Finishes Analysis: Kiva Q

Kiva Q has been analyzed and partially treated for the last four years by several researchers. The following observations are a combination of past and present field investigations, and previous (M. Slater, 1999) and current analysis. After initial archival and field investigations, representative samples from the major architectural elements of the kiva: lower banquette wall, pilasters, and upper inter-pilaster walls were selected for further study.

Panoramic photomontage of Kiva Q showing all sectors. Sample locations are indicated in red.

Architecture
Kiva Q is a large, well-proportioned kiva of slightly oval plan. Its masonry is finely constructed with dressed and coursed stones and precisely positioned pilasters. Although the western portion of the kiva is destroyed (since the earliest descriptions), the remaining walls are in excellent condition and retain their finishes, especially along the banquette.

Kiva Q was first indicated in plan by Nordenskiöld (1893, 60, Pl. XI) as Estufa 51. The earliest detailed photographs by Nusbaum in 1907 (MEVE 9523/#19) and Vreeland (in Fewkes, 1911, Pl.19) display well preserved interior walls. The current bichrome scheme is clearly visible and the recent masonry deterioration and plaster loss in the lower banquette (Sectors 10 & 11) is minimal. Two niches display original closure insets that were subsequently removed after 1954. The deflector, although partially collapsed on the southern end, displays a stone slab top (missing by 1954 and possibly not original) and plastered walls. The deflector's plaster on the northern end clearly exhibits a dark lower band that has been confirmed through field examination and cross-sectional analysis as a red dado with triangular embellishments on an upper white field on the east face. Collapse of the upper interpilaster walls in Sectors 3 & 4 are clearly visible and explain Fewkes' repairs in those areas as well as on the deflector and at the broken northern end (Fewkes, Pl.15). Stabilization work in 1934 by Morris and Lancaster repaired the broken end walls again, removing Fewkes' earlier stabilization. Later (date unknown), the floor was replastered.

Surface Finishes-Stratigraphy
Numerous simple and complex surface finish schemes are evident at Kiva Q including several with discreet (e.g., painted hands and animal paws) and repetitive (finger tip dado band) embellishments.

Representative banquette wall samples taken together constitute one complete sample. Sample 3 contains the three initial layers while Sample 1 continues the sequence with the most recent five layers. Thus, one can infer that the lower banquette wall of Kiva Q originally had at least eight finish layers. In contrast, the pilasters in Kiva Q have fewer finish layers. Samples 41 and 43 are representative of the lower pilasters and Samples 38 and 40 are representative of the upper pilasters. The lower pilasters display four layers while the upper pilasters have only three layers. As represented by Sample 46, the upper walls between the pilasters (interpilaster wall) have only two layers. These stratigraphies relate directly to the frequency with which individual architectural elements were finished during each scheme or campaign of finishing. As in most kivas, banquettes were refinished more frequently than pilasters, which were refinished more frequently than the upper, inter pilaster walls. This suggests a hierarchy of treatment based on element or sphere of activity, i.e., the banquette level being the zone of sitting.

The thickness of the finish layers of most of the samples in Kiva Q is between 200-500µm. On site investigation, however, revealed that many of the finishes vary in thickness and can be further defined as plasters or washes depending on the sample location and grain size distribution. In all the samples, the first layer appears to be a coarsely textured base or leveling plaster between 500µm-1cm thick. Most successive individual finish layers are separated by a thin soot layer suggesting the passage of time between applications. Plaster layers were applied sporadically throughout the kiva, presumably to repair or level areas of significant finish loss over time, especially around niches. Washes were used to both color and delineate the space as design elements and embellishments. Washes were also applied as discreet designs, rendered as repetitive or discreet motifs such as the finger-tipped dado border in Scheme 3 and the red painted handprints and white animal paw prints in Scheme 6. Surface characteristics of the finishes in Kiva Q vary from earlier smoothly polished or burnished surfaces to the most recent layers exhibiting distinct application striations made with the hand.

On-site investigation has revealed five successive schemes for Kiva Q. Microscopic investigation of the finishes stratigraphy has confirmed the on-site investigation with one possible additional scheme. All six schemes can be considered as main schemes, however Schemes 2 and 6 are transitional in that they incorporate finished elements from earlier schemes. A conjectural sequence of schemes was made in gouache based on the Munsell color match of each layer.

As in other kivas, the earliest scheme (Scheme 1) is a simple, although full finish covering all interior surfaces in a monochromatic brown plaster. All later schemes are complex, being a combination of two or more colored finishes delineating the banquette and pilasters from the upper interpilaster walls. In Scheme 3, spatial delineation has been elaborated through the use of a highly contrasting red dado and white field including the pilasters and upper walls. During Schemes 4, 5 and 6, a heavy black soot deposit was retained or intentionally smoked to create a uniform black upper inter-pilaster wall that was incorporated into the red and tan schemes. Unlike the telltale evidence of hand-application of the plasters and washes, these black-sooted areas are smoke-deposited and reveal no manual manipulation. The last (most recent) scheme (Scheme 6) is an unusual vertical BI-chrome treatment of the banquette into pink and tan hemi-circles created by the selective refinishing of the southern half of the kiva. The precise division and intent of this refinishing is evident in the clear termination of the applied wash through the center of the niche on the eastern banquette.

First (earliest) scheme:
The first scheme is a simple, full finish monochrome application. It consists of a thick brown (10YR4/4) leveling plaster that was applied all over. This scheme was probably extant for a short time as very little if any soot was found between the leveling plaster and the next finish layer.

