A Probable Prescott Branch Site
on the Lower Agua Fria River
Three years ago the Arizona Archaeological Society was incorporated under the laws of the State of Arizona. Soon after its inception the Society undertook the excavation of a site on the lower Agua Fria River, nine miles south of Lake Pleasant. The work was performed on weekends using volunteers from among the membership as labor. The excavations were originally under the direction of Mr. William Lahti. However he soon moved out of town and the job of dig director fell to Mr. Edward Pritchett. Mr. Pritchett had no previous experience in archaeology; however, having been educated and working as an engineer, he knew the basics of data collection and has done a good job despite the differences between his field of training and those of archaeology.
The biggest problem faced by the society has been the training of its membership, there being very few members who have had any previous experience at all, and of those that have very little time to contribute. Due to the lack of training, field notes are few in number and generally poor in quality. Therefore the bulk of the following is base upon the observations of Mr. Pritchett and of this writer. Mr. Pritchett has been working at the site since field work was initiated. The writer has worked a bare four days.
The site is commonly called Calderwood Site, or just plain Calderwood, being named after a prominent hill in the vicinity. The official designation is Ariz T:7:1. The site is located on land owned by Mr. David Rich, who has kindly permitted the society to carry on its operations.
Figure 9. Calderwood (AZ T:7:1 (PG?)) site map.
Medium resolution image, 580 x 1224 pixels, 90KB
High resolution image, 3550 x 7490 pixels, 386KB
Calderwood is located on a terrace on the east bank of the Agua Fria River. The terrace rises approximately forty feet above the present flood plain of the Agua Fria. The terrace seems to be formed predominantly of alluvium, containing gravels and boulders up to two feet in diameter. Roughly the top seven feet of the terrace is cemented by caliche. No apparent stratigraphy was noted. The surface of the terrace is covered by volcanic debris, possibly from volcanic activity which occurred about a mile to the east of the site.
Nobody has yet compiled a resume of the flora and fauna to be found in the region today. However, the region is of the same general ecological pattern as that presently found in the vicinity of Phoenix. The Agua Fria River, up to the creation of Lake Pleasant, had a flow for at least part of the year.
There are several other prehistoric sites in close proximity to Calderwood. To the south approximately 150 yards, lies a site (Ariz T:7:2) similar to Calderwood, but with several distinct differences. A little farther south and across a deep gully is the remains of a single room. This room is situated on a wide area of very thin trash. The trash gives the appearance of being of a different time horizon from the solitary room. Architecturally (from surface indications) the solitary room is more closely related to Ariz T:7:2 than to Calderwood.
There are sites on the west bank of the Agua Fria roughly across from Calderwood which may represent the same time and culture as Calderwood (Hiser: Personal Communication).
The site as determined by surface indications prior to excavation (see Fig. 9) consisted of twenty-five non-contiguous structures. Twenty-four of these lay within an incomplete compound wall, the twenty-fifth laying outside the northern end, to the east of the compound wall. With the exception of two rooms which proved to be rock piles, excavations to date has tended to confirm this, no additional structures being located.
There has been very little variability noted in the construction and orientation of these rooms. With two exceptions, rooms X and XXIV, the long sides consistently run in a general north-south direction with the door being found midway in the east wall. The long sides of room X run from northeast to southwest, the door facing Southeast (Pritchett: Personal Communication). Room XXIV is oriented east-west, the door probably being in the north wall.
Wall construction is of two types, both of which will be found in the same structure. In type 1, the masonry will be based at the same level as the floor. See fig. 10a. In type 2, the sides of the pit were lined with stone slabs to the original ground level. Above the slabs, a coursed masonry was used, being based partially upon the native soil and partially upon the slabs. See fig. 10b.
Figure 10a. Schematized drawing of block masonry used in structures set on the surface.
High resolution image, 503 x 666 pixels, 86.6KB
Figure 10b. Schematized drawing of block masonry used in structures set on the surface.
High resolution image, 503 x 666 pixels, 86.6KB
The masonry is crude, using volcanic blocks in the main, with an occasional river boulder finding its way in. The blocks were not worked in any way. They are set in a thick caliche mortar with no chinking noticed. Over the interior of the walls, a thick caliche plaster up to 10 CM. Thick had been laid. Exteriors of structures were also plastered. An area in Room XXIV where the plaster had fallen inward intact, indicate that the walls stood to a minimum height of five and one half feet above the floor level. There was enough stone found in the fill to indicate that the masonry also extended to this height. This general height had be estimated at by others using different methods (Pritchett: Personal Communication).
