A slab fire screen, 24 inches (0.61 m.) wide by 25 inches (0.64 m.) high by 1-3/8 inches (0.034 m.) thick, abuts the southeast wall 27 inches (0.69 m.) from the south corner. Eight inches east of this screen and 5 inches (0.13 m.) from the southeast wall is a fireplace whose dimensions average 16 by 18 by 7 inches deep (0.41 by 0.46 by 0.17 m.); it is lined and paved with slabs and its inner corners rounded with adobe. The ceiling willows above this fireplace had been plastered with mud—a simple, protective measure that might have been followed profitably by other prehistoric house builders.

Our repairs were confined largely to the southwest wall. In addition, we propped a cracked ceiling beam with a cedar post set on a sandstone slab; and placed, through the hatchway, a pole ladder fitted with oak rundles.

On the cliff above room 11 two holes, 1-1/2 inches apart, had been drilled through a thin layer of sandstone.

Room 12 is a storeroom, built against the outer northeast wall of room 11. Its floor is the solid sandstone of the uppermost terrace; its masonry bears no trace of smoke. A single northeast-southwest beam crosses the room slightly nearer the cliff than the southeast wall; four cross poles support a ceiling of willows, cedar bark, and adobe mud. Several potsherds were used as chinking in the northeast wall. One peg protrudes from the upper southeast side, 3 inches from the east corner.

An 18 by 24 inch (0.46 by 0.61 m.) door opens through the middle northeast wall. Its lintel consists of three pieces of split cedar, supporting a large stone; its sill is a thin sandstone slab. Both outer jambs are slightly grooved for the usual door slab; loops for fasteners appear on each side. A mano was incorporated in the masonry under the outer north jamb; in the wall above, two empty holes for wall pegs may be seen.

Court 13 separates room 11 from room 14; on the southeast stands an inclosing wall, now 2 feet high. Room 12 is a later addition, built in the west corner of the court against room 11; sheer cliff forms the northwest side. The rock surface in the north corner had been pecked down to approximate the level of an adobe pavement which covers a deep débris fill against, the southeast retaining wall. Open fires and the varied domestic activities pursued in the court have darkened its entire floor and left their mark on some of its surrounding masonry.

A shallow fireplace, rimmed with adobe (now much broken), lies at the base of the cliff 26 inches (0.66 m.) from room 12. Three slab fragments stand on edge. back of this fireplace; within it we found a dressed sandstone tablet measuring 10-1/2 by 9 by 3/4 inches (0.26 by 0.22 by 0.019 m.). Near by a mortar, 11 inches (0.27 m.) in diameter by 14 inches (0.35 m.) deep, had been pecked into the rock floor 9 inches from the cliff and 2 feet 7 inches from the north corner. Its concave bottom and vertical sides are stained by fire and ash. Three grooves on stones of the northeast wall near the north corner show where implements were sharpened. Here also a protruding beam end and three shallow steps pecked in the cliff gave access to the roof of room 14; a fourth pecked step, or handhold, appears a bit higher on the cliff. The roof of room 11 was doubtless formerly reached by a ladder in the alcove south of room 12.

To aid passage to and from court 13 we substituted a notched cedar (pl. 27, A) for the old pecked steps in the north corner and placed a pole ladder in the south corner. Minor wall repairs were made, especially on the southeast.

Room 14, apparently a ceremonial chamber, stands between court 13 and room 15. Its floor is entirely of native rock, partially worked down in the west corner. On the northwest side a ledge slopes from the upper west corner to the floor at the north. The west half of this ledge was built up with masonry, filled with débris and capped with slabs 31 inches (0.78 m.) above the floor to form a shelf 12 inches (0.31 m.) wide at the northwest and 24 inches (0.61 m.) wide at the northeast.

The roof is supported by two central northeast-southwest beams; three others, side by side, lie next the cliff. On these beams fifteen cross poles support a layer of willows, mud covered. Ceiling and walls are heavily smoked. The lower two-thirds of the southeast and northeast walls are plastered while the southwest and northwest sides remain unsurfaced. Several small ovals and circles, made by pressing a pointed instrument into the new plaster, appear on the southeast wall; also, several casually incised lines. Two pegs protrude from the upper northeast side. Apparently after completion of the room, a vent about 4 inches in diameter was cut through the southeast wall, 14 inches (0.35 m.) below the cross poles and 4 feet 6 inches (1.37 m.) from the east corner.

A door opens through the middle southwest wall with, its slab sill, slightly grooved where a stone ax was sharpened, at the court 13 level. Just within this door and continuing from its sill a sandstone slab which has settled 2 inches at the north tops a masonry platform 16 inches (0.41 m.) wide and 16 inches high. This platform lies between the wall and a masonry fire screen, 25 inches (0.64 m.) high by 27 inches (0.69 m.) wide by 10 inches (0.25 m.) thick, capped by a single stone. Pecked into the rock floor at the northeast base of this screen is an ovoid fireplace, 25 by 18 by 4-1/2 inches deep (0.64 by 0.46 by 0.114 m.). In the ceiling above the platform an unrimmed smoke vent measures 6 by 10 inches (0.15 by 0.25 m.). Northwest of fireplace and platform the cave floor slopes upward toward the cliff; in this sloping surface four steps were pecked.

Four feet six inches from the north corner and 8 inches (0.20 m.) above the floor a second door opens into room 15. Loops for door fasteners protrude at points 8 inches and 12 inches (0.31, respectively, from its grooved jambs.

Five pairs of pecked eyelets and others, unfinished, appear on the northwest cliff above the roof; here also is a narrow groove where some implement was pointed.

We repaired the walls, recapped them with a single course of stones, and patched several holes in the roof.

Room 15 may be identified as a storeroom from the absence of smoke stains on its inner walls and from the presence of grooves on its outer door jambs. Within the room, solid cliff sandstone slopes rather sharply downward toward the east corner, which was filled to provide a limited floor area. A rock ledge averaging 12 inches (0.31 m.) wide extends along the northwest wall base.

The roof is supported by a single beam seated 8 or 10 inches from the cliff and by three pairs of smaller timbers lying butt ends opposite. Upon these are five pairs of cross poles, covered with layers of willows and cedar bark. One peg protrudes from the southeast wall; five, from the northeast. A 17 by 24 inch (0.43. by 0.61 m.) door opens through the southwest wall 16 inches (0.41 m.) above the floor; two broad pieces of split cedar form its lintel. In the sloping sandstone beneath this door are five pecked steps.

On the northeast side a hole, approximately 20 inches in diameter and obviously of fortuitous origin, was repaired in 1917. In addition, the walls were recapped and holes in the roof mended.

Room 16 lies against the cliff, between room 15 and court 24. Its northeast and southwest walls are of masonry, but its southeast wall, now missing, was undoubtedly of wattle. Although the rock floor next the cliff had been worked down somewhat, it remains slightly higher than the front, filled portion. The east side of a low, narrow ledge at the base of the northwest, or cliff, wall was leveled with masonry to form a shelf. Smoke stains appear on the floor and at the base of all three walls.

The roof had burned, but one charred beam end remains in the northeast wall next the cliff, and three notches, pecked for beam rests, appear in the southwest wall. Above these, charred timbers protrude from room 15; the northeast wall masonry extends 18 inches (0.45 m.) above the roof level of room 16.

Room 17 is a possible storeroom situated between rooms 18 and 41; in front of and below rooms 15 and 16. Cliff forms its lower northwest side; its southeast wall is of wattle—posts supporting up right willows, bound at intervals by horizontal withes either singly or impairs, and the whole plastered with adobe. (Pl. 9, A.) Smoke stains are not perceptible.

Excepting a small section in the west quarter the floor is artificial and rests upon 19 inches (0.48 m.) of débris covering an earlier surface. No plaster appears on the southeast wall below the upper floor, although its willows and posts extend to the lower. In the north corner a 9 by 12 inch (0.22 by 0.31 m.) shelf was made by leveling a small rock ledge with adobe. Above this shelf, next the northeast wall, is a recess measuring 8 by 12 by 5 inches (0.20 by 0.31 by 0.13 m.) deep. In the west corner pecked steps lead up the sloping sandstone to a ceiling hatchway. Between floor and roof, one additional pecked step appears in the northwest wall; two in the southwest.

The roof is supported by four northeast-southwest beams. Upon these are 12 cross poles, only one of which extends the entire length of the room; then layers of split cedar, reeds (Johnson grass), and sand. In the west corner an 18 by 24 inch (0.46 by 0.61 m.) hatchway is reached by pecked steps. From the north middle beam a loop of knotted yucca leaves hangs 12 inches and is rather sharply curved at the bottom as though it once supported a pole approximately 2 inches in diameter. We recapped the walls and patched the broken edges of the roof.

Room 18 is a kiva, or ceremonial chamber, if one may judge from the character of its furniture. Its floor is mostly of native rock; its walls and ceiling are heavily smoked. The lower half of its northeast and southeast walls is plastered and in this appear miscellaneous scratchings—mostly vertical, parallel lines. Similar incisings may be noted on the northwest, or cliff, side. A sloping ledge along the face of this same wall had been leveled in the north corner to form an 18 by 30 inch (0.46 by 0.76 m.) shelf about 3 feet (0.91 m.) above the floor. During the initial examination of Betatakin in August, 1909, the present writer found two clay pipes and other articles on this shelf. They are now in the University museum at Salt Lake City.

The roof of room 18 includes 3 northeast-southwest beams, 10 pairs of crosspoles and layers, respectively, of willows and cedar bark. As usual, the willows are tied down to the cross poles by split yucca leaves. A door through the middle southwest wall has a slab sill and four oak lintel sticks about 1 inch in diameter; a 2-1/4-inch post stands against the inner south jamb. (Pl. 8, A.) Within the room, 20 inches (0.51 m.) from the southwest wall and separated from it by a platform lying 6 inches below the door sill, stands a masonry fire screen, 34 inches (0.86 m.) high by 27 inches (0.68 m.) wide by 10 inches (0.25 m.) thick. The platform extends a few inches beyond the south end of the screen; the latter is capped with a single sandstone block, 27 inches (0.68 m.) long by 13 inches (0.33 m.) wide by 3 inches (0.07 m.) thick. At the northeast base of this screen is a now broken, unrimmed fireplace approximately 2 feet (0.60 m.) in diameter; it was formerly lined with irregular stone blocks. In the roof above the platform, and obviously cut after completion of the ceiling, is a 7 by 10 inch (0.17 by 0.25 m.) smoke vent, rimmed with adobe-covered cedar bark. A second ventilator, 9 inches (0.22 m.) wide by 12 inches (0.31 m.) high, appears in the southeast wall 3 feet (0.91 m.) above the floor and 4 feet 5 inches (1.34 m.) from the east corner. Externally, this latter vent is almost round and about 6 inches (0.15 m.) in diameter.

During the snowstorms we experienced in 1917 my crew and I repeatedly sought shelter in this room and built fires in the south corner. Excepting minor patchwork along the roof edges and a new course of capping stones no repairs were necessary.

