University of Virginia Library

VI. POTTERY

Our knowledge of Pueblo del Arroyo pottery derives from studies
made at Pueblo Bonito. Here, four years after inauguration of our
explorations and two years after beginning work at Pueblo del Arroyo,
I called in Frank H. H. Roberts, Jr., and the late Monroe Amsden
and delegated to them the colossal task of analyzing the sherd collections
from both ruins. Two yard-square columnar cuts through 12
feet of previously undisturbed household rubbish under the West
Court at Pueblo Bonito revealed the local pottery types and their
sequence.

Practically all black-on-white potsherds from the lower 8 feet of
Test II, to limit these remarks, were types Amsden and Roberts called
"the Transitional" (because it looked to be a transition between prePueblo
and Early Pueblo wares) and a decadent variation, "the Degenerate
Transitional." However, in the upper 50 inches of that same
test there were 43 fragments (3.1 percent of all black-on-white sherds
from the 12-foot-deep cut) ornamented with straight-line hatching
and 31 fragments (2.3 percent) of a type designated "the Chaco-San
Juan."

Basing their classification on ornamentation, shape, and surface
finish, Roberts and Amsden separated our Pueblo Bonito potsherds
into 21 categories. However, in my earlier reference to their findings
(Judd, 1954), I focussed attention upon eight types only: Transitional,
Early Hachure, Solid, and Plain-banded culinary as representatives
of the Old Bonitians, the founders of Pueblo Bonito; Late
Hachure, Chaco-San Juan, Mesa Verde, and Corrugated-coil culinary
ware as equally representative of the Late Bonitians. Since no fragment
of Early Hachure was recovered below the upper 50 inches of
that 12-foot-deep rubbish pile I erred in placing it with Old Bonitian
types. I erred also in giving Mesa Verde ware the weight of a local
type rather than identifying it as a late and limited import. I should
have recognized the Degenerate Transitional for what it truly is, an
important derivative of Transitional and a long-lived product of the
older portion of the population.

Paraphrasing Roberts's field notes, our Transitional black-on-white
(pl. 20, A) was at first thickly slipped and well polished prior to
ornamentation. Later, scrapers and hand-smoothing gradually replaced
polishing stones; slips became thinner and thinner and finally

were omitted altogether on bowl exteriors. Ladles were of the halfgourd
type, pitchers were rather squat and wide mouthed, ollas were
round bottomed, tall and egg shaped, with a low neck and often with
a bulge between neck and shoulder. Bowls were deep, approaching
the hemispherical, with rounded bottoms, and direct, tapering rims.

Bowl, pitcher, and jar lips were ringed with a black line in which
there was an appropriately located line break. This interruption, sometimes
referred to as "the spirit path," is rarely overlooked even today.
Ruth Bunzel writes (1929, p. 69): "The Zuñi potter who closes the
circle or `road' around her pot feels that her own life road will end,
and she will die."

Decoration, in iron oxide paint, was generally banded in bowls,
zoned on pitchers and ollas. Design elements, in either curvilinear
or rectilinear composition, included stepped lines and triangles, interlocking
whorls, squiggled hachure, and ticked lines often running past
corners. Degenerate was merely an uninspired outgrowth of the
Transitional, with the same forms and finish, and the same decorative
designs carelessly accomplished. Together, Transitional and Degenerate
Transitional were not only the principal pottery products of the
Old Bonitians but the predominant ware of Pueblo Bonito.

The Solid (pl. 20, B) first appeared as a contemporary of the Degenerate
Transitional, with like shapes and surface treatment but with
new ventures in embellishment. Its designs, mainly precursors of
Early Hachure, favored broad, heavy lines and pennantlike elements,
usually in an all-over pattern and sometimes balanced with Early
Hachure figures. Indeed, the Solid was a bridge between the Degenerate
Transitional and Hachure A, but its popularity was relatively
brief. Production had ceased before introduction of Hachure C.

The famed hachured pottery of Pueblo Bonito, sherds of which
were not found below the upper 4 feet of the old subcourt debris heap,
developed in three stages (pl. 21, B). At first composing lines
were widely spaced and often heavier than their frames (A). Subsequently
this practice was reversed and framing lines became the
heavier (B). Bowls continued round bottomed with direct, tapering
rim and black-painted lip; ladles of the half-gourd form persisted but
the bowl-and-handle type began to appear; pitchers developed smaller
bodies and longer necks, with an emphasized line break inside the
neck.

An initial preference for curvilinear designs gave way before the
rectilinear; shaded tips and filled-in corners and solid elements shadowing
the hachured figures were favored successively. Hachure B,

with its heavy frames and lighter composing lines on a white or
cream-colored slip, has long been recognized as the most distinctive
of Chaco Canyon types. Next to the Transitional Degenerate group
it was the most abundant variety at Pueblo Bonito.

Hachure C, identified by its finer composing lines and more expert
brushwork, was a late creation in the occupancy of the village. New
vessel forms were introduced, and the cylindrical vase, presumably
created during Hachure B times, reached perfection in the Hachure C
period. Pitchers with absurdly small bodies and very long necks were
another outstanding shape. Ladles were entirely of the bowl-andhandle
form. Hachure C differs from its predecessors not only in its
superior brushwork and its thinner, closer composing lines, but also
in the complete absence of shaded tips and solid elements in its design
composition.

These several black-on-white pottery types, from the Transitional
to Hachure C, have certain features in common. They are thin-walled
and hard, with preponderantly sherd-tempered pastes blue-gray to
dirty white in color, hand smoothed and slip washed, decorated with a
mineral paint and fired in a reducing temperature. Rims are direct
and tapering and invariably painted black. This black-painted lip,
with a line break usually emphasized, is one of the diagnostics of Chaco
Canyon pottery. In doubtful cases I have identified iron oxide paint
by the simple magnetic test proposed by Colton (1953, p. 21).

Our controversial Chaco-San Juan variety (pl. 21, A) was so
named because it appeared to combine certain Chaco techniques of
manufacture with decorative designs more at home north of the Rio
San Juan. Like Hachures A-C it was not represented in the lower
8 feet of Tests I and II but did occur in the overlying 4 feet. While
some vessels were left unslipped on the outside, most exteriors were
surfaced with a conspicuous yellowish- or grayish-white slip and
stone polished. Bowls predominate in this group, and bowl interiors
are usually decorated with a repetition of the same unit, banded and
framed as on proto-Mesa Verde ware (Kidder, 1924, p. 67). Bowl
exteriors remain plain except, perhaps, for a belt of watered slip
around the upper half and a smear of slip paint across the bottom.
Unlike local black-on-white types, however, the Chaco-San Juan was
decorated with organic paint, a northern trait.

Shape provides another tie with the north. Chaco-San Juan bowls
are shallower than local types with straighter sides, flatter bottoms,
and rounded lips that are boldly ticked and provided with a line break

that is not only emphasized but often overemphasized. Jars tend to
be squat and, like large bowls, are normally equipped with downraking
handles; pitchers have small bodies and high, cylindrical necks as do
those of Hachure C. Pitchers and jars usually have concave bases.
The flat-bottomed, slope-sided mug, a proto-Mesa Verde form, was
introduced during this period. Irrespective of shape, Chaco-San Juan
vessels have thicker walls than do their Chaco counterparts.

Several years ago, to compare Pueblo Bonito pigments and tempers
with those in La Plata Valley pottery, Miss Anna O. Shepard examined
all sherds from two of our stratigraphic tests (II and IV)
together with representative lots selected by type. Her observations
have since been published (Shepard, 1939, p. 280; see also Judd,
1954, pp. 181-183, 236-238) but leave us with an intriguing problem,
the presence of sanidine basalt as a tempering agent. This igneous
rock, not known to occur in the Chaco country, was found as inclusions
in the sherd temper of our three hachured varieties (A to C) but
not in that of earlier mineral-paint types (Degenerate and Solid). It
also appeared secondarily in the organic-paint, Chaco-San Juan group,
a type predominantly sherd tempered but sometimes with bits of
quartz, feldspar, or other rock added. In contrast, sanidine basalt
appeared as a direct, primary temper in much of the culinary ware,
the proportion varying from about a quarter of the pot fragments
from the lower 8 feet of Test II to half of those from the upper 4 feet.

