University of Virginia Library

III. KIVAS

There are perhaps 17 ceremonial chambers, or kivas, within the
walls of Pueblo del Arroyo. All are circular and, in effect, subterranean.
On our ground plan, figure 2, they appear to have a studied
arrangement. One is conspicuously located in the south wing; two are
similarly situated in the north wing; seven crowd the middle portion,
and at least six are indicated between the house mass and the arc of
one-story rooms enclosing the court on the east. In addition we assume
the presence of a principal chamber, perhaps a Great Kiva,
situated within and at the west side of the court. For this latter we
had reserved the designation "A," but excavation plans for the final
year were changed and we made no search for it. Kiva B and three
others west of the pueblo are outsiders and will be considered as such
in the next chapter.

It is now generally accepted that the circular, subterranean ceremonial
room of the Pueblos evolved from the Basket Maker pit house
wherein were combined family living quarters and space for the observance
of family or clan rituals. The Basket Makers are best known
from cultural remains they buried in caves throughout the broad
drainage area of the San Juan River. Chaco Canyon lies on the south
margin of that area, but its prehistoric inhabitants included occupants
of at least one Basket Maker village. Here individual dwellings again
served for clan and family worship, but community ceremonies were
performed in a vast, semisubterranean structure 40 feet in diameter,
a forerunner of the Great Kiva (Roberts, 1929, pp. 73-80).

The major ruins of Chaco Canyon represent the very summit of
communal endeavor in the southwestern United States in prehistoric
times. Pueblo del Arroyo is one of those major ruins and the nearest
neighbor of Pueblo Bonito. Of the 17 or more kivas in Pueblo del
Arroyo we excavated and examined 7. Only one, C, apparently occupies
space originally set aside by the town planners. With the possible
exception of unexcavated A, all are individually enclosed within a
quadrangle of masonry walls. That many of these walls represent
houses vacated to make way for ceremonial chambers emphasizes the
dominant influence of ritual in local affairs. Kivas are still essential
in the life of modern Pueblo villages, but since Conquest times when
alien pressure forced the change, they have usually been rectangular
rather than circular and located inconspicuously among ordinary

dwellings. In consequence, historic kivas may have lost some of their
former specialized functions; several are known to have been deserted
as the group responsible for its construction and upkeep gradually
declined.

During his study of western Pueblo architecture, 1881-88, Victor
Mindeleff observed that although the kivas were built by religious
societies, none that he entered at Oraibi was then occupied exclusively
for religious purposes. Each was used, especially in winter, as a
meeting place for members of the society to which it belonged. "The
same kiva thus serves as a temple . . . as a council house for the discussion
of public affairs . . . as a workshop by the industrious and
as a lounging place by the idle." (Mindeleff, 1891, p. 130.)

Parsons (1939, p. 9) confirms Mindeleff's observation: "Kivas are
communal or partly communal buildings, clubhouses used by the men
as meeting-place, workshop, or as a place to dance or hold ceremonials."
Titiev, a later student of the Hopi, adds: "Kivas are owned
by the clans whose members took the initiative in building them. In
the event that a kiva has fallen into disuse, its ownership may be
transferred to whatever clan is most instrumental in its repair. . . .
Except on special occasions . . . females are barred from the kivas."
(Titiev, 1944, p. 104.)

Every kiva-owning group is the recognized possessor of one or
more distinctive rituals, each of which, it is expected, will be performed
at the appointed time. The entire community benefits. Major
ceremonies, usually of 9 days' duration, are conducted almost wholly
in the secrecy of the kiva. Since these rites differ from each other
and since each society is at pains to guard its secrets from other
groups, one would expect kivas to differ. But the contrary is true.
All are very much alike, inside and out. This is particularly true of
kivas in prehistoric ruins throughout the San Juan drainage, including
Chaco Canyon.

Half a century ago T. Mitchell Prudden examined a number of
small-house sites north of the San Juan River and propounded his
theory of a dwelling-kiva-burial mound complex as the nucleus, or
"unit," of all Pueblo ruins, irrespective of size (Prudden, 1903, 1914,
1918). He was particularly intrigued by the kivas, and we may paraphrase
his description of those he excavated in southwestern Colorado
as follows:

Small-house kivas are circular, 13 to 21 feet in diameter and wholly
subterranean, with an inner encircling bench about 3 feet high from
which six masonry pilasters rise flush as roof supports and with shelflike

banquettes or recesses between the pilasters; with an underground,
man-sized passageway leading from the north banquette to a
front room in the pueblo and a south banquette much deeper than the
other five; with an above-floor ventilator tunnel under the south
banquette connecting with an external, vertical air shaft; with a fireplace
near the middle of the kiva floor, a deflector to protect the fire
from in-rushing air, a sipapu between fireplace and north banquette
and, below the latter, one or more niches in the face of the bench.

Some years later Kidder (1924, pp. 67-68) pointed out that instead
of being simple and primitive, as Prudden thought, these smallhouse
kivas are as highly specialized as those of the famous Mesa
Verde cliff dwellings, many of which were still inhabited during the
twelfth century A. D. (Douglass, 1935, p. 52). A combination smokevent-hatchway
in the roof, a deep south recess with an above-floor
ventilator under it, six masonry pilasters rising flush with the face
of the bench and 2 to 3 feet above it, a deflector, fireplace, and sipapu
are among the architectural features to be expected in Pueblo III
kivas north of the San Juan River.

In contrast to this standard, ceremonial rooms in Chaco Canyon
ruins generally lack the sipapu and the deep south recess. Ventilator
tunnels lie beneath, rather than above, floor level. An above-floor
ventilator in Chaco Canyon is invariably evidence of reconstruction.
Instead of a deep south recess at bench level or above, Chaco kivas
ordinarily have a shallow recess in the bench itself. These and other
characteristics will be considered more fully in the following descriptions
of our Pueblo del Arroyo kivas. Abbreviations used in the
descriptions are: d. = deep; dia. = diameter; h. = high; l. = long; w. =
wide.

Kiva C (fig. 13).

Average diameter at floor, 25′ 10″; bench, 29″ w. by 22″ h.; south recess,
11½″ d. by 6′ w. in front, 6′ 4″ at rear; 8 wood pilasters averaging 17″ w. by
26″ 1. by 8″ h. set back 2-3″ from edge of bench; "wainscoting" of poles and
grass between pilasters at back of bench; 15″-by-16″ opening of subfloor ventilator
tunnel 16″ w. by 28″ d. is 5′ 5″ from back of recess; oriented S. 1° W.,
tunnel passes under middle of recess to connect with vertical shaft incorporated in
outer kiva wall; shaft intake, 15″ by 18″; semicircular fireplace 27″ by 35″ by
16″ d., flat side to south; subfloor vault west of fireplace, 8′ 16″ l. by 46″ w. by
18½″ d., with clay floor, crude masonry lining, and two partitioning walls about
9″ h., vault filled with clean sand; north of vault, near pilaster No. 5, single
post hole 7″ dia. by 9″ d.; ceiling burned, timbers salvaged, and kiva thereafter
used as neighborhood dump.

The facing masonry of the Kiva C main wall and bench includes
both dressed friable sandstone and laminate sandstone, the latter predominating

in the bench. Construction apparently began with the
bench, since its rubble core was widened to form a foundation for the
main wall. The facing of the latter does not extend below the bench
surface but rises above it 10 feet 2 inches to court level. Time had
left its mark on this masonry, so we rebuilt the upper east side in
hope of delaying further disintegration (pl. 18, right).

Fig. 13.—Kiva C.

On the bench as it encircles the chamber are the remains of eight
pilasters, each a squared timber. All the timbers had been burned and
three of them completely destroyed, but from their seatings or from
dimensions of the sockets they had occupied, we know them to have
measured from 15½ to 19 inches wide and about 8 inches thick. Each
timber came to within 2 or 3 inches of the face of the bench, and its
opposite end was socketed in the main wall for a foot or more. Here
a varying quantity of slaty coal had been packed about the log as
masonry was built up over it.

Pilasters are supports for a ceiling of cribbed logs. Kiva C had
eight, and the lower course of ceiling logs, in pairs, rested upon

neighboring pilasters, say 1 and 2, 3 and 4, 5 and 6, 7 and 8. The
second course of paired logs joined the ends of those in the first, as
2 and 3, 4 and 5, 6 and 7, 8 and 1. Beginning with the fourth or
fifth layer, a third log was added, then another and another until, as
courses multiplied and height was attained, the cribwork drew toward
the center creating a domed ceiling. The uppermost course was
bridged by logs laid side by side and covered by cedar splints, bark
or grass, and adobe mud. In the middle of this bridged area, flat and
8 or 10 feet square, a rectangular opening was left to serve both as a
doorway and a vent for fireplace smoke.

