Stonehenge Solved Its Three Observatories, Who built them, How and Why

How best to make a start? We must try to identify some aspect of the monument which seems to hold promise of offering a useful clue, or clues, we can work on. We can literally walk round the jumbled edifice with binoculars, or do so metaphorically by studying accurate plans of the site while examining close-up pictures, or preferably both. As we peer through the veils it seems likely we have an essentially practical structure. The task now is to examine it objectively to find out how the designers intended it to be.

One thing is clear. If we view the upright stones at the main entrance on the northeast when standing inside the Sarsen Circle, they seem to be undisturbed, regularly assembled and in relatively good condition, Fig 9. But as we walk round the whole circle the degree of collapse and the irregularities become progressively worse, until sharply they change back to orderliness at our start point. It is here we may find something useful, especially by scrutinising close-up illustrations in publications. Demonstrably, the best hope is the three-way joint at the top of an upright: one of these and two cross-beams, or lintels, resting on it, form a tee, Fig10, (i).

In the detailed, schematic, but realistic depiction of Fig11, it is possible to imagine how the upright and beams would have moved, or not, when pressures were applied to them from various directions. From this assessment we can form a view about the stresses they were designed to withstand when they were used, and in turn what may have been going on to cause such stresses. The design was as follows. A beam was locked to its fellow on each side by a vertical tongue down the centre of one end, and a groove similarly at the other. It was also anchored to the upright by a mortice, or recess, in its underside, fitting loosely over one of two tenons on the upright. The mating ends of the fellow beams formed a very slight taper, narrowing inwards when viewed from overhead. These are the features of the joint when viewed vertically. Now we must consider those we can see horizontally. These are designed to ensure the underside surfaces at the ends of the beam, and the top of the upright exactly mate. Firstly, after the upright had been made vertical and allowed to finally settle in its foundation hole, the top was reduced and made dead level with that of the upright last erected, but leaving two upright lugs of stone for the prospective tenons. Then the topmost few centimetres of the upright's four rough, vertical surfaces were made smooth and rectangular when viewed from above, to leave a sharply defined corner all the way round. From about 5cm (2 in) inwards of this corner, again all the way round, stone was removed to form a shallow tray, on the bottom of which was formed the tenons. The edge of the 'tray' is more obvious at the top of stone 56, where, on one side, erosion is almost absent, Fig12. We now come to important details.

Only the ends of a beam's underside were made flat and smooth and also aligned to one another, and in these the mortise holes were fashioned. The tongue and groove were fashioned at opposite ends. The rest of the beam's surfaces were left rough but all of its uppermost 'corners', including the fashioned ends, were radiussed. The outer bottom corner was chamfered, the inner one not. The vertical sides of the beam were curved to suit the circumference of the circle they formed. It had a thickness of 76cm (2.5 ft) and a width of 106cm (3.5ft), and it weighed about 6.25 tonnes. The beam's ends and the fashioned top of the upright, worked with precision, contrast with the bulk and roughness of the stones and leave no doubt that the joint was intended to cope with stresses in service. The roughly trimmed finish of the rest of the beam seemingly was of little moment, although the radii and chamfer extending along its entire length had functional significance, as we will see.

Two less obvious but important characteristics of the joint must be understood. It would not be possible for both the mortice and tenon, and the tongue and groove, all to be exact fits at one and the same time, therefore one or other must have priority. This is why the former were made a slack fit, and probably there were two reasons for the choice: locking the beams tightly together was more critical than their slight horizontal movement on the upright. The second reason was that at the final stage of assembly, while the tongues and grooves could be seen and watched, the mortices and tenons were inaccessible and out of sight.

The strengths and weaknesses of the joint, or more appositely, the assembly, can now be addressed and this is readily done by contemplating what would happen in extreme circumstances when it is about to fail. Suppose it is pushed, or 'loaded' as engineers say, horizontally from the inside of the circle outwards: after riding over the tenons both beam-ends would move outwards and crash to the ground. Conversely, when loaded from outside the circle inwards, the tongues and grooves would ensure the beams could do nothing but jam together in the tapered gap formed with their fellows. The tongues and grooves would also prevent twisting, or tipping inwards, or outwards. While an upward load under a beam could lift it off, increasing downward loads would make it more and more secure until at worst the narrow edge around the top of the upright crumbled.

From the foregoing it appears the joint is designed to withstand inward and downward loads, and possibly twisting, or tipping. We can now also understand the importance of the tray. Were particles of stone to become dislodged through inadvertent grazing of the stones during assembly, as an example, if the unseen mortice scraped the tenon during lowering, they would foul otherwise smooth, flat, mating surfaces. The tray avoids this possibility. If a particle fell on to the narrow edge, however, it could readily be seen and brushed off, which brings us to why the avoidance of fouling by particles is so important. It can possibly create 'point', or local loading, of the stone causing it to crumble. This in turn would increase the load on the remaining edge thereby increasing the chances of further failure elsewhere. An advantage of the narrow edge, or 'land', was that as the beam was lowered to mate with the upright, any unforeseen unevenness between the two could more readily be rectified, again avoiding local loads and incipient crumbling. It now appears that in addition to withstanding inward, downward, twisting and tipping loads we must add stability, the absence of rocking movements and convenience during assembly.

The intended, or Aim-Design, has now been identified for all of the sarsen circle, Fig 13, except for the two entrances. Here, beams with a rebate, or stepped underside were employed, thus preventing their two uprights closing together, Figs 9 and 57. This feature must be a further clue to the beam's duty requirements. At both entrances, at the centre of the outer vertical face of the beam, was a protrusion; in the case of the secondary entrance, appearing to be similar to a printer's double-curve bracket, when viewed from above. These are thought to be conventional signals of their functions. The tee joint is a sophisticated design, employing the principles of toolmaking practice in today's production engineering. What use of the structure did the designers have in mind?