Since sanidine basalt occurs as a primary temper in Pueblo Bonito
culinary wares and secondarily in both mineral- and organic-paint
black-on-white types, Miss Shepard reasons that the Bonitians probably
imported their cooking pots from the Chuska Mountains, 50 airline
miles to the west, the nearest recorded source of the rock and an
area wherein sanidine basalt was the customary temper for pottery.

With one exception, the pre-Pueblo, the 21 pottery types Roberts
and Amsden named at Pueblo Bonito are also represented at Pueblo
del Arroyo. Outwardly there are no visible differences in sherds from
the two ruins, and the data Miss Shepard obtained from our Pueblo
Bonito collections should also apply to those from Pueblo del Arroyo.
The latter ruin is the younger of the two, however, and consequently
fewer early-type sherds were found in it.

After eliminating all recognizable duplicates, Amsden and Roberts
tabulated 203,188 potsherds from Pueblo Bonito and 77,405 from
Pueblo del Arroyo. Among these the eight categories I selected to

distinguish between the older and later elements in the Bonito population
occur at the two sites in the following proportions:

                 

Pottery type	Pueblo Bonito	Pueblo del Arroyo
Transitional	5.6	0.1
Degenerate	3.7	0.4
Solid	5.0	1.8
Plain-banded	14.7	0.4
Early Hachure	7.6	4.5
Late Hachure	1.3	3.7
Chaco-San Juan	6.6	8.6
Corrugated-coil	33.5	54.7

The number of restorable earthenware vessels (118) recovered
during our excavations at Pueblo del Arroyo is surprisingly large in
comparison with other classes of artifacts. Some had been broken and
discarded with floor sweepings; others had fallen with collapse of
upper stories. Only three were complete when found. Thirty-three
came from six ground-floor rooms wherein household waste was
present in quantities justifying the designation "dump." Fourteen
restorable vessels were recovered from Room 15, 26 from Room 27,
and most, if not all, had been in the second story and were shattered
when its floor gave way. Twelve corrugated pots stored on the floor
of one-story Room 65 were crushed when the roof caved in; six
more, in fragments, were tossed in later with sweepings from nearby
homes. For comparative purposes these diverse utensils are most satisfactorily
considered according to shape.

Bowls.—Our 27 black-on-white bowls vary in diameter from 5 to
13½ inches and in depth from 2 to 7¼ inches. Averages are 10⅕ and 4⅖
inches, respectively. Nearly half are listed in our Chaco-San Juan
classification although several fall short of the Roberts-Amsden specifications
for that group. Brown-with-polished-black-interior and
black-on-red bowls were few in number. Sherds of foreign wares
were still fewer.

Figures a, b, plate 47, show two restored bowls from Room 5, one
of the casual structures abutting the outer south side of the pueblo.
Both are ornamented with vegetal paint; both are Chaco-San Juan
types. The first, apparently hand smoothed, has a rounded rim that is
irregularly ticked. In contrast, the second (U.S.N.M. No. 334620),
with its flattened, undecorated rim, is stone polished inside and out.

In shape at least the bowl interred with the Room 40 burial (pl. 24,
d) is Chacoan, but its dark-gray pigment is organic; as is that on the
accompanying Chaco-San Juan pitcher (pl. 24, e).

Among 14 vessels recovered in Room 15 are 3 black-on-white bowls
(pl. 22, d-f). One of these, e, a Chaco-San Juan creation, presumably
had been broken before collapse of the ceiling, since a third of its
fragments were not present. The other two, d, f, are local products
and bear characteristic Solid and Hachure B designs in iron oxide
paint.

Four bowls, all of Chaco Canyon manufacture, came from Room
27, noteworthy for the number of its earthenware vessels. Three are
decorated with iron oxide pigment: two in the familiar Hachure B
style (pl. 22, g, h,) and one, i, in Solid. Despite its blackened lip, the
first of these three, reflects northern influence through its better-thanusual
external finish and parallel brush lines pendent ¾ of an inch on
opposite sides. Finger-grip handles 2½ and 1¼ inches wide, respectively,
will be seen in h and i. Those on the former are cupped on
the under side. The fourth Room 27 bowl is brown-with-polished
black-interior (pl. 24, c).

In Room 32, in addition to the stone artifacts enumerated on page
25, we found four restorable bowls. Three are Chaco Canyon types
(pl. 23, c, e, f); the fourth, a Chaco-San Juan. This latter (pl. 23, d),
although ornamented with organic paint, has the wide-sloping sides
and blackened lip of Chaco bowls. Furthermore, the rim is thinner
and the lip less rounded than in more typical Chaco-San Juan vessels.
Strap handles, pressed in at the middle and fused to the vessel wall,
were attached 1½ inches below the rim.

The three Chaco Canyon bowls from Room 32 are decorated with
mineral paint and in three different but contemporaneous styles:
Hachure A (pl. 23, e), Degenerate Transitional, f, and Hachure B, c.
On the first of these, three parallel ¼-inch-thick rolls of clay, pressed
in at the middle, form double-loop handles. The pigment employed
on f is largely of a reddish hue—the result either of some peculiarity
of the mineral, improper preparation, or carelessness in firing.

Room 39 was one of those utilized as a dumping place. In the 4 feet
of blown sand and floor sweepings that had accumulated before the
ceiling fell we found a number of stone artifacts and 10 restorable
earthenware vessels, including 6 bowls. Three of these latter (pl. 23,
g-i) are decorated with mineral paint in Solid style. Their rim edges
are painted black, but only on the smallest is a line break positively
identifiable. The other three, j-l, fall in our Chaco-San Juan classification,
although k, in design composition and surface treatment, more
nearly approaches Classic Mesa Verde ware. Organic paint was used
on all three. The lip of j, although somewhat flattened, is blackened

after Chaco Canyon custom while the other two have rounded rims,
ticked at short intervals.

Among the 1,256 miscellaneous potsherds tabulated from Room 39
is a bowl fragment decorated inside and out with dissimilar, rectilinear
Hachure B designs (fig. 24). The design on the inside was
less carefully drawn and with thinner paint. Paired holes, perhaps for
a carrying cord, were punched through while the clay was still plastic.
Two fine Hachure B pitchers from Room 39 are shown on plate
28, i, k.

Room 43 was another dumping place from which we removed an
assortment of stone implements but only two restorable pieces of
pottery, both bowls. One of these is the small black-on-white porringer
seen on plate 23, b. In typical Chaco-San Juan style it is slipped inside
and for half an inch below the outside rim, decorated with organic
paint and carefully burnished over the decoration. The rounded,
slightly-incurved rim was ticked, and some of the paint was allowed
to run down irregularly over the slip band. Three conical, unperforated
lugs segment the upper outside wall.

Our second Room 43 bowl is that on plate 54—black-on-red inside
and thick red-on-creamy-white outside. It was an import from the
Houck area, east-central Arizona (Roberts, 1932, p. 112). Morris
(1917, p. 179) found a companion bowl in Aztec Ruin. Trade in
pottery and in the materials for pottery manufacturing, and gifts from
visiting friends are age-old Pueblo customs.

Architecturally, Room 44 was one of the most interesting in the
pueblo. It had three successive occupancy levels, the last two separated
from their predecessors by purposeful fills of sand and household
rubbish. Discarded artifacts and potsherds were collected from
each fill but only one restorable vessel, the pitcher shown on plate 28, h.

Three bowls recovered above the third and final floor in Room 44
merit comparison. One is a typical Chaco-San Juan specimen—darkgray
organic paint, flattened rim closely ticked, and a band of thin
slip outside (pl. 24, a). A companion piece, b, has the characteristic
Chaco finish and tapered rim but without the traditional black rimline.
On one side, pressed so close as to leave no opening behind, is a
pair of luglike strap loops, more decorative than useful. On either
side of these loops there remains an inch-wide smear of thin, black,
nonmetalic paint—the only trace of pigment on this particular specimen.
The bottom is much flatter than on the average Chaco bowl.
Drilled holes evidence ancient repairs.

A third bowl from late debris in Room 44 is illustrated on plate 25.

