DONATE TODAY
13
Mad Dog Tower
James H. Kleidon
Contents
- Pit Structure and Courtyard Area (Sampling Strata 2 and 3)
- Structure 2 (Kiva)
- Construction
- Stratigraphy
Dating
Interpretations - Sampling Unit 87N/92E
Sampling Units 88N/95E, 87N/99E, and 86N/97E
Sampling Unit 90N/92E
Outer Periphery (Sampling Stratum 6)
Introduction
Early documentation of Mad Dog Tower was made by T. Mitchell Prudden in 1903 and by Jesse W. Fewkes in 1917. In Fewkes's 1919 report, he refers to the site, illustrates the tower, and mentions Prudden's earlier visit to the site (Fewkes 1919:57). According to Gleichman and Gleichman (1989), a more systematic archaeological survey was undertaken by the University of Colorado Mesa Verde Research Center in the lower Sand Canyon area in 1965. It was then that Mad Dog Tower was recorded as Site 5MT181. In 1975, the Mesa Verde Research Center, under the supervision of Curtis W. Martin, conducted a ruins-stabilization inventory of the Sand Canyon area (Martin 1976). The tower at the site was stabilized in 1977 under the supervision of Betsy Tipps (Tipps 1978) as a result of the assessment of stabilization needs at 5MT181. Most recently, the site was recorded in an archaeological survey undertaken by Native Cultural Services in 1988 (Gleichman and Gleichman 1989). It was excavated in 1988 by the Crow Canyon Archaeological Center as part of the Sand Canyon Archaeological Project Site Testing Program.
Mad Dog Tower is located on the crest of a ridge on the first terrace of Sand Creek (Figure 1.7, Figure 13.1, and Figure 13.2). The main channel of the creek lies 460 m to the west. Elevation of the site is 1789 m (5870 ft). The site slopes gradually (5 percent) to the south, and the sparse vegetation cover is predominantly pinyon, juniper, and sagebrush. Historic land use includes grazing and nearby carbon dioxide drilling and extraction. The site has also been vandalized.
Remains visible on the surface before excavation consisted of a stabilized tower (Structure 3), a sandstone rubble mound, and an area of darker soil with the densest artifact scatter on the site, interpreted to be the midden (Figure 13.3). A looter's pit dug into the sandstone rubble mound had exposed masonry wall alignments.
On the basis of the surface remains, six sampling strata were defined at Mad Dog Tower (Figure 13.4). These are the surface structure, pit structure, courtyard, inner periphery, midden, and outer periphery areas, designated as Sampling Strata 1 through 6, respectively. The surface structure sampling stratum was defined by the sandstone rubble pile and partially standing tower. A pit structure depression was not visible on the surface, so the boundaries of the pit structure sampling stratum were defined on the basis of where one would expect to find a pit structure at a typical site--that is, south or southeast of the surface rooms. The midden sampling stratum encompasses the area of dense artifact concentration with slightly darker soil. The courtyard sampling stratum is bounded by the surface rooms and the area surrounding the suspected location of the pit structure. The inner periphery surrounds all these cultural units at the site. The outer periphery extends from the inner periphery to the edge of the surface artifact scatter at the site.
Twenty-nine randomly selected and four judgmentally located sampling units were excavated at Mad Dog Tower (Figure 13.4). The judgmental squares were excavated to facilitate the excavation and interpretation of the random squares.
Figure 13.5 illustrates the major cultural units defined during testing. These include a single surface room (Structure 1), a pit structure (Structure 2), the tower (Structure 3), and the midden. On the basis of the pottery recovered, these features are believed to date to the late Pueblo III period.
The grid at Mad Dog Tower was oriented toward wall alignments exposed in the looter's pit in the rubble mound. This grid is 28 degrees west of true north. Vertical Datum A was located at 80N/96E and arbitrarily assigned an elevation of 100 m.
The following descriptions are organized by sampling stratum. The cultural units investigated are discussed as part of the sampling stratum in which they were found.
Surface Structure Area (Sampling Stratum 1)
Sampling Stratum 1 includes the masonry tower and the sandstone rubble pile associated with a fallen room (Figure 13.4). Of 25 sampling units in Sampling Stratum 1, four were randomly selected for investigation. These are 92N/90E and 91N/92E, both located outside of cultural units, and 92N/94E and 84N/101E, located in Structures 1 and 3, respectively. Judgmental units were excavated adjacent to these two random squares to facilitate the interpretation of the two cultural units. These judgmental sampling units include a small portion of 91N/94E, located within Structure 1, and 83N/101E, located both inside and outside Structure 3. The sampling units contained within the cultural units are described below either under Structure 1 or Structure 3.
