4. Assemblages from the Well

4.1 General assemblages

Hodgson: Report on Roman stones (Digital Archive)

A range of artefacts was found in association with the well deposits (Table 1). Coarseware ceramics and faunal assemblages will be discussed in detail later as whole assemblages. However, certain other, particularly significant, objects are noted here first. A Roman roof finial, in coarse-grained yellow sandstone, was built into the well facing, positioned c. 1m down from its top surface and thus breaking up the symmetry of the coursed masonry. Its square base, 250mm across, rose into a cylinder surmounted by a truncated pyramid. Shallow, arch-like openings in two opposite sides of the base suggest that it was placed on the ridge of a roof, most probably that associated with large quantities of lozenge-shaped stone roof slates found in the vicinity (see archive). This artefact does not seem to fit any of the common finial categories, lacking the classic four-way arch openings and being simply and crudely executed. It does, however, suggest that a substantial structure was being dismantled when the well was built. This element of spolia may have been recycled opportunistically or, more likely, as a deliberate act.

Figure 5

Figure 5: Bucket as recovered in 2093

Allen: Report on a waterlogged wooden bucket (Digital Archive)

Several small fragments of undiagnostic leather were recovered from lowest fill 2093, along with a rather more significant, near complete bucket or pail (Figures 5 and 6) found on the north side of the well at the interface between 2093 and underlying fill 2109. Having an estimated capacity of c. 5l, it comprised twelve individual staves made of yew (Taxus baccata L.), the base being formed of a single piece of ash (Fraxinus excelsior L.). The staves were held in place by two iron hoops (see archive). This bucket is unlikely to have been discarded as waste, given that the metal fittings remain, and such a complete vessel might suggest deposition as an offering. Equally, a missing mount and handle could imply accidental loss.

Figure 6

Figure 6: Illustration of bucket © S. Allen. Used with permission.

Building material from the well comprised two main elements; remnants of the hexagonal sandstone roof tiles, also seen elsewhere on site, and a smaller group of ceramic brick and roof tile (see archive report). Sherd size in the latter assemblage reflects the depositional processes outlined above, with the largest fragments, by count and size, and range of types by function and fabric, in initial dumping 2093 and much smaller fragments from later fills, including a slightly larger number of what must be very residual material from post-Roman fill 1106.

McComish: The ceramic building materials and stone roofing tiles (Digital Archive)

Williams: Amphorae report (Digital Archive)

Several of the well deposits contained amphorae sherds, all derived from thin-walled Gallic examples (see archive). Numbers are low, but the larger sherds again came from early fills 2093 and, especially, 2109, while the absence of the earlier Dressel 2-4 or Dressel 20 forms testifies to the generally non-residual character of the well assemblages, something also implied by the lack of worked flint. Equally, the absence of medieval or post-medieval pottery and clay pipes suggests intact stratigraphy free from intrusion. Finally, coinage, metal finds and human bone, in any case rare on the rest of the site, are entirely absent from the well, suggesting that such items were not being selected for specialised disposal practices.

Lastly, turning to material that might indicate broader contexts of deposition, we can consider environmental samples (see archive). Early fill 2093 produced evidence for hemlock (Conium maculatum L.) and stinging nettle (Urtica dioica L.), likely denizens of abandoned waste ground, and of uncharred heather (Calluna vulgaris (L.) Hull) and bracken (Pteridium aquilinum (L.) Kuhn), potentially from stable litter. Twisted willow (Salix) withies might suggest such fencing near the well head. Shoots of some calcifuge mosses, along with weedy and scrub waste elements and also heather root and twig implying heathland, were derived from overlying 2046, along with flax seed, beet and a single uncharred spelt wheat glume base, suggesting the presence of three domesticated plants within this fill as it formed. The assemblage derived from contiguous deposit 1979 also contained an abundance of hemlock and nettle fruits with some hazel and willow roundwood. Thus stands of hemlock and nettles, with some rough grassland, may have surrounded the well when 2046/1979 were deposited.

Hall and Kenward: Assessment of plant and insect materials (Digital Archive)

Insect assemblages from 2093, 2046 and 1979 are all typical of well fills, with abundant remains of 'pitfall' species and others that probably entered in flight. They, again, suggest a weedy terrain, with some grassy areas and probably grazing in the general locale, given the dung beetles (scarabaeids). This is supported by the few molluscs recovered, the species being predominantly widespread generalists and those preferring moderately moist habitats, whether open, waste ground, woods or ditches. Importantly, the insect assemblage evidenced no clear synanthropic influence beyond the species that generally favour disturbance. This suggests general abandonment from fill 2093 onwards, and that conditions around the well either remained surprisingly constant over time or, more likely given other evidence, that the well filled quickly.


Internet Archaeology is an open access journal. Except where otherwise noted, content from this work may be used under the terms of the Creative Commons Attribution 3.0 (CC BY) Unported licence, which permits unrestricted use, distribution, and reproduction in any medium, provided that attribution to the author(s), the title of the work, the Internet Archaeology journal and the relevant URL/DOI are given.

University of York legal statements

File last updated: Fri Jun 21 2013