It is brown with glossy black interior, stone polished inside and out.
It is slightly flattened on the bottom and has a rounded rim. Our
prime interest, however, is in its ornamentation—a linear design in
matte paint that is typologically much older than the surroundings in
which the fragments were found. We do not know just when and
where this peculiar watery sort of pigment was first employed, but
800 years or more after Pueblo del Arroyo fell into ruin Maria and
Julian Martinez of San Ildefonso pueblo rediscovered the secret of
its preparation (Guthe, 1925, p. 24).

Our bowl had been broken while in use, repaired, and again broken.
We recovered its fragments from four separate rooms—the major
portion from 44, lesser sections from 39, 43, and 47. Five ceiling
timbers from these four dwellings gave cutting dates between A. D.
1066 and 1086. Fragments of similar vessels were unearthed during
our explorations at Pueblo Bonito.

Because there were three successive floors in Room 44, each overlain
by household debris, it is instructive to note the results of our
sherd analysis. After eliminating all recognizable duplicates, Roberts
and Amsden tabulated 969 miscellaneous fragments from the lowest
fill and of these 17 were Chaco-San Juan types and 664 Corrugatedcoil
culinary ware. Of 2,326 sherds from the second fill, between the
second and third floors, 46 were Chaco-San Juan and 1,486 Corrugated-coil;
of 2,820 sherds in the late fill above the third and last floor,
226 were Chaco-San Juan and 1,580 were Corrugated-coil. The makers
of these two wares obviously were increasing in numbers or productivity
during the years Room 44 was occupied.

Five more bowls may be cited: four of Chaco-San Juan type (pls.
22, a-c; 23, a) and the fifth a small local specimen with Hachure B
curvilinear design (pl. 23, c). The large bowl (pl. 22, b) is the only
restorable earthenware vessel recovered from Kiva F, a Chaco-type
ceremonial structure that had been remodeled to meet northern requirements
and thereafter was stripped of its timbers and briefly
utilized as a repository for neighborhood rubbish. In connection with
this remodeling and subsequent abandonment, it is pertinent to note
that we collected from the Kiva F rubbish 1,729 miscellaneous potsherds
among which Roberts and Amsden counted 301 of Chaco-San
Juan type and 8 of Classic Mesa Verde.

Part of an oval, flat-bottomed brown bowl with black interior, stone
polished inside and out, was recovered west of Pueblo del Arroyo,
outside Rooms 42-49 (U.S.N.M. No. 334677). Pepper (1909, p. 211)
reports a similar but slightly larger bowl, with corrugated exterior,
from Room 33, Pueblo Bonito.

A. Northwest corner fireplace, Room 3.

Plate 46

B. Fire screen and ventilator, southwest corner of Room 3.

(Photographs by O. C. Havens, 1923.)

Plate 47.—Two bowls from Room 5 (a, b) and Corrugated-coil pots from Room 3 (c, d).

Plate 48

Left: North-wall
recess in Room 4
and, to the right of
it, seating for the
single ceiling beam.
The upper rectangular
openings
held the ends of
paired beams extending
through
from Room 9A.

(Photograph by
O. C. Havens,
1923.)

Right: Milling
stone in its slabwalled
bin, Room 5.
Room 4 beyond.

(Photograph by
Neil M. Judd,
1923.)

Plate 49

A. South recess and blocked subfloor ventilator duct
(foreground), Kiva I.

(Photograph by Karl Ruppert, 1925.)

B. Fireplace, south recess, ventilator, and platform in Kiva B.

(Photograph by O. C. Havens, 1923.)

Plate 50

A. The subfloor vault in Kiva "c" as seen from the south.

(Photograph by Neil M. Judd, 1926.)

B. Covered outside ventilator shaft for Room 3. Above, two
empty holes once held poles tying the later stonework to the
west buttress.

(Photograph by O. C. Havens, 1923.)

Plate 51

Upper: Excavations west of Pueblo del Arroyo revealed the foundations of a triple-walled
McElmo Tower.

Lower: Fireplace, screen, and covered south ventilator in Kiva "b."

(Photographs by Neil M. Judd, 1926.)

A. A mass of rubble built over two rude walls between Kivas "c" and "d" provides an
unsolved puzzle.

Plate 52

B. Mr. Ruppert examines stratified accumulations against the outer west wall of
Pueblo del Arroyo.

(Photographs by Neil M. Judd, 1926.)

Plate 53

Plate 53.—Three Hyde Expedition photographs showing Jackson's
"old arroyo" and what is believed to be part of the silt-covered Pueblo
del Arroyo trash pile.

(Courtesy of Dr. Harry L. Shapiro and the American Museum of
Natural History.)

Figure 25 shows an "eagle" as part of the black-painted ornamentation
on the brick-red floor of a bowl sherd from south of the Little
Colorado River, perhaps the Petrified Forest area. The fragment is
slipped inside and out; externally, a distinct red slip wash comes well
down over a cream-colored slip that covers the flattened bottom and
lower sides. Coarsely ground sherd temper is visible in the blue-gray
paste.

Ladles.—Four of the five ladles recovered from the wreckage in
Room 27 are shown on plate 27. Figure b is a hollow-handled, ChacoSan
Juan type decorated with organic paint and a careless outside
rim band of white slip. Five dots appear on the lip above the handle,
and six more lie on the lip opposite; between the two groups is another
lot of five. The lip opposite this latter group has been worn
away, erasing its balancing series of dots.

Three of the Room 27 ladles are of the old P. II half-gourd form:
one dark brown with polished black interior (pl. 27, f) and two
smaller examples ornamented with mineral paint in Transitional style,
g, h. A fifth Room 27 ladle, if it can be so termed, is the 3-bowl creation
shown on plate 26, a. It is of a reddish-brown ware, slipped and
stone polished but unevenly modeled and undecorated. The handle,
flattened above and below, expands to either side beneath each bowl;
the bowl lips are flattened and unpainted.

A hollow-handled ladle and two ladle bowls were restored from
fragments found in the trash above the latest floor in Room 44 (pl. 27,
a, c, d). Four holes in a downcurving arc were punched through one
side of the handle of c and two on the opposite side. A horizontal
ring is missing from the handle end. This specimen and one of the
bowls, d, are finished and decorated in Chaco-San Juan style, while the
second bowl, a, follows Chaco Canyon practices.

A black-on-red ladle with hollow handle from Room 32 (pl. 27, e)
is ornamented with mineral paint, including the black line around the
lip. In restoring this specimen we replaced the handle ring.

Largest scoop-type ladle in the collection was recovered in Room 39
(pl. 27, i). Its full-length design and black rimline are in mineral
paint. In contrast, a miniature (pl. 27, j) found among late debris
at the east end of the Kiva "d" enclosure, and likewise decorated in
mineral paint, lacks the rimline. Part of an unusual Chaco-San Juan
ladle handle, solid throughout, white slipped and beautifully polished
(fig. 23), was recovered from the same rubbish heap; also, the end of
a white-slipped, longitudinally perforated handle that could be mistaken
for a pipe mouthpiece were it not for the holes punched through
the upper side (U.S.N.M. No. 334677).

Two other ladle handles, both stone polished and bearing organic
paint, may be noted. One (U.S.N.M. No. 334562) is solid, 5¾ inches
long by ⅞ inch wide and ½ inch thick, with two tail-like appendages
curled up over the distal end and four holes punched through vertically.
The second fragment (No. 334678) measures 3⅝ inches long by
2⅛ inches wide by ⅜ inch thick and thus recalls the broad handles on
certain ladlelike vessels from Pueblo Bonito (Judd, 1954, p. 202).
Irregularities on this second piece, top and bottom, have been abraded
away and so, too, the broken end. A bold checkerboard in black organic
paint enlivens the upper surface. Found in Room 9B-II, the
fragment presumably served as a toy.

Two perfectly absurd ladles were recovered in fragments at bench
level on the east side of Kiva C (pl. 26, b, c). Both are decorated
with iron oxide paint in Hachure B style. Their original length may
only be surmised, but we have restored no more than 1 inch on the
longer and less than that on the under side of the second. Considering
their similarity in other particulars, I assume the two were originally
of equal length or nearly so. The handle alone on c is 14⅝ inches.