Court 19. During occupancy of the village this open space may have been leveled with débris. But to-day its sandstone surface, slightly higher in the middle, drops abruptly to the sloping cave floor between room 20 and the retaining wall below room 7. On the southeast side of the court a few courses of masonry cap the cliff terrace that serves as northwest wall for rooms 20-21. One sees no evidence of a passageway to court 13, above and at the northwest, but six pecked steps in the cliff at the north corner gave access to the roof of room 18.

At the cliff base, 3 feet 4 inches (1.01 m.) from the north corner, a pot-shaped hole 11 inches (0.27 m.) deep had been pecked in the solid rock. Three inches below the surface the diameter of this receptacle is 11 inches, but the body diameter, like that at the orifice, is 13 inches. Open fires had burned between this hole and the north corner of the court.

Room 20, a probable storeroom, stands below and south of court 19. (Pl. 10, B.) The face of a cliff terrace, thinly plastered, forms its northwest wall; the others are of masonry. Smoke stains are not present. The floor is mostly artificial. A cedar log is embedded in the masonry of the southwest wall; about 2 feet below the floor level.

Four northwest southeast beams, two of which lie side by side in the middle of the room, carry 17 cross poles with overlying layers of willows and cedar bark. At the northwest all four beams rest upon a log which lies on a shoulder of the cliff, although notches for their individual support had been pecked in the cliff face above the log. Eight pegs protrude from the walls, close up under the beams; a hole for one additional peg is noted. The only entrance to the room is a hatchway, 20 by 24 inches (0.51 by 0.61 m.), in the south corner.

The terrace on which this storeroom was built had previously settled away from the cliff 14 inches; the resultant crack, extending lengthwise through the middle of the floor, had been filled with household débris. A 2-inch, fracture in the northeast masonry evidences further settling since abandonment of the room. To check this we tied the southeast wall to the cliff with two steel rods, equipped with turnbuckles and expansion bolts. In addition, a hole in the lower southwest side was closed, the walls were recapped, and the roof patched.

Room 21, the longest in Betatakin, may also have served for storage, since its walls bear no trace of smoke. The sandstone cliff, capped with about 14 inches (0.35 m.) of masonry, forms its northwest wall; those adjoining on the northeast and southwest have been cracked by settling of the ledge on which the room stands. The floor is mostly native rock.

Four northwest-southeast beams supported the roof. Of this, only two cross poles now remain; upon them, in the west corner, are willows and cedar bark. Five wall pegs are noted, broken off flush with the masonry. Lacking evidence to the contrary, we assume the hatchway—sole entrance to the room—was located in the south corner, as in room 20.

The southeast wall we anchored to the cliff with a single steel rod. The north end of this, provided with an expansion bolt, was set in a drilled hole and packed with cement. In addition, we repaired the upper walls and replaced a missing beam. No effort was made to complete the roof.

Room 22, probably a kiva, adjoins room 21 and occupies the same terrace. Its northwest wall is formed by the face of the next higher ledge, topped by 2 feet of masonry. The lower three-fourths of all four walls are plastered and heavily smoked; the floor, mostly native rock, was worked down in the south corner to within 6 inches of the general level. Without lateral doors the chamber necessarily was entered through a hatchway.

Two northwest-southeast beams supported the roof; seven pecked rests for cross poles appear in the masonry of the southwest wall. At the floor level a 9 by 7 inch (0.22 by 0.17 m.) ventilator opens through the southeast wall 38 inches (0.97 m.) from the south corner. Extending into the room from the east jamb of this vent is a deflector slab, 17 inches (0.43 m.) wide by 16 inches (0.40 m.) high; at its northeast base lies a broken, stone-lined fireplace.

Two pecked steps at the west end of the northwest wall were doubtless cut before construction of the room; another "step," 6 inches (0.15 m.) wide by 4 inches (0.10 m.) high by 1 inch (0.025 m.) deep, appears in the middle of the same wall, 21 inches (0.53 m.) above the floor. Near by is a partially demolished shelf, 5 inches (0.12 m.) wide by 8 inches (0.20 m.) long, built of adobe on a narrow surface. One wall peg protrudes from the northeast side.

A rock ledge extending the entire length of the southwest wall was leveled in front with masonry to form a bench, 16 inches (0.40 m.) wide by 25 inches (0.64 m.) high. Four sandstone slabs surface this bench, the face of which had been battered by stone hammers and subsequently plastered. The crack caused by settling away of the terrace on which rooms 20-23 were built had been filled throughout with rubbish. But in room 22 the southwest wall masonry continued down into this crack, thus leaving, under the paved bench, a recess 10 inches (0.25 m.) wide by 16 inches (0.40 m.) deep.

Incised designs appear on the plastered bench face and on all four walls. (Fig. 2.) An irregular rectangle about 5-1/2 inches (0.13 m.) high by 7 inches (0.17 m.) wide, with crossed lines forming 100 small squares, is seem on the southeast wall; a similar figure shows faintly on the northeast side near the east corner, and just to the left of it is a familiar pottery design. The best preserved of all these incised figures appear on the middle northwest wall, in the plaster covering the cliff face; in addition, there are faint waved lines and miscellaneous scratchings. A series of six "turkey tracks," with other figures, will be noted in the plaster above the southwest bench.

FIGURE 2.—DESIGNS CARVED ON THE WALLS OF ROOM 22. (THE BAR REPRESENTS 1 INCH)

To check settling of the southeast wall we anchored it to the cliff with two steel rods fitted with turnbuckles. Minor wall repairs were made.

Room 23 is a relatively small structure between rooms 22 and 42, south of and below passage 40. Its northwest wall is formed mostly by the perpendicular face of a ledge; the other three are masonry. Of the northeast wall only three stones remain and these lie in the subfloor crack continuing through: rooms 20-23; of the southeast wall, only a fragment in the south corner has survived. The floor is entirely artificial. Fires had burned against the southwest wall, near the south corner. Upon the floor at the base of the northwest wall lay 3 feet of old débris; above this, the looser earth and sand thrown out by earlier excavators.

Although the roof is entirely missing, two beam rests show in the stonework above the cliff and holes for five cross poles appear in the masonry of the southwest wall, 5 feet 6 inches (1.7 m.) above the floor.

The Douglass ground plan inadvertently shows the northeast wall of room 23 a continuation of that in room 39. Its correct position, evidenced by the masonry fragment in the subfloor crack, is indicated on Plate 3.

Court 24 lies between rooms 16 and 25. A cliff ledge on the northwest side forms a bench 18 inches (0.45 m.) wide at the west by 3 feet (0.91 m.) wide at the east; its middle surface rises nearly a foot above the remainder. At the east end, which stands about 3 feet high, a depression had been filled with spalls and débris; over this, a series of nine stone steps led from the court floor to the roof of room 26. At the opposite end of the bench four pecked steps gave access to the roof of room 16. Below this shelf is a second and narrower ledge from which the native rock floor slopes gently downward toward the south side of the court. Here a southeast wall supports a débris fill, approximating the level of the sandstone surface at the northwest. In the west corner two shallow mortar-like basins, 11 inches (0.27 m.) in diameter by 2 inches (0.05 m.) deep and 8 inches (0.20 m.) in diameter by 1 inch deep, had been pecked in the solid rock.

The southeast side and the adjacent wattled wall required minor repairs (pls. 11, A; 12, B) in 1917; three of the north steps were relaid (pl. 27, B) and a notched cedar was substituted for those pecked at the west.

Room 25, east of court 24, stands against the cliff immediately below room 26. Half of its wattled southwest wall was still standing; we reconstructed, but did not surface, the remaining portion. (Pl. 11, B.) The lower 3 feet (0.91 m.) of the three masonry walls had been surfaced; 10 superposed layers of plaster were counted in one place. Incised designs appear on all except the wattled wall. In the north corner, 2 feet 8 inches (0.81 m.) above the floor, is a triangular shelf measuring 7 inches (0.17. m.) on the northeast by 6 inches (0.15 m.) on the northwest. The charred ends of two beams, 8 inches in diameter, protrude on the northwest, 4 feet 5 inches (1.34 m.) above the floor; a lesser beam supported the top of the wattled wall. Seatings for 10 cross poles appear on the northeast side.

Three feet eight inches (1.11 m.) from the west corner a door, 18 inches (0.45 m.) wide and 10 inches (0.25 m.) above the floor, opens through the wattled wall into court 24. Against its inner north jamb stands a shattered masonry abutment, 9 inches (0.22 m.) thick. This joined the wall to a now missing fire screen which stood in front of the door to shield a fireplace, 24 inches (0.61 m.) in diameter by 5 inches (0.13 m.) deep, pecked into the rock floor. A subfloor cist 3 feet (0.91 m.) deep, about 2 feet 6 inches (0.76 m.) in diameter and lined with rough masonry, occupies the east corner of the dwelling.

In the northwest wall, 19 inches (0.48 m.) from the west corner and 13 inches (0.33 m.) above the floor, is a pecked recess measuring 4 inches (0.10 m.) square by 4 inches deep. A second recess of equal size and comparable depth lies in the northeast wall, 3 feet 6 inches (1.1 m.) from the north corner and 14 inches (0.35 m.) above the floor. A peg protrudes at an angle from the north half of the wattled southwest wall. In the upper southeast wall, near the east corner, is a 7 by 9 inch (0.17 by 0.22 m.) opening whose external dimensions are 5 by 6 inches. (0.13 by 0.15 m.). A former ventilator, now blocked, about 11 inches (0.27 m.) wide by 14 inches (0.35 m.) high, is discernible at the floor level in the middle southeast wall.

Room 26, built upon an upper terrace northwest of and overlooking room 25, was used for storage purposes. Its rock floor slopes unevenly downward toward the southeast. Although the roof is missing, two pecked beam rests are noted in the northwest cliff; timbers protrude slightly from room 25. In the south corner three jutting stones in the northwest wall and two in that adjoining served as steps to a former hatchway.

In 1917, we substituted other timbers for the two missing beams and laid a hewn plank lengthwise of the room to facilitate access between courts 24 and 34.

On the cliff about 6 feet (1.82 m.) above the roof level of room 26 are four pairs of holes—shallow, pecked depressions connected by drilling. Two of these pairs are still occupied by small willows which protrude at right angles from the upper member. Like similar holes elsewhere on the walls of Betatakin cave, these were obviously prepared for suspension of maize, herbs, or equally light articles. They could hardly have been intended for loom cords since they are unequally spaced and the smaller pairs would not have resisted the tension required.

Room 27, a dwelling between rooms 25 and 100, stands below and in front of court 34. The cliff, surmounted by 18 inches (0.45 m.) of stonework, forms its northwest side; the southeast wall is of wattle, mud covered. Only the lower half of the three masonry walls was plastered. The charred ends of three northwest-southeast beams are visible, and two of these lie side by side, near the west corner. At the southeast these three beams rested upon individual posts set just within the wattled wall. Seatings for five cross poles were pecked in the upper northeast masonry, but the former ceiling undoubtedly included 10 or 12 such timbers. The stonework upon the northwest, or cliff, wall was set back 5 inches (0.12 m.), thus forming a shelf the entire width of the room and about 4 inches (0.10 m.) below the beams. This shelf we repaired; at the same time the northwest and northeast walls were recapped.