So far, no evidence of roof support is available. Many of the floors have been too fragmentary to preserve posthole traces (Pritchett: Personal Communication.)
In the rooms where fire pits have been located to date, they are consistently found in front of the doorway, roughly one-third of the distance from the door to the back wall (Pritchett: Personal Communication). Data concerning all fire pit forms is not immediately available at the time of writing this essay, they shall have to be described at a later date.
The compound wall, as shown in Fig. 9, is incomplete. Whether this is due to it having never made a complete circuit of the enclosed area or having been ripped out at a later date has yet to be determined. Mr. Pritchett has been informed that people used to dig around the wall looking for "arrowheads" (Pritchett: Personal Communication). The compound wall, where it has been exposed, is basically a type 1 masonry as described above. It appears that it may be based in a shallow trench. It was plastered on both sides. No estimates as to the original height of the wall have yet been made.
Discussion: The field work and notes to date are insufficient to make extensive comparisons with other sites, but some points can be made.
The slab lining of the type 2 masonry has a parallel farther up the Agua Fria at the Fitzmaurice Ruin. This parallel extends to finding both type 1 and type 2 masonry in the same room (Spicer and Caywood 1936:96). At Fitzmaurice Ruin this was due to the building of the structure in the side of a hill where the uphill portion had been excavated. The uphill excavation had been slab lined to the ground level, from where a coursed masonry was built above. On the down hill slope, the coursed masonry was built directly on the ground level (Spicer and Caywood 1936:96). Whether the occurrence of both type 1 and type 2 masonry in the same room at Calderwood is due to the same reasons as at the Fitzmaurice Ruin has yet to be fully determined. The trait of slab lining may end up being of very little diagnostic value, as slab basing is quite a common characteristic in Southwestern architecture, being found at such diverse areas as Gran Quivira (Thomas Caperton: Personal Communication) and near Globe, Arizona (Vivian: Personal Communication).
Another parallel is in the use of rough blocks of stone in the construction of walls (Spicer and Caywood 1939:98). The materials are different, granite being used at the Fitzmaurice Ruin and blocks of a volcanic material at Calderwood, but this would probably be the difference of available materials.
One principle difference between the Fitzmaurice Ruin and the Calderwood Site is that there was a clustering of contiguous rooms at the Fitzmaurice Ruin along with outlying non-contiguous rooms. There is no clustering of contiguous rooms at Calderwood. However, Ariz T:7:2, just south of Calderwood and by surface indications a culturally related site, there appears to be both a clustering of contiguous rooms as well as several outlying non-contiguous rooms, as at the Fitzmaurice Ruin.
Ariz T:7:2 differs from both Fitzmaurice and Calderwood in that a much greater percentage of the stones which went to make up the walls are not unworked slabs, but rounded river boulders. In this respect it matches the architecture at King's Ruin. At King's Ruin the core of the walls were rounded river boulders. They were piled haphazardly and could not have remained standing except that a tremendous quantity of clay was used to hold them in place (Spicer and Caywood 1936:22). King's ruin did not have outlying non-contiguous rooms.
One feature which separates both Calderwood and Ariz T:7:2 from King's Ruin and Fitzmaurice Ruin is the presence of compound walls at the first two and their lack at the last two.
The writer has looked over only a comparatively small lot sherds from Calderwood, and is not in a position to discuss it in great detail.
From sherds submitted to Mr. J. Olaf Sund, several things of interest were learned. Mr. Sund's correspondence is reproduced in full below.
Messrs Don Dove and Ed Pritchett,
Phoenix Chapter, Arizona Archaeological Society.
Binocular Microscope Study of Pot Shards from the Calderwood Site on the Agua Fria River.
The following includes a brief description of the materials used in the clay pots manufactured at the Calderwood Site on the Agua Fria River.
Fragments of the Clay pots were merely broken and crushed with a common hammer and anvil and examined with a binocular microscope at various magnifications. The crushed shards were examined under water such that washing was effected but with no loss of constituent material.
