Does the A303 corridor record glacial till at Stonehenge?

Steve Hooker (BuyStonehenge) and now Brian John (Stonehenge and the Ice Age) have asked the fair question: a few NSIP evaluation sheets say “glacial” or even “glacial till” — is that ice on Salisbury Plain, or loose wording?

I’ve put the Highways / NSIP holes on a map with the elevations and deposit labels: A303 corridor gazetteer. Five trenches carry a “glacial wording” flag. Four of them (Report 4, Western Portal) sit together on Normanton Down high ground (~97–99.5 m OD): T241 “glacial coombe”, T247 / T248 “glacially affected coombe chalk”, T263 “glacial scouring / glaciation”. The same report’s results section and neighbouring sheets call the natural soliflucted / cryoturbated chalk with periglacial stripes (e.g. T214 / T215, p.71) — one coombe-crossing cluster, not four ice claims.

The fifth is eastern (Report 5, Countess West): T511 context 51132 has the famous “fluvial action on glacial till…” sentence. Immediately above, 51131 is “Soliflucted chalk. Same as (51103)”, and 51132–51136 are all equated to that natural. Sample 51138 is a 40 L bulk of 51135 (light green sand and degraded chalk), not of the till sentence — and in Appendix C (p.86) it sits in the charred-plant / flotation table. It was kept for environmental archaeology, not as a till micromorphology bag waiting for a glacialist.

Report 5’s own geoarchaeology defines coombe as material soliflucted under periglacial freeze–thaw. BGS Salisbury Sheet 298 maps Head / Coombe on the Plain, not ice-sheet till. Clarke & Kirkland (2026) likewise leave the Plain unglaciated. So the corridor answers Steve and Brian’s wording question: cold-stage Head / coombe, not lodgement till. 

The hummocky ground on the west side of Stonehenge Bottom is likewise not ice: LiDAR shows scrape hollows upslope and spoil heaps below, ordinary chalk diggings and slumps, the same pattern as on other dry-valley edges, not dead-ice hummocky moraines. As for “nobody has clast-counted the coombe for exotics,” river-gravel and terrace assemblages in this district have already been worked; if the ask is something else, it has not been said.

What about the OSL dates? Steve and Brian also ask after four Phase 7B cores recovered in black liners for luminescence dating (TR010025-002259: STP70501, STP70503, STP70404, DTP70704). Those holes were drilled and the cores sent to a specialist lab, but no ages appear in the public factual report — and with the tunnel scheme cancelled, I doubt anyone still had the cash or enthusiasm to push unpublished lab work into the NSIP pile. That is a dating loose end. It is not evidence of till waiting to be discovered.

We already know what sits in these dry valleys from the boreholes. At Stonehenge Bottom the logged stack is periglacial colluvium / head over chalk (see the corridor gazetteer, including BGS SU14SW62) — not an ice-sheet till. West of there, Report 7 BH1–BH6 across the Winterbourne Stoke coombe likewise put freeze/thaw coombe and Holocene colluvium on chalk rockhead (~71–76 m OD). Missing OSL numbers cannot turn those descriptions into lodgement till.

The OSL we do have is still interesting — just for Holocene landscape history. Report 5 (Eastern Portal) dated the Trench 504 buried soil and colluvium (Appendix D): roughly 2.08±0.19 ka (Late Iron Age–Romano-British), 1.07±0.11 ka (late Saxon), and 0.47±0.05 ka for the upper colluvium — slopewash in stratigraphic order, with the report’s bleaching caveats. Useful. Not a Middle Pleistocene till clock, and not those four unpublished Phase 7B cores.

(PDF page numbers are viewer pages; printed footers can differ by one.)