In the middle southeast wall is a broken door. Within it, 22 inches (0.56 m.) from the wall, stands a slab fire screen measuring 25 inches (0.64 m.) wide by 14 inches (0.35 m.) high; its west end is joined to the wattled wall by slabs embedded on edge and extending 4 inches above the floor. At the northwest base of the screen is a fireplace, 7 inches (0.17 m.) deep, and averaging 16 by 21 inches (0.40 by 0.53 m.).

Court 28 lies between rooms 27 and 29, southeast of room 100. Like other courts, it served as an open living room in which diverse domestic activities were pursued. Its northeast wall, 5 feet 3 inches (1.6 m.) long, is of wattle (pl. 12, A); on the southeast a low retaining wall, extends westwardly to form a narrow passage fronting room 27. In the middle northwest side a door opens into room 100; both jambs are grooved and loops are present for wooden fasteners. Minor repairs were made on the wattled wall and on that adjoining at the southeast.

Room 29-30. A 2-story dwelling northeast of court 28 and room 100; southeast of and below room 31. Little remains of the upper chamber; the following notes pertain to the lower. For the most part, the floor is artificial, but along its northwest side the native rock had been worked down to approximate the general level. In the west corner a hole, 10 inches (0.25 m.) in diameter by 12 inches (0.30 m.) deep and now fire stained inside, had been pecked from the solid sandstone. Three shallow grooves where axes were sharpened appear in the northwest half of the room; here also are three pecked holes, averaging 2 inches in diameter by 2 inches deep, one of which contains a loom anchor stick. The wattled portion of the southwest wall was entirely plastered; of the others, only the lower half was so treated. No door being present, this lower room necessarily was entered through a hatchway.

The southeast wall was built upon a horizontal log. At the northwest, lying on a low ledge of rock and supporting the upper wall masonry, is a similar log; above it, 4 feet (1.2 m.) from the sloping floor, is a second log (boxelder) and a foot higher, the charred ends of two pairs of northwest-southeast beams. A fifth beam, its ends only slightly embedded, rested close against the northeast wall. On this same side are recesses for, or the broken ends of, seven cross poles; three pole ends appear in the north 4 feet 3 inches (1.3 m.) of the southwest wall, the remaining 5 feet (1.5 m.) of which is of wattle. Extending along the base of this wattled section is a masonry bench 15 inches (0.38 m.) wide by 3 feet 4 inches (1.01 m.) long. An upright slab, 2 feet (0.60 m.) wide by 2 feet 8 inches (0.81 m.) high, adjoins the southeast corner of this bench as a screen for an ovoid fireplace. The latter measures 18 by 17 inches by 7 inches deep (0.45 by 0.43 by 0.17 m.); although pecked from solid rock, it is lined on the southeast and southwest by slabs. To further shelter this fireplace, a second upright slab, 12 inches wide by 2 feet high, extends eastwardly from the bench corner and the first screen. The bases of both these deflectors are supported by embedded slab fragments. In the west end of the south wall and at the floor level is an air vent, 8 inches (0.20 m.) wide by 12 inches (0.30 m.) high. The southeast masonry continues 2 feet beyond the abutting wattled wall. In the west corner narrow shelves on the northwest and southwest sides were capped with slabs. Only one wall peg is evident, on the southeast.

The second-story northeast wall apparently was added after completion of the lower room, for its basic stones are larger than the others. Although still in excellent condition, this upper wall shows no seatings for ceiling beams or cross poles; an unusual feature of it is the stepped construction at its south end, the "step" being 14 inches (0.35 m.) deep by about 30 inches (0.76 m.) long. (Pl. 11, A.) In this same second-story wall are two small recesses, approximately 5 inches (0.12 m.) wide by 4 inches (0.10 m.) high by 4 inches deep. One lies in the north corner; the other, in the middle wall and about 2 feet (0.60 m.) above the second-story floor level. The first of these recesses was plainly left during construction, but the second appears to have been made subsequently by removal of a few small stones.

Room 31, northwest of and above room 29, is another storeroom. Its floor is of native rock. A cliff terrace forms its northwest wall, but masonry was employed at each end to support the ceiling cross poles; between these seatings are three holes, pecked for other pole ends. A single northeast-southwest beam crossed the middle room. Through the southeast wall a 15 by 22 inch (0.38 by 0.56 m.) door opens upon the roof of room 29. The slab sill and both jambs are grooved; loops occur at either side.

After patching and recapping the walls, we presumed to reconstuct the roof. (Pl. 13, A.) Ceiling poles and willows were salvaged from other sections of the village; new willows were brought in from the banks of the creek near camp. Upon the poles a layer of willows was spread and tied with strips of yucca leaves; upon the willows we placed a layer of cedar bark and covered it with dry sand. Although in roof construction they invariably used mud as a covering for willows or reeds, the builders of Betatakin seem often to have employed dry sand only in completing a roof in which cedar bark formed the next but final layer.

Room 32, northwest of and above room 31, also was utilized for storage. Unworked native rock, quite uneven and sloping sharply downward toward the south corner, sufficed as floor. In this slanting surface a single pecked step occurs. A shallow, pecked beam rest is noted on the cliff in the north corner; a few inches lower, but outside the northeast wall, is a similar recess, 5 inches (0.12 m.) in diameter by 2-1/2 inches (0.06 m.) deep. It is conceivable that the builder's plans were altered during construction.

Through the middle east wall a 16 by 24 inch (0.40 by 0.61 m.) door opens upon an irregular bench. Although its lintel had disappeared both jambs are grooved and the remains of willow loops show at either side.

In repairing this room we placed two adjacent beams parallel with and next to the northwest, or cliff, wall; a single beam was embedded in the restored masonry near the southeast wall and three steps at the east end were enlarged.

Room 33, on a ledge west of room 32, is yet another storeroom. Its floor is of native rock, higher in the middle than at either end. One peg protrudes from the upper southeast wall; no doubt others were formerly present.

A door, 15 by 20 inches (0.38 by 0.50 m.), opens through the middle southwest wall. Its south jamb is a stone slab on end; the other, an upright stick embedded in adobe. Outside, both are rounded off with mud and grooved for a door slab; holes at either side were once occupied by willow loops. The lintel is a stone slab supported by six strips of wood; the sill consists of two slabs set at different levels. The uppermost of these, 11 inches (0.27 m.) wide by 13 inches (0.33 m.) long, lies 6-1/2 inches (0.16 m.) above the lower; the latter, 8 inches (0.20 m.) wide by 15 inches (0.38 m.) long, lies 7 inches above the general floor level. Outside this door the cliff ledge forms a flat space approximately 2 by 4 feet (0.60 by 1.2 m.).

We placed two salvaged beams lengthwise of the room, seating their ends in the restored upper northeast and southwest walls.

Court 34 lies between rooms 26 and 31, in front of and below room 33. The face of the narrow ledge on which rooms 32-33 stand forms its northwest side; at the west corner seven pecked steps lead up and over this ledge to the roof of room 26. In the north corner three similar steps and a fourth in the outer wall of room 31 gave access to the roof of the latter. The court floor is mostly artificial and continues at the same level with the roof of room 100.

We substituted a notched cedar for the west steps and placed a similar, though larger, ladder in the north corner.

Room 35, a dwelling, is situated between room 37 and court 45. We observed no evidence of the second story shown in Douglass's ground plan as reproduced by Fewkes (1911).

The northwest side is sheer cliff, surmounted by rooms 29 and 30; the southeast and southwest walls are masonry; the northeast, of wattle. While the latter was surfaced all over, only the lower half of the other three was plastered. The north part of the floor is native rock, worked down nearly 2 feet (0.60 m.) in the west corner and along the southwest side. A slab-lined but screenless fireplace, 18 by 20 inches (0.45 by 0.50 m.), occupies the middle floor. Three feet two inches above the floor a 5 by 9 inch (0.12 by 0.22 m.) opening extends diagonally through the middle southeast wall. In the west corner, 11 inches (0.27 m.) above the floor, is an 8 by 4 inch (0.20 by 0.10 m.) shelf,

A small beam lies against the northwest cliff; two larger northeast-southwest beams cross the middle room. At the northeast these three timbers rested upon individual posts set just within the wattled wall. Although five cross poles (1 single, 2 pairs) appear in the southeast wall, at least four others were formerly present.

In 1917 we repaired but did not replaster the wattled wall; we reconstructed the rounded adobe sill of its door and rebuilt the central fireplace.

Court 37. The northwest wall is sheer cliff, 8 feet (2.4 m.) high, surmounted by the retaining wall fronting court 28. No beam rests appear on this side. Two beam ends from room 35 protrude through the northeast wall, which apparently never extended more than 5-1/2 feet (1.7 m.) above the court floor. Of the southeast side only traces now remain in the east corner. At the northwest is the cleaved end of a rock ledge on whose sloping surface stands the fragmentary northeast wall of room 39, ending abruptly with the southeast face of the ledge. Court 37 lies several feet below passage 40, but connects with it by a single pecked step. Other similar steps may long since have disappeared with disintegration of the soft, friable sandstone.

Bone awls were pointed on the northwest cliff face; stone axes were sharpened on the sandstone floor. In the west corner four steps were pecked in the gently sloping surface where steps seem quite unnecessary.

We observed no suggestion of the second story indicated on Douglass's ground plan.

Room 39 is a dwelling, between rooms 41 and 37. Its northwest wall is formed by the face of a cliff terrace, surmounted by room 25; a fragment of northeast masonry remains at the north. The southeast wall, of wattle construction, was plastered all over; successive plaster layers show on the lower half of the northwest and southwest walls. While the north half of the floor is of native rock, the remainder consists of a débris fill surfaced with adobe. A protruding rock mass on the southwest side had been worked down 5 inches to the general floor level. Through the middle southeast wall a door gave access to passage 40, which in turn connects with court 37.

A narrow ledge lies at the base of the northwest, or cliff, side; on this ledge, in the north corner and abutting the northeast wall, stands a masonry column, 12 inches square, that probably once reached to the ceiling. The lower half of this column is plastered; its southeast face is smoked. On the same rock ledge and abutting the southwest side of the column is a masonry shelf, 8 inches (0.20 m.) wide by 2 feet 3 inches (0.68 m.) long. An adobe-rimmed fireplace, the north side of which lies 3 feet 5 inches (1.04 m.) from the middle southeast wall, had been partially destroyed with caving of the east quarter of the floor. Both firescreen and door-sill were lost at the same time.

Pecked into the rock floor at an average distance of 17 inches (0.43 m.) from the cliff are four loom anchor holes averaging 2-1/2 inches (0.06 m.) in diameter. The westernmost of these lies 2 feet 8 inches (0.81 m.) from the southwest wall; the second, 16 inches (0.40) from the first; the third, 16 inches from the second and 16 inches from the northwest wall; the fourth, unfinished, lies 8-1/2 inches (0.21 m.) from the third and 19 inches (0.48 m.) from the cliff. Another incomplete hole is noted on the north side, between the third and fourth. Hole No. 1 contained a 3/8-inch stick embedded 1 inch below the room floor; hole No. 2 held two quarter-inch sticks, placed one upon the other, with the uppermost 1-1/4 inches (0.031 m.) below the floor; hole No. 3 contained a single stick. Each anchor rested in an undercut at one side of a vertical groove (fig. 3); each had been introduced through the groove into the undercut, after which both groove and socket were packed with adobe mud. At some later time all five holes were filled with mud; their presence was disclosed by the difference in color between this adobe filling and the sandstone. From the uniformly discolored floor here we infer the loom anchors were in disuse long before abandonment of the room.

