Montezuma Castle Archeology - Part 2: Textiles


This series of 61 specimens was sent for analysis from the Montezuma Castle National Monument, Camp Verde, Arizona, to the Indian Department of the Denver Art Museum during the winter of 1937-1938.


The specimens were received in three groups and were numbered arbitrarily upon receipt, merely as a method of identification.1 (The PS numbers appearing on some of the pages are original National Park Service numbers.) Each piece was then cleaned as thoroughly as possible. Next, gross measurements were taken of each—in inches and centimeters. Slides were then made of the material or materials found in every specimen and checked by F. Martin Brown, Dr. Volney Jones and Dr. A. C. Whitford. Dr. Whitford made the photomicrographs shown in the paper.

1. These numbers appear in the photographs on the metal-edged, round disks.

Technical sketches were made after careful examination of each specimen under the microscope. One or more photographs of each piece were then taken.2 Finally, comparative literature was checked to determine the location and techniques of similar specimens. Haury's Sierra Ancha finds, and Hough's Gila River report proved to be the most valuable data in this respect.

2. Negatives are filed at the Danfer Art Museum.


The results of this laboratory research are set down in the following pages. Specimens are grouped together in classes (textiles proper, skirts, etc.), each piece within a class being described in detail. Preceding each class there are a few preliminary remarks coordinating the salient points regarding it. These summaries are intended to offer the casual reader a good view of the Montezuma Castle finds without necessitating a study of the piece analyses.



There are 61 specimens in the Montezuma Castle series. These are described under the following classifications:

Textiles331 specimens
Sandals10 specimens
Cords, braids, ropes8 specimens
Skirts6 specimens
Matting4 specimens
Net1 specimen
Twined-woven bag1 specimen

3. The term "textile" as used hereafter in this report refers to those fabrics which are products of the loom.

1. Textiles proper. Twenty-two of the 31 textiles are plain weave, 4 of these having paired wefts; 3, paired wefts alternating with woven strips or bundles of cotton string; and 2, single wefts alternating with woven strips or bundles of cotton string. There are 4 weft-wrap openwork textiles, 2 of which also show gauze weave. One piece shows gauze weave alone. Two of the fabrics are weft-float, 1 slit-tapestry, and 1 diamond twill.

2. Sandals. Of the 10 sandals, 7 are an over-one under-one braid.4 Four strands are folded to make 8 elements, which are braided from toe to heel. The eighth sandal is an over-one under-one braid of 6 cords folded to make 12 elements. Of the remaining 2, 1 is an over-two under-two braid of 27 strands, while the other is wickerwork made by weaving cords back and forth over 6 string warps.

4. An "over-one under-one braid" is also called "plain plaiting," and an "over-two under-two braid," "twilled plaiting." Braiding and plaiting are synonymous terms applied to the continuous intertwining of strands from one point downward.

3. Skirts. Four of the six skirts are made with a center pad of two string bundles, over which other strings are woven, their loose ends twisted to make a fringe on either side of the pad. The other two have a foundation unit of three strings, with other strings twisted over them forming a fringe on one side only.

4. Cords, braids, ropes. In this classification are 2 braids, 1 forming a coarse 3-strand rope, the other a flat over-three under-three braid of 49 strands. There is 1 decorative cord made of 6 strings fastened at intervals by other strings wrapping around them—probably a necklace. Three cords are made with a 2-strand twist, 1 with a 3-strand twist. There is 1 small ring of partly shredded yucca leaves, not twisted in any way. This brings the total of cords, etc., to 8. If the sandal ties be considered here, the number would be increased to 17. Six sandal ties are of untwisted, partly shredded yucca leaves, 1 is a 2-strand twist, 1 a 3-strand braid, and 1 an 11-strand over-two under-two braid.

5. Matting. Two of the four mats are an over-three under-three twill; one is a twill too badly damaged to determine the exact technique; and the fourth is a plain checker weave.

6. Bag. There is one fragment of a twined-woven bag.

7. Net. A portion of net, made by knotting strands together with square knots, is included in the series.


Cotton and yucca are the most common materials used. Cotton is found principally in the textiles proper, while yucca is the basic fiber in sandals, cord and ties, and skirts. Four varieties of yucca are found appearing in the frequencies:5

elata7 specimens
baccata12 specimens
macrocarpa4 specimens
mohavensis6 specimens

5. This follows Whitford's classification. He mentions the predominant type in a piece stating "... sometimes there might be Yucca elata and baccata in the same object as they did not seem to be religiously careful as to the species used so long as it was yucca. I will say, though, that in most cases the samples were pure."

