University of Virginia Library


Page 212


Two conspicuous trash piles lie in front of Pueblo Bonito (pl. 3),
and both are enclosed by contemporary masonry. Trash piles have a
peculiar fascination for an archeologist because, to a degree, they
reflect the lives and industries of the people responsible for their
being. Since materials at the bottom would have been discarded first,
a vertical section through such a trash pile should provide a partial
understanding, albeit limited to the least perishable materials, of the
local culture during the period of accumulation. Hence our initial
archeological undertaking at Pueblo Bonito was a stratigraphic examination
of the West Refuse Mound, larger of the two.

I was not aware at the time that such an examination had already
been accomplished. It was mid-May 1921 when I received, through
courtesy of Dr. Clark Wissler of the American Museum of Natural
History, partial page proof of Pepper's Hyde Expedition notes, 18961899,
and still later before I saw the results of Nelson's 1916 testing
of the two Pueblo Bonito mounds (in Pepper, 1920, pp. 383-390).
Nelson's findings are both confirmed and augmented by ours, as
will be seen.


A few days after we made camp I selected a previously undisturbed
site near the crest of the West Mound, cleared away the uppermost
12-14 inches because it was loose and trampled, and dug
a trench to clear sand at a depth of 20 feet (pl. 6, left). Our sampling
was limited to a 3-foot square at the end of that trench, and the
successive layers, irrespective of thickness, were determined by the
materials between. Lenses of ash, vegetal matter, and blown sand
provided clear-cut separations. We were seeking fragments of
pottery, then as now a handy gage of cultural progress, but were not
prepared to find early and late types—sherds typologically different
and distinct—intermixed throughout the full 20 feet. Nor were we
prepared to find, throughout, pieces of sandstone and chunks of
discarded mud mortar in quantities dwarfing village debris. The West
Refuse Mound was not a normal trash pile.

In that first attempt at stratigraphy we recovered 2,118 potsherds
from 23 separate strata (U.S.N.M. No. 334180), and early and late


Page 213
types were associated, top to bottom. Of the total, 820 fragments, or
38.7 percent, were black-on-white, and of these 345 were early varieties
—types that Roberts and Amsden later designated Transitional, Degenerate
Transitional, and Solid—while 48 percent were Hachure B,
a late and preponderant variety at Pueblo Bonito. In Stratum T, an
18-inch thick layer beginning at depth of 15 feet 9 inches, there were
24 Early Black-on-white fragments and 29 of Hachure B. There
were three proto-Mesa Verde fragments among the 2,118—one from
Stratum H and two from N—but none I would consider Classic
Mesa Verde. Only 93 sherds, 4.3 percent of the total, were of Plainbanded
Culinary ware; 36.7 percent were Corrugated-coil. We recovered
five Plain-banded and six Corrugated-coil fragments from
Stratum U, between 17 feet 3 inches and 17 feet 8 inches deep. Such
intermixture was illogical and frustrating.

There are those who still debate the distinguishing characteristics
of "early" and "late" Pueblo pottery. My own 1921 yardstick, a
combination of form and decoration, was based on several seasons'
fieldwork among ruins north of the San Juan River. I recognized as
"early" that pottery assemblage which included Banded-neck cook
pots, half-gourd ladles, squat pitchers with wide mouths, and deep
round-bottomed bowls slipped with white and bearing such familiar
black-paint designs as stepped squares or triangles, key-shaped figures,
interlocking whorls, checkerboard, and waved or squiggled lines.
"Late" pottery, to me, included the bowl-and-handle ladle, Corrugatedcoil
Culinary ware, ollas with down-raking handles, thick-walled bowls
with dotted rims, zoned decoration, and polished surfaces. But in
Chaco Canyon late black-on-white vessels had whiter slips and blacker
paint than those with which I was familiar and an endless variety
of rectilinear figures featuring straight-line hachure. To find these
two groups intermingled throughout 20 feet of village waste was confusing
in itself but the preponderance of constructional debris was
confusion confounded. We cut a second stratigraphic section without
clarifying the puzzle.

In 1922 we sectioned the West Mound a third time and made a
more determined effort in the other. Two years later we tried both
mounds again. In each of my seven attempts at Pueblo Bonito
stratigraphy the results were identical: A preponderance of building
waste intermixed with debris of occupation that contained both early
and late pottery. How this mystery finally was solved has been related
in a previous report (Judd, 1954, pp. 175-177), but, the better
to understand our present subject, I may add that the Late Bonitians


Page 214
were indefatigable builders. They were continually tearing down and
replacing perfectly serviceable structures and the waste from such
activities, along with their daily household sweepings, was carried
out and dumped south of the pueblo. But, as we later learned to
our astonishment, these discards had filled a broad, pre-Bonito
floodway several feet deep before they began to pile up to form the
two principal refuse mounds under consideration.

