A stone erected in 1932 (at SO89163208) is said to mark the site of Holm Castle. It bears an inscription stating that the castle was burnt down in 1140, afterwards rebuilt, and finally destroyed in the early 14th century. {Source Work 862.}

Holme Castle was certainly fortified as brattices are mentioned in connection with it and places its site a few hundred yards southeast of the Abbey Church possibly referring to the position shown as "Fishponds" on the OS plan (See HER 7426) {Source Work 862, 213.}

This is described as the site of an important castle of the de Clares Much of the earthworks and masonry were cleared away c1820 and only a circular ditch, enclosing the crest of a rise, surmounted by a modern monument remains. {Source Works 2873, 29.} There is much discussion over the site of Holm Castle neither this site nor HERs 4236 8089 can definitely be referred to as such {Source Work 5000.}

A secondary source from the OS Card reads Rigold SE, 1950, correspondence, and cannot be located further (AM 19/11/1999).

1972 - (Sites known as Holm Hill Burials (HH72) and King George Playing Fields (KGF Rew 72) were archived by Gloucestershire County Council in 2013. Pers. Comm. T. Grubb 10/2/2015.}

1974 - A geophysical survey was undertaken by Tewkesbury Archaeological Unit in 1974 in the western part of Holm Castle, Lower Lode Lane, Tewkesbury. The survey was aimed at locating the archaeological remains so the excavation could target the most sensitive areas {Source Work 5218.}

1974 - 75 SO 887321. Excavations by Tewkesbury Archaeological Unit at Holm (Windmill) Hill in 1974 and 1975, prior to Council development, revealed the plan of a 13th century stone built hall underlying which was a 12th century stone hall and beneath this the pits and post-holes of a possibly late Saxon or early Norman timber hall.

The site of a Late Saxon - Early Norman timber hall (Building A), was located during the excavations in 1974. Two broadly parallel rows of post pits and post holes 40m in length demarked the plan of the timber hall. The northern row of pits was continuous and on an east-northeast to west-southwest orientation, however the southern row was not complete, with the eastern end offset to the south by c. 1m. Various reconstructions can be suggested from this plan and these features may represent a single building, with a transeptal arrangement, but such a reconstruction has few parallels. A more plausible alternative is one in which the features represent two halls, either contemporary or in succession on an almost identical alignment, are a recurring theme of high status layout in the pre-Conquest period. As post pipes were not in evidence it seems as though the posts had been removed and the pits/holes backfilled with the surrounding clay. No evidence was found for dating the construction, use or dismantling of the halls, although a later stone founded hall dating to the Anglo-Norman period overlay the post structure. The alignment of the later stone buildings closely followed that of the timber building and points to the dismantling of the latter to make way for the stone constructions. A single sherd of late Anglo-Saxon pottery and a fragment of an Anglo-Saxon annular loom weight recovered from the site, although it is thought to be indicative of an early medieval date, they were not from a useful dating context. {Source Work 5218.}

Earlier reports and interpretations of the excavations at Holm Hill detailed the series of large and medium sized post holes underlying the later Norman buildings. Post pits and post-holes were uncovered over an area of 500 square metres although only 10 were fully excavated, some were nearly 2m deep. The failure to detect associated surfaces was due in part to later activity, but also to the use of mechanical stripping. The arrangement of the features suggested two periods of timber buildings and there was the possibility of a transeptal plan. These features could not be dated with certainty to the Anglo-Norman period, but this date was supported by the sealed find of late Saxon pottery within the area of the timber features and by the reference to an aula in the Domesday Book entry for Tewkesbury. {Source Works 1791, 1811, 5297.}

The period from the mid 11th century to the mid-14th century provided much of the evidence for activity on Holm Hill. It was characterised by the construction and occupation of two large stone structures following the alignment of the earlier timber hall with other buildings and spaces located to the south and south-east. There is a strong possibility that the excavated remains were those of Tewkesbury Manor. It was evident that there had been stone structures on Holm Hill and the remains of two major stone buildings emerged, one succeeding the other.

The medieval aisled hall (Building B) stood to the east of the later Building C and comprised an aisled hall with a chamber block to the east end. Foundations and lower wall coursing defined the ground floor of the chamber block which had a solar above it. The hall is defined by lengths of stone coursed foundations which formed a rectangle 18.5 by 13.5m. The width of the building suggests that post pits from Building A were re-used to erect arcades, creating an aisled hall. The entrances to the building from the southwest and northwest corners opened onto this passage.

It is though only the foundations of the building are stone, and that it was tiled with wooden shingles, although stone architectural details were used. The highest concentration of horseshoe nail occurred near the southwest corner of the hall, suggesting this was the main entrance for riders after dismounting.

In the mid/late 12th to 14th century - A subsequent hall (Building C) was built to the west of the earlier stone founded hall, and had foundations (1.8m wide and 0.70m deep) of proportions which suggest the provision of an upper floor, and the presence of a fireplace and chimneys where the foundations were re-enforced. On the lower floor was a wall with an entrance in the middle, which split the level into two rooms of 11 and 15m long. The evidence suggests that at the first floor level there was a large open space at the east end and a smaller space (an upper chamber) at the west end. The larger space was over the stone vaulting, which was intended to bear the greatest weight when the hall was the venue for communal events, and was heated by the larger fireplace which was placed on the north side to cause less restriction to sunlight entering the windows on the south side. The hall, occupying the eastern end of the building would have had a ridge roof following the east-west orientation of the building, in contrast to the smaller upper chamber which was ridge-roofed from north-south. This was demonstrated by the extent of the drip trench on the south side of the hall.

