Most reconstructions of how the Stonehenge sarsens were moved fall into two families: sledges on rollers, and sledges on a fixed timber track or greased rails. Both face the same problem on soft ground. A heavily loaded runner or roller presses on a narrow strip, and on wet clay-with-flints or waterlogged topsoil it sinks. This post proposes a third option, using a technology the Neolithic is known to have had: woven hazel hurdles, laid as portable ground-pressure mats, greased on the top face, and leapfrogged from the rear of the sledge to the front as it advances.
The aim here is not to show that it was done, but whether it could work: whether a hurdle that two people can carry survives a 25-tonne load, and whether the numbers for crew, pace and materials are realistic.
Neolithic hurdles
Woven panels of coppiced hazel are well attested in the Somerset Levels. The Neolithic hurdle trackway at Walton Heath was published by Coles and Orme (1977). Making them required skilled hurdle-making, managed coppice and a habit of laying hurdles as walkways over soft ground. Using them to carry a load is a small step from that: a short length of portable trackway, rather than a permanent one.
The method
A sarsen of about 25 t is lashed to an oak sledge with two runners, each 6 m × 0.3 m, set 1.4 m apart, with upturned noses. Hurdles of about 2.4 m × 1.2 m are laid across the line of travel, butted end to end, with tallow on the top face along the strips the runners will cross. Each 1.2 m the sledge advances, the rearmost hurdle is unloaded. A pair lifts it, carries it forward beside the track, lays it against the leading hurdle, and greases it again. Ten hurdles are in use at any one time: six under the runners, two laid ahead and two being carried.
Figure 1. Side elevation, schematic.
Figure 2. The relay cycle in plan.
Animation of the relay cycle, side and plan views.
The track moves with the stone. Nothing has to be laid along the whole route, and the heaviest item anyone lifts is a hurdle of about 40 kg.
Ground pressure
The total load, stone and sledge, is about 27.5 t (270 kN). On bare runners the contact area is 3.6 m², a mean pressure of about 75 kPa. On the hurdles the load spreads over the panels beneath the runners: about 19 kPa if the hurdles behaved as rigid mats, about 28 kPa allowing for the flexibility of the wattle. That is close to the pressure of a person standing (about 21 kPa).
The undrained bearing capacity of a surface strip footing is about 5.14 cu. For very soft, waterlogged ground (cu ≈ 12 kPa) that gives about 62 kPa, and for soft ground (cu ≈ 20 kPa) about 103 kPa. Bare runners exceed the first and would plough in. On hurdles the factor of safety is about 2–3 even on the weakest ground considered.
Figure 3. Contact pressure against ultimate bearing capacity.
Will the hurdles survive?
Crushing. With rods about 25 mm in diameter at about 30 mm spacing, the two runners cross about 400 rods, roughly 0.67 kN per crossing. If only one in four crossings bears its full share, because the weave makes the surface uneven, that rises to about 2.7 kN. Taking the compressive strength of hazel across the grain as about 7 MPa, the crushed contact band is about 1.3 mm wide. Crushing does not limit the method.
Bending. Bending does. A green hazel rod of 25 mm (modulus of rupture about 50 MPa) bridges a void under the runner of only about 0.1 m at the uneven-bearing load, and about 0.45 m at the mean load. Hollows along the line of travel are bridged by the stiff oak runner. Ruts running parallel to the track are not, and they would have to be filled with turf or brushwood before the hurdles were laid.
Staying in place. The method only works if the hurdle stays still while the runner slides over it. The runner drags the hurdle forward with a force of μtopN, and the ground resists with μground(N + w). With wattle bedded into turf (μground ≈ 0.55), ground grip exceeds the greased drag by about 4.2× at nominal friction (0.13), 3.1× with worn grease (0.18), and 1.6× even on dry wattle (0.35). This is why only the top face is greased. Tallow on the underside would undo the arrangement.
Figure 4. Rod loading and the force balance on a hurdle.
Crew and pace
The haul force is F = W(μ cos θ + sin θ). Taking a sustained rope pull of 0.3 kN per person, the crew for a 25 t sarsen on greased hurdles (μ ≈ 0.13, including resistance from the bumps of the weave) is about 120 on the level, about 160 on a 1 in 20 slope and about 210 at 1 in 10. On dry wattle the figures are about 315, 360 and 400.
For comparison, in the 2016 UCL trial ten people pulled a 1.1-ton stone on a sycamore sledge over a track of silver birch logs, at about ten feet every five seconds (Archaeology; see also the earlier sarsen.org note). At the same 0.3 kN per person, that implies an effective friction coefficient of about 0.25, between the greased and dry values used here. That is a cross-check of order of magnitude, not a calibration.
Pace is set by the relay, not by the hauling. One relay (walk about 10 m, lift, lay, butt, grease) takes about 35 s. With three relay pairs taking turns, a hurdle is laid every 12 s or so, which gives about 370 m an hour. At six hours a day, the 25 km from West Woods to Stonehenge takes about 11 working days per stone.
Figure 5. Haul crew needed, and relay logistics.
Weak points
- Wear. Over 25 km each of the ten hurdles in use would be loaded about 2,100 times. Abrasion of the rod crowns, not their strength, will set hurdle life. A reserve of 30–50 hurdles, with a team making replacements, is the realistic requirement.
- Lubricant. Assuming a 0.2 mm film on the runner strips, with 10–30 per cent lost on each pass, tallow consumption is about 16–49 kg/km, or 0.4–1.2 t per 25 km trip. This is the least certain number in the analysis. If the true loss rate sits at the top of the range, the supply of grease, rather than the supply of people, could limit how often stones were moved.
- Ground preparation. Ruts and hollows across the line of the runners must be filled before laying.
- Steering and slopes. A sledge on greased hurdles slides sideways as easily as forwards. On a cross-slope it will drift, so it needs guide ropes or levers. Downhill it needs a braking crew on rear ropes.
What an experiment should measure
Each of the uncertain numbers can be tested at modest scale: a single loaded runner of, say, 2–3 t on two or three hurdles. Four measurements would constrain the analysis:
- the haul force on greased and dry wattle, which gives μtop directly;
- whether, and at what force, a loaded hurdle creeps over turf, which gives μground;
- grease loss per pass, by weighing;
- wear on the rod crowns after a few hundred passes.
Assumptions
| Parameter | Value |
|---|---|
| Stone + sledge | 25 t + 2.5 t |
| Runners | 2 × 6.0 m × 0.30 m, gauge 1.4 m |
| Hurdle | 2.4 × 1.2 m, ~40 mm thick, ~25 mm rods; ~39 kg |
| Friction, greased top | 0.08 (fresh) – 0.13 (nominal) – 0.18 (worn) |
| Friction, dry wattle | 0.35 |
| Friction, hurdle on turf | 0.55 |
| Hazel, compression across grain / MOR (green) | ~7 MPa / ~50 MPa |
| Rope pull per hauler, sustained | 0.3 kN |
| Undrained shear strength, very soft / soft ground | 12 / 20 kPa |
| Relay cycle | ~35 s; three pairs |
All values are assumptions for estimating orders of magnitude, not measurements.
References
Coles, J.M. and Orme, B.J. 1977. Neolithic hurdles from Walton Heath, Somerset. Somerset Levels Papers 3, 6–29.
People Could Have Pulled Bluestones to Stonehenge. Archaeology, 27 May 2016.





No comments:
Post a Comment
Comments welcome on fresh posts - you just need a Google account to do so.