The Hyde Exploring Expedition in 1896 trenched both East and
West mounds in search for Bonitian burials (Pepper, 1920, p. 26),
and others on the same quest had been there before. Pepper does not
elaborate upon these operations, but his pre-excavation view from
the north cliff (ibid., fig. 3, p. 20) shows both trenches, the larger
crossing the West Mound opposite Rooms 138-140 and, on either
side of it, lesser trenches exposing the north enclosing wall.

In July 1916, two years after his important pioneering studies
in the Galesteo Basin, New Mexico, N. C. Nelson, of the American
Museum of Natural History, sought stratigraphic data from these
same Bonito mounds—the first inquiry of its kind undertaken here—
and his findings were hastily summarized four years later for inclusion
in the Pepper volume (ibid., pp. 383-385). Nelson apparently
took advantage of the two Hyde Expedition trenches for a vertical
cut was to be seen at the side of each in 1920, the year of the
National Geographic Society's Chaco Canyon reconnaissance (pl. 3).

As the one who introduced stratigraphy as a method of archeological
analysis in the Southwest and who elsewhere had built a
foundation for all subsequent research in this field, Nelson was both
surprised and disappointed with the results of his observations in
Pueblo Bonito refuse. He was disappointed because his sherd totals
disclosed less evidence than he had anticipated, and he was surprised,
as I was, by the incredible quantities of refuse from razed buildings.

Nelson's observations were restricted to two vertical columns, each
2 by 4 feet, which he divided arbitrarily into 6-inch layers for the
recovery and study of potsherds. From the first column, in the West
Mound and 16 feet deep, he recovered 1,083 fragments; from the
East Mound column, 11½ feet deep, 1,040 fragments. These he
sorted into four principal groups—corrugated, black-on-white, red,
and shiny black—and then subdivided each lot on the basis of ornamentation.
There were more than 20 varieties in the black-on-white
group alone.

In the upper layers Nelson expected to find only black-on-white
sherds bearing hachured designs, or a combination of hachured and


Page 215
solid. From the lower layers he expected only those with thin,
parallel lines and other early-type figures. To his surprise, however,
there was direct association of the two groups throughout both
columns. Mesa Verde-type sherds appeared first in the middle deposits;
red ware and shiny black, scarce or absent at the bottom, occurred
more frequently toward the top. Corrugated, present from
top to bottom, comprised less than a third of the total sherd count
(Nelson, ibid., p. 384).

This association of early- and late-type pottery baffled me no less
than it had Nelson. There seemed no explanation for the fact that
here, in these two great refuse piles, fragments of globular, shortnecked
pitchers lay side by side with fragments of those having
small bodies and high, cylindrical necks; that pieces of bowl-andhandle
ladles lay juxtaposed with ladle fragments of half-gourd
form; that sherds of bowls and ollas painted with stepped lines and
triangles, volutes and interlocking whorls, checkerboard patterns, and
squiggled hatching should occur with or even above those bearing
designs in straight-line hachure. In addition, there remained the
problem of building waste, unbelievable quantities of broken sandstone
and chunks of wall adobe spread irregularly through household

As stated above, we of the National Geographic Society's Pueblo
Bonito Expeditions were quite unaware of Nelson's earlier studies
here when we undertook in 1921 to learn the sequence of local pottery
development. During our first four summers we made altogether
seven serious attempts toward this end, and each time we were
thwarted by the same inexplicable mixture of unrelated pottery types
and by the abundance of waste from building operations. Not until
1925 when I invited Frank H. H. Roberts, Jr., and the late Monroe
Amsden to join my field staff and take charge of pottery research
was the puzzle solved. Early that season we had extended our West
Mound trench to the ruin and northward through the length of the
West Court (fig. 7). In so doing there were brought to light the
remains of a Great Kiva built in an excavation dug 10 feet deep into
an Old Bonitian trash pile and, nearby, an undisturbed remnant of
that old pile.