FIGURE 3.—CROSS SECTION AND PLAN OF LOOM ANCHOR

Room 40, a passageway about 21 feet (6.4 m.) long, furnished access from court 37 to rooms 39 and 41 and the roof of room 22. At its west end the alley is 18 inches (0.45 m.). wide; between rooms 39 and 41 it attains its maximum width of 35 inches (0.88 m.). The southeast wall probably did not extend more than 3 feet (0.91 m.) above the floor level and may have been stepped down at the east end. An earlier débris-covered floor lies 6 inches (0.15 m.) below that last utilized.

The masonry separating rooms 39 and 41 ends flush with their wattled southeast walls, instead of continuing, as shown on Douglass's plan.

Room 41, between rooms 17 and 39, opens into passage 40. As so frequently happens in Betatakin, its northwest wall is formed by the face of a rock ledge, in this instance surmounted by the southeast wall of court 24. A fracture in the cliff, 4 inches (0.10 m.) above the room floor, left a flat surface 12 inches wide by 3 feet (0.91 m.) long.

The northeast and southwest walls of the room are of masonry; the southeast, of wattle. The roof had been supported by at least, two large northwest-southeast beams whose south ends rested on posts incorporated within the wattle construction. Charred ends of nine cross poles protrude from the southwest wall. The floor, almost wholly of solid rock, had been pecked away more than 2 feet in the west corner. Such an excavation, made only with hammerstones, furnishes ample proof of the patience and industry of prehistoric Pueblo workmen.

A doorway, with stone sill slab lying 3 inches above the floor and wholly outside the door, pierced the middle southeast wall. Directly in front of and 15 inches (0.38 m.) within the door stands the remnant of a masonry fire screen, 30 inches (0.76 m.) wide by 7 inches (0.17 m.) thick; an upright stone slab connects its west end and the southeast wall. It is not improbable that this slab, like others present in houses with wattled south walls, was intended both as a secondary screen for the fireplace and as a check for sand carried by the winds which usually sweep through the cave from east to west. In 1917 we had ample opportunity to observe the force of these winds and the amount of sand they transported. The fire-pit, pecked from the solid rock floor, measures 2 feet 4 inches (0.71 m.) long by 16 inches (0.40 m.) by 5 inches (0.12 m.) deep.

When we repaired the upper northwest wall no effort was made to indicate positions for the principal ceiling timbers.

Room 42. The long room numbered 42 on Douglass's plan is not now traceable, although the ledge which marks its site, especially at the west end next room 23, is wide enough and not too sloping to have supported a room. Debris covering the floor of room 23 extended downward and across the sloping surface of this ledge. No portion of the southeast wall shown by Douglass is evident.

Room 43, a dwelling, stands between rooms 48 and 49, above court 45. Five medium-sized beams were covered directly with a thick layer of Johnson grass to form its ceiling; there were no cross poles. In the west corner native rock was worked out to a depth of 16 inches (0.40 m.); the remaining floor is a rubbish fill, surfaced with adobe. Part of the southeast wall had fallen, releasing some of the débris. Two feet four inches of the northeast wall consists of room 51 stone work; the remainder, of wattle (now mostly destroyed), apparently replaced a masonry wall 7 inches (0.17 m.) thicker. The southwest side, also of wattle, was built in after completion of the roof which covers both rooms 43 and 48. All four walls are heavily smoke stained.

A door through the middle of the northeast wattled wall has four thin strips of split cedar to support its stone lintel; as usual, a slab forms its sill. Another door, 16 by 20 inches (0.40 by 0.50 m.), opens through the middle southwest side; its sill is a slab on edge, 11 inches high. A fireplace, now broken but probably slab-lined, lies 15 inches (0.38 m.) inside the northeast door. We observed no trace of a screen.

In 1917 we propped a broken beam with a post, replaced two other beams, patched the roof edges with willows and cedar bark, repaired the broken southeast wall, and filled the floor break above noted.

Room 44, above and northwest of room 51, is a storeroom. On account of the sloping cliff at the northwest the floor area is reduced to half that of the ceiling. Two large slabs of rock had slipped down the cliff (pl. 13 B), breaking the northeast wall and pushing it outward 7 inches (0.17 m.) from its original position. The southwest side was built over a ledge which had been worked down within the room. Smoke stains appear on all four walls; more prominently in the south corner. Two narrow, groovelike notches were pecked in the cliff near the west corner, perhaps as beam rests. Only two cross-pole fragments, one at each end of the room, remained in place.

A former northeast door had been blocked; its jutting lintel slab and the recessed masonry above, form a shelf 6 inches (0.15 m.) wide by 2 feet 8 inches (0.81 m.) long. A later door, 15 by 25 inches (0.38 by 0.64 m.), opened through the southeast wall onto the roof of room 51; two stones composed its sill. Although the outer jambs of this second door are slightly grooved, fasteners were not provided.

Restoring the southeast door in 1917, we seated a single lintel slab, without supporting sticks. To suggest the original beam level, a pole was placed parallel with the cliff.

Court 45 lies in the left front of the village with room 35 ad joining on the southwest; rooms 46 and 50, on the northeast. The northwest side is sandstone cliff, sloping up from the floor to rooms 29, 43, and 48. The southwest wall, of wattle, is divided by a door whose slab sill lies 4 inches above the court floor, At each side of this door stands a post, joined to the wall by 6 inches of masonry. In the upper middle of the southeast inclosing wall was an opening (reproduced in 1917) slightly larger than the usual "windows" or ventilators.

Most of the court floor is artificial; fires had burned in various places. Several pecked steps lead up the sloping northwest side to a small flattish area in the corner formed by the outer walls of rooms 29 and 48. Two loom anchor holes, pecked in the north corner of the main court floor, each measure 3 inches (0.076 m.) in diameter by 2-1/2 inches (0.06 m.) deep; a third hole, unfinished, lies slightly to the west. A narrow bench, 8 inches (0.20 m.) wide by 14 inches (0.35 m.) high, extends nearly the entire length of the northeast wall and serves as a sill for the door to room 46. This door measures 18 inches (0.45 m.) wide by 25 inches (0.64 m.) high; staples appear at each side.

Where it abuts the nearly vertical cliff, the northeast wall was built into a dug groove about 3 inches deep. (Pl. 14, B.) We repaired the broken masonry here, repaired the wattled southwest wall, and did miscellaneous patching elsewhere.

Room 46 stands east of court 45 and southeast of room 50. Its northwest and southeast walls follow the irregular cliff ledges on which they were built. For example, the southeast wall extends 4 feet 9 inches (1.4 m.) from the south corner to a northwest-southeast jog, 2 feet 3 inches (0.68 m.) long; thence, 3 feet 7 inches (1.1 m.) to the east corner. The northwest wall, standing on parts of two narrow sloping ledges, has a similar offset 3 feet 10 inches (1.16 m.) from the west corner and 5 feet (1.5 m.) from the north corner. While the floor consists mostly of native rock, depressions along the southeast side were filled and surfaced with adobe.

A door opens through the southwest wall into court 45; 24 inches (0.61 m.) below its sill a step, 6 inches (0.15 m.) wide by 17 inches (0.43 m.) long, was pecked into a protruding mass of native rock.

The crude masonry of room 46 required much repair in 1917. The west end of its northwest wall was rebuilt to furnish support for the south corner of room 50, but no attempt was made to show beam seatings.

Room 47, adjoining room 46 on the east, occupies a sloping rock surface in the lower middle front of the cave. Only the southwest and part of the southeast walls remain. The pecked groove on which the northwest masonry rested continues to the probable north corner where we restored a small section in 1917. The east half of the southeast wall, built on a lower ledge, necessarily includes masonry much higher than that in its west half. Within and against this wall a very considerable fill of débris provided a floor the former level of which is not now evident.

Room 48 is a small chamber that originally formed part of room 43, which it adjoins on the southwest. Its two roof beams support numerous small cedar sticks and a thick layer of Johnson grass. Owing to the sloping cliff on the northwest side, the floor area is reduced to less than half that of its ceiling. For example, the floor measures 2 feet 8 inches (0.81 m.) by 16 inches (0.40 m.), while the ceiling measures 2 feet 8 inches by 4 feet (0.81 by 1.2 m.). The wattled northeast wall is not plastered as is its opposite face in room 43. A triangular space between the cliff and the northeast door is occupied by a single large stone, forming a low shelf 5 by 14 inches (0.13 by 0.35 m.). The inner walls are heavily smoked.

The lower southeast masonry, having fallen, was replaced in 1917 (pl. 14, A) and the floor brought up to its former level.

Room 49 stands on a ledge northeast of room 43, between rooms 50 and 51. The northeast wall has wholly disappeared; the southeast, a fragment only of which remains, formerly concealed a large, rounded rock mass that lay just below the floor level of the east quarter. Absence of seatings for ceiling timbers suggests that room 49 might have served as a court, its floor at the roof level of room 50 or possibly separated from it by low stonework.

Part of a slab-lined fireplace will be noted against the southeast wall near the south corner. Through the middle southwest wall a door connects with room 43; a second door, whose grooved west jamb only remained, opened into room 51 through the west half of the northwest wall. We partially restored this latter wall and its broken door in 1917.

Room 50 lies north of court 45, between rooms 46 and 49. Its northeast wall is wholly missing, while that on the northwest is represented merely by sections of masonry at the west corner and under the bowdlerlike rock noted below the east corner of room 49. Of the southeast side, a fragment remains in the south corner; the adjacent southwest wall is complete and shows seven holes left by cross poles. The lower half of this latter wall is plastered and smoke stained. Native rock at the base of the northwest side had been worked down to the general floor level; elsewhere, the floor was of abode mud over a débris fill. This rubbish covered three earlier steps, pecked in the sloping sandstone.

In 1917 we rebuilt an alcove in the west corner; reconstructed the northwest wall and thus again concealed all except the protruding east end of the rock mass under room 49. A beam was also replaced on top the northwest wall.

Room 51 may be found in front of and below rooms 44 and 56, northwest of room 49. North of the angle in its southwest wall, rude masonry holds back the débris forming a level for the unnumbered court marked "Bench" on Douglass's ground plan. Four shallow steps, pecked into the sloping rock floor of this west corner, suggest the rude masonry above them was added when room 51 was constructed in this space, originally used as a passageway.

The lower west half of the northwest side is sandstone cliff, with room 44 above and at the roof level; the remainder is the masonry foundation of the wattled southeast wall of room 56. Stones loosely piled in débris marked the northeast wall. In the middle southeast wall a door connected with room 49; east of this opening the masonry had fallen.

We supplied a new sill for the southeast door and partly reconstructed both the southeast and northeast walls. In the latter, we left an irregular opening to provide ready access to the ladder placed in the west corner. The north leg of this ladder stands in a shallow hole we pecked in the sloping cave floor.