It is only conjecture that the timber hall was built in the 11th century and replaced by the aisled hall in the early 12th century, although this hall probably incorporated the window and door mouldings identified as early 12th century in date, these may have been reused in the later first floor hall. It is suggested that the first floor hall may have been built when Tewkesbury became the caput (or administrative) centreof the estate of the Earls of Gloucester. {Source Works 5218, 5297.}

The medieval halls were part of a complex of agricultural, industrial, domestic, and religious building to the south. Among these would have been barns, a kitchen, a chapel and a dovecot. It is assumed that the overall layout of these buildings were established during the time of the earlier stone hall (B) and modified during the life of the later stone hall (C). The form of these lesser structures was difficult to recover, some were identified in a fragmentary form in 1974; others were seen in the earthmoving of 1975. One group extended almost directly south from the earlier stone hall (building B) and could have formed roofed buildings or a series of walled yards. The alignment of the surviving wall footings conformed to that of the main buildings but they were generally flimsy constructions. During earthmoving in 1975, fragments of floor tile with the appearance of having been in-situ, were exposed along the edge of a "corridor" of smoothed stone, which ran through these flimsy structures and was the main route into the hall complex. These tiles suggested the presence of a small chapel on the approach to the main building group. To the southeast lay the remains of a large rectangular building suggesting a barn with the traces of a circular structure beyond, probably a dovecot. Earthmoving in 1975 revealed the remains of a chamber to the east of the stone halls and linked to them by a wall. Beyond the barn an area of worn and smoothed stone led to the "corridor" where a narrow slot cut into the stone surface suggested an arrangement for an intermittent barrier akin to a portcullis. Beyond this entrance a large area of compacted, worn stone formed an open area to the south of Hall C. From observations made in 1975 there were no remains of buildings on the south side of this yard.

No evidence of an enclosing embankment or ditch was located during the excavation, although the resistivity survey indicated a pattern of responses on a north-west to south-east alignment, forming an oblique line across the hill (and enclosing the buildings), but no physical evidence was seen during earthmoving in 1975. Robber trenches of a late medieval to post medieval date are located across the site in particular in the area of Building C.

Finds included 20 sherds of Romano-British pottery; a single sherd of Anglo-Saxon pottery; and large medieval and post medieval pottery assemblages; several hundred glass fragments; a variety of iron and copper objects including arrowheads; knives; horse furniture and personal decorative equipment; and a number of coins, tokens and jettons. For full details see {Source Work 5218.}

2007 - Carrent Valley Landscape NMP

Holme Castle, a Saxon and Medieval manor, is supposed to have been destroyed in the early 13th century. The site was probably on Holme Hill, above the Vineyards where masonry and foundations were visible in 1836. Traces of a ditch are still visible.

The ditched feature published by the OS is described as the remains of Medieval Fishponds, probably belonging to the Abbey of Tewkesbury by Rigold. Bennett says that the Vineyard was probably a quarry for building material.

The ditched feature shown, and variously described as 'Holme Castle (site of) and 'Fish Ponds', by the OS is on low lying ground and has been almost completely overlaid by a large refuse heap. The present remains are insufficient to determine whether the feature was a moated site as suggested by earlier OS plans, or Fish Ponds as stated by Rigold. A stone erected in 1932 on higher ground to the south west (at SO 89163208) is said to mark the site of Holme Castle. It bears an inscription stating that the castle was burnt down in 1140, afterwards rebuilt, and finally destroyed in the early 14th century. It is possible that 'The Vineyards' is yet another reference to the ditched enclosure. There seems to be no other feature to which the OS name can be applied although Tewkesbury Abbey has a vineyard. In the Abbey's Records there is a reference to one tun of wine from Tewkesbury vineyard being sold to the king in 1238.

Tewkesbury was one of the principal residences of the Earls of Gloucester, and passed to King John after his marriage to Isabella, Countess of Gloucester, until her remarriage in 1214. He stayed there on many occasions during his reign and it was frequently repaired on his orders. Holm Castle was certainly fortified, as brattices are mentioned in connection with it. The site was leveled and backfilled in 1826 by a `utilitarian banker'. It was probably a few hundred yards SE of the Abbey church, possibly referring to the position shown as "Fish Ponds" on the OS plan Leland describes the castle ruins as standing on Holm Hill, by the bank of the River Swilgate at its junction with the Avon and SW of the abbey; Col Blyth suggests that the only place fitting Leland's description is the top of Holm Hill to the west of Holm hospital. He also mentions a 14th century kitchener's account for work done on the weir across the Swilgate, which supplied water for the fish ponds.

The Medieval fishponds or moated site noted above, sources, are visible as earthworks on historic aerial photographs but subsequently have been leveled. They extend over an area that is centered at SO 8936 3216 and comprise a linear drainage channel that forms an incomplete rectilinear circuit around two parallel rectilinear ponds that are orientated north / south. A further roughly rectilinear pond with two circular islands is situated immediately to the west. The pattern of inter-connected ponds corresponds more closely with a system of fishponds rather than a moated site. {Source Work 4249.}