Structure 1 (Surface Room)
Excavation within Sampling Stratum 1 demonstrated that only one room was present in the rubble mound. This surface structure is illustrated in Figure 13.6. A looter's pit excavated within the room uncovered portions of the south and west masonry walls. The limited testing consisted of excavation of random sampling unit 92N/94E and a judgmental sampling unit measuring 1 m by 20 cm adjacent to, and south of, the random square. Descriptions of Structure 1 are based on these excavations and the walls exposed by the looter's pit. The floor surface and a section of the interior face of the room's south wall were located within these sampling units.
Construction
Walls. Structure 1 is inferred to be rectangular, and it measures approximately 1.9 m long by 1.4 m wide. These dimensions produce a floor area of 2.66 m². Distance from the preserved top of the south wall to the room's surface is 23 cm. The south wall is semicoursed, two courses high, and a single stone thick. Sandstone used in the south wall is unshaped. The mortar joints found between the sandstone courses average 2 cm in size. The interior face of the south wall has no chinking.
Roof. There was no evidence of roof fall within the Structure 1 fill. It is assumed, however, that a roof did exist. The roof was likely supported by the walls.
Floor. The floor of Structure 1 is an unprepared, use-compacted surface. This surface lies at the same elevation as the base of the wall and is probably the unaltered prehistoric ground surface. No artifacts or features were located on this surface.
Stratigraphy
Fill inside Structure 1 consists of a single stratum. Stratum 1 is a loosely compacted, red brown sandy loam measuring a maximum of 35 cm from the modern ground surface to the floor of the room. Inclusions in Stratum 1 consist of minimal amounts of small charcoal pieces, artifacts, and many sandstone pieces of various sizes. The sandstone found, although unshaped, is probably wall collapse. Artifacts include 18 pieces of chipped-stone debris and 21 sherds (19 Indeterminate Local Corrugated Gray, one Pueblo III White Painted, and one Late White Unpainted). Stratum 1 is postabandonment wind and water deposits and structural collapse.
Interpretations
Structure 1 is a masonry room of unknown function. The unprepared floor and the mostly unshaped sandstone used in the walls suggest that only a minimal amount of labor was expended in the room's construction. The absence of roof fall in the fill may indicate that the room's roof was disassembled at abandonment to salvage construction materials.
Structure 3 (Tower)
Structure 3 is a partly standing, stabilized tower (Figure 13.7 and Figure 13.8). Two contiguous sampling units (84N/101E and 83N/101E) exposed the interior and exterior faces of the structure's south wall. A third contiguous unit (82N/101E), although located outside the tower and within Sampling Stratum 4 (the inner periphery), is more appropriately discussed here. Figure 13.9 illustrates Structure 3 and the units excavated within or adjacent to that structure.
Construction
Walls. Structure 3 has one continuous, circular, double-stone-wide wall that ranges from 35 to 45 cm thick. Maximum wall height is 3.1 m; minimum wall height is 33 cm. The interior diameter of the tower measures 2.55 m. The wall is semicoursed on both the interior and exterior faces and is probably of compound construction. The exterior face of the wall consists mostly of pecked-block masonry; the interior face has more tabular and flaked sandstone than pecked blocks. The base of the tower wall rests on native sediment. In the small section of wall that was exposed during excavation, almost no chinking stones were visible. The mortar appeared to be the same as the native sediment at the site, a brown sandy loam. Mortar joints were 2 to 4 cm thick. One wall feature--a cluster of three posts built into the upper wall--was observed.
Feature 1 (Posts): The ends of three wooden posts are visible between upper courses of masonry in the interior north side of Structure 3. These posts are located within a meter of one another and are at nearly the same elevation (about 2.25 m above the modern ground surface). The posts average 3.5 cm in diameter and extend a maximum of 4 cm out from the wall. The angle of projection from the wall was measured for each post. The two posts on the east side angled between 10 to 15 degrees east; the third post angled 10 to 15 degrees west. The function of these posts is unknown.
Roof. There was no evidence of roof fall to indicate how or with what materials the structure was roofed.
Floor. The floor of the tower was apparently the unprepared prehistoric ground surface. The surface of Structure 3 corresponds with the base of the tower wall, which rests on native sediment. No artifacts or features were located on the exposed surface.
Stratigraphy
Figure 13.10 illustrates the stratigraphic profile of Structure 3. Because these grid units were located both inside and outside the tower, fill adjacent to the exterior wall of the tower is also discussed here.