Room 52. The only traces of this structure were a few superposed stones of its southwest wall and several others under the north edge of the bowdlerlike mass upon which the southeast wall had rested. Room 57, at the northwest, may have opened onto the roof of this building, although we have no reason to say it did. Room 56 protrudes slightly into the west corner from which a low ledge, worked down several inches, extends the entire length of the room. Two cup-shaped depressions had been pecked into the cliff floor as rests for the northeast masonry.

We reconstructed portions of the southeast and northeast walls in 1917 and left an opening in the northeast side to facilitate access to the rooms beyond. (Pl. 15, B.)

The series of pecked steps under the southwest wall of room 52 is a continuation of the trail which passes up the slope westwardly from room 55. These steps and the fact that all partitions had abutted the major northeast-southwest walls seem to indicate that rooms 51, 52, 53, etc., were later additions to the village, built upon the narrow ledge previously used as a passage between the east and west house groups. It is certain that the ledge crossing the cave at this level was used as a trail for its course is marked by pecked steps and otherwise. (Pls. 6, B; 15, B.) Water seeps out just above the old path, providing lodgment for wind-driven sand and moisture for growing plants. From this area we removed a thick mat of columbine and other vegetation. Also we cleared the old trail and deepened some of its more weathered steps.

Room 53. In front of and below room 60-61, the pecked out cliff between room 52 and the projecting southeast wall of room 64-65 marks the former floor level of room 53 and its northeast-southwest length as 8 feet 7 inches (2.61 m.). A shallow groove, pecked in the sloping sandstone, served as a rest for the southeast wall, in continuation of that in room 52.

During our repairs six steps were cut in the cave floor of room 53 to facilitate access to the western portion of the ruin.

Room 54. Not indicated on Douglass's plan.

Room 55, obviously a ceremonial chamber, stands in the lower middle front of the cave. Like other Betatakin kivas, its special function is evidenced by certain furnishings never present in secular structures.

Although the south wall is now missing its former position is marked by a pecked rest on the cliff face below the floor level. The east and west walls stand more than 7 feet (2.13 m.) in height without trace of beam seatings; unlike any other in the ruin, they measure over 2 feet thick and include a core of sandstone spalls and adobe, faced on each side with masonry.

On the north the sharply inclined terrace face was squared up in front with masonry to form a shelf 18 inches (0.45 m.) wide at the east and 10 inches (0.25 m.) wide at the west, extending the entire length of the room, 3 feet 11 inches (1.19 m.) above the floor. In the middle face of this shelf, 22 inches (0.56 m.) above the floor, is a pecked hole, 7-1/2 inches (0.19 m.) in diameter by 3 inches (0.07 m.) deep; above the shelf, where the room length is a few inches less than that below, masonry continued to the roof level. The walls were plastered and smoke stained.

Filling was necessary adjacent to the south wall, but elsewhere the floor is the flat surface of the lowermost terrace. In this are four loom anchor holes, averaging 3 inches (0.07 m.) in diameter by 2-1/4 inches (0.056 m.) deep. They were pecked in the sandstone about 12 inches (0.30 m.) from each other and about 17 inches (0.43 m.) from the base of the north shelf; each hole lacks its former anchor stick.

Built in the lower northwest corner is a triangular bench, 2 feet 3 inches (0.68 m.) high by 2 feet 4 inches (0.71 m.) on the north and 8 feet 7 inches (1.09 m.) on the west. In the face of this bench, at the floor level, is a recess measuring 6 inches (0.15 m.) wide by 18 inches (0.45 m.) high by 11 inches (0.27 m.) deep; its unplastered interior is but lightly smoked.

At the beginning of our 1917 excavations débris sloped downward from near the present top of the north wall, covering the floor and extending beyond its south margin. In this rubbish were quantities of cedar bark, sticks, and adobe flooring from dwellings higher in the cave. We repaired the northwest corner bench and the broken edge of the north shelf; recapped the walls and replaced some of the missing plaster. (Pl. 23, B.)

Room 56 lies north of room 51, between rooms 44 and 57. Its northeast and southwest sides are of masonry, but its southeast wall, above a masonry base which supports a débris fill and the floor level, appears to have been of wattle construction. There is no trace of masonry abutting the outer east corner of room 44. Of the northeast wall only a small section remained; in this, next the cliff, was the charred end of a ceiling cross pole and, in the lower front portion, the north jamb of a door. An offset in this same wall forms a triangular shelf, 12 inches (0.30 m.) wide by 18 inches (0.45 m.) long, against the cliff and just below the ceiling level. The slab sill of the blocked door to room 44 protrudes 4 inches (0.10 m.), forming a shelf 2 feet (0.60 m.) above the floor.

From the built-in floor the cave surface slopes sharply upward to the rear cliff, which may have served as the northwest wall of a second-story room.

In the débris-filled front part of the room were three mealing bins. For two of these the inclosing slabs were still in position when our excavations began; those of the third bin had fallen. The two bins measured 28 inches (0.71 m.) long by 17 inches (0.43 m.) and 14 inches (0.35 m.) wide, respectively. Although the original metates were missing, their angle of inclination was evidenced by the adobe packing inside the bins.

In repairing this chamber we set three posts just inside the masonry base of the southeast wall to suggest the wattling which probably once stood there. Also we partly rebuilt the adjoining northeast wall while retaining its original irregularities. But it should be understood that the present dimensions of the restored door are approximations only. We reset the slab sides of the three mealing bins and substituted for those formerly employed, metates salvaged from other sections of the ruin. (Pl. 18, B.)

Room 57 lies between rooms 56 and 60; in front of and below room 58; above and behind room 52. From the rather precipitous cave floor a section 20 inches (0.51 m.) high by 12 inches (0.30 m.) deep and extending the entire width of the room had been battered away with hammerstones. This excavation doubtless marks the former floor level, the front portion having been filled in with rubbish and surfaced with adobe as in so many other rooms, similarly situated. Two charred beam ends protrude from the upper southwest wall, next the cliff. A long narrow groove, pecked in the sandstone, formerly supported the now missing southeast wall. This latter, as in the case of room 56, we assume to have been of wattle above a masonry base that rose only to the floor level. It should be admitted, however, that we have no real justification for the assumption.

We deepened the southeast wall groove and, upon it, reconstructed the masonry foundation to approximately its original height; the adjacent sides were also represented by limited stonework built upon deep steps we cut to replace the shallow, pecked seatings provided by the aboriginal masons. (Pl. 17, B.)

Room 58 is merely a platform fronting room 59. When examined in 1917 its retaining (southeast) wall extended only one course above the landing; there was no certain evidence of a former southwest wall, although several large stones had been piled in there, below the floor level.

Directly in front of the door to room 59 is a groove where a stone ax was sharpened. On the slanting cliff west of the platform is a single pecked step which, with similar steps on the large rocks that had slumped into room 44, suggests that room 59 was entered from above the roof levels of rooms 44 and 56.

We added a second course of stones on the southeast side and, perhaps erroneously, replaced the loose subfloor blocks at the southwest with masonry extending two layers above the platform level.

Room 59, a probable storeroom, situated high in the middle portion of the cave, above and northwest of room 60-61. The foundation of its southeast wall was supported by a log resting upon the precipitous cave floor with its ends embedded in the northeast and southwest walls of room 60-61. A post about 6 feet (1.82 m.) high and reaching from floor to arching cliff above was incorporated in the masonry at the north corner. Rough stonework fills the space between this post and the adjacent lower cliff, thus creating an angle in the northeast side as shown in Plate 3. Smoke stains are discernible on the lower walls. Although the floor is mostly of solid sandstone, a shallow fill was made in front. Pecked in the rock floor at the west end is a seemingly unnecessary step.

The roof was supported by two parallel beams next the cliff and a third, now missing, toward the front. Through the middle southwest wall, with two large cedar sticks as lintel, is a 17 by 30-inch (0.43 by 0.76 m.) door. Its outer jambs are not grooved; hence the space within may not have been utilized solely for storage purposes.

During progress of our work the southeast wall collapsed, for reasons explained below. It was subsequently rebuilt; the space behind its supporting log was filled with spalls and covered by slabs extending out over the timber thus to provide a sturdier foundation for our new stonework. A salvaged pole replaced the missing ceiling beam.

Room 60-61 lies northeast of room 57, in front of and below room 59. The front wall continues from that in room 57 and, like it, rested on a pecked groove several feet below the actual floor level. We observed nothing to indicate that this wall included other than masonry. The two sides were erected above series of shallow, pecked steps. Some notion of the degree of inclination here may be gained from the fact that the floor length of the lower room is less than half its ceiling length. It is possible, therefore, that the lower chamber was utilized primarily for storage while the upper provided living quarters.

Part of the northeast wall and a small section of that adjoining on the southeast were in position when our work began. But these fragments, cracked by previous settling (pl. 16, A), collapsed utterly on April 18 after their adobe mortar had softened under the clinging snow of successive storms. Twelve days later, while we were reconstructing these fallen walls upon more deeply cut steps (pls. 17 and 18), the upper northwest wall gave way owing to vibrations set up by the reverberating echoes of an unusually severe thunderstorm. This wall, which serves as a foundation for room 59, was also reconstructed and upon its original horizontal supporting log. Our experiences here afford evidence of two distinct factors contributing to the destruction of cave dwellings.

A third factor might be noted in passing, namely, the abrasive property of wind-blown sand. As blown sand contributed to the creation of Betatakin cave, so has it played a part in destruction of those prehistoric dwellings subsequently built within that huge Cavern. When the spring sandstorms were at their worst our work was repeatedly interrupted. The wind that leveled camp on April 9 and hung our spare clothes on far, high branches also drove us to seek shelter in the old rooms. Scouring sand cut pellets from the cave roof and showered them down with surprising velocity. Came, also, occasional larger stones. Storms of comparable temper persisted well into May. Blown sand has left its mark on both cave and house walls; masonry has been undercut and thrown down. Over a period of years the amount of destruction so caused might prove not inconsiderable.

Room 62. As shown on Douglass's ground plan, this room does not exist.

Room 63 is a small storeroom adjoining room 59, above and northwest of room 64-65. The cliff at the rear rises vertically to the gallery wall. As with rooms 58 and 59 a masonry foundation on the southeast side rose to the floor level; of the northeast wall, only a few stones remained in position. In 1917 we added several courses to these two fragmentary walls.

Northeast of room 63 the ledge upon which it stands bears several grooves where stone axes were whetted; farther along, this ledge narrows and disappears. But the line of cleavage continues as a seepage zone from which, in 1917, rippled a fluffy green band of columbine. A second similar band grew down slope, on a parallel seepage extending eastwardly from room 53. Over and below this lower seep, wind-blown earth and sand had gathered to a depth of 3 feet (0.91 m.); in this grew several oaks and box elders, from 3 to 5 inches in diameter. The decaying trunks of others were disclosed during removal of the accumulation.

Room 64-65. Of this structure only one wall was standing in 1917. (Pl. 16 A.) The lower story, quite V-shaped, might have been utilized for storage; with equal plausibility, it could have been packed with rubbish. Pecked grooves and steps as former wall rests and a difference in floor coloration marked the house site. Where once covered by masonry, the slanting sandstone remained unsoiled; elsewhere it was darkened by ash and débris of occupation. As in other structures similarly situated, the floor level of the lower room is now represented by a battered and partially excavated cliff section from which an adobe-surfaced, rubbish fill formerly extended to the front wall. This latter had been erected above a pecked groove 2 feet 8 inches (0.81 m.) lower on the slope than the southeast wail of room 60-61, which adjoins on the west.