Into that remnant Amsden and Roberts sank two yard-square
stratigraphic sections, the first 13 feet deep and the second, 12. A
few pre-Pueblo sherds were found near the bottom, but otherwise all
the fragments recovered from the lower 8 feet were early types:
Banded-neck cook pots; squat, round-bottomed pitchers; half-gourd


Page 216
ladles and water jars with low-neck, if any—the black-on-white
pieces decorated with mineral paint in a variety of rectilinear and
curvilinear designs. Only from the upper third of those two 12-footdeep
sections were sherds recovered that bear design elements in
Straight-line Hachure, a combination of Straight-line and Solid,
Chaco-San Juan and, less frequently, those characteristic of the
Mesa Verde plateau. Thus our West Court Trench revealed both
the sequence of pottery types at Pueblo Bonito and one very obvious
reason for the admixture of occupational and constructional debris
in the two principal refuse mounds.

Early during their residence here the Late Bonitians chose to build
a super-kiva where Old Bonitian housewives had long been accustomed
to throw their household sweepings. That Great Kiva, 50 feet
in diameter and 10 feet deep, served its purpose for a time and then
was completely dismantled. Thus, in turn, approximately 39,000
cubic feet of Old Bonitian rubbish were removed to provide space
for the kiva and subsequently, when the kiva was razed, all building
materials not suitable for reuse were carried away and discarded.
Much of this waste went to the south refuse mounds where, with
like material from other sources, the two piles gradually increased in
height as they spread out laterally—east, west, and south.


In successive efforts to confine these mounting accumulations and
to limit their dispersal, the Late Bonitians erected rock fences or
barriers around both mounds—fences that were raised or replaced
as necessity required. But the interesting fact is that neither enclosure
was begun until village waste had filled the old watercourse
fronting the pueblo and piled up 7 or 8 feet above the valley floor as
it existed at the time.

The retaining walls about Pueblo Bonito's two principal refuse
mounds are not unique but they are more extensive, and more purposeful
than any other known to me (fig. 23).[1] Architecturally they
vary from nondescript stonework faced on the outside only to good
second- and third-type masonry finished on both sides. Steps for the
convenience of burden bearers led over the wall nearest the pueblo.

Because a sand-filled trench showed that the West Mound enclosure
had been examined previously we turned our attention to that

No Page Number

Remnants of earlier second-type masonry walls were buried at the northeast corner of
Kiva G enclosure.


Plate 70

The north bench recess of Kiva G and, to left of it, crude masonry resembling
Old Bonitian stonework.

(Photographs by Neil M. Judd, 1922.)

No Page Number

Great Kiva A and its surroundings after excavation. The West Mound trench is
visible at middle left; stones were piled for wall repairs.


Plate 71

The south pillars, vaults, and other features of Great Kiva A were buried under
water-soaked sand and silt.

(Photographs by O. C. Havens, 1921.)

No Page Number

Two masonry pillars and adjoining vaults occupy the north floor of Great Kiva A.
Above, stones piled for wall repairs.


Plate 72

Five masonry steps with wooden fore-treads led from the north benches of
Great Kiva A to Room 148.

(Photographs by O. C. Havens, 1921.)

No Page Number

Plate 73

Left: Repairwork
in progress,
northwest arc of
Great Kiva Q, with
fireplace and west
vault in the foreground.

(Photograph by
O. C. Havens,


Right: A 4-foothigh
built in a dug hole
and packed about
with shale supported
the northeast
pillar of Great
Kiva A.

(Photograph by
O. C. Havens,

No Page Number

Great Kiva A from the north, showing fireplace, deflector, south pillars, east and
west vaults. Stones for wall repairs are stacked in Rooms 217-218.

(Photograph by O. C. Havens, 1921.)


Plate 74

The floor of Room 336 concealed older, partially razed walls and was itself covered by
an intermural kiva, subsequently demolished.

(Photograph by O. C. Havens, 1924.)

No Page Number

Plate 75

Upper: Overlooking the south side of Great Kiva Q was a small alcove with 4 steps to
West Court level.


Lower: The south pillars, fireplace, side vaults, and 3-pole ladder of Great Kiva Q.

(Photographs by O. C. Havens, 1924.)

No Page Number

Plate 76

Upper: Western flight of steps over north retaining wall, East Refuse Mound.


Lower: East steps, East Refuse Mound. North retaining wall shows behind workman.

(Photographs by O. C. Havens, 1924.)

No Page Number

Plate 77

Upper: The northeast corner of a second Late Bonitian retaining wall, East Refuse Mound.


Lower: A walled area at the northwest corner of the East Refuse Mound was occupied by a
crude stone circle, purpose unknown.