In Kiva C the middle of the first pilaster, reading counterclockwise,
was 24½ inches from the south recess and, center to center, the 8
averaged 9.9 feet apart. At the rear of the bench, next to the kiva
wall, post holes about 2 inches in diameter and varying in number
from 13 to 15 between pilasters marked the position of a sort of
wainscoting. Since the lower members of the cribbed ceiling rested
directly upon the pilasters, no more than 8 inches of this wainscoting
would have been visible to occupants of the chamber. Presumably
grass had been crowded in between posts and wall, as in the case of
several kivas excavated at Pueblo Bonito. Such a feature seems quite
superfluous since the main wall was invariably plastered before construction
of the ceiling began.

The cribbed ceiling of Kiva C was at least partially destroyed by
fire, yet most of the logs were apparently salvable since few had been
left on the floor. Although the squared pilaster timbers were consumed
more or less completely during the conflagation, we recovered
from their respective seatings the following sacrificial offerings:

Pilaster No. 1:

• 30 olivella beads and fragments

• 27 oblong and figure-8 beads and
  fragments

• 9 discoidal beads and fragments

• 10 turquoise fragments

• 1 standstone cover 2¼″ diameter
  by ¼″

Pilaster No. 2:

• 12 oblong and figure-8 beads

• 2 discoidal beads

• 1 turquoise pendant fragment

Pilaster No. 3:

• 20 olivella beads and fragments

• 11 oblong and figure-8 beads

• 5 discoidal beads

• 7 abalone shell fragments

• 3 turquoise fragments

• 1 rib fragment, small deer or antelope

Pilaster No. 4:

• 2 discoidal beads

• 1 turquoise pendant

Pilaster No. 5:

• 18 olivella beads and fragments

• 26 oblong and figure-8 beads and
  fragments

• 6 discoidal beads

• 5 abalone shell fragments

• 2 turquoise pendants and fragment

• 3 turquoise fragments

• 1 cylindrical shell (?) bead with
  longitudinal and transverse
  drilling

Pilaster No. 6:

• 16 olivella beads and fragments

• 32 oblong and figure-8 beads and
  fragments

• 9 discoidal beads and fragments

• 1 Chama bead fragment

• 2 shell fragments

• 4 turquoise pendants and fragments

• 6 turquoise fragments

Pilaster No. 7:

• 13 olivella beads and fragments

• 13 oblong and figure-8 beads and
  fragments

• 1 discoidal bead

• 1 turquoise pendant fragment

Pilaster No. 8:

• 5 olivella beads

• 70 oblong and figure-8 beads and
  fragments

• 7 discoidal beads and fragments

• 1 oval shell bead fragment

• 8 discoidal turquoise beads and 7
  fragments

• 4 turquoise pendants and fragments

• 1 turquoise tessera

• 47 turquoise chips

Pilaster No. 8 was least burned and the fact that we found in it a
number of turquoise beads and tiny chips leads me to the belief that
like offerings might have been overlooked among the charcoal and ash
of other pilasters. Similarly, the absence of stone covers for all except
No. 1 suggests the use of wooden disks. The pilaster furnished a
tree-ring date of A. D. 1067+x.

The subfloor ventilator tunnel or duct is masonry lined, paved with
sandstone slabs, and roofed with poles of uniform diameter, split
cedar, and adobe mud. The floor slabs are part of a triangular pavement
that more than spans the south recess and extends over 10 feet
toward the middle of the room (fig. 14).

A subfloor vault west of the fireplace requires an additional word.
Its rather crudely constructed sides apparently were never plastered;
packed clay provided an indefinite floor. Five and one-half inches of
clean sand had been spread upon the floor and, upon that sand, two
secondary walls were built. The first parallels the west side for part
of its length, and the second abuts both the first wall and the east side
of the vault. Both walls are about 9 inches high which brings them to
within 4 inches of the kiva floor. Excepting a small quantity of wood
ash against its east side the vault was filled with clean sand.

After destruction, Kiva C served briefly as a neighborhood dump.
Ashes were not conspicuous but floor sweepings were present for we
recovered a quantity of potsherds, three restorable pieces of pottery
and a few other objects. Two unusual ladles (pl. 26, b, c) lay
shattered at east bench level but not on the bench; a sherd pottery
scraper (U.S.N.M. No. 334672) was found in the fireplace. Of 518
miscellaneous potsherds collected, 1 was of Mesa Verde ware; 121,
or 23.4 percent, were Corrugated-coil, and 115, or 22.2 percent,
Chaco-San Juan. Antelope, mule deer, and turkey bones were present,

Fig. 14.—South recess and ventilator, Kiva C, with underlying pavement.

as always; also, the incomplete skeleton of a puppy (U.S.N.M.
No. 334957) and several bones of a young coyote. Part of the skeleton
of a prairie falcon (Falco mexicanus) lay about a foot above the
bench between pilasters 7 and 8. Only three manos, fragments of two
metates, and four hammerstones were unearthed—ample evidence
that the dump was short-lived.

The most puzzling of these finds is a piece of glasslike pumice
(U.S.N.M. No. 334800), as light and frothy as seafoam. It lay a
few inches above the floor on the east side of the chamber, an obvious
discard. As reported elsewhere (Judd, 1954, p. 293), two mineralologists
identified the substance as rhyolitic pumice having a high
silica content, while a third expressed the opinion that it was a fragment
of perlite or pitchstone altered by fire. Near this specimen
was a miniature jar (U.S.N.M. No. 334640), blistered by heat.
Broken masonry from second- and third-story rooms had fallen upon
this household debris to complete the fill of Kiva C.

After concluding our examination we dug an exploratory trench
from pilaster No. 1 north to pilaster No. 4. At a depth of 44 inches
we came upon an earlier floor or, more likely, a work surface. This
abuts the bench foundation, here composed of adobe mud and chance
blocks of sandstone, some of which protrude several inches. Where
we bared it, the foundation stands out an inch or more from the face
of the bench and is unbroken throughout its full height of 44 inches.
At the south the foundation does not follow the lines of the recess
as one would expect but continues across the front of it. Whether
intended or not, no foundation stones protrude at this point.

The floor at a depth of 44 inches abuts the bench foundation while
another, 3 inches lower, continues under. We were interested to note
that this second surface is on the same level as the floor of Room 32.
Twenty-two inches deeper, or 5 feet 9 inches below the kiva floor,
we came upon a third pavement and, 46 inches from the north end of
our trench, an unfinished or partly razed east-west wall 40 inches
wide and 7 inches thick, built upon an 8-inch-high foundation. This
third pavement does not appear on the north side of the 40-inch wall.
From a sandy fill between the second and third pavements we recovered
a single bone bead and a number of potsherds decorated with
late, fine-line hachure.

To explore the quadrangle in which Kiva C stands, test pits were
dug in the northeast, northwest, and southwest corners. At the
northeast the butt of an 8-inch beam extends through from the adjoining
unexcavated room and, on the north side of it, a comparable

timber crosses the corner at an angle, joining the east wall with the
outside of the kiva. Ceiling poles resting upon this beam (pl. 10,
lower) might suggest occupancy of the space beneath, but I incline to
the belief they were intended merely to level the court. Ceiling poles
in pairs were noted repeatedly in excavated rooms, but this is the only
instance where we found them trebled.

Above the poles, masonry of the north and east walls consists of
carefully laid laminate sandstone, while that below is careless and
crude in comparison. This lower stonework, like the external wall of
the kiva, was made with large quantities of mud and large blocks of
unshaped sandstone, some of which at irregular intervals protrude
6 or 7 inches. As exposed in our 6-foot-deep test, the fill of the corner
consisted almost wholly of adobe spalls from razed buildings. In this
were a few potsherds of late type. The beam from the adjoining room
(field No. 146) gave Dr. Douglass a date of A. D. 1102+; that
across the corner, 1103. These two are more likely to represent the
time of construction than is the No. 8 pilaster log, dated 1067+x.

The northwest corner was equally interesting but without the remnant
of roofing. Both the north and west walls had originally been
built of superior, laminate sandstone masonry; they are tied and
several layers of plaster still cling to that on the north. But—and
here is the interesting fact—sometime after completion the west wall
facing had been torn out and replaced with a crude stonework that
abuts the plastered north wall. Then a masonry brace, 22 inches wide,
was built in against the new corner union to extend diagonally to tie
with the outer curve of the kiva. From this diagonal, as from both
the reconstructed west wall and the exterior of the kiva, building
stones protrude sporadically. Debris of reconstruction filled the angle.
A second, shorter brace ties with the west wall about midway of its
length.