Before construction of the dwelling several dissociated steps had been pecked on this lower slope. These were never actually used as wall seatings, although they may have been intended as such. There is also the possibility that they served as foot rests during building operations or as earlier trails across the precipitous cave floor.

Upon specially prepared cuts we reconstructed so much of the front and side walls as would indicate the original position and size of this room. (Pls. 17 and 18.)

Room 66 is identified as a second-story dwelling in the lower, middle portion of the cave. We find here a certain discrepancy in the Douglass ground plan. Therein, room 66 is shown as a 1-story house separated from a 2-story structure (R. 68-69) by narrow room 67. As a matter of fact, room 68 is barely traceable; room 66 is the only one of the three chambers having two doors in its west wall. One of these occurs in the first story; the other, in the second. If one takes these two openings as the distinguishing feature of room 66, its correct position relative to its neighbors will be that shown by our Plate 3.

The upper story of this structure had burned during occupancy for the charred beams and willows of its original roof still show in the wall masonry 3 feet 10 inches (1.2 m.) above its floor. Subsequently, a new ceiling was constructed approximately 6 inches (0.15 m.) above the remains of the earlier. Only one door occurs and that in the west wall, near the southwest corner; it measures 18 inches (0.45 m.) wide by 31 inches (0.78 m.) high. Two feet inside this door and 2 feet 6 inches (0.76 m.) from the south wall is a slab-lined fireplace, 13 by 24 inches (0.33 by 0.61 m.) by 9 inches (0.22 m.) deep; it was protected by a wattled firescreen whose charred willows are still visible. Between this screen and the door, 4 inches (0.10 m.) below the sill, is a platform measuring 20 inches (0.51 m.) wide by 8 inches (0.20 m.) high; on the south side of this, joining screen and west wall, one notes the fragment of an upright slab which once served as a secondary deflector. In the middle south wall, 3 feet 5 inches (1.04 m.) above the floor, is a 6-inch square smoke vent.

During our 1917 repairs we replaced one of the missing upper beams and added split sticks and a stone lintel to the broken west door. The adobe floor was patched where needed; the walls, recapped.

Beneath room 66 is an unexcavated, débris-filled chamber whose floor area, on account of the sloping cliff, is considerably less than that of the upper story. In the north half of the lower west wall is a partially blocked door. (Pl. 9 B.) While some of the closing stones may have been removed within recent years, the writer seems to recall that this particular opening was in much the same condition when Betatakin was discovered by Professor Cummings in 1909. We presumed to close it completely in 1917.

Room 67 is a narrow building, east of room 66. Only a decayed fragment of its south wall remained and this we replaced in 1917. The adjoining sides had entirely disappeared; no grooves were present to mark their former positions. The north wall, about 5 feet (1.5 m.) long, was doubtless a continuation of that in room 66.

Room 68. Of this structure no trace survived except the battered cliff edge, east of room 67. Here the south wall had stood; the east and west sides doubtless continued from those in room 122, which lies next on the north.

Room 70 formerly rested precariously on the cliff edge, east of rooms 68 and 122. Pecked wall seatings give the approximate dimensions recorded in the table on page 73. The stonework had entirely disappeared with the exception of a small north section supporting the artificial platform adjoining room 122, at the foot of the steps leading to room 73 and beyond. Upon the seatings they formerly covered we partially reconstructed this retaining wall and those adjacent.

Room 71. Although no masonry remained at the time of our work, the former presence of this room, on the lower cliff terrace east of room 70, is well evidenced by the series of steps pecked for support of its north wall. These steps may be noted on the upper edge of a slight break in the sloping cave floor; they extend thence, eastwardly toward the cliff edge. Below the break mentioned the native rock had been battered away to mark the floor level.

Room 72. Not shown on Douglass's plan.

Room 73 is a small storeroom built midway of the old step series leading from room 122 to the upper east end of the cave. (Most of these steps were enlarged and deepened in 1917.) The east, west, and south walls are of masonry; the north side is an acclivity, about 2 feet high, on which a number of horizontally pecked steps may represent footings for former masonry. But these steps continue to the west of room 73, along the crest of the same slanting ledge.

Abutting the outer southwest corner of the storeroom a short retaining wall supports a platform, partially made with débris, at a turn in the old trail. Fires had burned on this platform. A crack in the north, or cliff, side had been enlarged by pecking and continued, in snakelike fashion, downward toward the floor.

Construction of room 73 caused annoyance to some of the villagers, since it formed an obstacle on their much-traveled trail between the east and west house groups. Those most deeply irritated thereafter used another step series at the very top of the sloping sandstone; the others soon made a new path close on the north side of the granary.

In clearing room 73 we found numerous corn cobs and a small circular basket. (Pl. 42, 1.) Also, a stone ax was discovered in a recess created by removal of a stone from the lower south wall.

Room 74. The former presence of this room is evidenced by a fragment of masonry at the northeast corner; by the battered cliff at the floor level and by pecked grooves on which the north and south walls were erected.

Room 75 lies east of room 74, in the lower northeast portion of the cave. Most of its west wall still stands and the lower half of this bears traces of plaster. The north side, now missing, stood upon a well-marked groove; the south wall, also missing, had been built on the very edge of the cliff. Of the east wall a small section found in position rested upon several inches of loose household rubbish.

The north half of the floor had been worked out of native rock, while the remainder was filled in and adobe surfaced. Pecked in the sandstone near the east wall is a shallow hole, 2 inches in diameter; just below it is a step that may have been cut before the house was built. A stone in the outer east wall bears an incised, swastika-like design. (Fig. 4.)

FIGURE 4.—DESIGN INCISED ON OUTER EAST WALL OF ROOM 75

In 1917 we partially restored the north and east walls. (Pl. 22, B.)

Room 76 was built on the sharply inclined cave floor between rooms 75 and 77. A pecked groove marks the former position of its south wall; its east and west sides, now represented by a few courses of masonry, had been erected upon loose débris. We saved the west fragment by building in new foundations as the rubbish was removed; the north end of this same wall and the east side were partially reconstructed.

Room 77 is a small chamber between room 76 and the retaining wall in the upper northeast section of the cave. Its missing south wall formerly stood upon a narrow, pecked groove; a worked-out area marked its floor level. The east and west walls, having disintegrated beyond hope of repair, were partially restored in 1917, as was also the adjacent retaining wall. (Pl. 20, B.)

Room 78, also in the upper east end of the cave, was designed for storage purposes. (Pl. 19, A, B.) Sheer cliff forms its northeast wall; masonry, the others. Externally the stonework of this room is among the best in the ruin; inside it is crude and irregular. The level floor is wholly of native rock. Four beams supported the roof; two of these, of aspen and still present, lie side by side next the cliff.

Thirteen inches above the floor in the middle northwest wall is a 15 by 24-inch (0.38 by 0.61 m.) door, whose heavy stone lintel is supported by five strips of split cedar. The outer jams and lintel of this entrance are deeply grooved for the door slab so typical of Betatakin storerooms. In the southeast wall, 22 inches (0.56 m.) above the floor, one notes a former opening that measured about 16 by 20 inches (0.40 by 0.51 m.). Its irregular sides and the absence both of lintel and sill slabs suggest that this prospective door was broken through after completion of the room and then closed again before it was actually finished.

Rude masonry, 2-1/2 feet (0.76 m.) high, caps the southeast wall; through this a single beam end protrudes from room 79.

Room 79, a dwelling, stands southeast of and next to room 78. Its ceiling of cross poles, willows, and Johnson grass rested upon two northwest-southeast beams. But the roof had been crushed by large sandstone slabs, fallen from the vaulted cave. (Pl. 21, A.) The northeast wall is cliff; the southwest, of wattle—willows and Johnson grass bound to posts and surfaced inside and out with adobe. No plaster appears on the other three sides; all four are heavily smoked. The northeast half of the rock floor has been pecked out to a depth of 12 or 14 inches (0.30-0.35 m.); a 3-inch fill covers the remainder.

A door opened through the middle southwest wall with its adobe sill at the outside terrace level, 4 feet 3 inches (1.3 m.) from the retaining wall. Just inside this door is a platform, 4 inches high; at its southeast side a slab on edge, 10 inches (0.25 m.) high by 13 inches (0.33 m.) wide, abutted the wattled wall and a now missing masonry firescreen. Against the northeast base of this former screen one notes a slab-lined fireplace, 23 by 11 by 7 inches deep (0.58 by 0.27 by 0.17 m.). On a narrow ledge in the east corner, 3 feet 6 inches (1.06 m.) above the floor, lies a boat-shaped shelf measuring 5 by 12 inches (0.127 by 0.305 m.), made of adobe and chinked with small sandstone chips. The stonework blocking the unfinished opening into room 78 was laid flush with the northwest wall face.

Room 80 will also be found at the upper east end of the cave, between rooms 79 and 81. Its ceiling height may not be determined, for the upper northwest and southeast walls were broken by great slabs falling from the cave roof. The northeast wall is formed by the cliff; the now missing southwest wall, of wattle, stood 4 feet 6 inches (1.4 m.) from the retaining wall. The adobe floor, spread over a shallow fill, lies 3 inches below the outside terrace level. In the west corner of the room is one stone of the former slab-lined fireplace.

Room 81, adjoining room 80 on the south, also was utilized as a dwelling. Like its neighbor, its northeast wall consists of the sandstone cliff; its southwest wall, now missing, was of wattle. The rock floor at the northeast had been slightly reduced by battering; on the opposite side two grooves evidence the whetting of stone axes. A door undoubtedly opened through the wattled southwest wall.

The retaining wall which forms a terrace fronting this and near by dwellings stands 2 feet 6 inches (0.76 m.) from the south corner and 4 feet 3 inches (1.29 m.) from the west corner of room 81. On the cave floor just south of this room are two grooves where axes were sharpened. Here the cliff drops sharply away, perhaps 25 feet (7.62 m.), to rooms 83-85; down this slope débris had been thrown from the upper dwellings.

Room 82 stands on the cliff edge in the lower east part of the cave. (Pl. 21, B.) Its east wall is of wattle; the others of masonry. All four walls were plastered their full height; all are heavily smoked. The ceiling beams lay north and south.

A door with the usual slab sill divided the wattled east wall. Against each inner jamb stands an upright slab, that at the north being 18 inches (0.45 m.) high by 18 inches wide; that at the south, 7 inches (0.17 m.) high by 22 inches (0.56 m.) wide. These appear to have abutted a slab fire screen, now missing. A fireplace, 18 inches (0.45 m.) square by 8 inches (0.20 m.) deep, lies 2 feet 6 inches (0.76 m.) from the east side. In the middle south wall, 3 feet (0.91 m.) above the floor, is an 8 by 11 inch (0.20 by 0.27 m.) opening whose outer dimensions are reduced to 4 by 10 inches (0.10 by 0.25 m.).