Pitchers.—We have only seven specimens to represent all the
pitchers made and used at Pueblo del Arroyo. Five are of local manufacture,
the other two are Chaco-San Juan types, decorated with organic
paint. One of these latter (pl. 28, j) is slipped and stone polished;
its rim, mostly restored, has three broad ticks above the handle.
In our restoration we have allowed a concave base. The second ChacoSan
Juan pitcher, an offering with the Room 40 burial (pl. 24, e), is
less carefully finished but it has the customary concavity on the
bottom.

Our five Chaco Canyon pitchers, each cupped on the bottom, are
ornamented with mineral paint but in as many different styles. The
two from Room 39 (pl. 28, i, k) are alike in shape and decoration
but the design on one is zoned while that on the second is in an all-over
pattern. On each the handle is concave transversely. Figures f, g,
plate 28, are later types, both in form and finish. The work of a tyro,
h, was found in the southeast corner of Room 44, in the 17-inch
accumulation separating the second and third floors.

Squash pots.—So called because, as Morris (1939, p. 144) suggests,
the form obviously originated in a pumpkin or squash from which the
stem had been cut away, leaving a hollow vessel with a slightly depressed,
rimless opening. "Seed jar" is a synonym frequent in archeological
literature. In B.M. III times squash pots were commonly used
over the fire; from P. I onward they were nonculinary and customarily
decorated.

Our Pueblo del Arroyo collection includes five black-on-white
squash pots, or seed jars, and four of redware. They vary in diameter
from 3½ to 11¾ inches. The two largest (pl. 29, d, e), both from
Room 27, are of Chaco-San Juan type and ornamented with organic
paint. On the first, surface finish is noticeably inferior to that of the
second, but its basal concavity is four times as deep. The other three
black-on-white specimens, a-c, were decorated with mineral paint although
the design on a is barely traceable. Both a and c were provided
with flat bottoms, but the base of b is deeply cupped.

Of four redware seed jars (pl. 29, f-i) the first three were treated
with a thin red slip and stone polished; horizontal striae left by the
polishing stone show clearly. The first is flattened on the bottom, the
next two rounded. Figure i was both slipped and carefully smoothed,
but it lacks the visible toolmarks and the luster of the other three.
Under its slip the paste is dark gray all the way through. The vessel
is plainly overfired; its bold, black pigment has not been identified.

As to color, the first, f, and third, h, are a reddish brown; g and i
are of a brighter, livelier red. The first, browner below the waist,
wears above its middle a design in thin brownish-black pigment, presumably
iron manganese, applied before polishing.[1]

One might find the ancestor of these nine vessels among the seed
jars Roberts (1930) recovered from Pueblo I ruins in the Piedra
district of southwestern Colorado or among those described by Morris
(1939, p. 161) from La Plata Valley, 50 miles farther west.
After P. I times, squash pots lost their popularity along the La Plata,
and redware was "almost absent" in Early P. III (Morris, ibid.,
pp. 206, 212).

"Kiva jar" is a term proposed by Kidder (1924, p. 62) for a special
form of earthenware vessel, examples of which had previously
been found in San Juan area kivas. Because their shape is similar,
kiva jars and squash pots are obviously close relatives, but instead of
the depressed rimless mouth of the latter, the kiva jar is provided
with a flange for support of a discoidal cover and a rim that rises
outside and above the flange. Kiva jars are customarily attributed to
the proto-Mesa Verde and Classic Mesa Verde cultures.

At Pueblo del Arroyo we recovered only one fragment of a kiva
jar, a [fraction 5 by 16]-inch-thick white-coated but unslipped, coarsely sherd-tempered
piece decorated outside above flange level with a curvilinear

design in iron oxide paint (U.S.N.M. No. 334680a). The remaining
arc of that design represents a circle at least 2 inches in diameter, and
the rim must have stood a bit higher. Pepper (1920, p. 124, fig. 48, b)
illustrates a low-rimmed kiva jar from Room 32, Pueblo Bonito.

Canteens.—Five specimens are before us, no two alike (pl. 28, a-e).
Diameter varies from 6 to 8¼ inches; height, from 5 to 6¾. Loop
handles high on the shoulder for a carrying cord occur on four specimens;
knobbed handles on the fifth.

Plate 28, c, of Chaco-San Juan surface treatment and decoration,
was restored from fragments recovered in Rooms 2 and 3 south of
the pueblo. It has the low, squat profile, the concave base, and the
triangular scroll design common to northern canteens. The others
are all ornamented with iron oxide paint. Plate 28, a, one of three
unbroken vessels in the entire collection, found on its side, orifice
close in the southeast corner of Room 65, bears a Degenerate Transitional
pattern; Hachure B appears on b and e, however indistinctly
in the latter case. Each has a concave bottom except d, which is
somewhat flattened.

Cylindrical vases.—In contrast to the numbers unearthed at Pueblo
Bonito, only one black-on-white cylindrical vase is known from
Pueblo del Arroyo (pl. 24, f). It leans a bit to one side; its simple
Hachure B ornamentation is divided into two zones by a plain band
occupied by four upturned loop handles.

In addition to that just described three redware vases were restored
from fragments found in Room 15 (pl. 55). Each is flat bottomed
with a basal diameter about half that at the rim. Each is coated with
a thin red slip that was stone polished to the point of lustrousness,
inside and out on the first (left figure), and to a less degree on the
inside of the other two. Longitudinal channels left by the polishing
stone remain visible upon the outer surfaces of all three, encircling
striae within. None of the vases is decorated; none has handles or
perforations for suspension cords.

The first, 6 inches in diameter at the rim and 3¼ inches on the bottom,
has a height of 9½ inches. The bottoms of the next two have
been restored but their indicated heights are 10¼ and 10 inches, respectively;
their rim diameters, 6⅝ and 6¼ inches. Spalled surfaces
and leftover sherds show pastes varying from gray in limited areas
to reddish all the way through. On the basis of two small fragments
submitted, Miss Anna O. Shepard reports a sherd temper with "a noticeable
preponderance of white paste particles" plus "a scattering
of well rounded quartz grains."

Despite a predominantly sherd temper our three redware vases
from Room 15 are unique in Chaco Canyon. In surface treatment if
not in color the two associated squash pots (pl. 29, g, h) are identical
with these vases. Origin of the five remains unknown.

What may be the fragment of a vertical-walled vase, white slip
washed inside and out but unpainted, was found in the general digging.
Its rim is thinned from the outside and the lip rounded; indicated
inside diameter is about 5½ inches. The unusual feature of this
fragment is that its exterior had been horizontally incised at ⅛-inch
intervals with an awl or like pointed tool prior to being whitened.
Paste is uniformly light gray throughout; sherd, quartz grains, and
dark rock particles are visible as temper.

Water jars or ollas, 15 in number (including two unrestored fragments)
vary in maximum diameter from 6¾ to 19¾ inches. Height,
bottom to base of neck, varies from about 6½ to 18¼ inches. In each
case where the bottom is intact it is concave. Of six necks present
and measurable, height of five is 1½ inches; of the sixth, 1¾. Rims,
usually with a slight flare, range from 2¼ to 9¼ inches in over-all
diameter. The largest orifice occurs on an olla foreign to Chaco
Canyon (pl. 31, a).

Comparatively small mouths and short necks identify this group as
vessels for transporting or storing water. Omitting one of the unrestored
specimens (neck and shoulder only), 2 of our 15 ollas are
provided with lugs for lifting or for embellishment, 2 have horizontal
strap handles at or below maximum diameter, 3 are equipped
with inset fingerholds, and 7 have no handles at all. Their individuality
in other respects is readily apparent from the illustrations.

Two water jars from Room 15 differ in size and decoration but
both were provided with inset handles just below greatest diameter
(pl. 30, a, b). Both bear zoned ornamentation in iron oxide paint,
but that on the first is in Degenerate Transitional while that on the
second is in contemporary Hachure B. Neither neck was recovered;
the broken edges were not tooled. As restored, the bottom of this
latter was left concave but within an erroneously flattened area.