The fill outside the tower is the result of natural deposition. This deposit was excavated as Stratum 1 and divided into two levels. The upper level consists of a 20- to 30-cm-thick layer of naturally deposited sediments containing many shaped sandstone pieces. This is interpreted as wall fall from the tower. The lower level, about 10 cm thick, contains little sandstone. This lower fill may represent a period of natural erosion preceding the collapse of portions of the tower.
Fill inside the tower was excavated in two strata. The upper fill, Stratum 1, is a mixed deposit 12 to 20 cm thick. It contains disturbed sediments indicative of looters' spoil: aluminum foil, a .22 caliber shell, numerous charcoal pieces, ash, some sandstone, and 11 pieces of chipped-stone debris mixed within a loose, gray brown, sandy loam matrix.
The lower fill, Stratum 2, is undisturbed. It consists of a layer 20 to 30 cm thick of loose, red brown sandy loam. Stratum 2 contains a few artifacts and small pieces of charcoal; moderate amounts of various-size sandstone pieces were found as well. Seven pieces of chipped-stone debris and a Late White Unpainted sherd were collected. Stratum 2 covers the tower floor and is interpreted as naturally deposited sediment and structure collapse.
Interpretations
Structure 3 is a masonry tower, the main architectural feature on the surface of the site. It is presumed to be connected to the adjacent kiva by a tunnel. This tower-kiva connection, according to some archaeologists (Ferguson and Rohn 1986:42), may signify a ceremonial complex.
The use of many well-prepared, pecked sandstone blocks in the wall of the tower suggests that a relatively greater amount of labor was invested in its construction than in the construction of other buildings at the site. This includes the kiva, which was not masonry lined, and the surface room.
The original height of Structure 3 was at least two stories, which would have offered a commanding view to an observer in the upper story. One view would have been 1 km west across Sand Canyon to a tower located above Saddlehorn Hamlet, another site tested during the 1990 field season. This suggests that inhabitants of the two sites could have communicated with one another via the towers, if the two sites were occupied at the same time. Other sites with towers that are possibly contemporaneous with 5MT181 are 5MT2797, located .3 km to the southeast, and 5MT261, located .72 km to the southwest.
Sampling Units 92N/90E and 91N/92E
These two sampling units lie in or near the sandstone rubble pile near the northern end of the site. Prior to excavation, these two units were believed to be within a surface structure. Excavation proved they were just outside of Structure 1.
The two units were excavated in two strata. The upper fill, Stratum 1, consisted of a naturally deposited, soft, brown sandy loam with charcoal, artifact, and sandstone inclusions. The amount of sandstone in the two units differed. Sandstone found in 91N/92E was minimal and of small size. The sandstone in 92N/90E was more abundant and varied in size; it included both shaped and unshaped pieces. The position of the sandstone rocks primarily in the southwest corner of 92N/90E suggests that they were structural collapse from Structure 1.
The lower fill, Stratum 2, was also similar in both units. Sediments consisted of a slightly hard, red brown sandy loam to silty clay loam. The clay was probably the mortar from the structural collapse. Both strata were deposited after site abandonment by erosional processes. Excavation in Stratum 2 stopped at the contact with the sterile B horizon. A pit feature was found in 92N/90E at the contact with the sterile sediments.
Feature 1 (Pit)
Feature 1 is an oval, basin-shaped pit prehistorically excavated into native sterile sediment. The pit measures 58 cm long, 49 cm wide, and 15 cm deep. Fill within the feature was a naturally deposited, soft, dark brown sandy loam. Inclusions consist of minimal amounts of small charcoal pieces and larger amounts of medium- and large-size sandstone pieces. A piece of chipped-stone debris and four sherds (two Local Indeterminate Corrugated Gray and two Late White Unpainted bowl rims) were recovered. The function of the feature is unknown.
Pit Structure and Courtyard Area (Sampling Strata 2 and 3)
Sampling Stratum 2 is the pit structure sampling stratum, which contains 16 1-×-1-m sampling units (Figure 13.4). Five randomly selected (84N/93E, 84N/94E, 85N/94E, 86N/92E, and 87N/92E) and one judgmentally selected (84N/95E) sampling units were excavated in Sampling Stratum 2. All but one unit (87N/92E) landed inside the pit structure and were therefore provenienced as Structure 2.
Sampling Stratum 3, which contains 57 1-×-1-m sampling units, is the courtyard sampling stratum. Excavation took place in five randomly selected sampling units (90N/92E, 88N/95E, 87N/99E, 86N/97E, and 84N/96E). One landed inside the pit structure (84N/96E) and was therefore provenienced as Structure 2. The other units were located either north or northeast of the pit structure.