Our 1917 efforts included repair of the wattled east wall and replacement of the adjacent ceiling beam (pl. 22, B); miscellaneous patching and resurfacing of the three masonry walls; restoration of the fireplace and partial reconstruction of the north wall as a support for an extramural platform or walk, 2 feet (0.60 m.) wide and 3 feet (0.91 m.) above the room floor.

Court 83 lies between rooms 82 and 84, on the cliff edge in the lower east end of the cave. Its south wall was relatively low; opposite this the cliff slopes down to form part of the court floor, the remainder being filled with débris and surfaced with adobe a few inches above that of room 82. A fireplace, 19 by 23 inches (0.48 by 0.58 m.) by 8 inches (0.20 m.) deep, lies against the middle southeast wall. At each end of this wall a single step had been cut into the masonry. In the north corner a series of four pecked steps leads to the platform between room 82 and the cliff.

We repaired the wattled west wall by tying horizontal willows to those still standing between the posts; no plaster was added.

Room 84, a storeroom, adjoins court 83 on the southeast. All four walls are of masonry; that at the northeast stands on a narrow ledge, 3 feet 6 inches (1.06 m.) from the cliff, the space between being filled with débris. Two ceiling beams support nine cross poles with layers of willows and cedar bark. On the northeast side is a 16 by 20 inch (0.40 by 0.51 m.) hatchway, once covered by a door slab; in the masonry beneath this opening are three pecked steps. No wall pegs are present. The most unusual feature of this granary is the fact that it is floored with sandstone slabs, set in adobe mud.

We replaced the stone slabs that rimmed the ceiling hatchway.

Room 85, another storage chamber, adjoins room 84 on the southeast and stands on the very edge of the cliff, at the base of the precipitous slope below rooms 78-81. (Pls. 21, B; 25.) Two parallel ceiling beams near the outer wall and a third next the cliff support seven cross poles with superposed layers of willows and cedar bark. No wall pegs were noted. Like its neighbor on the northwest the room was entered through a 16 by 20 inch hatchway. Beneath this opening two steps were pecked in the cliff. The northeast half of the floor consists of native rock, sloping and uneven; the remainder is paved with slabs set in adobe.

On the roof between rooms 84 and 85 and next the cliff is a slab-lined fireplace measuring 13 by 27 inches (0.33 by 0.68 m.) by 8 inches deep; it is rimmed by slabs and surrounded by an adobe pavement.

We placed two posts in this chamber to support a broken cross pole and the northeast beam.

Room 86, under the cliff at the lower southeast end of the cave, has not been excavated. It is represented by several wall fragments, mostly concealed by rocky débris.

Room 87, also unexcavated, abuts the north corner of room 86. It is filled with large rocks; between these one notes the jamb of a north door. We observed no evidence of the curved west corner shown on Douglass's plan.

Room 88 is a small storeroom, adjoining room 87 at the base of the southeast cliff. Its fragmentary walls disclose no trace of beams; its entrance was probably a ceiling hatchway. A rock ledge on the east cliff about 3 feet (0.91 m.) above the floor forms a shelf 3 feet 6 inches by 4 feet 6 inches (1.06 by 1.4 m.).

On the cliff above one notes a 26-inch circle, painted in dull red except for the upper left quarter. Near this is an indistinct white spiral 3 inches (0.07 m.) in diameter. At the outer southeast corner a rock contains five deep grooves where axes were sharpened.

The space between rooms 88 and 89 has not been excavated. There are no visible indications of a west wall. Smoke stains on the cliff suggest a probable open fireplace.

Room 89 served as a dwelling. Vertical cliff forms its east wall; masonry, the others. Its floor is of native rock, pecked out to a depth of 6 inches (0.15 m.) in the southeast corner.

Part of one jamb places a former door in the middle south wall. Just within the room, 5 inches (0.13 m.) below the door sill, stone slabs form a step 2 inches (0.05 m.) high; close on the west is a slab-lined fireplace measuring 14 by 24 inches (0.35 by 0.61 m.) by 6 inches (0.15 m.) deep. In the west wall at the southwest corner a blocked door, 12 inches above the floor, measured 19 by 22 inches (0.48 by 0.56 m.); its protruding lintel slab formed a shelf 4 inches (0.10 m.) wide by 22 inches (0.56 m.) long. Below the sill and next the south wall is a ventilator, 8 inches (0.20 m.) square, apparently cut through after the door was blocked.

On the cliff above the former roof level is a white painted figure which Cummings (1915, p. 277) identifies as a "Slayer god"; near by, a typical Pueblo conception of a mountain sheep. (Pl. 29, A.)

Room 90, occupying a narrow ledge north of room 89, was utilized for storage purposes. Of its walls only a few courses remain. At the north end two beam rests had been pecked in the cliff 3 feet 6 inches (1.06 m.) above the ledge surface; no others were noted.

On the same ledge, but north of room 90, three grooves were worn by patient sharpening of stone axes. Two shallow steps, without apparent purpose, are noted on the cliff face.

Room 91. Only a small fragment of the north wall remains. The room is still unexcavated; undisturbed and much-weathered rocks lie at the southeast, close below the northwest corner of room 90.

It seems doubtful that this room could have been so large as Douglass represents or that the retaining wall he shows could have extended all the way to room 94.

Room 92 lies below and slightly south of room 85, on a narrow ledge which has settled outward 10 or 12 inches. The west wall, built on the terrace edge, has fallen; sheer cliff forms the north and east sides. A pecked beam rest may be noted in the north wall, 6 feet (1.82 m.) above the floor; on the adjacent cliff one observes the marks of rubbed implements and numerous daubs of variously colored mud, but no smoke stains. Near by, just above the south wall, is the white imprint of a small hand. The room floor is the unworked ledge surface. A door opened through the south wall at the southwest corner.

Room 93, a small dwelling, was built south of room 92 and on the same ledge. The north and south walls are of masonry; the west has disappeared. Smoke stains are discernible. The south half of the floor had been pecked down to a depth of 23 inches (0.58 m.); in the face of this cut, under the south wall masonry, are two vertical grooves about three-fourths of an inch wide. Marks left by sharpening of stone tools show plainly in the rock floor.

A north door gave access to room 92; near by a circular fireplace, 19 inches (0.48 m.) in diameter, was pecked into the ledge surface. From the near door jamb a masonry wall extended to the east cliff, forming a triangular bin in the northeast corner; against the inner face of this wall is a thin, upright slab.

Room 94. No walls are now visible, but seven pecked beam rests appear in the east side, below the ledge on which room 92 stands. There are no similar seatings on this same bench face below room 93, which fact, together with the presence of large rocks, suggests that unexcavated room 94 is smaller than Douglass has indicated.

Room 95 is a dwelling on the same ledge with, and south of, room 93. Its north and south walls are of masonry; the west has vanished. Traces of smoke adhere to all standing walls. The ledge surface had been pecked down to a depth of 22 inches (0.56 m.) at the northeast and 27 inches (0.68 m.) at the southeast, but the west side of the floor was filled in. Near the west end of the south wall is the east jamb of a former T-shaped door, the only one observed in Betatakin. Its upper portion, about 20 inches (0.51 m.) high, is set back 4 inches (0.10 m.) from the lower, 13 inches high. A protruding rock east of this door had been partially severed with stone saws.

Just within the door is a platform, 23 inches (0.58 m.) wide by 10 inches (0.25 m.) high; at its north side stands a masonry fire screen, 10 inches (0.25 m.) thick, 3 feet 5 inches (1.04 m.) long and now 21 inches (0.53 m.) high. Pecked in the rock floor at the north base of this screen is a fireplace measuring 13 by 18 inches (0.33 by 0.45 m.) by 7 inches (0.17 m.) deep. Three feet 5 inches from the north wall and 37 inches (0.94 m.) from the cliff a pecked hole, 3 inches (0.07 m.) in diameter by 2-1/4 inches (0.05 m.) deep, contains a single stick as a loom anchor.

A large rock lies south of room 95 and beyond it is space for two rooms. The ledge surface here has been leveled by battering with hammerstones, but no beam rests appear in the cliff and only one masonry wall fragment, at the southwest corner of the floor area, is evident. Daubs of colored mud had been thrown against the cliff; here, also, is a rudely painted white square with sides 2 inches wide and, high above, the rude representation, also in white, of a mountain sheep.

Room 96. Visible wall fragments on the talus below the two probable chambers on the ledge south of room 95 suggest a former irregular retaining wall extending to the outer southwest corner of room 94. This section remains unexcavated.

Room 96 is the last indicated on the Douglass ground plan. In 1917, however, while clearing Betatakin of débris preparatory to our work of repair, a number of additional structures or traces of them were disclosed in the cave proper. These we have numbered 100 and following.

Room 100, lying between rooms 27 and 29 and entered from court 28, could have answered only for storage. Its north wall is the face of a cliff terrace, with superimposed masonry; its floor is of native rock, seamed and uneven. Three northwest-southeast beams support eight aspen cross poles with layers of willows, Johnson grass, and sand. The roof lay in continuation with the floor level of court 34.

A section of the smoke-stained northwest wall of room 29 extends through into room 100; the upper portion of this protruding masonry had been raised somewhat to form an 8 by 20 inch (0.51 m.) shelf, 22 inches (0.56 m.) above the floor. Below this shelf is the embedded end of the lower log supporting the northwest side of room 29. Through the middle southeast wall a 15 by 23 inch (0.38 by 0.58 m.) door opens into court 28. Six small aspen branches support the stone lintel; its sill slab lies about 12 inches above the floor.

Room 101 will be found in the north middle portion of the cave, adjoining room 64-65. Only a portion of its northeast wall remained in 1917, and this was so disintegrated and so insecure we partially replaced it with new masonry, built upon deepened steps. The southeast wall had rested on two rather shallow steps, slightly north of the southeast corner of room 64-65. Other pecked steps, varying in size and depth, are present on the abruptly sloping cave floor in front of and below rooms 101 and 102. (Pl. 16, B.) While some of these doubtless served as wall rests, it is quite possible the others merely furnished temporary footing for the builders.

Because the space between it and the slanting cliff is so restricted, the southeast wall rose several feet to the floor level. This space may have been filled with débris or it may have served for storage purposes. Above it, however, there must certainly have been a living room, corresponding with that next on the west.

An old step series leads northwesterly up the sloping cliff, across the middle of room 101 and the upper end of room 64-65, to room 63.

Room 102. Of this room, northeast of and adjoining room 101, no walls remained in 1917. But their former positions were plainly evident from the pecked steps and grooves commonly used as rests for masonry. The southeast wall formed a continuation of that in room 101. High on the upper slope great slabs of sandstone have separated from the ledge. Since these slabs are grooved by the grinding of stone axes we endeavored to hold them in place by several steel drills, set on the lower side.

It is improbable that other rooms formerly stood on the sloping cave floor northeast of room 101, but disintegration here has removed all trace of them, if any.

Rooms 103-106. Four buildings northeast of rooms 49 and 50 and northwest of room 47 may be postulated. With but few interruptions a pecked groove continues from the west corner of room 55 to the north corner of room 50; this undoubtedly once served as a wall rest. A similar groove, extending from the north corner of room 46 to the east corner of room 52, is supposed to be that on which the northeast walls of room 49 and 50 once stood. Other wall seatings in this section are less certain.

