White Dog Cave, Arizona

Atlatls and Darts of White Dog Cave, Arizona

By Devin B. Pettigrew and Justin Garnett

In The Atlatl, volume 28, number 2, pages 1-5

April 2015


White Dog Cave, location in space, time, and history

It was the early 20th century, and the Southwest artifact mining boom was in full swing. Commercial excavations in the Grand Gulch country of southeastern Utah by the such as the Wetherill brothers and other professional agencies had generated a great interest in Southwestern relics in general, especially those of the so-called Cliffdwellers, whom we now know as the ancestral Puebloans.  These uncontrolled, profit driven diggings had begun to notice significant differences in the material culture within dry cave sites, with deeper digging producing not typical Cliffdweller fare, but another sort of material entirely; the stuff of an earlier people, who came to be known as the Basketmakers for their lack of a well-defined ceramic tradition.

Figure 1. Pettigrew holding replica White Dog Cave atlatl equipment. In the background is Marsh Pass in northeastern Arizona.

Figure 1. Pettigrew holding replica White Dog Cave atlatl equipment. In the background is Marsh Pass in northeastern Arizona.

Although the Wetherill diggings had recognized the distinctiveness of this early culture, their methods, while good for amateurs, were still lacking, and the race was on for archeological institutions to locate and excavate Basketmaker sites in controlled, scientific fashion.  It was this race that brought a Harvard Peabody Museum excavation team, led by Alfred Vincent Kidder (later to develop the Pecos Classification Scheme of 1919, which formalized the Basketmaker designation) and Samuel James Guernsey to the Four Corners region.  The summer of 1916 brought Kidder and Guernsey to the Marsh Pass region just east of Kayenta, Arizona (Figure 1).  The Peabody team was impressed with the rugged, nearly impassable landscape of Marsh Pass, which runs between Black Mesa and Tyenda Mesa.  Many blind curves and sudden drop-offs make the land treacherous, while high spires of stone stand in stark contrast to flat alluvial deposits.  There were many caves in the Marsh Pass region that the team investigated, but none yielded so many well preserved artifacts as what they named White Dog Cave, after the hair color of a mummified Basketmaker dog that would be uncovered in association with a human burial within the cave.

White Dog Cave is unobtrusive, possible only because of the deep folding and gullying in the area, for the cave is of immense size, but one must approach closely to get any inkling it is there at all.  The cave proper is found at the rear of a massive rock shelter, and is approximately 70 feet deep by 120 feet wide.  As is typical of such caves, White Dog Cave is mostly filled with rubble from rockfalls from the ceiling, the remainder of the floor being composed of clean, wind borne sand.   An ancestral Puebloan period kiva ruin was found against a wall of the cave, but the rest was devoid of signs of ancient Cliffdwellers.  However, beneath the sands of the cave, test pits found basketry and human bones, and the cave was marked for excavation.   The artifacts were of Basketmaker origin, and have been dated to approximately 2,000 B.P. (Coltrain et al. 2007). The artifacts from this site were extraordinary, well preserved, and revealed a rich cultural tradition of great complexity.  The Peabody team had found an important site.

Figure 2. A replica of the cist 27 atlatl by Pettigrew.

Figure 2. A replica of the cist 27 atlatl by Pettigrew.

Excavation revealed a large number of human burials in shallow stone lined graves known as cists.  The atlatl gear with which we are concerned in this article were found in association with specific burials. A 24” long, triple weighted atlatl (Figures 1, 2 and 3) (Peabody # A2809) was recovered from Cist 27, in association with a male approximately 25 years of age.  A shorter complete atlatl (Peabody # A2951) was found in Cist 24, in association with the burial of a female.  Dart fragments, mainshaft fragments and foreshafts were found in various cists.  Many other atlatl and dart fragments were also found, but can’t be considered further in this article because their highly fragmentary nature prevents their replication.

Replicating the White Dog Cave atlatl and darts

In terms of the atlatl material alone the White Dog Cave (WDC) site is important.  The discovery of two complete atlatls and three complete darts offers a rare glimpse at the specifics of an ancient Basketmaker atlatl hunting kit.  In our experience, entire systems need to be replicated so that darts and atlatls are matched, since throwing the wrong darts with a replica atlatl—or vice versa—can seriously alter our interpretation.  One gets the best understanding when all details of material, construction and dimensions are closely followed.  Windows can exist in which slight variations in some of these parameters are acceptable, but it often takes experience to identify these windows.