(Photographs by O. C. Havens, 1924.)

No Page Number

Fig. 23.-The South Refuse Mounds with enclosing walls and
transverse trenches. (From the original survey of O.B. Walsh.)

No Page Number


Page 217
surrounding the East Mound. Here a rock wall originally 206 feet
long but now much disintegrated, screened the north side. At 80 feet
from its west end a flight of eight stone steps led up and over
(pl. 76, upper). The lowermost step, outside the barrier and without
foundation, measured 28 inches wide with a 24-inch tread; the
upper steps averaged about 15 inches wide and were protected from
encroaching debris by casually placed end-stones. At this stairway
the enclosing wall is double: an outer one of third-type masonry
18 inches thick and now 33 inches high built against second-type
masonry 24 inches thick.

A second series of eight stone steps crosses the same wall 85 feet
farther east (pl. 76, lower). In this case the retaining wall, 37 inches
high on the east side of the steps and 28 inches on the west side,
clearly had been breached to allow for the crossing. The upper steps
average 25 inches wide and, like those of the first flight, rest directly
upon mound refuse and are screened at the ends by dry-laid stones.
The bottom step, 43 inches wide with a 22-inch tread, lies outside
the wall and upon 4 inches of blown sand that covers a hard and
fairly uniform adobe surface.

Excellent third-type masonry, built upon a 12-inch foundation and
still 5 feet high, stands at the northeast corner of the enclosure
(pl. 77, upper). The corner foundation, in turn, rests upon an adobe
surface apparently continuing from that below the second series of
steps, 35 feet to the west. However superior the stonework at the
corner, it deteriorates rapidly southward along the east side: First,
two to four carelessly laid upper courses; then carelessness to a
greater depth (pl. 78, left). Halfway along its length the wall takes
a 5-foot outward jog, then south again to a low barrier, an apparent
improvisation, that extends east about 30 feet and there is
reduced to two uncertain courses. The opposite end of this extension
is lost under the earth dam of Wetherill's reservoir, but
farther west a few salvaged sandstone blocks again provide a rude
barrier (pl. 78, right).

The west end of this East Mound enclosure was originally constructed
of second-type masonry. Subsequently a nondescript addition
of three or more courses was piled on top. Still later, after the
middle section of this addition had collapsed outwardly, a new and
equally nondescript substitute was built 3 feet outside the original
(fig. 23). At the new northwest corner a rock-fenced area was provided
for some unknown purpose, its adobe floor dipping unevenly
toward the west and the middle of it occupied by a crudely built
stone circle 45 inches in diameter (pl. 77, lower).


Page 218

An exploratory north-south trench at this point (fig. 24, b) revealed
stratified village debris sloping down from east and west.
To keep it in bounds, retaining walls had been erected at the near
end of each mound but later debris had overflowed them. Subsequently
both walls were raised a few courses and were again overflowed.
Nevertheless, under these accumulations and apparently on
the same adobe stratum as that in the fenced area noted above, we
happened upon evidence of limited, unscreened domestic activity—
a slab-lined fireplace 19 inches below the surface and 15 inches above
the base of the West Mound enclosure.

At 85 feet south of its northeast corner this West Mound enclosure
is abutted by a foundationless, nondescript wall 19 inches high
by 22 inches wide—a wall that extends east and south until lost among
broken stonework projecting from the East Mound. Although some
of these fragmentary walls may have been built to check the outward
spread of mound rubbish others clearly were designed to confine
floodwaters and direct them past the village. Here, again, the accumulated
floor sweepings, debris of demolition, and wind-blown
sand slope down and away from the mound crest.


In preparing this drawing from two separate surveys an error of approximately
3 feet was disclosed. This difference has been adjusted in the space
between mounds.


Our West Mound Trench (fig. 7), which largely provoked this
chapter, held many surprises. Foremost, as previously explained,
was the admixture of early and late pottery and the presence of
constructional waste throughout. Our stratigraphic columns of 1921,
1922, and 1924 were each cut to clean sand at a depth of 20 feet or
thereabout (pl. 6, left). Early in 1925, still seeking an explanation
for the confused nature of the mound debris, we extended the trench
northward to Room 136 and thence through the West Court to
Kiva Q. Later in the season, at Kirk Bryan's suggestion, it was
extended southward to intercept his buried channel, a subject of
much interest that summer (Bryan, 1954, p. 33).