Room 107 stood next on the east of room 53, but our excavations exposed only a single fragment of its northeast wall. This abutted the southeast side of room 101 near the outer east corner of room 64-65. Subsequently we restored this small section. (Pl. 15, B.) The former position of the southeast wall has been lost with disintegration of the cave floor.

Rooms 108-109. A pecked groove and the apparently worked appearance of the disintegrated sandstone below rooms 101 and 102 suggests the former presence of two rooms no wall of which remained in 1917. Masses of adobe-covered grass flooring had slid down slope from rooms 64, 101, and 102, and were removed from this area during our preparatory excavations.

Rooms 110-113. Fragments of flooring in place above the north wall of room 55 indicated the former presence of dwellings, but we observed no pecked holes or grooves that would serve to mark the position of masonry walls. There had been much seepage through out this portion of the cave with resultant disintegration of the sandstone; blown sand had lodged here and in it trees and shrubs had taken root. (Pl. 6, A.)

A narrow ledge between the steep, upper cave floor, and the probable north line of this room series marks the old trail which led past rooms 117 and 122 to room 73 and the upper east end of the village.

Rooms 114-115. The ledge face which formed the north wall of room 55, continuing to the eastward, underlay the south walls of two rooms whose respective floor levels are indicated by pecked-out areas. The floor of room 114 was approximately 2 feet (0.60 m.) below that of room 115. A portion of the east wall of room 115 remained in position; the other walls were not traceable in 1917.

Buried in the accumulated débris removed from this area was an infant's skeleton, the only burial we found.

Room 116. Disintegrated portions of the east and west sides were found, resting on steps pecked in the sloping sandstone. Although the south wall was missing, its former position was plainly marked by a groove along the cliff edge. Two other pecked grooves, apparently incomplete, extend northwestwardly up the slope from the southeast corner of the room. Fully 3 feet of débris against the inside south wall was necessary to complete the floor level, represented by a worked-out section on the cliff face.

Following excavation we partially reconstructed the fragmentary east and west walls.

Room 117, north of room 121 and northwest of room 66 in the middle north portion of the cave, was set aside for the grinding of maize and other foodstuffs. When exposed by our excavations only sections of the north and west walls remained, and both these were covered by blown sand which had settled about the roots of trees. Pecked grooves identified both the east and south wall positions. Behind the north wall a débris fill marked the cross-cave trail heretofore mentioned. In the middle floor were three broken, slab line mealing bins from which the milling stones had been removed.

In our work of repair we restored the lower walls of this room, rebuilt the grinding bins, and fitted them with metates recovered from other portions of the ruin. (Pl. 23, A.)

Rooms 118-119. Built upon the lowermost terrace and adjoining room 55 on the east was a room approximately 15 feet (4.57 m.) long, subsequently divided by a wattled partition. The fairly level, unworked ledge surface formed a floor several feet above that in room 55. The south and upper north walls had wholly disappeared. Incorporated in the north end of the east side is a projecting block of masonry, 21 inches (0.53 m.) long by 10 inches (0.25 m.) wide, the purpose of which remains unknown. When inhabited, the two rooms were doubtless connected by a door through the dividing, wattled wall.

Room 120 is situated in the lower, middle front of the cave between rooms 119 and 121. Portions of its north and west walls stand at the northwest corner below these fragments the sloping sandstone has been slightly reduced. The room floor, however, was doubtless lower than this pecked area, and must have rested on a fill supported by the now missing south wall.

Room 121, a large dwelling, adjoins the west side of 2-story house No. 66. (Pl. 9, B.) The west and part of the south walls were missing; at the north, on the next higher terrace edge, masonry which we replaced in 1917 separated this structure from room 117.

In the southeast corner is a subfloor, masonry-walled fireplace measuring 3 feet 4 inches (1.01 m.) north and south by 22 inches (0.56 m.). Its west side had slumped with collapse of the southwest quarter of the dwelling. West of this fireplace the room floor had rested on a deep débris fill; sloping cliff, worked down in the northeast corner, occupies the north half of the room. Between the base of this terrace face and the fireplace and 16 inches (0.40 m.) from the east wall is a pothole pecked into the solid rock. It measures 10-1/2 inches (0.26 m.) in diameter by 12 inches (0.31 m.) deep. In this hole we found the earthenware colander illustrated in Plate 46, 1, a charred hairbrush (pl. 39, 1), and a short mano. A door blocked with rude masonry formerly connected this dwelling with the storage chamber below room 66.

It is not unreasonable to suppose that rooms 118-121 were originally 2-story buildings.

Room 122. Obviously a later addition, this structure was built across the old trail on the sloping cave floor north of room 68. The north half of its floor area, of native rock, was worked down somewhat; the remainder consists of adobe mud spread over a débris fill. The masonry walls, which had entirely disappeared, were partially restored in 1917. (Pl. 24, A.)

East of room 122 and abutting the north wall of room 70 an artificial terrace forms a landing at the foot of the step series leading past room 73 to the upper east house group. Following construction of room 122 its roof may have been crossed by aid of ladders or the old trail may have been abandoned in favor of the pecked steps one notes higher up the slope.

Room 123 is situated on the front edge of the slope, east of room 71 and below the trail past room 73. Seatings for its missing walls are not at once apparent, but the rock surface has been battered away somewhat and still shows the stains of smoke and ashy deposits.

The rather precipitous cave floor between rooms 123 and 76 was mostly covered at the beginning of our work by large masses of sandstone, fallen from the cave roof. On this slope is space for at least six large rooms.

Room 124, in the upper east end of the cave, is evidenced by smoke stains and wall adobe adhering to the north, or cliff, side. Owing in part to seepage at this point the masonry has wholly disintegrated.

The narrow ledge extending west from this room would seem to have offered desirable house sites, yet no indication of masonry was found beneath the huge masses of sandstone we removed. (Pl. 20, A.) One might infer from this omission that at least part of the broken rock had accumulated prior to or during occupancy of this upper east terrace.

Room 125. Portions of the east wall were present; masonry which doubtless once rose to the ceiling level stood against the north, or cliff, side. The south wall appears to have been of wattle; through it a door must have opened upon the artificial terrace above room 77. Next the cliff the rock surface had been worked down to a uniform floor level.

In front of room 125 is a narrow walk, supported by a retaining wall which meanders from this point to the cliff beyond room 81. (Pl. 19, B.)

Room 126 lies between rooms 125 and 127. The masonry of its east and west walls abuts the cliff; the south wall is of wattle with a door in the middle. At the base of the cliff a masonry bench, 15 inches (0.38 m.) wide by 17 inches (0.43 m.) high, extends the entire width of the room. The artificial floor lies 8 inches (0.20 m.) below that of room 125; in its middle is a slab-lined fireplace. Three of the slabs bordering this pit extend 5 inches (0.13 m.) above the floor level.

Room 127. The east wall of this dwelling had been demolished by falling blocks of sandstone and wholly disintegrated by seepage. But smoke stains on the cliff indicate a room larger than those inmediately to the west. Portions of the west and wattled south walls were present; against them lay broken flooring, 6 inches (0.15 m.) above the walk outside.

Between rooms 127 and 78 is space for two dwellings, but the level terrace surface here, although smoothed with adobe, discloses no evidence either of masonry or wattled walls. The outer jambs and lintel of the door into room 78 are deeply grooved for the door slab characteristic of storerooms. At one side of this opening two pecked steps gave access to the roof.

Room 128 stood on a low detached ledge west of room 74. A cut out area marks its approximate floor level and indicates the probable wall positions.

From above the northwest corner of this room pecked steps extend northward for several feet up the sloping cave floor, as seatings for a former wall.

Room 129 was built on the cliff edge in the lower east end of the cave, south of room 75 and west of room 82. The cliff face had been worked down in two terraces; the lower of these forms part of the floor, while the upper extends nearly the entire width of the room as a bench 9 inches (0.22 m.) high by 8 inches (0.20 m.) wide. The curved southeast wall, now reduced to 2 feet (0.60 m.) in height, follows the contour of the cliff. No beam holes are evident in the smoke-stained walls.

We repaired the north end of the curved wall, but did not attempt desirable restorations.

Room 130 is one of possibly six large rooms formerly situated on the abrupt slope southeast of room 73 and north of rooms 74 and 128. Pecked grooves and steps as probable wall seatings are noted at intervals, but it seems likely that whatever masonry once stood here was erected upon débris thrown down from dwellings above, as was the case east of room 76.

Rooms 131-135. Excavated areas approximating floor levels, with pecked grooves and steps as rests for masonry walls, indicate the former presence of six or more rooms in that section east of room 76 and north of room 82. At least some of these structures had been erected upon loose débris. No walls were standing in 1917. At beginning of our operations this entire slope was buried under fallen masonry; great blocks of sandstone from the cave roof seemed largely responsible for the destruction.

A series of steps leads up the cliff from the northeast corner of room 75 to the retaining wall opposite the northwest corner of room 78 and was undoubtedly used before houses were built in this quarter. Above rooms 83-85 the cliff appears too precipitous for dwelling sites.

Lack of time precluded even partial restoration of the demolished buildings we have numbered 131-135. But we renewed the old step series from the northeast corner of room 75 down to the platform back of room 82, thus facilitating access to court 83 and its near-by storerooms.

Retaining wall. Rooms 78-81 and 124-127 stand on a level ledge at the upper east end of the cave. Erected upon a 2-inch groove, pecked on the outer edge of this terrace just where the cave floor drops abruptly down to rooms 82-85, is what we have called "the retaining wall." It is the counterpart of that which inclosed court 10 and continued southward to form the walk leading to room 3.

The east retaining wall widened the natural terrace and thus formed in front of the dwellings a walk that varies somewhat in width as it extends from room 81 northward to room 124. Opposite the south corner of room 81 this terrace walk is 2 feet 6 inches (0.76 m.) wide; at the south corner of room 80 it is 4 feet 3 inches (1.3 m.) wide; at the south corner of room 78, 4 feet (1.2 m.); at the west corner of room 78, 3 feet 9 inches (1.14 m.). Three slab steps, totaling 15 inches (0.38 m.) in height, connect the upper terrace level fronting rooms 79-81 with the lower level northwest of room 78.

We repaired and recapped the retaining wall throughout its greater length; rebuilt a missing segment above rooms 132 and 134 after widening and deepening its seating groove; omitted restoration of the extreme ends. (Pls. 19 and 20.) As noted under the description of individual rooms, certain minor repairs were also made on the houses hereabout.

Gallery wall.—In the high, upper portion of the cave, on the very edge of the cliff and about 25 feet (7.62 m.) above room 59, is a long straight wall approximately 4 feet (1.2 m.) high. Viewed from below, this wall appears to stand alone, but there is a bare possibility that other walls lie buried in the loose shale which has accumulated on the slope behind the wall. About 2 feet (0.60 m.) above the front base of the masonry 11 poles protrude to suggest a probable floor level behind the wall.

Access to the gallery was formerly gained by means of the long pine pole still standing at the north end of court 10. Only one who has shinned up this splintery pine and felt it tip threateningly outward from the sheer cliff can fully appreciate our lack of knowledge concerning the upper gallery and the structures, if any, it shelters.