Much of our experience in these matters has resulted from efforts to replicate the WDC artifacts.  Pettigrew’s first attempt at artifact replication was with the WDC darts.  Initial attempts with river cane (Arundinaria gigantea) mainshafts failed.  The shafts were too stiff and produced horrible flight.  While Pettigrew knew better, frustration mounted and he couldn’t help but consider the possibility that these ancient hunters had a poorly refined projectile kit!  This of course is the worst possible result of an experimental approach to archaeology.  We hope that our experiments will teach us something about ancient people, whose tools we try to replicate, but good insight can take long periods of trial and error, and often only result after adequate skill in constructing and using ancient tools has been developed.  Later, reproducing the darts from wooden shoots revealed them to actually have excellent flight (Pettigrew 2008).  The first throw in fact sent a fletchless WDC dart sailing straight away with low oscillation and good trajectory.  A subsequent effort to reproduce the Cist 27 atlatl and use it with the darts was also informative (Pettigrew 2009). Since then we have refined our replications further.

Figure 3. An effective gripping method used with the cist 27 atlatl and WDC dart.

Figure 3. An effective gripping method used with the cist 27 atlatl and WDC dart.

Anyone interested in these artifacts and the context of their discovery should seek the original report (Guernsey and Kidder 1921), which can be found online with a simple search.  However, several details about the artifacts are missing or were not given an adequate description.  Additionally, some of the original measurements are imprecise.  Chuck LaRue examined the Cist 27 atlatl at the Peabody in 2006 and provided us with detailed measurements and color photos.  Garnett also visited the Peabody in 2010 and examined the Cist 24 atlatl and the darts.  Their measurements are provided in Tables 1 and 2.  Photos of the artifacts aren’t published here due to copyright, but our replicas are pretty close, and should satisfy general interest.

Darts              Close replicas are encouraging, because they show the WDC cist 27 atlatl paired with WDC willow darts to be an effective system, and one that was obviously refined by generations of people who relied on their weapons daily.  A close inspection of the leaf scars on the original darts suggests to Garnett that they were made of coyote willow (Salix exigua) which is common in the Southwest, and works well at these dimensions (see Pettigrew 2013 for construction methods).  Remnants of pigment indicate designs in red and black, with short spirals under the fletchings and broad bands along the shaft, or shafts painted entirely red.  Sockets can be drilled out with stone drills and were often left with a narrow shoulder where the foreshaft connects, which leaves a stronger socket, but does not help with penetration.  Foreshafts are constructed of oak, about 15 cm long, and given roughened spirals at their socket tangs to create friction for the joint.  Wetting the tang and inserting the foreshaft into the socket with a twist produces a solid connection for a few throws, but the foreshafts can disengage fairly easily.  This probably results in preservation of mainshafts when prey is struck and runs—the mainshaft will disengage before the animal dives through brush.  A variety of foreshafts can also be used, and fewer mainshafts need to be carried by mobile hunters.  One WDC dart still has a wooden blunt inserted in the socket, and blunts were found at other Basketmaker sites composed of the cut off distal ends of deer or sheep tibiae with wood inserts for socket tangs (Pepper 1902).

The method of fletching the mainshafts is confusing, and warrants better description.  Splints of wood or yucca spines were inserted into the trimmed off, hollow quills of whole feathers and the quills were wrapped individually with sinew or gut before being lashed to the shaft.  At their distal ends the feathers are simply attached with another wrapping over the untrimmed barbs and vane.  Three feathers seem to have been lain along the shaft tangentially using this method (Figure 3).  It’s important when doing this that the feathers are of the right size, and the quills are properly aligned so that the feathers lay close to the shaft.  On other dart shafts in the arid West a tuft of smaller feathers is attached above the distal wrapping.  We think this curious fletching method was used to accommodate whole feathers that had symbolic significance.  The method of inserting a spine or splint into a wrapped quill was also used in the attachment of feathers into hair ornaments (Guernsey and Kidder 1921). In Basketmaker rock art we often see darts depicted with dramatically large fletchings; some even showing the distal tufts, and some with interior designs that appear to mimic the tail feather coloration of the golden eagle (Aquila chrysaetos) (Figure 4) (Chuck LaRue, personal communication, 2015).  In short, fletchings probably meant more to Basketmaker people than simply functional additions to improve dart stability in flight.

