Harescombe Castle

Description

There appears to be considerable confusion between Harescombe and Haresfield, possibly initiated by William Worcestre in 1480 and perpetuated and enlarged upon in the 19th. century. Worcestre's entry in his Itinerary is simply "Harescombe Castle, near Painswick, in the county of Gloucester". Harescombe was well off his route from Cirencester to Malmesbury and there is no reason to suppose he visited it.
Rudge notes that after the Fitz Allens Harescombe belonged to the Rous family who built the church in 1273, and to the Mills from 1360 until the mid 17th. cent. He states that "an old mansion, said to be the castle of the de Bohuns, was situated a little to the westward of the church. The materials were carried to Elmore for the purpose of building the court-house (SO 71 NE 5) in the reign of Elizabeth. The appearance of the moat etc, is still to be observed".
He considers the story may be apocryphal and if remains were removed they were probably from the Mills' house. He then says that there are neither remains nor a tradition of a castle at the de Bohuns manor of Haresfield; on both these counts he is incorrect, (see SO 81 SW 7).
Melland Hall, Rector of Harescombe, wrote a lengthy, well documented but seemingly muddled article. He shows that formerly Harescombe was a tithing of Haresfield and that Harescombe church was daughter to that at Haresfield. The castle at Harescombe was probably built in Stephen's time and was the residence of the FitzAllens who preceded the Rouse family. The castle was demolished possibly in the reign of Edward II. Hall also quotes Worcestre and the story of stone being taken to Elmore Court.
Both Harescombe and Haresfield were originally de Bohun property, the Harescombe tithing being leased.
If the Fitz Allens built a castle in Stephen's time it would probably have been a motte, but there is no documentary evidence for this or for its slighting in the reign of Edward II. It is reasonable to assume that a manor house or a succession of these existed at Harescombe.
There is no trace of a motte or any defensive work at Harescombe. The published 'Castle' site falls in a pasture field. Two or three small platforms probably represent depopulation though there is no trackway leading to the area. The sites of the Fitz Allen and Rouse houses are speculative. The two stretches of water described as "Moat" on the OS 25" are unimpressive and doubtful. The southernmost is a broad deeply cut watercourse, probably entirely natural, and the L-shaped stretch at the north end is little different though it may have been enlarged to form a pond for the local corn mill (F1 NVQ 03-MAR-70).
The area of the supposed manor house was investigated as a student survey exercise in September 2001. There is a large rectangular platform defined by slight scarps generally no more than 0.4m high. The platform, immediately S of the churchyard, measures at least 50m N-S by about 25m transversely. If it is indeed a building platform it suggests more than merely peasant occupation. The millpond, and some fishponds in the close to the W, the existence of which is suggested by the field name 'Fishpond Orchard', have been largely filled in (Mark Bowden/6-Sept-2001/EH student survey exercise). (PastScape)

Comments

Recorded by David King as a 'possible' site, which generally meant he had considerable doubts about the site. The site does seem to be that of a manor house but the evidence that this was fortified, or even moated, is very weak. The suggestion it was a castle does seem to an under critical reading of a historical source which probably referred to Haresfield Mount.
Gatehouse is kindly informed that resistivity surveys done prior to 2007 (Ecclestone 2007) indicated “buildings within an enclosure” which could constitute a manor house, but gave no indication of whether the site was fortified.