Bison Experiment 2: A summary of interesting results

We are very close to submitting initial publications from our recent bison experiment. In the meantime, I want to give an update of some of the more interesting discoveries.

The coupling of the powerful radar gun we acquired with funds from our backers and the high-speed camera that we've used in a number of experiments to measure velocity and deceleration was a great success. Readings from both instruments were in strong agreement, legitimizing our velocity measurements from past experiments as well. The last paragraph mentions further use of the radar gun.

The bison bull presented a tougher target than the bison cow that we tested previously. The average force (mass*acceleration) met by stone points that penetrated the skin over the torso was 37% greater than the average force penetrating the cow. This resulted in a slight decrease in penetration relative to the cow. But we still documented deep penetration in soft tissues of the bull. Ten shots with Clovis dart points on average penetrated the width of the torso and one shot passed through both sides of the thorax. The wound surface areas from these shots far surpassed the recommended wound surface area to hunt a bison (see Friis-Hansen, 1990, "Mesolithic Cutting Arrows").

 

Some archaeologists and bowhunters have used momentum (mass*velocity) to predict penetration depth, but in our data across five very different carcasses (two goats, two bison, and a hog) and a range of stone-tipped atlatl darts and arrows, kinetic energy is the best predictor of penetration. This is not surprising, since in classical mechanics kinetic energy can be thought of as the work a projectile can do to a target (both kinetic energy and work are measured in Joules). Changing a solid target by creating new surface areas within it (i.e. by a projectile piercing or cutting through it) requires work to accomplish. Next to kinetic energy, our second-best predictor of penetration is the aforementioned force cutting through skin. The latter provides a relative measure of the efficiency of the stone tip.

We tried several measures of stone points to capture their efficiency, including length and width ratios, length and thickness ratios (i.e., frontal angles), edge angles measured through 3d models, and various different measurements of cross-sectional size. The cross-section (imagine looking straight down the point from the tip and seeing the shape of the cross-section) is frequently used to model bullet penetration and has been suggested to be an important predictor of stone point penetration as well. But this measurement has been validated in homogenous target simulants (e.g., ballistics gelatin and pottery clay) and seems more important for fluid models of penetration, which can be relatively effective for modeling bullet wound ballistics given their very high velocity.

However, ballistics gel tends to underpredict penetration for darts and arrows with cutting tips and does not capture the same characteristics of efficiency (e.g., a smaller and duller point may penetrate deeper in gel than a larger sharp one, whereas the opposite result occurs in carcasses). Subsequently, cross-sectional size performs quite poorly for capturing either penetration depth or force cutting through skin. Surprisingly, a better measure seems to simply be the material the point is made from. For example, two large obsidian Clovis points penetrating the bison bull averaged 83% less force cutting through the hide of the torso than two Clovis points of Texas cherts. However, obsidian is quite fragile, so a balance of sharp and durable stone has been recommended (Loendorf et al., 2018, “Raw material impact strength and flaked stone projectile point performance”). This requires further investigation, but it may give some insight as to why high-quality stone was so sought after in ancient times, sometimes being traded or carried hundreds of miles. Clovis people almost always used quality "exotic" stone for their points and would settle with rougher local stone for other tools (Bamforth, 2009, "Projectile points, people, and Plains Paleoindian perambulations").

 

A screen capture showing analysis of an obsidian Clovis-tipped dart penetrating the bull. In most shots, a fast drop in velocity occurs when the point enters the carcass, but initial deceleration (a) is not rapid for this point.

There has been some question recently regarding the capacity of Clovis points for hunting ice-aged megafauna (e.g., mammoths). For instance, a recent paper described an experiment shooting ground stone Clovis points into pottery clay and documented ~20 cm average penetration (Eren et al., 2020, "North American Clovis point form and performance"). However, their projectiles carried limited kinetic energy. It follows that archaeologists need access to more data on the ballistic potential of the atlatl and dart, with a focus on potential kinetic energy. We performed a small experiment using some heavy darts, strong throwers (Donny Dust and Carlton Shield Chief Gover), and the new radar gun, and we've now managed to document well-over 100 joules of kinetic energy with the atlatl.

Although modern folks probably aren't as skilled, or on average as strong, as Pleistocene hunters, it is useful to show what we can accomplish. But we should also be hesitant to draw insights from experiments like these about the upper ends of ballistic potential. If Clovis people adapted atlatl darts specifically for hunting big animals and were well-practiced with them, it follows that if they performed such an experiment today, they could achieve deeper average penetration than we can.

Look for more details on ballistic performance after publication of our papers.