The typical pots include an assortment of larger fragments a very fine-grained material. The larger fragments are of three predominant constituents that include pure quartz, pure feldspar and a schist. The quartz pieces range from 1/4 to 1 mm, and the feldspars more generally about 1 mm in diameters. The schist fragments are consistently larger and more uniformly the same size, being about +/- 5 mm in diameter.
The quartz fragments are a clear glassy variety that have an angular shape and very "fresh" surfaces. That is to say, there are no pieces that have rounded corners and edges, or a "frosted" surface that would be expected from normal weathered and transported sand grains.
The feldspar fragments are milky white crystals that display good angular cleavage surfaces. These feldspars should readily breakdown and decompose from the normal processes of weathering and transportation.
The combined quartz and feldspar materials make up about 20% of the pot constituents.
The more predominant fragments are muscovite schist that makeup approximately 40 to 50 % of the entire pot constituents. These are broken fragments that are generally tabular with semi-round and angular shapes. All fragments are relatively fresh and non-weathered. The nature of this schist is such that it would readily weather and decompose. However the uniform fragment size, the shapes, and the lack of any real decomposition suggest a lack of normal weathering.
The finer-grained material or groundmass is a powdery substance that makes up about 30-40% of the pot constituents and is undoubtedly the clay and sand mixture normally used in pot manufacture.
Very little can be deduced from the fine-grained clay other than its source was probably from the nearby Agua Fria River. The fine round sand grains would attest to this.
Certain ideas can be guessed at with regard to the coarser fragments. It seems a reasonable assumption that this coarser material was physically crushed from fresh rock immediately prior to the pot manufacture. This inference is based on the angular, non-frosted quartz, the feldspar crystals with good cleavages etc. and the angular uniform sized schist fragments.
These latter materials are apparently what are termed the "temper".
A follow-up field study would be required to properly establish the material source. Despite the fact that this has not been done, certain speculation is possible however. A rock that would yield such materials as are used in the temper would be a granitic intrusive. In the immediate vicinity of the site there are large masses of both Precambrian and Laramide "Granite and related Crystalline Rocks".
J. Olaf Sunday
June 10, 1968
The sherds examined by this writer for the most part (aided only by a 10X lens) fall within the description given by Mr. Sund. However, there is another group which was not apparently represented in the sample examined by Mr. Sund. This group is also characterized by a very heavy temper concentration, however the temper is a quartz sand.
Upon the initial inspection, the writer received the impression that the technique of manufacture was coil and scrape. However, further scrutiny revealed that the method more closely may be described as being a coil paddle and anvil. The initial observation of a coil and scrape was due in fact to efforts to smooth the surface after the use of the paddle and anvil. There are several sherds which demonstrate this quite clearly.
In sherds with a predominant mica schist temper, the temper regularly appears on the surface. In the sand temper sherds, this is generally not the case.
Decorated pottery has fallen into two main groups. The first is a Black-on-gray, which will probably upon an examination of a larger sample be found to have variants, and the second, Casa Grande Red-on-buff. Although no figures are available, the Black-on-gray seems to be more common. The Black-on-gray also seems to be made of the same paste as the schist tempered plainware. An interesting find is that of an olla sherd decorated on the interior.
The pottery from Calderwood, both plain and decorated, except [sic] Casa Grande Red-on-buff, falls into the general descriptions for King's and Fitzmaurice Ruins. That is it crudely manufactured and decorated (Spicer and Caywood 1936:28). The method of manufacture is also found at King's Ruin (Spicer and Caywood 1936:30-1). The decorated pottery being made of the same paste as the plainware was also found at Fitzmaurice Ruin (Spicer and Caywood 1936:102).
The single most diagnostic ceramic feature at this time however is the occurrence of ollas decorated on the interior. This feature is found in the Prescott Branch sits (Spicer and Caywood 1936: 41, 103) and to this writer's knowledge, no others.
This section will not be complete, not enough time being available to do the job. Some of the material from Calderwood will be compared to material from other Prescott Branch sites and where references are immediately available, to other sites. This will not be done in terms of detailed analysis, but only on the obvious general traits.
Metates: Two types at Calderwood, 1) Basin, shallow; 2) Trough open at both ends, deep. This condition exists at the Fitzmaurice Ruin (Spicer and Caywood 1936:111). Basin at Nantack Village, Point of Pines; full trough not represented, only closed trough (Breternitz 1959:39). Basin and trough at Forestdale Valley (Haury and Sayles 1947:64-5).