In the southwest corner the exterior wall of Room 21, below its
second-story ceiling offset, seems to have been planned and built in the
coarser type of stonework noted in the lower northeast and northwest
corners. And here also, as in the northwest angle, the original west
wall facing had been replaced. I believe this cruder uneven-stonework-with-protruding-block
technique was intended to be concealed.
It invariably occurs on the outside of kivas, and since the space between
a kiva and its enclosing walls was always filled to roof level, it
seems possible that the reconstruction we have noted on the west side
of the Kiva C square was forced by a tradition of concealment.

At a depth of 6 feet 5 inches our southwest corner test came upon

an uneven pavement, overlain by charred timbers and charcoal. This
wreckage must have been gathered elsewhere and dumped here, for
the degree of smoking on the walls was insufficient to indicate a local
fire.

The three subfloors in Kiva C and the 40-inch-wide crosswall exposed
by our trench at a depth of 5 feet 9 inches remain unexplained.
On a work surface 3 inches above the floor level of Room 32 a 44-inchhigh
foundation was built for the Kiva C bench. The need for a
foundation twice the height of the bench it supports is not apparent
unless it be a desire on the part of the builders to lift up their creation,
to place it on a higher plane in relation to its surroundings. The
flat roof of Kiva C and the filled-in corners of the square formed a
courtlike area at the second-story ceiling level, a convenient dooryard
for inhabitants of the third-story rooms adjoining. There are no
second-story doors facing Kiva C.

Kiva E (fig. 15).

Average diameter at floor, 14½′; bench, 20″ w. by 24″ h.; south recess, 12″
d. by 4′ 3″ w. in front, 4′ 5″ at rear; 6 pilasters average 15⅔″ w. by 18½″ l. by
7¾″ h., set back 2-2½″ from edge of bench; 6′ from back of recess, 13″-by-15″
outlet of subfloor ventilator tunnel with 2″ overhang on north; tunnel, oriented
N. 10½° E., masonry lined, slab floored, 2″-dia. roofing poles 19½″ above floor;
ventilator shaft 26″ south of recess has 12″-sq. intake; 7″ beyond north end of
ventilator tunnel is masonry-lined fireplace 22″ dia. by 14″ d.; ceiling logs and
pilaster offerings removed.

Both the main wall and bench facing consists almost wholly of
laminate sandstone, plastered and replastered. We counted no fewer
than 20 successive layers on the front of the bench, the total thickness
being only 1 inch. An unusual feature of the subfloor ventilator
tunnel is that three courses of stonework 5½ inches thick overhang the
north side of the outlet as though designed to deflect the draft
backward.

On the bench at fairly uniform intervals are six pilasters each consisting
of a section of a round log, walled at the sides with small-stone
masonry and plastered over. The logs are tenoned into the main wall
a foot or more; their forward ends are not covered with masonry
but stand free, square cut, smoothed, and plastered. A few logs were
countersunk into the bench surface; none was flattened, top and bottom,
as were those in C.

Inasmuch as the Kiva E pilasters are in a better-than-usual condition
it seems desirable to record their constructional details, often
incomplete elsewhere. The pilasters are numbered counterclockwise

from the south recess and the distances between are front center to
center.

Pilaster No. 1: Front center 19″ from corner of recess; 19″ l. x 15″ w. x 8″ h.
enclosing 8½″-dia. log set back 2″ from edge of bench. Hole for sacrificial
offering in top of log 11½″ from wall is 1½″ dia. x 1½″ d. and grooved around
for countersunk stone disk 2″ dia. x ¼″ thick. Disk in place but cup empty.

Fig. 15.—Kiva E with outlines of earlier kivas beneath.

Pilaster No. 2: 7′ 6″ from No. 1; 19″ l. x 15″ w. x 7½″ h. enclosing 7″-dia. log
set back 2½″ from edge of bench. Offering hole 12½″ from wall is 1½″
dia. x 2″ d.; cover not found.

Pilaster No. 3: 7′ 4″ from No. 2; 18½″ l. x 15½″ w. x 8½″ h. enclosing 8″-dia.
log set back 2″ from edge of bench. Offering hole 12½″ from wall is 1½″ dia. x
1½″ d. with ¼″-d. groove around top; cover not found.

Pilaster No. 4: 7′ 7″ from No. 3; 18¼″ l. x 16″ w. x 6½″ h. enclosing 8½″-dia.
log set back 2½″ from edge of bench. Offering hole 10½″ from wall is 1¼″
dia. x 2″ d., grooved at top; cover not found.

Pilaster No. 5: 7′ 4″ from No. 4; 18¼″ l. x 16″ w. x 8″ h. enclosing log 7½″ dia.
set back 2″ from edge of bench. Offering hole 11½″ from wall is 1¾″ dia. x 2″
deep.

Pilaster No. 6: 7′ 4″ from No. 5; 17½″ l. x 16″ w. x 7″ h. enclosing log 7½″ dia.
set back 2¼″ from edge of bench. Log decayed; offering cup not found. From
No. 6 to No. 1, 7′ 2″.

Fig. 16.—Cross section A-A′ from the McElmo Tower through Kiva E and Kiva G.

(From the original survey by Oscar B. Walsh.)

Not a single bead or bit of turquoise
came to light and, since
broken masonry above each pilaster
evidences removal of the ceiling
logs, I would guess the six offerings
were salvaged at the same time.
Absence of covers, excepting the
one of stone, suggests that they too,
were reclaimed.

The elevation of Kiva E in relation
to nearby structures (its floor
is only 15 inches below that of
Rooms 60-61 [fig. 16]) naturally
prompted further inquiry. A test
pit in the southeast corner of the
enclosing quadrangle revealed part
of an earlier kiva, its floor 10 feet
below that of Kiva E. The stonework
of this earlier chamber seems
comparatively rough, but it is plastered
and smoke blackened. The
customary bench is 12 inches wide
and 25 inches high and on it, incorporated
in the foundation of Kiva E,
we bared one side of a masonry
pilaster 11 inches long and 15 inches
high. This pilaster, which ends
an inch from the face of the bench,
was thickly coated with adobe plaster
and, embedded on top, were the
broken ends of a pair of ceiling logs.
Despite discrepancies in bench width
and height of pilaster this earlier
chamber appears to be part of that
previously discovered underneath
Room 47B.

Following abandonment, the old
kiva walls were partially pulled down
and a new dwelling was erected, or
reconditioned, at the second-story
level. A relic of that later house is
an adobe pavement, 9 feet 7 inches
above the floor of the old kiva, that
was cut through upon construction

of Kiva E and the remainder left to support the second-story wall
built in to complete the Kiva E square and, at the same time, provide
a north side for Room 47B.

Where we exposed it, the arc of the old kiva abuts sections of two
straight plastered walls the significance of which was not apparent
within the narrow limits of our test pit.

Part of a second subfloor kiva, perhaps the intended successor to
that last described, was unearthed in the northeast quarter of Kiva E,
below pilasters 2 and 3. Here, again, was a confusion of foundations,
walls, and fragmentary walls.

First, immediately under the floor of E, we came upon a 3-footlong
section of curved wall, 15 inches wide, with Type III masonry
on its concave face, unplastered but smoke stained. Two feet lower
was what we judged to be the main wall of the chamber. It, too,
was of Type III but, unlike the upper segment, was 18 inches thick
and well plastered. We followed this to a depth of 9 feet 4 inches
without finding an associated bench, floor, or identifiable work surface.
As in the case of the kiva under Room 47B, at least two small
poles had been laid in horizontally to bind the facing masonry to
whatever lies behind.

Both the main wall and the 15-inch-wide section above it had been
built against the plastered face and end of a straight wall, constructed
of conspicuously large blocks of sandstone and extending in a northwesterly
direction. The north end of this construction was partly torn
out when the Kiva E bench was built and here, too, underlying the
bench and abutting both the straight wall and the kiva curve is a
foundationlike fill of sandstone and mud that we uncovered to a depth
of 5 feet 3 inches.

At the opposite end of this straight wall the convex face of the
3-foot-long section first mentioned is abutted by a comparable length
of east-west wall, 12 inches thick and of undetermined height. The
angle between the two had been carefully rounded with adobe and
replastered to a depth of 20 inches. When uncovered, this angular
recess was filled with sandstone spalls and mud; across the top of it
lay a hewn plank, 33 inches long by 3½ inches wide by 1½ inches thick.

