More on the well

Excavation on the well continued through last week and over the weekend with much of the bluey grey mud from the lower layer being sampled and either saved in tubs for later analysis or flotted on site over last weekend revealing some interesting iron finds. We are now down 1.6m below the top course of flint conglomerate that forms the upper layer of the lining. I think this is as far as we can go. We have had some cow bones but no votive dog burial (yet). Below is a brief initial report on the construction of the well as we observed it on Monday.

Nails, an eye and a plate from the well

The construction of the well in the NE corner of Trench 6 at Bridge Farm

The structural elements of the well were originally discovered at the end of the 2016 excavation at the base of a large pit just to the NE of the NE corner of the outer enclosure ditch, some 2m below the probable surrounding Romano-British ground level. The pit has sloping sides from an irregular sub-circular surface which had either smaller pits dug into its NE edge or shows signs of edge collapse; the latter being quite likely in the sandy silt through which it was cut. It is unclear whether this pit was dug for the construction of the well or much later to help remove the upper courses of the stone well lining after it had become redundant. The depth of the structure against the remaining section baulk restricted excavation of the interior to four/five course comprising of large blocks of conglomerate over slabs of hard ferruginous sandstone.

The upper area of the well pit was filled to a substantial depth of  by the dark brown ‘demolition’ layer (6050) which has a distinct layer of burnt clay fragments at its base and covers the entire NE corner of the enclosure ditches plus an adjacent area of the London road. Below this layer, which was extremely rich in metal finds including coins, was the grey-brown layer (6063). Both these layers gave good pottery recovery which should facilitate dating and phasing.

The well-pit’s relationship with the enclosure ditch is not clear due to disturbance of the ditch in this area. It would appear that the well itself is just outside the ditch line and constructed after the ditch/ditches had been back-filled, as the above mentioned layer (6095 same as 6050) overlays the flint surface of the London road which itself overlays the refilled ditches.

Stuart standing in the well

The recurrent rain storms of the 2017 season greatly hampered the uncovering of the well’s stone lined structure and the removal of the adjacent baulk until the last week of the season and forced an extension of the season into mid-August . The baulk itself had to be excavated carefully due to the high content of metal special finds contained in its main fill (6050). Dryer weather finally allowed the interior of the well to be excavated, by members of the CAP committee, site supervisors and a limited number of our regular volunteers, to a depth of 1.6m below the top of the remaining lining. This equates to a total depth of approximately 3.6m to the probable Romano-British ground level.

The various layer of the lining

This revealed a complex structure  comprising layers of various materials which in descending order are: irregular flint rich conglomerate blocks, ferruginous sandstone slabs, smaller chalk blocks, 4 substantial oak beams, and larger chalk blocks at the base on the various faces of the well as defined by the large oak planks surrounding the well which form an 850mm sided square at between 350mm to 600mm above the excavated depth.

The well pit and the lower well lining were both added to the site plans, with sections having already been drawn of the pit to the upper lining courses in 2016. The interior of the well was extensively photographed and a series of measurement taken. However we decided this was just too important a feature not to be drawn and so measured elevation drawings were produced by David and Stuart despite the restricted space of the well interior.

The construction of the well lining is intriguing with the oak planks forming a level square frame on top of a substantial chalk base with chalk coursed layers above. These layers are then capped by the hard sandstone slabs and finally be the irregular lumps of conglomerate. The occasional use of softer sandstone seems to have been either a symptom of shortage of chalk blocks or later repair as they are clearly not as resilient to the damp and potentially waterlogged conditions at the upper chalk layer. Does this change from chalk to sandstone suggest the normal/expected level of water within the well? Chalk, particularly the Lower Downland Chalk from deep quarrying is a robust material if used below ground level even when waterlogged and not exposed to frost. It was used for the foundations of flint walls in the locality during the Roman period, i.e. Barcombe villa and bathhouse, and also for the walls of vernacular buildings in later periods.  The surrounding sandy silt alluvium and the chalk well-lining would have provided a basic cleaning filter for the well water.

The thin planks revealed at the second layer

A group of thinner planks found within the well could possibly be the remains of a cover or roof which would have prevented the growing of algae. Smaller hard sandstone blocks and sizeable Downland flint nodules found within the well and in the surrounding context could be a remnant of the, robbed-out, upper well-lining giving an indication of the upper structure which may have resembled that of the well at Barcombe villa.

David Millum, ACIFA, Deputy Director of the Culver Archaeological Project  – 15/08/17

Comments are closed.