Structure 2 (Kiva)
Structure 2 is an unburned earthen pit structure with masonry pilasters (Figure 13.11). Figure 13.12 illustrates Structure 2 and the excavation units within it. The presence of surface rubble to the south of the structure suggests that the southern recess walls may have also been constructed of sandstone. Six sampling units excavated inside the kiva uncovered a wall niche, the majority of two pilasters, and portions of three benches, a hearth, and a tunnel. Structure 2 dimensions can be inferred from this limited testing. The main chamber appears to be circular with an exterior diameter of about 4 m. The main chamber floor is 2.27 m deep below the modern ground surface.
Construction
Walls. With the possible exception of the southern recess, the upper lining wall and bench face of Structure 2 are earthen. They were defined during excavation by the contact with sterile sediments. Evidence of some plaster on the bench face suggests that the entire earthen lower wall may have been plastered. The lower lining wall varies from 92 cm to 1.04 m high from the floor to the bench surface. Three segments of the bench, two pilasters, a tunnel, and a niche are features associated with the Structure 2 walls.
Feature 1 (Bench 2): The portion of Bench 2 discovered within grid square 86N/92E measures 55 cm long and 46 cm wide. The entire width, but only part of the length, was exposed; the total length is inferred to be about 1.5 m. From the bench surface to the top of the adjacent pilaster measures 56 cm. The surface of Bench 2 consists of use-compacted sterile sediment. A thin layer of soot covers the bench surface, and only one piece of chipped-stone debris was found upon it.
Fill above the bench surface consists of two strata. The upper layer, Stratum 1, is composed of naturally deposited sediments and measures 38 to 40 cm thick.
Stratum 2 is a 15-cm-thick roof fall layer that covers the bench surface. It contains a dark-colored fill with minimal inclusions of small burned adobe and charcoal pieces. Some of the charcoal pieces were large enough to collect as tree-ring samples.
Feature 2 (Bench 3): A portion of Bench 3 was exposed in sampling unit 85N/94E. It measures 64 cm long and 44 cm wide. Inferred length and width measurements of Feature 2 are 1.5 m and 46 cm, respectively. The distance from the bench surface to the floor of Structure is 1.04 m. The bench surface was unprepared and consists of sterile sediment. No artifacts lie on the exposed surface, but, like Bench 2, a thin layer of soot covers the surface.
Two strata cover the surface of the bench. The upper layer, Stratum 1, is a naturally deposited sediment measuring 42 to 45 cm thick. Stratum 2 lies beneath Stratum 1 and covers the bench surface. Stratum 2 is a roof fall layer containing minimal amounts of small charcoal pieces and small adobe chunks. It measures 5 to 8 cm thick.
Feature 3 (Bench 4): Excavation within sampling unit 84N/95E revealed a portion of Bench 4. The exposed portion measures 1 m long and 46 cm wide. The width is the actual size of the bench measured from the inner to the outer wall of Structure 2 along the bench surface. The length of the bench is inferred to be about 1.5 m. Distance from the bench surface to the pit structure floor measures 92 cm. A tunnel, which is described below, intersects the surface of Bench 4. This surface, like the surfaces of Benches 2 and 3, is unprepared and consists of sterile sediment. One Indeterminate Local Corrugated Gray sherd and five pieces of chipped-stone debris were found on the bench surface. Fill covering the surface is unburned roof fall.
Feature 4 (Pilaster 3): Pilaster 3 consists of well-shaped, pecked sandstone blocks mortared into a semicoursed masonry column. The pilaster is seven to eight courses (58 cm) high, three courses (58 cm) wide, and two courses (35 cm) thick. A thin (1 cm) mortar joint separates Pilaster 3 from the bench surface. The pilaster abuts the earthen outer wall of Structure 2.
Feature 5 (Pilaster 4): The construction of Pilaster 4 was similar to that of Pilaster 3. Pilaster 4 is five courses (40 cm) high, two to three courses (54 cm) wide, and two courses (39 cm) thick (Figure 13.13). The height measurement does not reflect the actual height of the pilaster, which would have been greater. Pilaster 4 rests on the bench surface and abuts the outer wall of the kiva.