Six Room 27 jars provide an even more interesting comparison
(pls. 30, c-f; 31, a; 32, b). All were recovered between 2 and 3 feet
above the floor, clearly fallen from the second story. All were decorated
with iron oxide paint except figure e, plate 30, which is stone
polished over a design in nonmetallic pigment. When painting this
vessel, the potter first outlined the checkerboard frames and individual
negative squares; the dotting and filling-in followed. The base

is concave. Strap handles immediately below maximum diameter were
pressed tightly against the wall for more than half their length, leaving
protuberant ends.

Figure a, plate 31, is unquestionably a stranger to Chaco Canyon.
It is slightly indented on the bottom, but its sloping shoulder, wide
mouth, and outflaring rim are not to be seen on any other jar in the
collection. Its design, in iron oxide pigment on a well-smoothed surface,
is reminiscent of that on a bowl from Room 266 at Pueblo
Bonito and I would guess the two came from the same distant but
still unknown village. Two hornlike lugs below the rim, their pointed
ends turned counterclockwise and touching the neck, were finger lifts
more decorative than useful. Broader lugs (pl. 30, d), cupped on the
under side, seem more practical as handles.

Among the restorable vessels recovered from household rubbish in
Room 28, or perhaps fallen from the second story, are the two great
ollas shown on plate 31, d, e. Both were decorated with mineral paint,
the first in a rusty-brown Degenerate Transitional pattern on an unslipped,
hand-smoothed surface; the second in a Transitional design
over a highly polished gray slip. Inset handles or finger grips appear
on the first; no handle of any sort on the second. In our restorations
we erred—to judge from other ollas in the collection—in making both
bottoms round instead of concave. The neck of figure e, plate 31, had
been broken out and the irregularities abraded away to leave a 10½-inch
orifice. So altered, the vessel was still serviceable for storage purposes.

Two other ollas from Room 28 were found to be incomplete when
unpacked in the Museum laboratory: (a) our smallest, 6¾ inches in
diameter, upper third missing, concave base, decoration an unsure
Hachure B or C over a polished slip (U.S.N.M. No. 334563), and
(b) the neck-and-shoulder fragment mentioned above (No. 334566),
with a bold Hachure B design.

A restored water jar from Room 51 (pl. 31, c) is a colossus among
Chaco Canyon pottery. In size, 19¾ inches in greatest diameter by
18 inches in height from concave bottom to base of neck, it surpasses
every other jar revealed by our excavations. The maker of that
enormous vessel must have been proud of her creation but, once
finished, it was not often moved. It lacks handles; it is bulky even
when empty; it would barely pass through existing doorways. In our
restoration we merely suggested the cylindrical neck which very likely
was between 2 and 3 inches high.

A second Room 51 olla, likewise with neck missing, is shown on
plate 31, b. Its ornamentation, in a brownish-black pigment, is in a

fine-lined, boldly framed, over-all rectilinear hachure I would classify
as C. The design is carried well down the side, an apparent characteristic
of late hatching. Handles are heavy straps, slightly indented,
placed below maximum diameter and carrying an appropriate portion
of the decoration. In restoring this specimen we have again erred in
providing it with a flat, rather than a cupped, base.

The design on figure a, plate 32, is a composite that does not look
Chacoan although the pigment is an iron oxide over a lightly slipped,
well-smoothed surface that is decidedly local. The fragments were
recovered from Rooms 43 and 44.

It is a curious fact that, while we recovered 13 restorable water
jars at Pueblo del Arroyo, only 4 were unearthed at Pueblo Bonito,
residence of a much larger population.

Culinary ware.—The 25 cook pots recovered at Pueblo del Arroyo
are shown on plates 32, 33, 34, 47. All have been restored: 2 from
Room 3, abutting the outside south wall of the pueblo, 3 from Room
27, 1 each from Rooms 15 and 64, and 18 from Room 65. As a group
they average 3 inches larger, both in diameter and height, than 29 like
vessels we exhumed at Pueblo Bonito and their rimflare is more
evident.

One-story Room 65 obviously was last used for storage, since 12
pots were ranged along its walls, 5 on the north side and 7 on the
south (pl. 14, A). Several stood upright; others, bottom up. All were
crushed when the ceiling fell in and then, after still useful timbers
had been salvaged, the resultant depression became a neighborhood
dump. From this rubbish we collected fragments of six more cook
pots.

These 18 Room 65 vessels differ appreciably in shape and size, as
will readily be seen from the illustrations and from the measurements
given in Appendix A. Those standing on the floor (pl. 34, a-l)
vary in maximum diameter from 9¾ to 14¼ inches; in height, from
10¾ to 17 inches; and in over-all diameter at the mouth, from 7⅝ to
11⅞ inches. In contrast, the six recovered from the overlying household
debris (pl. 33, a-f) vary from 7⅝ to 12¾ inches in diameter, from
8⅛ to 15 inches in height, and from 5⅛ to 12¼ inches in rim diameter.

The third vessel in this second group (pl. 33, c), with its shortened
rim and plain base, had been rubbed over before its corrugations were
entirely dry. This is also true, but less noticeably, of one of those
from the floor (pl. 34, d). A companion pot, plate 34, a, had been
smoothed with a moist hand until its coiling was half obliterated.

Where they occur, handles on these 18 pots likewise differ. Solid,

downturned lugs or finger grips occur 2½ inches below the lip on
opposite sides of figure k, plate 34. Similar grips, somewhat flatter, lie
1¾ inches below the lip of b. On one side of d a small solid lug, its end
turned to the right, lies 2¼ inches below the lip, while on the opposite
side of the neck is a wider finger grip, punched through from above.
The solid loop handles on figure e, attached by the riveting process,
and slightly indented, are unique in this collection. Again, one of the
six pots from the overlying trash pile (pl. 33, e) has a pair of conical
nodes, ¼ inch long and an inch apart, positioned 1¼ inches below the
rim on one side of the neck with nothing opposite by way of balance.
Clearly these were only decorative. The large pot with unusually
wide mouth and throat (pl. 33, f) looks to me more P. II than P. III.

It would be instructive to learn whether these two lots differ in
tempering materials as much as they do in size and outward appearance.
Lacking the means for thorough analysis, I have depended upon
a 10X lens. With this I recognize grains of rounded quartz sand in
both lots and angular bits of crushed rock. Undoubtedly our Pueblo
del Arroyo pots were variously tempered as were those at Pueblo
Bonito where Miss Shepard (1939, p. 280) found sanidine basalt,
directly or indirectly, in a noteworthy proportion of culinary ware.
It is her belief that the presence of this unusual rock in Chaco Canyon
pottery is evidence of trade from the Bennett Peak district, 50 miles
to the west, where corrugated ware was a speciality and where sanidine
basalt was the customary temper.

Temper is the most important diagnostic in analyzing corrugated
pottery. Structural variations such as coil width, depth of corrugations,
embellishment, and presence or absence of handles seem to me
more indicative of potter preference—and perhaps only a fleeting
preference at time of manufacture—than of established practices that
may measure cultural or time advances.

With few exceptions our 25 corrugated pots, irrespective of size
and shape, would have been equally at home in almost any other
Pueblo III ruin throughout the San Juan country. Only the tall
vessel with plain body and broad, smoothed-over coils on neck and
shoulder and downturned lug finger grips high on the sides (pl. 34, k),
seems out of place in this assemblage. With this exception and perhaps
the two next to follow, every cook pot in the collection comes
between the extremes of what Morris (1939, p. 196) regards as the
"standard" early P. III corrugated jar.

The squatter, more globular of the Room 65 pots (pl. 34, e) and
the lone example from Room 15 (pl. 32, d), each with high, rounded

shoulder, constricted orifice, sharply recurved rim, and manipulated
coiling far superior to that on any of the others, are superb examples
of the prehistoric potter's art. Both are probably to be identified in
point of time with the Mesa Verde phase of Pueblo III culture.

