henge

Geometry and alignments

Surveyed geometry — Established

The figures the app computes with, as encoded in HengeGeometry/Monument.swift (Cleal et al. 1995; Petrie’s and later surveys). Since September 2026 the stones themselves are placed by row from the 1989–90 Rees plan (HengeGeometry/StonePoses.swift, Daw’s digitisation, CC BY-SA 4.0); the ring figures below are the fallback slots for the stones the plan has lost, and the fitted ring the surveyed uprights stand on.

Feature	Figure
Monument axis	azimuth 49.9° (NE, to the midsummer sunrise quarter)
Sarsen circle	30.9 m diameter to the uprights’ centres (fitted through 13 locked uprights; the customary 33 m is to the outer faces), 30 uprights
Aubrey ring	87.0 m diameter, 56 holes, drawn as an even ring phased on AH18 a foot west of Station Stone 92 (Hawley 1923 p 16; AH1 at 55.2°); a reconstruction, ±1 hole. Today’s markers are round, 1.06 m across (the app’s choice: no source gives the slabs’ size)
Enclosure earthwork	~110 m across; the ditch’s centre line 107 m (Cleal p 24: radius 53.5 m; the app’s circle, fitted to the vendored Fig 36 segments, 53.8 m; the two readers’ own circles of the excavated half average 53.9 m)
Heel Stone	77 m out along the axis, 4.7 m tall
Avenue	22 m between its ditches’ centre lines (PP2020 p 256; 22.4 m at the elbow), the banks inside them; straight for 530 m from the terminal at the counterscarp to the elbow (Cleal p 292; Darvill p 1035), then east and south to the Avon, about 2.8 km in all (PP2020 p 32). The overlay draws the straight stretch on the axis, and past the elbow Cleal’s Plan 3 stretch by stretch: solid where an earthwork survives, dashed where the line is a cropmark plotted from the air, nothing where none is recorded, nor across the last c 167 m to the river (Cleal p 292). On the 2022 LiDAR the straight stretch’s ditches are 22–23 m apart and their midline runs about 2.5 m north-west of the axis through the circle’s centre
Station Stones	four, on the Aubrey circle, forming a rectangle

• The station rectangle’s short sides parallel the main axis; its long sides run toward the major standstill moonrise/moonset quarters — this much is geometry anyone can redo from the plan. Whether it was designed for the moon is Debated (see the lunar page).
• The “rectangle” is not quite one. In Daw’s plan the long sides bear 138.88° (94 → 91) and 139.95° (93 → 92), a degree apart. The app’s lunar row measures against their mean, 139.42°, and labels it provisional twice over: 91–94 are Daw’s placeholders (seed_only, plan_digitised), and which side — or what combination — is “the line” is an owner decision not yet made (CP1, #57). StationLine derives both sides from the stones as placed, so correcting a stone moves the line.
• The lunar row is solved against the Moon (#57): its rise azimuth (upper limb, parallax, refraction, over the skyline) and its own diameter as the tolerance. The extreme is the southernmost moonrise in the year centred on the major standstill, found from the node’s crossing of the equinox rather than a date. Over the baked skyline (0.21° along the line) that rise comes 2.5° south of the mean line in the builders’ era (4 April 2499 BC, Julian) and 1.2° south of it at the 2024–25 standstill (22 March 2025). The lore note’s “within a degree or so” is under review against those numbers; it has not been changed to agree with them.
• In the almanac strip the lunar row reads today’s moonrise against the line, tagged provisional, and two rows below it give the Station Stone answer for both epochs: the degrees from the mean line, and in brackets the range across the two long sides. VoiceOver reads the caveat and both sides aloud. The rows are solved once per civil day, and the two standstill searches off the main thread. The two epoch rows are always the monument’s answer, at its site over its skyline, whichever viewpoint is shown: switching to “here” does not move them.
• The axis alignment itself — midsummer sunrise up the Avenue, midwinter sunset the other way — is Established: it is the plan, not a theory about the plan.
• Cardinal directions: the app’s overlay draws true north from the ephemeris frame, not magnetic.

The stonehengeology.com constructions — Speculative

Source: stonehengeology.com/stonehenge-geometry (fetched 2026-07-31, at the owner’s direction). One author’s geometric reconstruction in the tradition of Alexander Thom’s “Megalithic Yard” (MY); enthusiast work citing real surveys (John Wood 1740, Petrie 1880, Gowland 1901, Hawley and Atkinson) but presenting derived constructions as demonstrated fact, without peer review. Every claim below is Speculative and must be badged wherever the app shows it.

Claims, as made:

• Sarsen circle diameter = 36 MY; 30 stones at 12° spacing with 4° gaps; primary axis azimuth 50°.
• An outer bluestone oval “dedicated to the moon”: two 27-MY-diameter circles with centres 2 MY apart (envelope 27 × 29 MY, sum 56 — read as the 56 Aubrey holes and a “56-year moon cycle”), its axis offset 24 megalithic inches from the sarsen centre, aligned on the southernmost moonrise / northernmost moonset.
• An inner bluestone oval “dedicated to the solstice”: 14.5 × 18 MY, built from four near-Pythagorean triangles (3.5 : 7.25 : 8 MY), on a secondary axis through trilithons 3 and 4.
• A central 3:4:5 triangle (18:24:30 megalithic inches); the Altar Stone 6 MY long, one sixth of the sarsen diameter; the station rectangle “1,920 times” the core triangle’s volume.
• Aubrey holes 56 and 28 said to have held timber posts fixing the axis five centuries before the sarsens.

What survives contact with the established register: the axis azimuth (50° against the surveyed 49.9°), the stone count, the 56, and the station rectangle’s lunar bearing — all of which the app already carries from better sources. The Megalithic Yard itself is rejected by mainstream archaeometry (Thom’s unit does not survive statistical re-analysis); the ovals and triangle constructions are the author’s own.

In the app: the geometry overlay draws the established table above with labels and ground-mapped measurements; the speculative constructions, if shown, ride the same hypothesis badging as the Hoyle markers.

The lunar hypotheses — Debated

See research/lunar-markers.md: Hawkins (1963/1965) and Hoyle (1966) on the Aubrey ring as an eclipse predictor. The app’s gold-marker mode acts out Hoyle’s machine and says so on screen, every time, with the honest scorecard (42 caught, 0 missed, 9 false alarms over a test decade). Hawkins’s specific mechanics still need primary-source pinning before any in-app prose uses them.