Feature 6 (Tunnel): A prehistorically excavated tunnel was discovered in two contiguous sampling units (84N/95E and 84N/96E) in the kiva. This feature was identified by the presence of fill within an area surrounded by sterile sediment. Excavation of the fill defined the shape of this feature. It is linear in plan view and basin-shaped in profile. Feature 6 measures an average of 53 cm wide by 87 cm high. The width measurement is complete; the height measurement is the distance between the surface of the tunnel and the inferred roof. Feature length (140 cm) measures the extent of the excavation. This extends from the 97E grid line to the inner lining wall of the kiva where Feature 6 intersects, and it was prehistorically excavated slightly through the surface of Bench 4. Because this tunnel goes toward the tower (Structure 3), it is assumed that the tunnel connects the two structures.
Fill within the tunnel was primarily naturally deposited sediments with the exception of a roof fall layer found at its intersection with the bench. Some sooting was also observed on the use-compacted surface of the tunnel. Sandstone pieces, including large slabs, were observed in the upper fill. The sandstone may have roofed the tunnel. This would have been accomplished by placing the large slabs over the width of the tunnel. The sides of the tunnel would have supported the slabs.
Artifacts in the tunnel include 15 Indeterminate Local Corrugated Gray and two Pueblo III White Painted sherds, a complete peckingstone, and 12 pieces of chipped-stone debris. The only artifact on the tunnel surface was a piece of chipped-stone debris.
Feature 7 (Wall Niche): A wall niche was constructed within the lower lining wall below Bench 4. Feature 7 is rectangular and measures 38 cm wide, 35 cm high, and 25 to 31 cm deep. Its base is 29 cm above the floor. A sandstone slab served as a lintel for the niche. Evidence suggests that the interior of the niche was once entirely covered with plaster. Roof fall materials filled the niche. One tree-ring sample collected from the niche provided a noncutting date of A.D. 932. The only artifacts in the niche were three pieces of chipped-stone debris.
Floor. The floor of Structure 2 is constructed of a red brown adobe plaster. A thin layer of a dark gray soot covers the plaster. Artifacts found on the floor and 5 cm above it include eight pieces of chipped-stone debris and seven Indeterminate Local Corrugated Gray sherds. Testing uncovered a hearth and a pit feature that may be a sipapu on the floor of Structure 2.
Feature 8 (Pit): Feature 8 is a shallow, basin-shaped pit located in sampling unit 85N/94E, near the north end of Structure 2. It was excavated 5 cm below the floor into sterile sediment. The pit feature is nearly circular, measuring 12 cm in diameter. No artifacts were found in the pit. Feature 8 was probably deliberately filled to surface level with sediments that were naturally deposited elsewhere on the site. The function of Feature 8 is unknown, although its position in the kiva suggests that it may have served as a sipapu.
Feature 9 (Hearth): Because most of the hearth was outside the sampling units, it was necessary to excavate portions of adjacent grid squares to better document the feature and collect appropriate samples. The hearth is a circular, basin-shaped pit prehistorically excavated into the floor of Structure 2. The pit is 18 cm deep and has an inferred diameter of 68 cm. A 3- to 4-cm-thick layer of red plaster covers the hearth rim and suggests remodeling. Fill within this feature consists of a light gray ash with no artifacts. All the ash within the excavated segment of the feature was collected for either flotation or fine-screen samples. The flotation samples contained Juniperus, Pinus, Pinus edulis, and Cercocarpus. An archaeomagnetic sample was taken from the sides and base of the hearth. That sample yielded a range of A.D. 1010-1325 for the hottest fire within the hearth.
Stratigraphy
Figure 13.14 illustrates the stratigraphic profile of Structure 2. Structure 2 was excavated in two strata. The upper fill, Stratum 1, is postabandonment and erosional sediment. It consists of a soft, brown sandy loam. Inclusions in Stratum 1 are minimal, except above the tunnel, and include small charcoal pieces, artifacts, and lesser amounts of small sandstone pieces. The charcoal pieces become more frequent and larger near the base of Stratum 1 (the area labeled 1A on Figure 13.14). The naturally deposited sediments above Bench 4 and the tunnel are the same as Stratum 1 found in the main chamber except for the increase in sandstone inclusions. These are mostly large sandstone blocks and slabs, as well as irregularly shaped pieces. Their presence is attributed to the collapse of a roof that covered the tunnel feature. Stratum 1 thickness ranges from 1.72 m near the center of the kiva to 1.06 m above the inner lining wall. Above Bench 4 and the tunnel, Stratum 1 measures 70 to 90 cm in thickness, becoming thicker toward the inner wall of the kiva.