A small, coarsely coiled vessel from Room 27 (U.S.N.M. No.
334647) was not included in the foregoing review because it did not
seem to belong. It is not sooted; has never been on a fire. Its globular
body and flaring rim suggest affinity with the two squat vessels from
Rooms 15 and 65 considered in the previous paragraph. Its coiling
is counterclockwise; thumb or finger indentations are infrequent;
coils are uneven and curled at the outer edge as though formed by
a dragging finger on a too-firm paste. A strip of clay, now missing,
had been pressed handlewise flat against the topmost coil. Overfiring
is evident.

A final corrugated specimen is the redware vessel seen on plate
32, c, restored from fragments gathered from floor sweepings dumped
into Room 39. Its neck had been broken while in use and the irregularities
leveled by abrasion. Coils average six to the inch and were
smoothed by scraping.

Feather box (?).—We have no pigeonhole into which to fit the rectangular
vessel seen on plate 24, j. Its slaty-gray slip is decorated with
vegetal paint and stone polished over all. Its bottom is flat and its
sides slant upward and inward to a flat-rimmed, rectangular opening
that averages 2⅝ by 4 inches. At either side of this opening paired
holes an inch apart, punched through from the under side, presumably
were the means for securing a cover. The vessel obviously was a
receptacle for treasured objects, perhaps feathers. It has the feel of
a Classic Mesa Verde piece.

Bird-shaped bowl.—The small, bird-shaped container with iron
oxide pigment in a Degenerate design (pl. 24, h) is the only one of
its kind in the collection. A tau-shaped opening occupies almost the
entire back; the ornamentation overlies a thin white wash; the concave
bottom is not slipped. The vessel was found in the southwest corner
of Room 27 about 3 feet above the floor and among wreckage of the
second story.

The tail from a similar vessel, decorated in Hachure B, was recovered
from a pile of late rubbish outside the west wall of the pueblo,
opposite Rooms 42-49. Here, also, was a second bird-shaped bowl
fragment—a flat, tail-like piece painted above and below with brownish-black
pigment and polished. It is 1⅜ inches long by 1⅛ inches
wide by ¼ inch thick— too long and too thin to be a bowl handle.

In this same late rubbish west of the pueblo we found two fragments
of a redware double bowl, one half black inside and the other
red. It was burnished inside and out but otherwise undecorated. Bits
of pulverized sherd, numerous white particles, and quartz sand grains
in the paste are visible under a 10X lens. In surface finish also the
fragments parallel the three redware vases from Room 15 (p. 156).

Figurine fragments.—From the late rubbish pile outside and west
of Rooms 42-49 we recovered seven fragments from six earthernware
figurines of human form and what may be the leg of a bear
effigy (pl. 36, A). This latter, with a broad toeless foot and short
shin, is twice ringed about in iron oxide paint over a thin white slip.
Two other fragments are also mineral painted, front and back. They
are the shoulders of, apparently, a single figurine despite the fact
that one piece is a little smoother than the other. Both are of a gray
paste, thinly slip washed, and well fired; the arm stubs are solid.

Of the five remaining fragments—two small feet tooled between
toes, a head fragment preserving the left ear and cheek, a braceletted
right forearm, with hand cupped to rest upon a rounded knee, and a
lower left leg—all but the latter are decorated with a brown to dullblack
pigment that does not respond to my magnet. All but two are
stone polished. The last fragment, flatfooted and its toes lightly indicated
by incising, is from a figure seated with knees elevated and calf
free from thigh; the knee is unslipped where a cupped hand had
rested.

The foregoing eight figurine fragments were all found in late
household debris outside of the pueblo. Within the ruin we recovered
only one comparable sherd (U.S.N.M. No. 334675) and that from
rubbish dumped into Room 40. It is the 2½-inch-long fragment of a
doubled right leg from a figure seated or squatting with heel against
buttocks. The shin rises with a slight outward bend; the calf is
pressed close against the thigh and the right hand is spread across
just above the knee—a rather awkward position. Breakage occurred
at the wrist, above the ankle, and at the union of thigh and torso. The
broken ends have been partially abraded.

Among random sherds is a short left hind leg with broad rounded
toe, vertically striped outside, at the rear, and underneath—the only
fragment of a cloven-footed figurine in the collection (U.S.N.M.
No. 334680a).

Earthenware effigies of bifurcated baskets.—Two restored specimens
(pl. 35) found among second-story wreckage a couple of feet
above the floor in Room 27, were briefly considered in my description

of comparable objects from Pueblo Bonito. Both have the vertical
incurve, front and back, that appears to be a characteristic of Chaco
Canyon bifurcated baskets and their earthenware representations.
Both are thinly slipped, presumably sherd tempered, and ornamented
with iron oxide paint. The larger was hand smoothed and the smaller,
stone polished.

On each effigy the rim is painted black and is surmounted by a
transverse loop attached inside at the rear. Each has been pressed
in along the median line until the inner walls, front and back, almost
meet; on figures b, b′, plate 35, this pressure has caused an open crack
vertically in the upper half of the inside front wall. The larger of the
two is hollow legged, the smaller is not. Both are equipped on the back
with horizontally perforated lugs representing tumpline attachments.

A fragment that may be part of a larger basket effigy with a still
more complex superstructure (U.S.N.M. No. 334677) was recovered
from late rubbish at the east end of the Kiva "d" enclosure. As I
interpret it, the fragment represents two loops, one rising above the
other, arching across the rear width of the basket cavity. The loops
are solid, [fraction 5 by 16] inch in diameter and molded together at the back; they
are slip washed except between their feet and where they were pressed
and firmed together. A single black line adorns the face of the lower
loop. Modeling at the rear is less finished, and a slight flare at the
bottom of each leg suggests a basket with upcurving sides.

My tentative identification is prompted by a similar but smaller
and previously undescribed fragment from the east refuse mound at
Pueblo Bonito (U.S.N.M. No. 336065). This latter sherd consists
of what I judge to be part of the right front rim of a basket effigy
with a solid ⅜-inch roll molded to the inside and arching low across,
and outside, the vertical groove that is a peculiarity of Chaco Canyon
bifurcated baskets and their effigies. Part of a second and higher loop
stands behind the first and is molded to it; the black rim line of the
basket is continued across the base of the lower loop. There can be
no doubt that the double loops of this second fragment were attached
to the front rim of a basket effigy— the only instance of the kind that
has come to my attention. In every other case the arching superstructure,
if any, was molded to the inner rear wall of the effigy.

Miniature earthenware vessels such as that represented by figure 26
are made by modern Hopi potters, according to Katharine Bartlett
(1934, p. 53), and left at clay pits as offerings to the spirits resident
there. Our example, found in household rubbish, is fired but undecorated;
a height and diameter of ¾ inch explain its want of symmetry.

The neck was added from the outside; there are no toolmarks visible
within the [fraction 5 by 16]-inch orifice.

A second miniature, larger and more successfully modeled, lay
among debris of occupation near floor level in Kiva C (U.S.N.M.
No. 334640). It is a bit lopsided, cupped on the bottom and undecorated.
Measuring 1⅝ inches in diameter by 1¾ inches high, the
piece was burned and blistered by the fire that gutted the kiva following
removal of its ceiling timbers.

Our third and final miniature (pl. 24, g) portrays a vessel with
wide shoulder, low neck, and flaring rim. It is flat bottomed, thinly
slip washed outside and within the rim, and ornamented by a black
rimline and eight encircling lines on neck and shoulder.

Spindle whorl.—Half of an earthenware spindle whorl, 2½ inches in
diameter by [fraction 3 by 16] inch thick, has a black-painted periphery and a halfinch-wide
circle in Hachure B on each side (No. 334666).

Worked potsherds.—Among a small quantity of potsherds with
abraded edges, eight are identifiable from the angle of wear as
smoothers employed in pottery manufacture and one as a possible toy,
the ring-ended fragment of a small Chaco-San Juan ladle handle,
round and solid, worn off at both extremes. Four sherds are discoidal,
½ to 2⅛ inches in diameter. Of these four the two largest are fragments
of black-on-red bowls, their edges on the concave side flattened
by abrasion. Three smaller sherd disks from rubbish thrown over the
west wall outside Room 48, one each of Hachure A and B and Early
Black-on-red, are less discoidal and quite unabraded on the sides
(U.S.N.M. No. 334662).