4. An "over—one under-one braid" is also called "plain plaiting," and an "over-two under-two braid," "twilled plaiting." Braiding and plaiting are synonymous terms applied to the continuous intertwining of strands from one point downward.

5. This follows Whitford's classification. He mentions the predominant type in a piece, stating ". . . sometimes there might be Yucca elata and baccata in the same object as they did not seem to be religiously careful as to the species used so long as it was yucca. I will say, though, that in most eases the samples were pure."

The first two varieties are the only ones used in the yucca textiles. All four varieties are used for skirts, cords, sandals, etc.

Agave parryi is used in four specimens, all cord; Dasylirion wheeleri in two mats; and juniper bark in one mat.

Photomicrographs of the four types of Yucca, Agave parryi, and Dasylirion wheeleri are shown on Plate 1.

It should be noted that no traces of apocynum were discerned by Whitford or Jones in this series. It may be that this plant was extensively used only in the Anasazi area. Whitford says, concerning this:

"In a rather extensive collection from the American Museum of Natural History of materials from the Southwest I have found some considerable apocynum but . . . chiefly from specimens found in Utah and Colorado."

He adds:

"I have looked over all the information I have obtained in my investigations of Southwest fibers and find that in northeast Arizona and south Utah there is some apocynum used by prehistoric Indians, but those are the only places outside, of course, of Nevada and that region (mentioned above.)"

Table I in the Appendix lists specimens by number and gives all data on materials used in each.


Fourteen of the 61 specimens show coloring, produced either by dyeing the fibers, or by rubbing on paint after the piece was completed. In the latter case the color shows only on 1 surface of the piece. Red, black and yellow are the 3 colors used. The first, red, is by far the most common, appearing on 10 of the specimens. Black is found on 5 specimens, and yellow on 1. Red and black are combined on 1 piece, and black and yellow on another.

Twelve of the 14 colored specimens are cotton; the remaining 2 are yucca. This favoring of cotton as a vehicle for color is probably because textiles proper are woven from it, and merited decoration in the opinion of their weavers, whereas sandals, skirts, etc., made from coarser materials, were considered too rough for decoration.6

6. Although in Basketmaker times coarse fiber sandals were brightly colored.

Yellow and black are always applied as a dye, but 5 of the 10 red specimens have color rubbed onto their surfaces.

The sources of dyes cannot be determined in the majority of cases. The red in two specimens reacts to a test for iron. According to McGregor:7

"Dyes or pigments used in coloring yarn or fabrics may be divided into two general classes: organic and inorganic. In prehistoric cotton fabrics the inorganic dyes are by far the most common, and consist largely of three colors: red, produced from hematite or some other iron oxide; yellow, from a yellow ochre; and blue or green, produced from copper sulphate. These inorganic dyes may be readily determined by a superficial examination with a medium-powered microscope, for the dye material does not penetrate the fiber but clings to it in the form of grains. Organic dyes seem to consist of only black, dark brown, and a light blue. These dyes are relatively permanent and cannot be washed out, as can the inorganic types."

7. McGregor, 1931, pp. 3-4.

Jones notes substantially the same thing in his report on the pigments in this series:

"The pigments seem to fall into the two general groups of internal or absorbed pigment, and external or surface pigment. The former are usually of vegetal nature and the latter are usually mineral. The surface pigments appear as granular, uneven material applied to the fibers. There is considerable likelihood that surface soil adhering to the fibers might be confused with such pigment."

In line with these statements it is interesting to note that the red and yellow dyes in the Montezuma Castle series are apparently all external, or inorganic (mineral), whereas the black dye is internal, or organic (vegetal).

There are many cases in which fibers look dyed when viewed under the microscope, whereas the specimens from which they were taken show only darkening through weathering or scorching. In such cases the appearance of the piece as a whole has been considered as the determining factor. This decision checks well with Jones' statements made after examination of cotton textiles from several archeological sites in the Southwest. He finds:

". . . considerable variations in color of the fibers from textiles which seem to be undyed. Some have sufficient coloration, due to ageing and weathering, that under the microscope they appear to have been dyed."

He notes, further, on examining lint from seeds of cotton from a Basketmaker site, that the fibers have a lovely yellowish-green color. According to Jones ". . . scorched cotton fibers have a yellow-brown, evenly distributed color, suggesting pigmentation."

The Montezuma Castle specimens have been subjected to weathering and charring, which factors have undoubtedly modified their colors.

Table III, p. 93, summarizes information on color.


