You want me to go down that vertical crack in the cliff-face, hanging on to only a rope, Tenzing?

The kids inveigled me into visiting the Sculptor's Cave on the Moray Firth, almost suicidally inaccessible, but well worth the risk once you get there... the figures in one of the images of the interior lend some scale to it. Hundreds of skulls of children were found in it, in the 1928 excavations. It dates back to the Bronze age, and it is thought it was considered an appropriate spot for them to be remembered in religious rites.
Extract from Canmore: This gloomy and near-inaccessible cave lies in a small bay on the S shore of the Moray Firth. It takes its name from a series of Pictish symbols carved on its entrance walls, and may have served a predominantly ritual function throughout its period of use.
In 1929-30, excavation by Sylvia Benton revealed evidence for two major periods of activity. The earliest cultural deposits are dated by their associated metalwork to the Ewart Park phase (c. 1000-800BC) of the Late Bronze Age. A later occupation layer contained a rich assemblage of Roman Iron Age material (notably coins, rings, pins, beads, bracelets and toilet instruments) ranging from the 2nd to the 4th centuries AD. In the late 1970¿s, Ian and Alexandra Shepherd conducted rescue excavation on the limited remaining deposits, which Benton had left within the twin entrance-passages; further Late Bronze Age metalwork was revealed.
The most striking feature of the cave is the (formerly) substantial assemblage of human remains that was revealed in both programmes of excavation. Benton apparently recovered around 1800 human bones scattered throughout the deposits. Examination (by Alexander Low) was cursory, being limited to the recognition of a `large proportion of bones of young individuals¿ among the material discarded. The few bones that were retained were limited to fused cervical vertebrae from adults, showing cut-marks indicative of decapitation. The Shepherds found further bones, including several mandibles from juveniles, which may indicate the display of severed heads at the cave entrance.
Notwithstanding the limited stratigraphic control and lack of dating evidence, the entire human bone assemblage was long assigned to the Late Bronze Age. However, re-examination of the cervical vertebrae raised doubts, at least some of the cut-marks having been made by sharp and heavy blades, tentatively seen as iron. Further examination using scanning electron microscopy is planned. An initial bone sample from one of these vertebrae has yielded an AMS date of 1738+/- 33bp (UB-6930), which may be calibrated (at 2 sigma) to AD 231-395, firmly within the Roman Iron Age.
Is it possible that the human bones derive from at least two distinct episodes over a millennium apart? The possibility may be considered that there were two periods of deposition, that in the LBA seeing the deposition of the remains of children, with some emphasis on the placing of heads at the entrance, and that in the RIA represented by the remains of several decapitated individuals. Concern with the removal, curation and display of human heads is a persistent trait across prehistoric Europe. To this end, a further programme of AMS dating and osteological analysis is in hand.