RECAPITULATION AND COMMENT

Roberts and Amsden, who conducted our pottery analyses, based
their study upon the stratigraphy of two yard-square tests cut through
12 feet of previously undisturbed household sweepings underlying the
West Court at Pueblo Bonito. With the exception of a few Pueblo I
sherds all black-on-white fragments found in the lower 8 feet of that
rubbish belonged in three stylistic groups: Transitional, Degenerate
Transitional, and Solid. In the overlying 4 feet of that same stratified
debris, however, in addition to sherds of the three groups just mentioned
there were fragments of straight-line hatching and of ChacoSan
Juan. Sherds of Corrugated-coil culinary ware were also recovered
from the upper 4 feet but only those of Plain-banded cook
pots below that depth. Clearly two distinct pottery assemblages
separated at the 8-foot level.

The several pottery types identified with Pueblo Bonito occur also
at Pueblo del Arroyo, although not in the same proportions. There
were fewer sherds of Transitional, Degenerate, Solid, and Plainbanded
culinary and more of the later types. The Chaco-San Juan
was especially conspicuous both from excavated rooms and from the
mound area west of the ruin, including remains of the McElmo
Tower. If not the predominant local type, it was nearly so. Of 77,405
sherds tabulated by Amsden and Roberts 6,614, or more than 8 percent,
were Chaco-San Juan. No other black-on-white variety rated
half as high. Only Corrugated-coil culinary fragments were present
in greater number.

Sherds of Classic Mesa Verde ware provide evidence of time and
trade. After eliminating all recognizable duplicates Roberts and Amsden
counted 930 such sherds at Pueblo Bonito but only 58 from
Pueblo del Arroyo. Of these latter all but two were found in household
waste that had been dumped into Kivas F, G, and J, each of
which had been altered during occupancy to conform with northern
tradition.

These divers data suggest that the inhabitants of Pueblo del Arroyo
were closer, socially and economically, to the more recent portion of
the population of Pueblo Bonito than they were to the older portion,
and that Pueblo del Arroyo was first of the two villages to be vacated.
The latest recorded bracket of tree-ring dates for the Mesa Verde
cliff dwellings is A. D. 1019 to 1274; for Pueblo Bonito, 807+ to
1130 (Smiley, 1951, pp. 19, 22). Trade from the Mesa Verde came
to Chaco Canyon late and continued at Pueblo Bonito after Pueblo
del Arroyo had been abandoned.

The black-on-white pottery of Chaco Canyon has long been praised
for its superior qualities. Kidder, first to analyze those qualities,
pointed to "its very white, almost paper-white, slip and the unusually
fine lines of its black decoration" (Kidder, 1924, p. 52). The tapered
or rounded rim, the black-painted lip, and the line break were other
conspicuous characteristics. The apparent conflict in form and decoration
that puzzled Kidder at Pueblo Bonito is readily explained by the
presence of P. II and P. III peoples as joint occupants.

Chaco Canyon black-on-white ware, from the Transitional to Hachure
C, is generally hand smoothed, sherd tempered, and decorated
with iron oxide paint. Some individual vessels are whiter and
smoother than companion pieces, some are harder, some exhibit a
coarser temper than others and even grains of sand or rock mixed
with the ground sherd. Equal firing temperatures were not always

realized, hence pastes and surfaces vary in color. These dissimilarities
are the natural consequence of a freehand process in which individual
potters followed accepted but flexible processes. Shapes changed
from time to time, but our several pottery types distinguish successive
fancies in painted ornamentation rather than in form. Some of these
fancies were short lived, others persisted for generations. They were
at the same time sequential and contemporary. Present-day potters
likewise have their preferences (Guthe, 1925, p. 78; Chapman, 1936,
p. 15).

Our Chaco-San Juan group offers a different challenge. Like the
local black-on-white series it is sherd tempered but is decorated with
an organic rather than a mineral paint and polished over the decoration.
From the first it appeared to embody Chaco techniques and
northern designs, components that were sometimes elusive. Occasionally
the decoration was done in mineral paint while designs and surface
treatment followed northern custom.

As recorded by Roberts, Chaco-San Juan bowls have straighter
sides than local bowls and thicker rims, variously ticked. Bowls, predominating
in our collection, are usually unslipped or partially slipped
outside but carry an interior, banded decoration framed above and
below. Jars are globular with downraking handles and zoned ornamentation
on body and neck. Pitchers have small bodies and long necks.
Both pitchers and jars are indented on the bottom; mugs, which first
appeared with the Chaco-San Juan, are flat based and slope sided.
Ladles are all of the bowl-and-handle variety.

In shape, finish, and painted design our Chaco-San Juan group
most nearly approaches the "proto-Mesa Verde" as initially described
by Kidder (1924, p. 67). Decoration favors a repetition of such wellknown
elements as the fret, the interlocking key, stepped figures, and
checkerboard—sometimes balanced by other units having the heavy,
widely spaced type of Mesa Verde hatching. Together, the protoMesa
Verde and our Chaco-San Juan equally foreshadow true Mesa
Verde. Gladwin (1945, p. 149) regarded our Chaco-San Juan type
as a "blending of Chaco designs and the Kayenta techniques of painting
and polishing . . . from the region between Toadlena and Shiprock,"
but a northern affinity seems more likely to me.

Proto-Mesa Verde black-on-white is the pottery of Prudden's "unit
type" structures (Prudden, 1903; Kidder, 1924, pp. 65-68). The
mesas and valleys where those structures occur is likewise the home
both of Mancos and McElmo black-on-white—names that appear
repeatedly in the literature, the former more than the latter.

Mancos black-on-white was first described by Martin (1936, pp.
80-94) following researches at Lowry Ruin, southwestern Colorado.
He described it as a chalky-white to gray ware, generally unslipped
and unpolished. Bowls were flat, or sometimes round-bottomed, with
direct rims and banded interior decoration including frets, solid or
hatched figures, pendent triangles, and pendent lines reducing in
length toward the left. Jars were globular with square shoulders,
downraking handles and zoned decoration; ladles were of the bowland-handle
form. Mancos, a sherd-tempered ware, was decorated
with iron oxide paint. McElmo, on the other hand, was a carbon-paint
type but also sherd tempered.

Colton and Hargrave (1937, p. 230) closely follow Martin but
misread him in recognizing Mancos black-on-white rather than
McElmo as synonymous with the proto-Mesa Verde. Mancos blackon-white
is widely distributed throughout the San Juan country and
beyond. It was the dominant type at Lowry Ruin (Martin, 1936,
p. 94) and at the Turner-Look Site, 15 miles northwest of Cisco,
Grand County, Utah (Wormington, 1955, p. 74). It occurred repeatedly,
sometimes preponderantly, at small-house sites in Mancos
Canyon, southwest of Mesa Verde National Park (Reed, 1944). From
a 3-room house plus kiva identified as Unit 1, Site 13, on Alkali Ridge,
southeastern Utah, Brew (1946, p. 199) reported "the black-onwhite
sherds were mixed Pueblo II (Mancos black-on-white) and
Pueblo III (Mesa Verde and McElmo black-on-white) types with the
latter predominating." Again, summarizing the subject for Alkali
Ridge, Brew wrote (ibid., p. 285): "The Mancos was technically
advanced and had begun to show Mesa Verde features. The Mesa
Verde was for the most part of the kind that could be called McElmo."

Nearly all who have written of Mancos black-on-white have remarked
its close affinity with McElmo black-on-white and the "Chacolike"
qualities of both.

During his initial work at Lowry Ruin Martin was perplexed by
this resemblance and "sometimes found it difficult to decide whether
a sherd was Mancos black-on-white . . . or Chacoan" (Martin, 1936,
p. 112). Reed, after balancing his own observations against those of
other students, concluded (MS., p. 127): "There seem to be fairly
definite indications that a distinct form of carbon-paint pottery, decorated
in what I have referred to as `McElmo style' occurs with
Mancos black-on-white as well as with Mesa Verde black-on-white
. . . and that it may appear late in the occupation of Mancos sites."
Nevertheless, Reed failed to convince himself of the reality of a

McElmo black-on-white bridge between the mineral-paint Mancos
variety and the carbon-paint Mesa Verde. The transition between
these two latter "evidently was a rapid but irregular process" (ibid.,
p. 118).