Manos: Round and rectangular at Calderwood. Neither round type nor rectangular specifically mentioned at Fitzmaurice Ruin, just fact that some two-faced, but most single surfaced (Spicer and Caywood 1936:112). Round and rough rectangular found at Nantack Village, Point of Pines (Breternitz 1959:40-1) and at Forestdale Valley (Haury and Sayles 1947:66-7). Round manos found in Perkinsville area (Fish: Personal Communication). The author does not know of any round manos found in Classic Hohokam sites.
Three-quarter Groove Tools: Two found at Calderwood. Both may be broken axes reused as mauls. One has the border of the grooved raised, the other has neither border raised. At King's Ruin three types found, one with neither border raised, the second with both borders raised, and the third with only one border raised (Spicer and Caywood 1936:55-6). Fitzmaurice fits in with those of King's Ruin (Spicer and Caywood 1936:112). Only type without raised border found at Nantack Village, Point of Pines (Breternitz 1959:48). The author knows of no axes from the Classic period Hohokam which exhibit a raised border around the groove.
Modeled Spindle whorls: Several specimens found at Calderwood. Two items described from Fitzmaurice Ruin which fall into description of modeled spindle whorls, though not specifically called so (Spicer and Caywood 1936:111). Modeled spindle whorls are quite common and well known to the Classic Hohokam by the author.
The principle conclusion is that Calderwood does not represent a Classic Hohokam site as Hohokam is defined today. There are too many traits found at Calderwood which are missing from Classic Hohokam, such as stone masonry with slab basing, Black-on-gray pottery, ollas decorated upon the interior, round manos, and three-quarter groove axes with raised borders around the groove. Among Classic Hohokam traits either missing or not yet found are cremations, Gila or Salt redware, and shell work.
It is true that some shell work has been found, however this writer believes that these represent trade items. This is based both upon the relative paucity of shell items and the almost total lack of waste shell found at Calderwood. Waste shell is very a common ingredient in all Hohokam trash tested by this writer, while yet finding any at Calderwood. Mr. Pritchett has said that there was some found in the trash mound at Calderwood, but very little (Pritchett: Personal Communication).
The dating of Calderwood has not yet proceeded very far. Paleomagnetic samples have been taken from one fire pit, but the results have not yet been returned.
Tradeware is little better. Tusayan Black-on-red, Cibola Black-on-white, and Casa Grande Red-on-buff the only types so far known so far. Casa Grande Red-on-buff is only of limited value, it having a fairly long history, extending through both the Soho and Civano phases of the Hohokam Classic period. Recent dates on Casa Grande Red-on-buff place an early date in the neighborhood of A.D. 1050 (A.E. Dittert: Personal Communication), although there is some evidence to indicate that Sacaton Red-on-buff, the precursor of Casa Grande Red-on-buff, was manufactured as late as A.D. 1250 in some areas (Ruppe 1966:12). This writer is not ready however to accept the later date at this time, it being based on survey work while the excavations of this writer do not bear it out. Also, Casa Grande Red-on-buff is found in association with Tusayan Black-on-red at Calderwood, which is dated A.D. 1050-1150 (Colton 1956). No Sacaton Red-on-buff has been seen from Calderwood by this writer.
Calderwood apparently was abandoned prior to the introduction of Jeddito Black-on-yellow at about A.D. 1325 (Colton 1956). The basis for considering Jeddito Black-on-yellow to be of value, is that Schroeder considered the Agua Fria River to be one of the trade routes for this type into the Salt River Basin (unfortunately this is from a secondary source, being from Hayden 1957:196). Therefore if Calderwood had been occupied into the time that Jeddito Black-on-yellow had been manufactured, then it should be represented among the tradeware, which it is not. I therefore place a tentative length of occupation for Calderwood of from A.D. 1100 to A.D. 1300.
The date for the site just south of Calderwood, Ariz T:7:2, is not yet even to be guessed at, although the writer feels there is a strong possibility that the two sites are at least partially contemporaneous.
Currently the Calderwood Site is of great importance for three reasons. The first being that more information is being gathered on a culture which has not been solidly defined in the past. The second reason is that we have an apparently non-Hohokam group located in a area previously occupied by the Hohokam. The third reason is that the site would apparently place the Prescott Branch at the time of about A.D. 1200, farther south than it has previously been defined.