Stratum 2 is an unburned roof fall layer found immediately above the Structure 2 floor. It consists of a soft, dark brown sandy loam with minimal inclusions (small pieces of charcoal, artifacts, adobe roof casts, and a few larger pieces of sandstone). Artifacts found in Stratum 2 include four modified bone fragments, a projectile point, a hammerstone that was recycled into a core, a biface fragment, a gizzard stone, a mineral sample, 38 pieces of chipped-stone debris, and 32 sherds (27 Indeterminate Local Corrugated Gray, and one of each of the following types: Indeterminate Plain Gray, Mancos Corrugated Gray, Piedra Black-on-white, Pueblo III White Painted, and Late White Painted). Two pieces of charcoal were large enough to collect as tree-ring samples. They provided noncutting dates of A.D. 932 and A.D. 1047. In the main chamber, Stratum 2 ranges from 24 cm thick in the center of the kiva to 93 cm thick adjacent to the inner lining wall. Stratum 2 is not present in the tunnel, but it measures 17 to 20 cm thick above the surface of Bench 4.
Dating
Two tree-ring samples collected from the roof fall stratum yielded noncutting dates of A.D. 932 and 1047. The archaeomagnetic sample taken from the hearth yielded a wide range of A.D. 1010-1325.
Interpretations
Structure 2 is an unburned, earthen pit structure with masonry pilasters. The stratigraphy of Structure 2 suggests that the roof of the kiva was dismantled and the timbers removed for use elsewhere. Evidence for this is seen in the lower fill of the kiva, which seals the floor surface. There, naturally deposited sediment mixed with adobe and small pieces of charcoal suggests the discard of unusable roofing materials. The absence of larger timbers suggests that they were salvaged. The scarcity of artifacts on the floor and in the roof fall suggests that usable tools were also salvaged.
Sampling Unit 87N/92E
This grid square was the only unit in Sampling Stratum 2 that did not fall within the kiva. It was located just north of Bench 2 and Pilaster 3 of Structure 2 and contained 20 to 31 cm of naturally deposited sediment with only a few artifacts.
Sampling Units 88N/95E, 87N/99E, and 86N/97E
These excavation units are located in Sampling Stratum 3, close to one another and to the kiva and the tower. Excavation removed a single stratum of fill down to sterile sediments. All three units contained sediment deposited naturally after abandonment. Stratum 1 was a loose, brown sandy loam with minimal charcoal, artifact, and sandstone inclusions; an exception is 86N/97E, in which the sandstone was larger and more abundant. This was likely a result of the structural collapse of the tower into that grid unit but not into the others. Stratum 1 varied from 7 to 23 cm thick.
Sampling Unit 90N/92E
This grid unit is located in Sampling Stratum 3 adjacent to the surface structure stratum and sampling unit 91N/92E, described above. Stratum 1 consists of a soft, red brown sandy loam with a few small charcoal pieces and artifacts. Stratum 1 varies in thickness from 35 to 55 cm and consisted of naturally deposited, postabandonment sediments.
Inner Periphery (Sampling Stratum 4)
Sampling Stratum 4 is the inner periphery sampling stratum which surrounds the cultural units at the site (Figure 13.4); it consists of 257 1-×-1-m sampling units. One judgmental and five randomly selected sampling units were excavated. Three of the sampling units (91N/82E, 93N/86E, and 93N/90E) lie west of Structure 1, and the others are located either northeast (88N/105E and 87N/105E) or south (82N/101E) of Structure 3. Because 82N/101E is located adjacent to two other sampling units excavated within the tower, it was described above under Structure 3.
These sampling units contain naturally deposited sediment that accumulated after site abandonment. The fill contains few artifacts, charcoal inclusions, or pieces of sandstone. The thickness of the deposit varies between 4 and 46 cm.
Artifacts
The following artifacts were recovered from Sampling Stratum 4 units: 100 pieces of chipped-stone debris, 52 sherds, a complete core, a gizzard stone, a piece of eggshell, and a piece of historic glass. The sherds were 42 Indeterminate Local Corrugated Gray, six Late White Unpainted, three Pueblo III White Painted, two of which are carbon painted, and a Late White Painted.
Midden Area (Sampling Stratum 5)
Sampling Stratum 5 is the midden sampling stratum (Figure 13.4). This area is provenienced as Nonstructure 1 because it is a culturally bounded, nonarchitectural study unit. Nonstructure 1 was defined on the basis of its dark soil color and the densest surface artifact concentration anywhere on the site. Of the 172 sampling units in Sampling Stratum 5, six were randomly selected for excavation. These were 76N/89E, 84N/83E, 71N/92E, 75N/88E, 75N/93E, and 89N/84E. Excavation demonstrated that the midden was located exactly as predicted by surface evidence.