Clearly this segment of subfloor kiva is part of that underlying
Room 52B. In both cases the masonry is Type III and plastered; in
both cases the exposed portion abuts a straight wall that lies at variance
with the studied regularity of Pueblo del Arroyo. In both cases
the fill is predominately debris of reconstruction.

On the west side of Kiva E, slightly underlying the bench below

pilaster No. 5, is a masonry-lined, ash-filled firebox measuring 16½ by
14 inches by 10 inches deep. It belongs to a third kiva, perhaps never
completed, that immediately preceded E since its floor is on the same
level. Six and one-half inches south of the firebox is the end of a
subfloor ventilator duct, 13 inches wide and 15 inches deep, masonry
lined and paved with slabs. We made no effort to clear this duct as
it passed under the bench of Kiva E, but later unearthed the partially
dismantled intake of its connecting shaft.

The eastern half of this third predecessor had been completely
razed, but sections of its foundation are still to be seen under the
Kiva E bench. We traced one section to a depth of 47 inches and, in
doing so, laid bare the southeast corner of the dwelling that formerly
stood here.

Remnants of this earlier kiva also survive outside and to the west.
But, to our disappointment, the work of demolition had been extensive.
The salvaging of suitable building stone seems to have been the driving
motive. In only one place did the main wall hearting extend as much
as 40 inches above the bench; what remained of the latter showed a
height of 22 inches and a width of 15. The masonry was chiefly of
laminate sandstone with sporadic banding, the same as that of the
kiva under the northeast quarter. The pilasters, if any, had been
demolished without trace. We estimate the floor diameter at 12 feet.

A 3-by-4-foot test pit in the northwest corner of the Kiva E quadrangle
shows 32 inches of masonry still standing above the secondstory
floor level. There is a 2-inch offset on the west side while on
the north the second-story wall actually overhangs the first by 2 inches.
Upper and lower walls had been plastered. A rotted 8-inch beam on
the west side 44 inches from the corner and the decayed ends of
poles embedded in the north wall are remains of a former first-story
ceiling.

In the southwest corner of the square, second-story masonry stands
to a height of 58 inches. It bears two coats of plaster: the first plain
adobe color and the second, white. Plaster below the floor offset is
not whitened. Six feet from the corner a 25-inch-wide wall juts
eastward, another relic of the rooms that once stood here. In that
wall fragment, 3 feet above the second-floor offset, we noted remains
of a log built in horizontally for added strength.

The south wall is secondary, crudely constructed, and faced on the
Room 46 side only. It abuts blocked second-story doors at each end
and, about 4 feet below ceiling level, rests upon debris of occupation
with which the space was filled. From the floor two clumsy buttresses

brace the outside of the old kiva. The second and wider of the two is
5 feet high and abuts the contemporary ventilator shaft which, presumably,
was at least partially erected in advance of the kiva.

Kiva owners obviously are persistent builders. Reviewing the results
of our several tests here we conclude that the first of the four
kivas on this site is that in the southeast corner, 10 feet below the

floor of Kiva E. This was followed by a larger chamber, represented
by the segments under the northeast quarter of E and under Room
52B. In time this second chamber was replaced by a third after the
two rooms east of 48 and 49 had been razed to make way. Then it,
too, was abandoned, perhaps even before completion, in favor of
Kiva E. Kiva E, therefore, was built where four dwellings formerly
stood after three earlier kivas had, in succession, occupied portions of
the same site.

Kiva F (fig. 17).

Average diameter at floor, 15′; bench, 19″ w. by 19″ h. from latest floor; south
recess, 19″ d. by 4′ 5″ w. in front, 5′ at rear; no shelf at back; 6 pilasters of

squared cedar average 9⅔″ w. by 16″ l. by 7½″ h., set back 3″ from front of
bench; subfloor ventilator tunnel 12″ w. by 24½″ d. ends with 12″-sq. outlet
5′9″ from back of recess; tunnel, oriented N. 14° W., with lintel poles 17½″
above its floor, had been filled, floored over, and replaced by lateral ventilator,
12″ w. by 15″ h., through middle of south recess to connect with vertical shaft
of subfloor tunnel; shaft intake 13″ sq.; crude masonry deflector, 34″ l. by
10″ w. by 15″ h., built over north end of subfloor tunnel with rock fill beneath;
2′ beyond deflector is fireplace 26″ sq. by 14″ d., upper 4″ lined with masonry,
offset on south side 5″ d. by 5″ h.; vault west of fireplace on 2d subfloor, 56″
by 27″ by 9½″ d.

Kiva F, like E, was built where dwellings formerly stood. Its
masonry is generally of Type III and its highest standing wall, on the
west where a stairway comes up from Room 44, rises 6½ feet above
the bench. An uncommon feature is that the main wall was finished
first and the bench then built against it. The customary south recess
extends to the main wall, the full width of the bench, without the
narrow shelf at the back.

Upon the bench are six pilasters, each a section of stone-ax-hewn
red cedar resting upon a masonry base an inch thick. Each section is
socketed in the kiva wall, and its front end lies about 3 inches from
the edge of the bench. Instead of being enclosed at the sides with
small-stone masonry, as were those in Kiva E, these pilaster logs stand
exposed except for a thin coating of adobe mud. At the sides successive
layers of bench plaster round up to the lower edges of the logs
but no farther. Each log had been provided with a cuplike repository,
grooved around the top for a ¼-inch-thick discoidal cover; each
repository had been emptied, and its cover reclaimed, when the ceiling
timbers were withdrawn. To present the details we again number the
pilasters counterclockwise from the south recess and measure the distances
between from center to center of the forward end.

• Pilaster No. 1: 23″ from recess; 16″ l. x 10″ w. x 8″ h.; repository, 7½″ from
  kiva wall, is 2″ dia. x 1″ d.

• Pilaster No. 2: 7′ 7″ from No. 1; 16″ l. x 10″ w. x 7½″ h.; repository, 7¾″
  from wall, 2″ dia. x 1″ d.

• Pilaster No. 3: 7′ 8″ from No. 2, 16″ l. x 10″ w. x 7½″ h.; repository, 7″ from
  wall, 2″ dia. x 1½″ d.

• Pilaster No. 4: 7′ 7″ from No. 3, 16″ l. x 9½″ w. x 7″ h.; repository, 7″ from
  wall, 1¾″ dia. x 1¼″ d.

• Pilaster No. 5: 7′ 8½″ from No. 4, 16½″ l. x 8½″ w. x 7½″ h.; repository, 4″ from
  wall, 1¾″ dia. x 1¼″ d.

• Pilaster No. 6: 7′ 9″ from No. 5, 16½″ l. x 10¼″ w. x 7″ h.; repository 5½″ from
  wall, 2″ dia. x 1″ d. From No. 6 to No. 1, 7′ 8″.

The bench abuts the kiva wall and the south recess interrupts the
bench for its full width. Five feet nine inches from the back wall is

the end of a subfloor ventilator duct or tunnel 12 inches wide by
24½ inches deep; its lintel poles, 17½ inches above its floor, are overlain
with sandstone slabs covered with adobe mud. The duct lies
N. 14° W. and passes under the middle of the south recess and 39
inches beyond to connect with its vertical air shaft, 13 inches square
at the intake. At its north end the duct is blocked by a barrier of irregular
sandstone blocks introduced as foundation for a rather crudely
constructed masonry deflector built upon the walls of the duct and its
12-inch-square outlet. This deflector is associated with a ventilator
outlet that had been cut through the back of the recess, 3 inches above
its floor, and directly above the subfloor duct, to connect with the
same vertical shaft.

Two feet beyond the deflector is an unusual fireplace, 26 inches
square by 14 inches deep. The lower 10 inches of it is lined with
slabs on edge; the remainder, with masonry. A second unusual feature
is a steplike projection on the south side with a 5-inch tread and a
5-inch riser to the kiva floor.

A broken area beside the deflector revealed three earlier floors at
a depth of 3, 4, and 5½ inches, respectively. The deflector stands upon
the first subfloor, that at a depth of 3 inches. On the floor next below,
we came upon one of those puzzling sunken vaults that often occur
on the west side of Chaco kivas. This one measures 27 by 56 inches
and 9½ inches deep; its clay lining rounds off neatly with its associated
floor. The vault is puzzling because, like many of its kind, it was
filled with clean sand and concealed beneath a 1-inch-thick adobe
pavement laid throughout the chamber. It was the existence of this
overlying pavement that compelled us to represent the vault as subfloor
on figure 17.