Figure 4. Depictions of atlatl darts in rock art in San Juan County, Utah. Left) an interior design that may depict the coloration of golden eagle tail feathers. Middle) a dart painted in white with tufts depicted in front of the fletches. Right) typical depictions of darts, with fletches indicated by simple circles. Photos by Chuck LaRue.

Figure 4. Depictions of atlatl darts in rock art in San Juan County, Utah. Left) an interior design that may depict the coloration of golden eagle tail feathers. Middle) a dart painted in white with tufts depicted in front of the fletches. Right) typical depictions of darts, with fletches indicated by simple circles. Photos by Chuck LaRue.

Atlatl              We have focused our efforts on what we initially saw as the nicer of the two WDC atlatls, the cist 27 atlatl with the three small weights attached. The weights are composed of swirly green and white rocks, worked down and polished, that have been described as the petrified teeth of some Triassic aged creature from the Chinle formation.  A thin layer of black was lain down on the wood within the section of the weights before they were attached, and then covered over some of the wrappings.  Our best attempts to reproduce this have resulted from covering the shaft in pine resin, chipping off the resin, reheating the then tacky surface and rubbing in crushed charcoal; or rubbing the wood with a non-heated glob of resin to produce a coated surface, then heating it and rubbing in the pigment.

The wood of the atlatl is probably gambel oak (Quercus gambelii); also used for many other atlatls in the Southwest.  This is a dense and tough, but springy wood that is easily bent to a desired shape when heated, but can also warp with occasional exposure to moisture.  Many atlatls found in the Southwest are highly warped.  The shaft of the cist 27 atlatl was snapped above the distal most weight, perhaps to “kill” it before placing it in the grave.  The loops were attached with a simple but effective method.  A piece of buckskin was folded in half to increase its rigidity, then perforated in the center and slid over the handle to the proximal end of a shallow set of finger notches.  The ends were brought above the notches, perforated with two holes in each side and “sewn” to the shaft (Figure 5).  A lashing around the threaded portion between the wood and leather tightens this lashing (this method is also used to tighten down the lashings around the weights).  A smaller set of notches creates lugs just above the finger notches to keep the loops from sliding down.  Hide glue holds the sinew together, but does not keep it affixed to the shaft with extended use, so these lugs are necessary.  The proximal end of the loops is retained with a “retention strap”, composed of a piece of leather threaded through the perforation on the underside and glued and lashed to the shaft above the loops.

Figure 5. Close-up of the loops on the Cist 27 atlatl replica.

Figure 5. Close-up of the loops on the Cist 27 atlatl replica.

The width of the loops and length of the handle works well for the authors when used with the grip shown in Figure 3.  The fingers are inserted to the middle phalanges, right up to the middle knuckle, and the lower handle often “floats” above the palm.  The atlatl balances well with the WDC darts right around the handle.  We find this works well when used with a light grip on the dart and a flinging technique at the wrist and fingers.  Some torque can also be introduced by the ring and index fingers on the lower handle.  This can be a powerful and accurate method to launch the dart.  Something we’ve found interesting to ponder over; using this grip we notice that the hand has a tendency to cant inwards in the holding position (Figure 1).  This typically doesn’t affect the throw unless absolutely no attempt is made to correct it.  It can be corrected more easily if the loops are twisted.  On other Southwestern atlatls the loops are straight, but an original photo of the WDC cist 27 atlatl in situ (Guernsey and Kidder 1921:Plate 10e), shows that the loops may have been twisted around when it was uncovered.  This could have also resulted from the excavation itself, or post-depositional settling, we can’t be sure, but it is possible to twist the loops a little on our replicas.

Conclusion     Many artifacts associated with human remains in the Southwest are now being reburied, so this may be our last chance to study many of them.  Replicating and testing them offers important insight on multiple levels.  We gain insight into how they were used, but also the information that we need to gather from them to make accurate replicas.  The White Dog Cave artifacts offer a rare glimpse at a complete ancient atlatl kit from North America.  Several aspects of the atlatls and darts are echoed in fragmentary and complete artifacts out West and even as far east as Arkansas and Missouri.  By replicating and using these artifacts, modern atlatlists can develop a detailed understanding of the operational characteristics a specific atlatl system that was developed and relied on by ancient hunters.  We hope others will find this article useful, and try their hand at replicating the WDC atlatl equipment.