The conditions under which the specimens were found are set down in a letter of September 22, 1937, from Earl Jackson, then custodian of Montezuma Castle National Monument, as follows:

"These specimens were found in the cave rooms of a cliff ruin about 300 feet west of the structure known as Montezuma Castle. The cave rooms were at one time the back rooms of a cliff dwelling which towered five or six stories against the vertical cliff, and contained more than 40 rooms. This dwelling collapsed entirely, probably in prehistoric times, leaving a great mound of disintegrated wall material at the foot of the cliff on a ledge about 12 feet above the river terrace level. I believe the dwelling fell as the result of a fire, because all traces of ceiling structure found in the excavation of this fallen material were completely charred.

"The cave rooms of this ruin were excavated or cleaned out (see following letter) in 1927; this work yielded practically all the textile material which you are now studying. Unfortunately, the work was not done scientifically, and no record was kept of details encountered in cleaning out the rooms.

"In the winter and spring of 1933-34, I was in charge of a project under CWA to clean out the material at the base of the cliff. This work uncovered the bases of the walls and the floor outlines of the houses which had adjoined the caves in front. While we found practically nothing in textiles in this CWA dig, since the material had been exposed to the weather, we were able to ascribe a tentative date for beginning of occupation as around 1100 A. D. A tentative closing date for this ruin would be 1425 A. D."8

8. Jackson and Van Valkenburgh, 1954.

George L. Boundey, formerly custodian of Gran Quivira National Monument, New Mexico, in a letter of April 11, 1938, states:

"In looking over the many caves in the vicinity of the main ruin, we found that vandals had excavated through the debris in practically all of them. Many modern digging tools were found, some dating back to the days when Camp Verde was an army post. In a pile of debris overlooked by the pot hunters we found several small objects. Mr. Pinkley [the late Frank Pinkley, then superintendent of Southwestern National Monuments] pronounced these museum specimens and suggested that I, with the two engineers, start a thorough cleaning out of the debris.

"We started work in each cave at the opening, digging down to the stone floor and passing back each shovelful through the opening where it was dumped on the ledge. In the light every fragment of pottery and anything which suggested man's work was placed in a paper sack and numbered.

"The textiles were taken from the largest of the caves, which evidently housed several families. Here, owing to the darkness and the depth of the debris, vandals had left the lower portions undisturbed. This room has several raised portions of the floor and we labeled the bags and boxes according to the different levels on which the material was found."

An analysis of pottery sherds from this site places it in the Pueblo III and IV periods, and indicates an infiltration into the region of trade influences from other areas. It places the most important period, culturally speaking, towards the end of occupation—in Pueblo IV—which probably lasted as late as 1425 A. D.


The dating of the perishable specimens checks well with the preceding remarks as to the period of occupation of the site. Weft-wrap pieces, matting, and cord sandals with braid ties, like the Montezuma Castle specimens, are placed in Pueblo IV by Haury in his Canyon Creek paper.9 Other specimens similar to ours are late Pueblo III or IV according to his classification. It is assumed, therefore, that the Montezuma Castle finds fall in these two periods. The great exception to this is the small fragment of what could be a Basket Maker III (or II) twined-woven bag. The technique of weaving and decoration check absolutely with Guernsey and Kidder's description of Basket Maker bags.10 The only difference is the use of yucca here, as opposed to the apocynum of their northern specimens.

9. Haury, 1934.
10. Guernsey and Kidder, 1921, pp. 65-74. Hereafter this publication is listed as PMP VIII, 2.


In the case of the textiles proper the rag-weft plain weaves and the weft-wrap and gauze fabrics indicate that the weaver's art at Montezuma Castle was closely akin to that at Tonto National Monument, Canyon Creek and the Upper Gila which flourished at about the same time. Weft-wrap openwork and gauze are not unknown from Anasazi sites, but are much more characteristic of central and southern sites, and seem to be earlier in the south. Rag-weft is not definitely established as one of the Anasazi techniques. Slit-tapestry, on the other hand, appears to be a very early technique in the Anasazi area, dating back into Basketmaker times. The diamond twill, found at Montezuma Castle, ties in more closely with the northern tradition than with central and southern sites, also. Weft-float has not been reported from other prehistoric sites.

Matting and sandals are much like those from the Sierra Ancha and the Upper Gila. Two sandals described by Mera,11 from the Guadalupe Mountains near Carlsbad, New Mexico, are also very similar to those described here.

11. Mera, 1938.

Further comparative data will be found in the preliminary notes before each weave class. Much of this material was gathered after the Montezuma Castle report was written. It will be more fully dealt with in a study of prehistoric Southwestern textiles to be published at a future date.

<<< Previous <<< Contents>>> Next >>>

Last Updated: 04-Mar-2008
Copyrighted by Southwestern Monuments Association
Western National Parks Association