When excavating for the vault its builders cut through a clay-lined,
ash-filled fireplace 23 inches in diameter, a feature of the third subfloor.
Five inches lower was a second and smaller fireplace. We believe
both to have belonged, successively, to an earlier kiva whose wall,
razed to within a few inches of its accompanying floor, immediately
underlies the Kiva F bench.

The masonry of this earlier chamber is of carefully selected laminate
sandstone and is plastered. From the short section of it exposed by
our trench we estimate the diameter at nearly 18 feet, a figure exceeding
that of Kiva F itself. Therefore, unless our observations are
in error, the western arc of that earlier kiva must flatten sharply
against the outside east wall of Room 44. We have assumed the east
wall of 44 to be one of those originally laid down by the village planners.

Its opposite was razed by the builders of Kiva F but, in their
effort to use as much as possible of the old wall, the line of its west
face was preserved in the kiva masonry above the east corner of the
south recess and, again, near pilaster No. 3.

With substantial walls standing on the west and north of Kiva F,
its required enclosure was completed by erecting new walls on the
east and south. The latter brought into being the long narrow room,
62; the former is clearly an improvisation.

As stated above, all ceiling timbers and all pilaster offerings had
been removed from Kiva F. During this operation chunks of roofing
adobe collected almost bench high. Thereafter the empty chamber
briefly became a depository for neighborhood rubbish. From this
trash pile we recovered a number of interesting artifacts, some of
which will be described in another chapter. In addition there were
two dog skeletons, one lying on the floor at the east side (pl. 12, B)
and the other (field No. A-485), about a foot and a half higher. Elsewhere
in the fill were the incomplete skeleton of a young dog (field
No. A-486), several jack rabbit bones, and one bone identified as that
of a bobcat. Of 1,729 potsherds tabulated by Amsden and Roberts,
8 were of Mesa Verde type and 301, Chaco-San Juan.

Kiva G (fig. 18).

Average diameter at floor, 18½′; bench 27″ w. by 26″ h.; south recess 12″ d.,
6′ 2″ w. in front, 6′ 7″ w. at rear; 6 pilaster logs averaging 8⅔″ dia. exposed
21¼″, sided with masonry to average 19″ w. by 23½″ l. by 9″ h., set back 2-3″ from
face of bench; repositories empty; subfloor ventilator duct 7½′ north of recess is
16″ w. by 33″ d., stripped of lintels 27″ above floor, filled, and floored over;
duct at angle of S. 2° E. continues 39″ beyond recess to connect with ventilator
shaft having 15″-by-19″ intake; in recess directly over subfloor duct a 16″-w. by
15″-h. above-floor ventilator outlet cut through to connect with same vertical
shaft; associated masonry deflector, 5′ 2″ l. by 10″ w. by 14″ h., with concave
side to south, stands upon subfloor ventilator duct 5′ north of recess; masonrylined,
plastered fireplace 32″ by 37″ by 28″ d., 2½′ beyond deflector is divided by
slabs on end, has 5-inch-deep steplike offset on each side.

Among those we examined at Pueblo del Arroyo, Kiva G is exceeded
in diameter only by C. Its masonry, predominately laminate
sandstone, stands to a height of 5½ feet above the bench on the west
side, 5 feet 2 inches on the south. On the east side the bench masonry
ends, without foundation, 5 inches below the floor. Both bench and
main wall have been plastered repeatedly.

Six pilasters are present and, like those in Kiva E, each consists
of a round log enclosed at the sides by small-stone masonry and
heavily plastered. Unlike those in E, however, the masonry siding

extends beyond the end of the log, leaving a hollow to be filled with
mud at plastering time. Each log was provided with a shallow hole
for a sacrificial offering but the holes were not rabbetted for covers.
All six repositories had been emptied. Reading counterclockwise
from the south recess, pilaster dimensions and details are as follows:

• Pilaster No. 1: 22″ from recess, 22″ l. x 19″ w. x 9″ h. Log, 9″ dia., extends 18″
  from wall; repository, 5″ from wall, 2″ dia. x 1″ d.

Fig. 18.—Kiva G.

• Pilaster No. 2: 9′ 4″ from No. 1; 24½″ l. x 19″ w. x 7″ h. Log, 8½″ dia., extends
  23″; repository, 6½″ from wall, 2″ dia. x 1″ d. Log decayed and siding fallen.

• Pilaster No. 3: 9′ 6″ from No. 2; 24″ l. x 16½″ w. x 10″ h. Log, 8″ dia., extends
  22″; repository, 5½″ from wall, 2″ dia. x 1″ d. Siding had fallen.

• Pilaster No. 4: 9′ 6″ from No. 3; 22½″ l. x 18½″ w. x 9″ h. Log, 9″ dia., extends
  20½″; repository, 3½″ from wall, 2″ dia. x 1½″ d.

• Pilaster No. 5: 9′ 6″ from No. 4; 24″ l. x 20″ w. x 9½″ h. Log wholly decayed
  but was 8½″ dia. and extended 22″.

---

76





• Pilaster No. 6: 9′ 5″ from No. 5; 23″ l. x 20″ w. x 8½″ h. Log, 9¾″ dia.,
  extends 22″; repository, 4½″ from wall, 2″ dia. x 1″ d. This pilaster was best
  preserved of the 6; on its front, 11 coats of plaster totaled 1″ thick. From
  front center to center of No. 1, 9′ 4″.

The south recess is 12 inches deep, leaving a 17-inch shelf between
it and the main wall. Seven feet six inches from that wall is the north
end of a subfloor ventilator duct that measures 16 inches wide and
33 inches deep; it is oriented N. 2° W. and its masonry sides rise to
within 6 inches of the kiva floor. Most of the lintel poles had been
removed and the duct filled and floored over. Among the fill was a
pile of irregular chunks of sandstone, as in Kiva F, providing a
foundation for a block of masonry built upon the sides of the duct.

That masonry block is a 14-inch-high deflector, rather crudely put
together with salvaged building stones, slightly curved lengthwise and
the concave side toward the recess. The deflector is the accompaniment
of an above-floor ventilator whose outlet, 16 inches wide, had
been cut through the back of the recess at floor level to make connections
with the vertical shaft belonging to the subfloor ventilator.
Secondary jambs 5 inches inside the opening were designed to support
a door slab, now missing. Repairs above the shelf are doubtless incident
to construction of the above-floor ventilator tunnel and its airshaft
connections (pl. 19, A).

Two and one-half feet beyond the deflector is an exceptionally large
kiva fireplace, 32 by 37 inches and 28 inches deep. It is floored with
clay, lined with masonry, and plastered. On the east side width is
increased by a steplike offset 3 inches wide and 5 inches deep. Opposite,
on the west side is another, likewise 5 inches deep but with a
4-inch tread. After considerable use the fireplace had been divided
lengthwise by slabs on end and, subsequently, the eastern half was
divided, its northern part continuing in use as an ash receptacle while
the remainder was filled with rock and sand.

The change in ventilating systems was made when the floor lay
1¾ inches lower, for the deflector stands upon this latter. Beneath
is a 3-inch layer of whitish clay and then a fill of adobe and windblown
sand. On the east side, at least, the foundationless bench
masonry was begun upon that sand and adobe fill.

Although its ceiling timbers had been salvaged and its pilaster repositories
emptied, Kiva G did not become an accepted community
dump. Among the fallen stonework within its walls we found only 1
broken metate, 4 manos, a few lesser artifacts and scraps. Of 1,362
miscellaneous potsherds tabulated, 46, or 3.4 percent, were of Mesa
Verde ware.

We attempted no exploratory tests within the Kiva G quadrangle
but our observations nearby suggest earlier utilization of the site. For
example, earlier structures underlie Kiva E, to the west, and the floor
of Kiva G lies 8 feet below that of E.

Kiva H (fig. 19).

Average diameter at floor, 15′; bench, average 11½″ w. by 20″ h.; no pilasters;
south recess in main wall 33″ d. by 7½′ at rear; recess in bench 20″ d. by 4′ 8″
at rear, sides curve in from outer width of 7′ 3″; subfloor ventilator outlet 50″
north of bench recess is 17″ on south, 15″ on west, 23″ on north; east side
opens into masonry-lined duct 21″ d., oriented S. 37° E. leading to vertical shaft
outside east wall of enclosing quadrangle; 4″ north of ventilator outlet is slablined,
clay-floored fireplace 30″ dia. by 4″ d.