Stratigraphy
Figure 13.15 illustrates the stratigraphic profile of 84N/83E, which is representative of the midden stratigraphy. The following description of the stratigraphy in 84N/83E applies to all the midden units. Stratum 1 is a loosely compacted, structureless, brown to red brown sandy loam. Stratum 1 ranges from 11 to 49 cm in thickness. Cultural inclusions, other than artifacts, were found in small quantities and consist of small charcoal and sandstone pieces. Artifacts were found in relatively greater quantities. Stratum 1 is interpreted as secondary refuse.
Stratum 2 is similar to Stratum 1, except for being more compact and containing fewer artifacts. Stratum 2 ranges from 14 to 24 cm in thickness. The prehistoric ground surface probably exists somewhere within Stratum 2. One pit feature was located within sampling unit 89N/84E at the contact with sterile sediments.
Feature 1 (Pit)
Feature 1 was discovered in sampling unit 89N/84E and extends farther to the south and east. Only the portion of the pit feature within 89N/84E was excavated. Feature 1 is irregularly shaped in plan and profile. Its excavated dimensions are 44 cm long, 32 cm wide, and 25 cm deep. The pit filled naturally after abandonment. The feature contained no artifacts. The function of this feature is unknown.
Artifacts
Midden excavations yielded a total of 382 sherds, 48.5 percent of the site total. The gray ware sherds, which make up 64.7 percent of the midden total, were Indeterminate Local Corrugated (N = 242), Mesa Verde Corrugated (N = 4), and Indeterminate Plain Gray (N = 1). The white ware sherds (35 percent of the midden total) were Pueblo III White Painted (N = 42), Late White Unpainted (N = 80), Mesa Verde Black-on-white (N = 11), and McElmo Black-on-white (N = 1). Only one Nonlocal Red sherd was found in the midden. The presence of 11 Mesa Verde Black-on-white sherds and only one McElmo Black-on-white sherd in the midden suggests that most of the trash dumping at Mad Dog Tower occurred after A.D. 1180.
Stone tools found in the midden accounted for 30.5 percent of the total tools found at the site. Complete and fragmentary stone tools include one mano, one core, two peckingstones, three projectile points, one modified flake, and two other modified stones/minerals. There was also one bead, a modified bone, a piece of eggshell, a gizzard stone, a piece of shell, and a number of animal bones. In addition to these tools, 266 pieces of chipped-stone debris (25.9 percent of the site total) were recovered in midden excavations.
Three flotation samples collected from the midden were analyzed. Botanical remains are limited to Juniperus charcoal interpreted as discarded fuelwood.
Interpretations
The midden, Nonstructure 1, is a distinctive cultural feature with well-defined boundaries. It covers about 170 m² of area and is a maximum of .5 m thick. The midden consists of secondary refuse that was deposited to the south and west of Structure 2. Its relatively shallow depth may indicate that the site occupation was short or seasonal. This interpretation, however, is highly conjectural; the relatively shallow depth may be a result of erosional deflation rather than duration of occupation. The four Mesa Verde Corrugated Gray and 11 Mesa Verde Black-on-white sherds suggest that refuse was deposited between A.D. 1180 and 1300.
Outer Periphery (Sampling Stratum 6)
Sampling Stratum 6 is the outer periphery of the site (Figure 13.4). It surrounds the inner periphery and extends to the far edge of the artifact scatter. Of the total 1,366 sampling units in Sampling Stratum 6, four were selected for excavation. All these sampling units contained naturally deposited sediments. Deposits were shallow and artifacts scarce. Sediments consisted of a brown sandy loam with few inclusions of small charcoal and sandstone pieces.
Mad Dog Tower Artifacts
The artifact data for Mad Dog Tower are presented in tables at the end of the chapter (Table 13.1, Table 13.2, Table 13.3, Table 13.4, and Table 13.5). Fewer artifacts were collected from Mad Dog Tower than from any of the other tested sites. Only 787 sherds were recovered from the site. Early (pre-Pueblo III) pottery is scarce. Early decorated white wares include one sherd each of the following types: Piedra Black-on-white, Early White Painted, Mancos Black-on-white, and Pueblo II White Painted. Combined, these early sherds make up only 0.5 percent of the total sherds found at Mad Dog Tower. Late (Pueblo III) pottery is more common at the site. The late pottery recovered from Mad Dog includes McElmo Black-on-white (N = 3), Mesa Verde Black-on-white (N = 15), Pueblo III White Painted (N = 81), Late White Painted (N = 4), and Late White Unpainted (N = 117). The presence of only four early white ware sherds, as compared with over 200 late white ware sherds, suggests that the site was occupied later than A.D. 1180. Only two red ware sherds were found at the site. They were identified as nonlocal.