References Cited

Coltrain, Joan B., Joel C. Janetski, and Shawn W. Carlyle

2007  The Stable- and Radio-Isotope Chemistry of Western Basketmaker Burials: Implications for Early Puebloan Diets and Origins. American Antiquity 72(2): 301–321.

Guernsey, Samuel James

1931  Explorations in Northeastern Arizona. Papers of the Peabody Museum of American Archaeology and Ethnology, Harvard University, vol. 8(2).

Guernsey, Samuel James, and Alfred Vincent Kidder

1921  Basket-Maker Caves of Northeastern Arizona. Papers of the Peabody Museum of American Archaeology and Ethnology, Harvard University, vol. 8(8).

Pepper, George Hubbard

1902  The Throwing-Stick of a Prehistoric People of the Southwest. International Congress of Americanists, 13th, New York, 1902. Proceedings.: 107–130.

Pettigrew, Devin B.

2008  White Dog Cave Darts. The Atlatl 21(3): 7.

2009  Throwing Rocks, Not Darts with Atlatls. The Atlatl 22(3): 2–3.

2013  Making a Basketmaker Atlatl Dart Using Modern Tools. Electronic document, http://basketmakeratlatl.com/?page_id=819, accessed April 4, 2015.


Table 1. Measurements of the Cist 27 atlatl taken by LaRue. An asterix indicates measurements from digital photos taken by LaRue, and scaled with his measurement of a nearby feature.

Length: of shaft overall 60.4i  (cm)
From spur tip to distal end 5
Of loading groove to spur tip 2.5
Of handle to center of finger notches 9.9
Of first weight from handle 2.55
Of second weight 3.12ii
Of third weight 4.05
Of leather piece that forms the loops 12
Width: of shaft at distal end 2.32
At spur tip 2
At center of handle 1.5*
At center of finger notches 1.29
Of groove at spur tip 1.3*
Of first weight from handle 1.3
Of second weight 0.82
Of third weight 1.5
Of leather piece that forms the loops (after being folded in half) 1.5
Thickness: of shaft at center 0.7iii
Of shaft at distal end 1.1
Of first weight from handle
Of second weight 0.7
Of third weight 0.9


Table 2. Measurements of dart shafts from White Dog Cave taken by Garnett.



Mean S.E. n
Length overall 1400iv (mm)
Distal diameter 14.83 14.86 0.29 12
Socket width 8.16 10 0.91 11
Socket depth 22.29 22.24 2.37 8
Proximal diameter 9.91 9.66 0.3 4
Nock depth 3.26 2.1 0.77 2


End Notes:

i  This overall length is shorter than that of 25 inches given by Guernsey and Kidder (1921), but is closer to a length of 24 inches, which was provided on the Peabody Museum website. LaRue took this length by measuring along the wooden shaft, which was broken and bent.

ii  This weight is chipped at its distal end. We are not sure if this occurred before or after it was lashed to the shaft. If after, an original length of 3.5 cm is estimated by LaRue.

iii  Unfortunately the shaft’s thickness is only an approximation and is taken from the replica shown above. We have not seen a recorded thickness for the Cist 27 atlatl. The shaft is slightly thicker at the distal end where the spur sits at the head of a slight ridge. These measurements are close to the cist 24 atlatl, another fragment from WDC, and the Kinboko atlatl from a cave in the same area (see the resources section on Basketmakeratlatl.com for measurements).

iv  This length is an average from Guernsey and Kidder’s (1921) measurements for three complete shafts of 52.5, 55, and 55.5 inches. The darts were originally in fragments but could be pieced back together. The rest of the measurements for the complete shaft belong to one of these three darts (Peabody # A2813). The others could not be located within the narrow time window available to Garnett.



Additional Image (not part of the original article):


Update (9-27-2015):

Chuck LaRue thinks Basketmaker foreshafts are typically constructed of willow, rather than oak as stated above. Other woods may also have been used, and can work on replicas. We have had luck with willow, oak and others, but willow may be more solid in the mainshaft socket. Willow is also plenty strong enough for stone pointed foreshafts when good binding methods are used for the hafts.

Comments are closed.