Kiva H, 15 feet in diameter, was built inside a former dwelling
that measured 17 feet 8 inches wide and 25 feet 5 inches long. The
kiva masonry abuts three sides of the old house and, perhaps in consequence
of a miscalculation, it was necessary to incorporate the
middle 4 feet of the straight north wall in the kiva curve. Opposite,
at the south end of the appropriated house, otherwise waste space was
filled with debris of reconstruction and three small second-story
chambers built upon it. One of these, Room 56, became a vestibule
for a stairway leading up from Room 55.

The bench in Kiva H is 20 inches high and varies in width from
10 to 13 inches. Its masonry, like that of the main wall, is chiefly of
laminate sandstone, but soft friable sandstone predominates at the
south side and in the south recess. The bench foundation protrudes
5 inches. Successive coats of smoke-blackened bench plaster total
2 inches. At the northwest the kiva wall stands 6 feet above its bench
and that doubtless is close to the original height. Here, a foot or more
above the bench, two small poles had been tenoned into the wall,
tying it to the older stonework. Lacking pilasters, the chamber necessarily
had a flat ceiling supported by logs reaching across from side
to side.

Kiva H is unique among those we examined at Pueblo del Arroyo
in that both bench and main wall are recessed at the south. The main
wall is set back 33 inches at bench level creating a recess 7½ feet
wide at the rear. The masonry of the bench curves in gradually on
either side to form a lesser recess and leave a 20-inch-wide shelf at
the back of it.

Four feet two inches north of the bench recess is the outlet of a
subfloor ventilator tunnel. The outlet is irregular, measuring 15 inches
on the west, 23 inches on the north, and 17 inches on the south; it is
masonry lined, 21 inches deep, and floored with adobe. A large sandstone

slab, 16 by 18 inches and an inch thick, has been set up against
the north side of the opening, 5 inches above its floor. From the open
east side of the outlet a tunnel extends southeast, its north wall

Fig. 19.—Kiva H.

oriented N. 37° W., to pass under the bench and kiva wall at the
corner of the south recess. Here the tunnel is 19 inches wide but this
width had been reduced by two 2-inch posts, 8 inches apart, standing
on the tunnel floor and rising 13 inches above the floor of the kiva,
their upper ends embedded in the bench masonry and plastered over.
Back of the posts was a single lintel pole with a flat slab resting
upon it.

Outside the kiva the ventilator tunnel continues, its south side razed
for some reason, passing under the east wall of the former dwelling.
The tunnel had been cut through the foundation of the wall directly
beneath a blocked T-shaped door and large lintels with improvised
stonework inserted to support the masonry above. The vertical shaft
of the ventilator was built against the east side of this blocked door.

In the middle of the kiva floor a slab-lined fireplace 30 inches in
diameter partly overlies an earlier one measuring 27 by 33 inches and
2 feet deep, lined with sandstone slabs and filled with ashes and sand.
The portion beyond the later, circular fireplace had been floored over.

Our only exploratory pits hereabout were two: in the southeast
corner of the former dwelling to expose the ventilator tunnel previously
described and in the northwest corner where smoke-smudged
plaster still adheres to both walls and where, 45 inches below the
second-story floor level, we came upon a solid mass of rock and mud.

We found Kiva H filled chiefly with fallen masonry and the everpresent
blown sand. There were no artifacts here but, at bench level
and a foot from the south recess, we unearthed a shattered adult skull
and, on the north side half a foot above the bench, parts of a child's
skeleton (field Nos. 465, 466).

Kiva I (fig. 20).

Average diameter at floor, 14′ 10″; bench, 18″ w. by 28″ h.; no pilasters; south
recess, 18″ d. on west, 6″ d. on east, 4½′ w. at rear; subfloor ventilator duct
14″ w. by 25″ d. with 14″-by-17″ outlet 5′ from back of recess, continues 3′ 4″
south to connect with 12″-sq. vertical air shaft; duct filled, floored over, and
replaced by above-floor ventilator cut through back of recess 2″ above floor;
outlet, 12″ w. by 17″ h. to lintels, provided with 3″ secondary jambs 3″ inside
opening; tunnel oriented S. 4° W., connects with shaft of older duct; 12″ from
subfloor ventilator outlet, fireplace 33″ by 43″ and 12″ d., clay floor, masonry
lined and plastered; kiva fill, debris of reconstruction to bench height, debris of
occupation above.

Kiva I stands within the quadrangle adjoining H on the east. Its
walls, built predominantly of friable sandstone, had weathered to such
an extent we refilled the room after excavation, the better to preserve
what remained. Because more laminate sandstone was utilized in the
bench masonry, it is somewhat better preserved. As in Kiva H, there
are no pilasters.

The south recess continues to the main wall, without the customary
narrow shelf at the back of it. An unusual feature at this point is that
from a width of 18 inches most of the way around the bench narrows

down to 6 inches at the east side of the recess. Five feet north of the
recess we came upon the 14-by-17-inch outlet of a subfloor ventilator
duct, 14 inches wide and 25 inches deep, that continues under the
south wall 3 feet 4 inches to its vertical air shaft. The duct had been
stripped of its lintel poles, filled, and floored over, as were those in
Kivas F and G.

Fig. 20.—Kiva I.

Replacing that described above, a lateral ventilator was cut through
the rear wall of the recess 2 inches above its floor. The opening is
12 inches wide by 17 inches to its lintels. On the left, 3 inches inside,
is a 3-inch secondary jamb for retention of a slab door; its opposite
was lost with collapse of the stonework. The tunnel connects with
the vertical shaft of the earlier, subfloor system. This shaft is a foot
square and near the top of it we found a spalled sandstone disk,
15 inches in diameter by 3 inches thick, that may have been the shaft
cover.

There can be no doubt that Kiva I, after abandonment, was utilized
as a neighborhood dump. Debris of reconstruction filled the lower

part, household rubbish the upper. Upon this accumulation and about
2 feet above the north bench, an uneven pavement of limited extent
and a 10-inch-deep firebox evidence temporary occupancy. The firebox
is exceptional for its shape and size. It measures 4½ feet wide as
it abuts the kiva wall and has an intruding 22-by-18-inch angle in the
southwest corner (fig. 21). Three sandstone firedogs were standing
among the ashes in the southeastern extension.

Fig. 21.—Intrusive fireplace on Kiva I fill.

A number of artifacts and discards were recovered from the fill of
Kiva I, including the two stone axes shown in plate 41, f, h. Near the
floor we uncovered the skeleton of a dog and part of another, skull
missing (field No. 487). Elsewhere we found the skull of a badger
(Taxidea taxus; field No. A-582) and several feet away, the mandible.
Part of an infant's skeleton lay among debris of reconstruction 6 feet
from the south wall and about 15 inches above the floor. Among
miscellaneous bird bones recovered were those of the raven (Corvus
corax), a rarity in local trash piles.

An exploratory trench in the northeast corner of the enclosing
square showed the rude, unplastered stonework characteristic of such
corners. Two small poles tied the east side to the kiva curve. A beam
lying close against the north wall, with one end seated upon the east

side of the square, supported a fragment of coarse masonry that is
possibly part of the foundation of a later second-story wall erected
above.

In the northwest corner of the square we unearthed a puzzle. The
northern 4 feet 11 inches of the west wall ends in a distinct vertical
line, like a neatly finished corner, that continues below our 6-footdeep
pit and had once extended above two 9-inch beam holes near the
present irregular top of the wall. This section is composed of relatively
small laminate sandstone blocks, obviously selected with a view
to uniformity of size. The remainder of the west wall to its union
with the kiva curve is made up of larger, often undressed blocks and
is capped with small-stone masonry similar to that in the northern
part. There is nothing to indicate the original height of this capping,
but presumably it equaled that adjoining, which still stands a foot
above the two beam holes. In either case, after the capping was laid,
it and the coarser stonework below were twice plastered—first natural
clay and then a whitened coat—and each, in turn, was extended over
the two layers previously applied on the superior masonry of the
northern 4 feet 11 inches. Larger blocks of dressed sandstone, with
small chips between, predominate in the lower north wall (fig. 39).

Debris of reconstruction and blown sand filled the corner. On that
fill and built against the original north wall is the 20-inch-high foundation
of a secondary north wall—a foundation that rests upon the
irregular top of the superior west-wall masonry and had once continued
westwardly above the roof level of Kiva H. Presumably this
foundation had also extended the other way to join with the fragment
above the northeast corner of the square.