In this 2-way extension we came upon evidence of a former eastwest
watercourse, an open gully under the south rooms of the pueblo,
that invited the dumping of village waste. As this waste increased
in bulk it not only filled the one-time channel but forced its transient
floodwaters southward into a succession of substitute channels.

Between mound and ruin our trench averaged 11 feet deep. The
bottom of it was a bed of clay-streaked sand with a scattering of
gravel. The dumping of village rubbish began early and on top of
one early pile we cut through a small, clay-lined hearth, made to


Page 219
meet a temporary need (fig. 7, above Sta. 158). Village debris extended
both north and south, some of it continuing northward under
Room 136 as though dumped there intentionally to provide space for
additional rooms. This thought may seem fantastic, but, whether
intentional or otherwise, it is a fact that additional space was

At 45 feet from Room 136 our trench crossed the north side
of the West Mound enclosure. This clearly was an attempt to confine
the accumulating village waste, and a tardy attempt at that. The
lateness of the effort may be judged from the fact that 8 feet or
more of such debris had collected here before the Late Bonitians
began the barrier. They began it with their well-known secondtype
masonry but later, when the need for more height arose,
third-type stonework was employed. The base of this replacement
lies 2 or 3 feet above floor level of Room 136 and about 4 feet above
the present plain in mid-valley which, as noted elsewhere, lies from
2 to 5 feet above that in existence when Pueblo Bonito was inhabited.

Between Stations 95 and 115 is a purposeful dump of constructional
debris that has been eroded, both north and south, by water
action. Across the top of this pile is a puzzling silt surface for which
I have no convincing explanation unless it be that chunks of adobe
mortar among the discards had been rained on and trampled and then
left bare for a considerable period. Other such surfaces exposed by
our trenching operations were, in general, far less extensive.

Two masonry walls at Stations 35 and 50, the latter the earlier
of the two, clearly were erected to confine channeled floodwaters
(fig. 7). Layered sand, clay, and gravel on the north side of each
barrier establish the recurrent presence of running water. To the
right of Station O, strata of laminated and clay-streaked sand
slope down and toward the south, there to be lost among the unconformities
of Bryan's buried arroyo. Farther to the right, between
-25 and -45, miscellaneous potsherds were recovered at depths of
9 to 10 feet in the refill of that ancient arroyo. Beyond the end
of our drawing and a short distance to the east more sherds, including
Late Bonitian types, were recovered in Test Pit No. 3, 18 feet 3
inches deep, as described by Bryan (1954, p. 58).

To repeat, I made an unsuccessful stratigraphic test in 1922 at
the upcanyon end of the East Refuse Mound. Later a second test
was attempted at the crest of the mound and about 75 feet west of
the first but it, too, proved a failure. However, this second effort
brought to light an entirely unsuspected midmound feature, a partially


Page 220
razed and buried wall of excellent second-type construction,
23 inches thick (fig. 24, c above Stations 85-105).

We lengthened our test pit to expose more of this wall and found
a neat corner 20 feet from the inner north side of the enclosure (pl.
79, left). The wall had been built upon a compact adobe surface
that dips west 10 or 12 inches in our 5-foot-wide trench, but continues
southward above the diversified fill of a former broad watercourse.
When razing the corner the Bonitian demolition crew allowed
more waste to fall inside the angle than out.


In 1925 we extended this East Mound Trench in both directions,
south to Bryan's "post-Bonito channel" (Bryan, 1954, pp. 33-36) and
north to the foundations of Room 171 (fig. 24, c). Between mound
and ruin the trench averaged 10 or 11 feet deep and was irregularly
floored with clay-streaked, stratified sand. This water-laid deposit
at once identified itself as part of the same east-west watercourse
profiled by the West Mound Trench. Here, as there, the former
channel had become a common dumping place at an early date and
gradually was filled with occupational debris from Bonitian dwellings
and with waste from razed walls. Some of this channel-fill
extended north beneath the ruin and south beneath the mound.
Throughout, from bottom to top, successive layers of silt and windblown
sand remain to evidence the passing years.

Ten feet below the original surface and beneath a curious body of
compact, sandy clay interspersed with clay pellets, occasional potsherds
and sandstone spalls, we came upon a shallow, ash-filled
hearth (Station 125, fig. 24, c). It was larger and more conspicuous
than the hearth we bisected above Station 160, West Mound Trench,
but, like the latter, had been built on dry, stratified sand for limited

Our East Mound Trench ended against the south wall of Room
171, the foundation of which rests 4 feet below the apparent surface
at time of abandonment. A layer of constructional waste immediately
beneath the foundation extends southward a short distance and then
is buried under a greater quantity of occupational refuse. A foot
lower, or 8 feet below the probable original surface, our trench-end
halved a second, clay-lined ash-filled hearth, this one measuring 34
inches wide by 4 inches deep (pl. 79, right).