Early gray ware types are nearly nonexistent at the site. Only one Mancos Corrugated and two Indeterminate Plain Gray sherds were found. Late gray ware types consist of four Mesa Verde Corrugated Gray rims. The majority of sherds found at the site (70.4 percent of the total) were corrugated body sherds (N = 554).
Complete and fragmentary tools recovered from the site include one mano fragment, three cores, one hammerstone, two peckingstones, eight bifaces, nine projectile points, four modified flakes, two other chipped-stone tools, and five other modified stones/minerals. Five pieces of modified bone, a modified sherd, eggshell, a shell fragment, several gizzard stones, and 1,026 pieces of chipped-stone debris were also found.
Nonlocal materials were scarce at Mad Dog Tower. These consisted of two nonlocal red ware sherds, one piece of jet, and 14 pieces of nonlocal chipped-stone debris. The scarcity of nonlocal materials indicates that the inhabitants of Mad Dog Tower did not exchange goods nonlocally.
Site Summary and Conclusions
Mad Dog Tower is a single-component Pueblo III site. The most striking architectural feature at the site is a partly intact and stabilized masonry tower. It stands over 3 m high on the crest of a small hill on the first terrace overlooking Sand Creek. Surface indications of a masonry roomblock and a trash midden are also present. Although not visible prior to excavation, a pit structure was also believed to exist at the site. A stratified random sample was excavated to test the site. Six sampling strata were identified, and 29 random sampling units were excavated.
Excavation revealed the presence of a single masonry room, a relatively shallow midden, a masonry tower, an earth-walled kiva, and a tunnel which probably connects the kiva and the tower. This architectural layout differs from the classic "Prudden Unit" in that a larger roomblock is absent and the midden is thin. The site layout and shallow midden suggest that Mad Dog Tower may have functioned differently than other unit pueblos interpreted as year-round habitations. How it functioned is unknown, but one possibility is that it served as a religious or ceremonial complex for a larger community. Other kiva-tower sites have been documented. Interestingly, most excavated examples occur north of the San Juan River and after A.D. 1100. Excavators of these sites often discover an underground tunnel connecting the tower to an adjacent kiva. Tunnels connecting towers and kivas have been documented at the Grinnell Site (Luebben 1983) near Yucca House, at Cedar Tree Tower and Far View Tower at Mesa Verde (Ferguson and Rohn 1986), at the Mustoe Site (Gould 1982) near Goodman Point, and at Troy's Tower (Varien 1990b; also see Chapter 7, this volume) in the Sand Canyon locality. Ferguson and Rohn believe that tower-kiva units functioned as community religious complexes. They argue, for instance, that Far View Tower at Mesa Verde probably served the whole early Pueblo III (A.D. 1100-1200) Far View community (Ferguson and Rohn 1986:42). If this was the case at Mad Dog Tower, what community did this tower-kiva complex serve?
Settlement-pattern data from the area surrounding Mad Dog Tower are available. Gleichman and Gleichman (1989) suggest that a community of dispersed, contemporaneously occupied settlements did exist during Pueblo III times. Sites possibly contemporaneous with Mad Dog Tower in the lower Sand Canyon drainage include 5MT2797, located .3 km to the southeast; 5MT261, located .72 km to the southwest; and 5MT262 (Saddlehorn Hamlet, Chapter 12 in this volume), located 1 km to the west. These sites share a common trait--the presence of a tower or "lookout." From these observatories, each site could view at least two, if not all, of the other three sites.
Absolute dates from Mad Dog Tower are not very informative. Two tree-ring samples dated, and the latest of these provided an A.D. 1047 +vv date. An archaeomagnetic date from the pit structure hearth has a date range of A.D. 1010-1325. Although the absolute dating at Mad Dog Tower is rather tentative, the pottery and architecture indicate that the site was occupied during the late Pueblo III period; the presence of Mesa Verde Black-on-white indicates that the occupation postdates A.D. 1180 (Wilson 1991). The stratigraphy of the kiva suggests that the roof timbers were salvaged when the site was abandoned. This implies that site abandonment slightly preceded the abandonment of the Sand Canyon locality and the Mesa Verde region as a whole, which took place between A.D. 1280 and 1300. Thus, the occupation of Mad Dog Tower probably occurred between A.D. 1180 and 1280.