As for the two 9-inch beam holes, it is obvious they mark the positions
of timbers that once roofed either Kiva H or the room in which
it was built. Had they extended eastwardly, a possibility in view of
the finished masonry in the northwest corner, we must suppose a former
dwelling here, one razed upon construction of Kiva I. This supposition
seems implausible for two reasons: (1) In the northeast
corner of the enclosing square both walls are roughly faced, and (2)
the distance, 23 feet, from east to west exceeds the length of timbers
customarily used by the builders of Pueblo del Arroyo.

Like its neighbor on the west, Kiva I lacked pilasters as bases for
a cribbed ceiling. Its roof, therefore, necessarily rested on beams
bridging the above-bench diameter of 17 feet 10 inches.

Kiva J (fig. 22).

Average diameter at floor, 11′ 8″; bench, 8″ w. by 23″ h.; no pilasters; south
recess altered; subfloor ventilator duct 12″ w. by 18″ d. extends north 45″ from

Fig. 22.—Kiva J.

back of recess at angle of N. 21° W., masonry lined, clay floored; filled, floored
over, and replaced by lateral ventilator opening 13″ w. by 16″ h. and 2½″ above
floor of recess; 7″ within opening, it is reduced to 10″ w. by 8½″ h., and floor
raised 1¼″; was cut through solid masonry S. 24° E. to reach vertical shaft of
subfloor ventilator, 12″-sq. intake; fireplace, 15″ w. by 31″ N.-S. by 11″ d., 5″
from end of subfloor duct. Briefly used as a dump.

Smallest of the seven we examined in Pueblo del Arroyo, Kiva J
occupies about two-thirds of a former rectangular dwelling while
Rooms 64 and 65 crowd the remainder. The main wall, which still

Fig. 24.—Bowl fragment from Room 39, inside and outside.

stands to an average height of 7 feet above the bench, is composed of
large blocks of dressed, friable sandstone with casual chinking between.
There are no pilasters; no narrow shelf at the rear of the
south recess. We counted 19 coats of smoke-smudged plaster on the
upper wall, a total of 1½ inches, all being the natural yellowish brown
of Chaco adobe except the second, which had been whitened.

In contrast to that of the main wall, the bench masonry is predominantly
of laminate sandstone. Its rubble core ties with that of
the main wall; its facing extends 4 inches below the kiva floor without
foundation. An additional bond, a pole 2½ feet long, had been embedded
in the stonework on the east side just below bench level. Here,

in the southeast quarter, part of the original bench has been removed
to create an abnormally wide recess (pl. 19, B).

There can be no reason for this enlargement other than planned
alterations that were never completed. Incompleteness is evident from
the rough, unplastered stonework at the back and east end. It is my
guess that the original recess was about 4½ feet wide while its enlarged
successor is more than twice that. The usual narrow shelf at
the back of the recess is lacking here, as in Kiva I. The Kiva J bench,

without pilasters, is 8 inches wide. We found
only one narrower bench and that is in the partially
razed kiva underlying Room 64 wherein
the indicated bench width was 3 inches; height,
25.

Three feet nine inches from the south wall
of Kiva J is the end of a subfloor ventilator
duct 12 inches wide by 18 inches deep; oriented
N. 21° W., it had been stripped of its lintel
poles, filled, and floored over. From the back
of the recess the duct passes under the kiva

wall and apparently bends to the southeast to meet its vertical air
shaft in the corner of the former dwelling.

An above-floor ventilating system supplanted the one described
above. The new outlet, 13 inches wide by 16 inches high, was cut
through the wall of the recess 28 inches from its west side and 2½

Fig. 27.—
Resonator.

inches above its floor. Seven inches inside the opening,
however, dimensions are reduced to 10 by 8½
inches and floor height increases to nearly 4 as the
tunnel continues, more directly than its predecessor,
toward the shaft in the southeast corner.

Half a foot beyond the north end of the subfloor
ventilator duct is a fireplace, 31 inches long by 15
inches wide and 11 inches deep. Its shape is unusual
and so, too, the manner of construction. Both ends
are concave and lined with masonry, while the sides
are of slabs that project a couple of inches above the
floor. Slab fragments partition off 5 inches at the
north end.

Like most of the others, Kiva J had in time been
abandoned and thereafter used as a convenient dumping
place for rubbish. Abandonment evidently followed
shortly after initiation of intended alterations
to the bench and south recess, for these proposed
changes were never completed and all roofing timbers
had been carefully removed and carried away.
Chunks of dried mud from the roof were allowed to
lie where they fell and upon them windblown sand
and household sweepings gradually accumulated.
From among this waste we recovered a number of
discarded artifacts, a cupful of charred corn, and
three large sandstone slabs, presumably doors, which
we left in the kiva. Largest of the three, with rounded
ends and abraded edges, measured 27 by 21½ by 1½
inches; the second measured 24 by 15 by 1 inches;
the third and smallest, 14 by 10 inches, had seen
limited service as a metate despite a thickness of only
½ inch. Under the debris, on the floor in the northeast
quarter, was a fragment of a large copper bell
(U.S.N.M. No. 334766). A long-used dump would
have produced a larger proportion of wornout stone
implements and other evidence of household activities.

A recessed stairway (fig. 7) once led up out of
Room 41 to the roof of Kiva J.

Kiva B is described in the next chapter, together
with its associated rooms, but we may note here, for
comparative purposes, that it was enclosed by walls

built especially for the purpose; that it had an above-floor ventilator
and a narrow, encircling bench with a shallow south recess but no
pilasters. The bench lacks masonry except in the south recess.

SUMMARY

In these eight kivas, including B, one finds similarities and dissimilarities.
All eight are circular in floor plan and masonry lined—
cylinders of masonry sunk into a quadrangle of straight walls re-
claimed from appropriated dwellings
or purposely erected to produce a
subterranean effect. All eight are
provided with the traditional bench
encircling the floor, but four of those
benches bear low pilasters for support
of ceiling timbers and four do
not. All eight have a centrally located
fireplace, and two once had
subfloor vaults west of the fireplace,
subsequently filled and floored over.
The seven kivas within the main
walls of Pueblo del Arroyo were
originally equipped with subfloor
ventilating systems although four
of these were later converted into
ventilators opening above floor level.
Kiva B alone was furnished with
an above-floor ventilator at time of
construction, but B is not a typical
Chaco Canyon kiva.

The masonry of B is quite nonChaco
in appearance; like that of its
associated rooms, it consists of salvaged
rocks carelessly put together.
The Kiva B bench is of earth, left
in place when the pit was dug; its
only stonework is that lining the
south recess. In contrast, most of
the seven kivas we excavated inside
the pueblo were walled with laminate
sandstone intermittently banded. For
these seven, diameter at floor level

Fig. 29.—Sandal effigy of wood.

varies from 11 feet 8 inches to
25 feet 10 inches; bench width
varies from 8 to 29 inches.
Bench masonry invariably excells
that of the main wall. In
two instances, C and J, the
rubble foundation of the bench
merged with that of the wall
above.

In relation to its surroundings,
each of the eight kivas
occupied a simulated subterranean
position whatever its
actual elevation. The five
grouped at the west side of
the court, Kivas F to J, were
so constructed as to bring their
roofs at the second-story level,
while E and unexcavated D
lie a story higher. Kiva C also
was purposely raised until its
flat roof provided a dooryard
for occupants of third-story
houses adjoining.

Each of the eight kivas, including
B, has a south recess
in its encircling bench but these
recesses vary in width and
depth. In three instances, F, I,
and J, the recess extends the

A. Bowl and jar fragments of Transitional ware.

Plate 20

B. Sherds bearing typical Solid-type designs.

A. Bowl and jar fragments of typical Chaco-San Juan ware.

Plate 21

B. Sherds illustrating straight-line hachure in styles A, B, and C.

Plate 22.—Bowls from Room 14 (a), Kiva F (b), Room 28 (c), Room 15 (d-f),
and Room 27 (g-i).

Plate 23.—Bowls from Rooms 28 (a), 43 (b), 32 (c-f), and 39 (g-l).

Plate 24.—Bowls from Rooms 44 (a, b) and 27 (c); bowl and pitcher from burial in
Room 40 (d, e), and miscellaneous vessels (f-j).

Plate 25.—Bowl with polished black interior and an old design in matte paint.

Plate 26.—Three eccentric ladles from Room 27 (a) and Kiva C (b, c).

Plate 27.—Ladles and fragments from various rooms.

full width of the bench, with
no shelf at the rear. In one
case only, Kiva H, the main
wall of the chamber is also
recessed—an echo of the
"deep south recess" in kivas
of Prudden's unit-type ruins
and others north of the San
Juan.