Silt-streaked, water-laid sand spread out beneath the hearth; debris
of demolition, above. A dozen feet to the south against the eroded


Page 221
bank of a small sand- and gravel-filled gully, was the discarded fragment
of a troughed metate, relic of the Late Bonitians. At intervals
throughout the 10-foot fill, village waste lay in more or less isolated
bodies as though gathered at one place and dumped at one time.
Some of this material showed unmistakable signs of water action.
On top of all lay broken masonry fallen from the upper stories of
Room 171, the uppermost courses cast outward 25 feet or more.
Partly overlying the outer limit of this fallen stonework, and therefore
of more recent origin, was a quantity of burned sand, sticks, and
cedar bark.

There can be no doubt as to the actuality of the former watercourse
exposed by our East and West Mound trenches. Depth is
comparable in both cases and the materials of the fill are the same:
wind- and water-borne sediments and debris carried out from the
village. The East Mound was already 8 or 10 feet high when some
debris-conscious individual conceived the idea of a retaining wall.
Eventually the mound accumulations crested out 10 feet higher, 12
feet or more above the present level of the plain. Our 1922 stratigraphic
column at Station 64 (fig. 24, c) was dug to a depth of 8 feet
below the plain without reaching the bottom of underlying constructional
debris, clearly reworked by water action. A nearby 1924 test
(U.S.N.M. No. 334181), 10 feet 7 inches deep, ended upon the
adobe layer at the base of the partially razed, second-type wall previously
described. Nevertheless, sherds from these two columns
show that the East Mound is later than its companion since earlytype
fragments are noticeably fewer in number. Also, the proportion
of constructional debris seemed larger—a factor that led Nelson
(in Pepper, 1920, p. 385) to identify the East Mound as the later
of the two.

In contrast to the north retaining wall that on the south is a makeshift
affair built upon what appears to be a man-made heap of
jointed sandy clay piled above an adobe bank at Station 50 (fig. 24, c).
Both the bank and the clay pile have been eroded, north and south.
On the north, water-laid sand and silt mark the course of a 30-footwide
waterway filled with village waste and a southward-sloping
mass of compacted sand containing numerous clay pellets and a scattering
of sandstone spalls. Then follows a round-bottomed channelfill
of laminated silt, a sort of mud conglomerate balled and soaked
by water. Beyond the barrier, in the midst of nonconforming beds
of silt, sand, and village rubbish, is another apparent watercourse
probably antedating Bryan's post-Bonito channel.


Page 222

Sand is everywhere in Chaco Canyon and floodwaters leave silt
layers. But it was the sheer bulk of village debris rather than silt
and sand that defeated our early efforts to read stratigraphy in
Bonito's south refuse mounds. We dug a deep trench through
each—two trenches that laid bare the composition of both mounds
and the manner of their development. Laminated sand, gravel, and
round-bottomed waterways show where water once flowed. The
leveling influences of wind and water are everywhere apparent.
Pockets of blown sand occur throughout the two piles; silt streaks
and puddled adobe remain as evidence of seasonal showers. And as
this village waste continued to pile up, sandstone walls were built
to curb its dispersal. Walls nearest the pueblo were highest and
strongest. End walls, wherever we examined them, were noticeably
weaker and those on the south were weaker still.

It is the presence of these enclosing walls as much as mound content
that determines the age of the two south refuse piles. Both content
and walls are primarily products of Late Bonitian industry.
Old Bonitian housewives habitually dumped their sweepings immediately
in front of their dwellings—at least until the Late Bonitians
took over. A mixture of Old Bonitian and Late Bonitian rubbish
from bottom to top of both mounds fixes their beginnings as subsequent
to arrival of the Late Bonitians. It was this latter group that
filled the old floodway immediately south of the pueblo, a fact evidenced
by the abundance of their distinctive Hachure B. pottery
which Roberts found in two 5-foot-wide stratigraphic sections, 3
and 4, between Room 136 and the West Mound. It was the Late
Bonitians that built retaining walls about both mounds and then continued
to pile up village waste until it overflowed the barriers and
attained a height of 20 feet or more.