Picken’s Hole is a small partially unroofed cave on Crook Peak, Somerset. It was excavated during the 1960s when significant Palaeolithic faunal assemblages, a small collection of Mousterian stone implements and some Human teeth were discovered. This paper gives the background to the site and to the work carried out there.

The nature and sequence of deposits encountered during the excavation of Picken’s Hole are described based on field observations during excavation and available records. Six units are recognised from Unit 6 at the base to unit 1 at the top which includes the present-day soil. Unit 6 presents an irregular top surface and a range of deposits from clays and free sandy material to densely cemented limestone breccia sometimes with significant voids, but often transitional to in situ bedrock. Unit 5 is a relatively thin unit and becomes patchy or absent to the south. It comprises a variable assemblage of reddish clay loams in places with weathered limestone, and generally abundant heavily mineralised and very broken bone. Species present include brown bear, wolf, red fox, reindeer and a large bovid, but hare, gregarious vole, northern vole, bird and frog bones also occurred. Unit 4 comprises an angular limestone breccia with limited weathering, and a grey yellow to brown sandy matrix. Where it overlay Unit 5 the base contained blackish animal bones and teeth, including wolf, red and arctic fox and reindeer. Elsewhere, Unit 4 rested directly on Unit 6, from which it contained cemented clasts. Unit 3 is a maximum of 80 cm thick and comprises yellowish clay to sandy loams as matrix within highly weathered limestone breccias. There is evidence for local topographic control of some contexts, with dips towards the south, and infill of a depression formed against the cliff. Substantial amounts of fragmented yellowish and largely un-weathered animal bone and teeth were present, mainly in the lowest 50 cm of the unit. Finds were dominated by spotted hyaena, woolly rhinoceros and horse, but apparently included 2 human teeth. Unit 2 comprises a variable sequence of matrix rich breccias and more clast free reddish brown clay silts. The angular limestone clasts lie horizontally and are more weathered at the top of the unit. In square D the unit overlies a thin yellow silt over bedrock. Unit 1 includes the modern soil. Unit 1 lies unconformably over the eroded remains of Unit 2 and 3 to the west and north of the site, but appears to be conformable on Unit 2 in the south of square B. The soil comprises turf and topsoil above a somewhat bleached horizon, with a brown more clay rich unit including weathered limestone clasts at depth. The deposits have been affected by the burrowing activity of badgers, the distribution of which is described in an Appendix.

The various radiometric dates obtained from specimens from Picken’s Hole are listed. The later, AMS, 14C dates and the two U-series dates and their utility in delimiting the ages of the faunal assemblages, including the Human teeth are discussed further.

Picken’s Hole (NGR ST 3964 5500) is a mainly collapsed limestone cave on the northern slope of a ridge of the Mendip Hills near Compton Bishop, Somerset. It is one of a small series of caves at about 50 m AOD in a ridge of carboniferous limestone. At the time of its excavation in the 1960s, the accessible cave consisted of a short passage 1-1.5 m across opening onto a roughly level platform about 6 m across at the foot of a low cliff. Today the site is almost invisible as the result of weathering of the limestone and the growth of vegetation but it is unlikely ever to have been more than a small shelter. It is an important site in a number of respects. It was excavated with great attention to documentation. Several horizons were identified and although few artefacts were found, abundant vertebrate remains were recovered. Those from two levels in particular warranted the detailed study presented here. They present a rare occurrence of two large assemblages of bones believed to have been brought to the site at different periods of the Late Pleistocene principally by two different predators. The earlier accumulation is ascribed to wolves and the later to hyaenas.

The British Late Middle Palaeolithic (LMP) is numerically extremely small compared to the Continental record, and the majority of lithic assemblages suffered extreme collector bias due to early excavations. A few site were dug during the later half of the 20th century, and provide much more reliable assemblages in terms of debitage and concomitant technological analyses. The lithic assemblage from Picken’s Hole, Crook Peak (Mendips) was excavated during the 1960s– when techniques were generally improved from 19th and early 20th century work– and is therefore potentially important at a national level, but it has not been published until now. A technological analysis of the lithic material is presented. This finds that much of the material can be placed within a general LMP classification, including centripetal reduction and significant use of the immediately available chert; features that are shared with regional LMP sites like Hyaena Den. On the other hand, other aspects are unusual and may point to mixing through later prehistoric activity (the presence of which is suggested by the Neolithic human teeth). This includes a lack of any bifaces, unequivocal biface production or maintenance waste, or ‘classic’ Mousterian retouched objects or waste from their retouching. Additionally,a tiny rotated core with multiple platforms appears unusual within the broader British LMP, although a diminutive chert biface is known from Hyaena Den. It is concluded that at least some of the assemblage is likely to be LMP, and therefore this assemblage is an important addition to understanding of later Neanderthal techno-economic strategies within the south-west of Britain.

A catalogue listing and brief description is given of the twelve human teeth recovered from the excavations at Picken’s Hole. One lower premolar was submitted for 14C dating and returned a date within the early Neolithic period

An account is given of the discovery and excavation of this small cave in the 1940s. William (Willie) Iredale Stanton’s (WIS) caving diaries, excavation notes and archaeological finds are curated in the Wells and Mendip Museum. Field visits to the museum and Scragg’s Hole were undertaken to better understand its archaeological significance, topography and stratigraphy. Whilst the excavation and written reports are remarkable for someone aged 13 years the archaeological excavation methods fall a long way short of what is expected today. In his notes WIS informs us that at the beginning of the excavation he and his fellow students were excavating soil at interval depths of six inches and documenting this accordingly. However, for some reason not explained, they later started excavating and recording at a depth of every 12 inches. This presumably was to reach the lower levels of the cave and the Roman floor level sooner. It may be that in their youthful exuberance they missed possible post Roman to Medieval levels, although some artefacts from this period are mentioned in his excavation record. The bones that were excavated were not all kept. In addition, the site was exposed, well known in the local area and therefore subject to site contamination. WIS notes that they frequently lost equipment due to theft and that the site was subject to other acts of mischief