Four kivas, C, E, F, and
G, have low pilasters spaced
at regular intervals upon
the bench as supports for a
ceiling of cribbed logs. All
are rectangular and built
either of squared timbers
thinly coated with mud, as
C and F, or of round logs
walled at the sides with
small-stone masonry and
thickly plastered, as E and
G. In both methods part
of the log was built into
the kiva wall at time of
construction leaving the
remainder thrust forward
upon the bench. Whether
bare or incased in masonry
the pilaster logs did not
extend to the edge of the
bench but stopped about 2
inches short. This was true
even of the partly razed
kiva deep under the floor
of Room 47B, with four
logs in each pilaster instead
of one. On top, nearer the
wall than the forward end,
was a small hole, made to
receive a sacrificial offering

Fig. 32.—Obsidian knife fragment, abraded
and rechipped.

—repositories permanently sealed when the cribbed ceiling was
constructed upon the pilasters. Kivas lacking pilasters, as B, H, I,
and J, were roofed with logs reaching from wall to wall.

Fig. 33.—Tip of
argillite blade.

Ventilation was provided by an external air
intake and an outlet inside the kiva. Air heated
by the central fireplace would rise to escape from
the overhead hatchway, drawing fresh air down
the outside shaft and through a tunnel to the
outlet, thus creating a circulation that seems to
have been adequate.

Each of our eight kivas is furnished with such
a ventilating system. In B the outlet is in the
south recess and above floor level; the other seven
have, or had, their ventilator outlets in the floor
6 or 8 feet in front of the recess. The below-floor
air passage or tunnel is a characteristic of the
Chaco-type kiva. In four of those in Pueblo del
Arroyo, however, F, G, I, and J, this subfloor tunnel had been dismantled
and replaced by a lateral air passage cut through the rear wall
of the recess to connect with the vertical shaft belonging to the abandoned
system. In F and G, masonry deflectors were built in to shield
the fireplace from side drafts, but in I and J comparable results were
apparently realized merely by reducing the size of the outlet.

Although the same data are included in Appendixes B and C, it
seems desirable to brief them here in order that our eight kivas may
be compared the more readily.

                   

	Diameter		Bench			Ventilator
Kiva	On floor	Above bench	Width	Height	Pilasters	Below	Above
B	11′ 6″	13′ 6″	12″	34″	0		X
C	25′ 10″	30′ 8″	29″	22″	8	X
E	14′ 6″	17′ 10″	20″	24″	6	X
F	14′ 11″	18′ 1″	19″	19″	6	X	X
G	18′ 6″	23′ 0″	27″	26″	6	X	X
H	15′ 0″	16′ 10″	11″	20″	0	X
I	14′ 10″	17′ 10″	18″	28″	0	X	X
J	11′ 8″	13′ 0″	8″	23″	0	X	X

Whether or not kiva builders took bearings on the stars when laying
out their ceremonial chambers, no two of ours are oriented alike, the
variation of the subfloor air passage being from N. 10½° E. to
N. 30° W. As regards fireplaces, two are square or nearly so; two
are decidedly rectangular, two are circular, and one semicircular. In
Kiva H a fireplace 30 inches in diameter and four inches deep had
replaced one 27 by 33 inches and 2 feet deep. One hearth is lined
with slabs, three are masonry lined, two are lined with a combination
of masonry and slabs, and my notes are indefinite in the case of two.

Among the seven inside Pueblo
del Arroyo, C is perhaps the
purest example of a Chaco Canyon
kiva. It is the only one having
eight pilasters, to be sure, but
these are thoroughly typical. Its
south recess is less than half the
depth of the bench, leaving a
shelf between recess and wall.
Its subfloor ventilator duct is
paved with slabs, walled with
uniformly sized blocks of laminate
sandstone, and roofed with
selected small poles, a layer of
slabs, and a thick adobe floor. The
duct vents 5½ feet north of the
recess but continues south, under
the kiva wall, an additional 18
inches to meet its vertical shaft.
The subfloor vault in C measures
8½ feet by 3 feet 10 inches by 18½
inches deep, almost twice that in
Kiva F.

With the possible exception of
C, none of the kivas we examined
had been included in the original
plan for the pueblo. Kiva H was
squeezed into a former dwelling
and so, too, the partially razed

kiva under Room 47B.

The depth, or apparent depth, of Pueblo del Arroyo kivas caught
Jackson's eye. One near the east end of the north wing "27 feet in
diameter, was three stories in height. . . . The interior is nearly
filled up, but it was originally over 25 feet in depth." (Jackson, 1878,
p. 443.) In the south wing, the roof of Kiva C provided a dooryard
for occupants of third-story rooms adjoining, but we estimated its
ceiling height at only 10 feet. Apparently floor and walls had been
raised several times.

Subfloor walls, if any, have been considered in our description of
each kiva. There can be no doubt of the degree of priestly or kivagroup
authority after studying the sequence of residential sacrifices
leading to Kiva E, for example: first the original Room 47 was

taken over, then the room next on the north and, finally, the pair
immediately west of these two. Even before the third one had been
completed it was replaced by E, its floor 10 feet above that in the
first of the series.

Fig. 35.—Head of "mountain lion."

Kiva C was at least partially
destroyed by fire; the
cribbed ceilings in E, F,
and G had been deliberately
removed and their pilaster
offerings reclaimed. Ceiling
timbers in Kiva I likewise
were salvaged. As
usual in such demolition
work chunks of roofing
adobe, bark, spalls, and
other waste was allowed to
lie where it fell. In each
instance the depth of the
fallen debris approached
bench height, and upon it
household sweepings from
nearby dwellings soon began to accumulate. The quantity of this
domestic debris was surprisingly limited, however. Potsherds were
always present but the number of bone awls, broken and discarded
stone implements, bird and mammal bones recovered falls far below
what one would normally expect from such dumps. One gathers the
impression that not many people were living thereabout at the time.

Kiva B has a narrow bench with a shallow bench recess at the
south and an above-floor ventilating system. Each of the seven we
examined within Pueblo del Arroyo had originally been equipped
with subfloor ventilators, but four of these were subsequently replaced
by ventilators like that in B. Above-floor ventilators are standard
equipment in P. II-P. III communities north of the San Juan River,
and they are usually accompanied by a deep south recess above bench
level and by a sipapu in the floor north of the fireplace. There are
exceptions but not many. Although the narrow bench without pilasters
occurs sporadically, most northern kivas were provided with masonry
pilasters rising flush with the face of the bench and 2 to 4 feet higher
(Kidder, 1924; Martin, 1929, 1936; Brew, 1946). In ruins with
marked Chaco affiliations to the south, the below-floor ventilator and
the sipapu again occur together (Hodge, 1923; Roberts, 1932).

None of our eight kivas was
provided with a subfloor layer
of shale, a standard provision
at Pueblo Bonito, but such a
layer was present in the razed
kiva partially underlying the
north end of Room 55.

Low pilasters set back an
inch or more, a south recess in
the bench rather than in the
wall above it, a subfloor ventilating
system, and absence of
the sipapu are earmarks of
the Chaco kiva. In none of
our eight at Pueblo del Arroyo,
including B, did we find an
indubitable sipapu. Three, with
benches 8 to 18 inches wide,
are without pilasters. Four
have pilasters on benches varying
in width from 19 to 29
inches, and in each case the
pilaster is less than a foot
high and is set back from the
edge of the bench. From P. II
and Early P. III ruins in the
La Plata Valley, southeast of
Mesa Verde National Park,
Morris (1939) reports a number
of kivas with shallow south
recesses in the bench, some-

times accompanied by a deeper banquette above, and pilasters set back
an inch or two. Unlike those of the Chaco, however, La Plata Valley
kivas usually have the sipapu; the subfloor ventilator is rare.

Northwest of the Mesa Verde, Martin (1930, 1936) reports both
above-floor and below-floor kiva ventilators even in the same ruin. The
associated pottery he describes as "Mancos" and "McElmo" and
sherds of those two wares predominate in late P. II and Early P. III
ruins north of the San Juan and they are conspicuous in our Pueblo
del Arroyo collections. Migrant makers of "Mancos" pottery may very

Fig. 37.—Figurines of clay (a-c) and sandstone.

Fig. 38.—Incised fragment of sandstone.

well have settled at partially vacated Pueblo del Arroyo and been responsible
not only for construction of Kiva B but also for conversion
of the ventilating systems in F, G, H, and I. Both types of ventilators,
shallow recesses in the bench and deep recesses above bench level,
sipapus and no sipapus, occur in Bc 51, a small-house Chaco ruin
(Kluckhohn and Reiter, 1939).