Second scheme:
The banquette and entire pilasters were next finished with a very pale brown (7.5YR 5/3) wash. Inter pilaster upper walls retained the brown plaster of the first scheme thus creating a complex bichrome scheme. Partial diagonal red (10YR 4/8) bands on the upper banquette walls were also found as discrete embellishments on this layer.

Third scheme:
The banquette was subsequently divided into a lower red (5YR5/6) dado with grouped fingertip embellishments extending into the white (10YR8/1) field above. Entire pilasters and upper inter pilaster walls were finished white as well. The deflector displays the same red plaster with a fragment of a small rectangular niche centered on the inward (east) face. Evidence of embellishments of red triangles at the border of the dado on the white field was also found on the deflector.

Fourth scheme:
The banquette and pilasters next received an overall red wash (5YR 5/6) similar to the color of the previous dado of Scheme 3. The upper inter pilaster walls were blackened with a thick layer of soot that appears to have been consciously retained or created as part of the scheme.

Fifth scheme:
The banquette and lower pilasters received a light brownish pink (10YR 7/4) wash. The upper pilasters and inter pilaster walls remained sooted black. Embellishments of two red (10 YR 6.5/4) painted handprints on Pilaster 3 and a white animal paw print on the dado below were also applied.

Sixth scheme:
Scheme 6 is Scheme 5 modified into an unusual hemi circular BI-chrome design by the application of a tan (7.5YR7/4) wash to the southern half of the kiva, dividing the space in half directly through the central niche on the eastern banquette wall. The northern half retains its slightly sooted earlier finish from Scheme 5. The interpilaster walls and upper pilasters remain sooted black. The southern niche retains its is red plastered interior from the fourth scheme.

Fabric Analysis
Quantitative analysis of the micro fabric of the plaster finishes of Kiva Q reveals both similarities and differences between layers. This is evident in the ratio of matrix to aggregate and the arrangement, size, shape and frequency of individual particles and voids within the layers as a whole, and within features themselves.

The ratio of the matrix to the aggregate indicates the percentages of matrix, aggregate and voids. The ratios determined for the leveling plaster (Layer 1 on most samples) reveal the highest percentage of aggregate to matrix of all the layers. With the exception of Layer 7-Sample 1, all other wash layers display a higher matrix to aggregate ratio. This confirms the distinction between plasters and washes based on the mineral matter ratio. Further analysis of the leveling plaster samples (Samples 3, 38, 41 & 46) suggest two types of leveling plasters. That on the upper walls of the kiva has a nearly equal ratio (1:1) of aggregate to matrix while the plaster on the lower walls has a higher ratio (greater percentage of aggregate). This anomaly was observed in most kivas and may be due to differential weathering of the matrix in the upper areas.

The aggregate grain size distribution patterns for the layers have also been plotted and superimposed with a representative best-fit curve. The aggregate is generally a fine sand and the grain size distribution shows a similar exponential distribution pattern; the washes possessing the best sorting. The differences in the graphs are mainly in the amount of aggregate in each layer, which can be explained by the variable thickness of the layers and therefore more mineral material.

The shape factor of an aggregate is the ratio of an aggregate's area to the area of a circle with an identical perimeter. Effectively, this is a measurement of an object's "roundness" on a scale from 0 (straight line) to 1 (perfect circle). Image analyses reveals that the average shape factor of the aggregate found in each consecutive layer (Samples 1 & 3) varies between 0.43 - 0.68. The boundaries or perimeters of the aggregate from each plaster layer were also measured and an average of the perimeter of the aggregate for each layer was derived. In the plaster layers from Samples 1 and 3, the perimeter of the aggregate varies between 140µm - 500µm. While the visual change from the aggregate to the matrix is abrupt, many other changes are gradual or diffuse such as calcification. Clarity depends on the magnification. The average perimeters of the aggregate from the leveling plasters in Kiva Q further suggest that there are two types. The lower wall plaster consists of an average of smaller perimeter size aggregates. A small fraction of the lower wall aggregate tends to be rounder than that found in the upper walls. This similarity in type is confirmed also by a higher ratio of aggregate to matrix in the lower walls.

Finally, the aggregate is found to be anhedral (mineral without crystal faces) and irregular with very few subhedral (crystal faces) grains. This may suggest that the aggregate was artificially prepared by grinding and possibly sieving to create the desired grain size.

Qualitative X ray diffraction (XRD) analysis identified the matrix of the red and brown plasters and washes as kaolinite and illite/smectite with calcite, feldspars and hematite. Red earth mineral pigments may have been mixed into the soil to enhance the color for certain washes. The matrix of the white colored washes was identified as gypsum, although other white pigments such as calcium carbonate and kaolinite have been found at other sites used alone or in conjunction with gypsum. Tan, buff and pink layers were also found to contain gypsum, however this could be the result of post-application contamination from ground water.

XRD revealed that the aggregate found in the plaster and washes is mainly quartz and plagioclase feldspar. (Slater 1999, p.58) Traces of nitrate sulfur were also found confirming the use of gunpowder in the late nineteenth century to blast open areas of the ruin during early pot hunting expeditions. Organic matter such as plant fibers and charcoal, crushed plaster and sandstone have also been found in leveling plasters, banquette tops, and bedding mortars. This is most likely midden material in the mud sources used either intentionally or through contamination during mixing.