Deric O'Bryan also has made a thorough analysis of the Mancos
and McElmo phases and balanced one against the other. He recognizes
distinctive qualities in each, the first culturally P. II and dated
about 900 to 1050; the second, P. III and about 1050 to 1150
(O'Bryan, 1950). Each phase has its own peculiar pottery, differing
not only in character of the pigment used, but also in surface treatment
and choice of design elements. Chaco Canyon influences were
particularly evident on Mancos Phase pottery according to O'Bryan
(ibid., p. 108); McElmo black-on-white, although often present on
Mancos sites, was more akin to Mesa Verde black-on-white. Mancos
sites are legion north of the Rio San Juan but O'Bryan knows of no
pure McElmo site (ibid., p. 109).

This Mancos-McElmo association is not accepted by all workers in
the field. Lancaster and Pinkley (1954) question O'Bryan's Mesa
Verde dating and his postulated McElmo phase because at Site 16,
in a P. II cultural level with a terminal date near A. D. 1100, some
44 percent of the pottery was Mancos black-on-white, yet no sherd
either of Mesa Verde or Corrugated-coil was present. Reed (1944,
p. 51) also doubts the reality of the McElmo; sees no need for an
intermediate ware between Mancos and true Mesa Verde. That there
was a merging along the Mancos-McElmo line is perhaps substantiated
by Miss Shepard (1939, p. 254), who points out that the mineral
paint of P. II potters throughout La Plata Valley was being gradually
replaced in Early P. III and then, quickly, was entirely superseded by
reintroduction of organic paint.

I have cited these several authors because it is my belief that many
of our Pueblo del Arroyo black-on-white vessels are really "Mancos"
and that our abundant Chaco-San Juan type is essentially "McElmo";
hence the "Chaco-like" quality various investigators have seen in pottery
north of the Rio San Juan. The producers of Mancos ware are
the older, but if they drifted south and arrived in Chaco Canyon first,
they were soon followed by the McElmo strain. Migration or exchange
of ideas was southward, in my opinion, rather than the reverse.

These thoughts find support in the recorded observations of other
students. In the Pagosa-Piedra region in 1922 Jeancon and Roberts
(1924, p. 214) first noted marked similarities to Chaco ceramics and
architecture. Roberts (1930, p. 18; 1932, pp. 12-13) later emphasized

Plate 54.—Bowl from the Houck district, east-central Arizona.

Upper: exterior; lower: interior.

(Kodachrome by Willard R. Culver.)

Plate 55.—Three redware vases from Room 15.

(Kodachrome by Willard R. Culver.)

this fact and suggested a migration from southwestern Colorado to
Chaco Canyon in "fairly early times." Miss Dutton (1938, p. 94)
echoes this possibility in concluding her monograph on Łeyit Kin, a
small-house ruin in Chaco Canyon.

It was a combination of Chaco features at Lowry Ruin that
prompted Martin's work there (Martin, 1936, p. 22). Reed (1944,
p. 49), adopting Mera's type designations, regarded Chaco 2 and
Chaco 3 sherds from Site 1, Mancos Canyon, as twelfth-century intrusives.
O'Bryan (1950, p. 108) recognized a "predominantly
Chaco influence" upon the culture of Mancos Mesa between A. D. 900
and 1050. Morris (1939) repeatedly mentions Chaco-like and pure
Chaco pottery in La Plata Valley and accepts both varieties as contemporaneous
with "non-Chaco" (ibid., p. 205).

Morris (ibid., p. 53) observed that sherds from Building I, Site 39,
"clearly belong to the Chaco complex" but conform more closely to
sherds from the lower level of the West Pueblo at Aztec than to those
from the great houses of Chaco Canyon. Morris quotes Kidder (1924,
p. 56) to the effect that "Aztec ruin is architecturally a perfect example
of a Chaco pueblo" and then points out that the "pottery made
and used by the builders of the Aztec ruin is more Chacoesque than
Chaco. . . . It is representative of the black-on-white ensemble that
was in general use north of the river at the time when the best wares
of the Chaco center were being made, and I believe is the result of
local expression of the generalized Chaco urge far more than of direct
influence from the Chaco itself" (Morris, 1939, p. 205). In form
and decoration many of the vessels Morris (1915, 1919) exhumed at
Aztec are indistinguishable from those we collected at Pueblo del
Arroyo.

This is equally true of Chaco-like pottery from Lowry Ruin and
elsewhere. Martin's descriptive text and illustrations of Mancos and
McElmo black-on-white vessels (Martin, 1936, 1938; Martin and
Willis, 1940) are equally applicable to specimens herein reported under
our several designations. Martin's data do not always distinguish
clearly between Mancos and McElmo and I am not sure they can.
Some of his illustrations, by the definition, look more McElmo than
Mancos and vice versa.

The ancestry of the brilliant Chaco Canyon culture has intrigued
every archeologist who has studied it. Kidder directed attention to the
wide distribution of Chaco-like pottery beyond the San Juan—pottery
that "may indicate a northwestern spread or a northwestern origin of
the Chaco culture" (Kidder, 1924, p. 56). Morris (1939, p. 204)
seems in complete agreement.

Morris seems to answer all those who see a northward spread of
Chaco Canyon culture and, at the same time, to summarize his own
thoughts on the subject when he wrote (ibid., p. 204): "The Chacolike
remains north of the San Juan, both architectural and ceramic,
are so widespread and so numerous that I consider it untenable to
view them wholly as an extension of or a backwash from, the Chaco
Canyon center. . . . The most Chaco-like of the vessels from the
north country, which seem so significant when viewed singly or selectively
grouped, become far less so when viewed as the minor component
that they are of the totality of wares among which they occur."

My own field experiences in southwestern Colorado and southeastern
Utah are too far removed to be useful in this study, but
memory of them finds me in complete accord with Morris and Kidder
if not, indeed, a bit more certain than they that the cradle of Chaco
culture lies among the P. I and P. II remains north of the Rio San
Juan. Nevertheless, I must leave to others the task of bringing together
the data that will substantiate or refute this conviction. All
elements of design indicated for the mineral-paint Mancos and for the
organic-paint McElmo wares will be found on Pueblo del Arroyo
vessels illustrated herein.

Florence M. Hawley, perhaps the first after Kidder to essay a detailed
description of Chaco Canyon pottery, divided the whole into
types and undertook to name each on the basis of its internal and external
features. From the east refuse mound at Chettro Kettle she
described (1936) "Escavada Black on White," "Gallup Black on
White," and "Chaco Black on White." These, together with "Puerco
Black on White" and "Red Mesa Black on White"—all sherd-temper,
iron-paint varieties—were subsequently reported from Bc 50-51,
a small P. II-P. III house on the south side of Chaco Canyon, opposite
Pueblo Bonito (Hawley, 1937; Kluckhohn and Reiter, 1939,
table 2). Gladwin (1945, p. 118), than whom none has studied southwestern
pottery more intensively, doubted that anyone could distinguish
between these five types, and I find myself equally helpless.
Other published descriptions (Hawley, 1936, 1939; Colton and Hargrave,
1937) reveal no basic type differences that I detect except,
perhaps, in paste color, surface finish, and design—features in which
no two vessels are precisely alike.

Summarizing her study of pottery from Tseh So (the Bc 50-51 of
Kluckhohn and Reiter, 1939), Miss Hawley (1937, p. 86) remarked
upon the prevalence of McElmo black-on-white and its evidence of
trade from the Mesa Verde country. McElmo black-on-white and our

Chaco-San Juan group are identical as far as I can determine. Both
are carbon-paint types and predominantly sherd tempered; both are
synonymous with proto-Mesa Verde, the precursor of true Mesa
Verde.

The great Chaco villages were at point of eclipse when true Mesa
Verde pottery began to appear in Chaco Canyon. Our sherd collections
from local small-house sites such as Tseh So indicated that some
were earlier, some later, than the great houses in their heyday. Only
a handful of Mesa Verde sherds was recovered at Pueblo del Arroyo
where the Chaco-San Juan, or McElmo, was the dominant black-onwhite
type.