Site Report

WoSASPIN 9246

(NS 6093 5953) Castlemilk House (NAT) and remains of Tower (NR)
OS 6" map, (1957).

Castlemilk, formerly a mansion house of a branch of the Stewart family, is now used for institutional purposes. Its nucleus is a massive tower dating probably from the late 15th century. The thick, roughcast walls rise from a basement course three main storeys to a parapet, with a modernised garret above. Windows have been enlarged, and a high arched doorway inserted in the W wall (See Tranter 1962-70).
N Tranter 19662-70.

Castlemilk House (Information from Matron, Children's Home) is generally as described. The remaining external features are modern, but portions within date from the 17th century. The building is in use as a Children's Home run by Glasgow Corporation.
Visited by OS (J L D) 31 March 1954.

This building has been demolished.
Information in letter from T C Welsh to OS 14 September 1973.

A small trial excavation was undertaken to investigate the remains of the tower house and to look for a barmkin or any courtyard structures in the surrounding area in advance of the redevelopment of the site into an adventure playground. The depth of the trenches was limited by the architects requirements for the project. The excavation was carried out by Archaeology Projects Glasgow in close collaboration with the Castlemilk Local History Group.
Castlemilk House dates to the 15th century when it was built as a tower house for the Stewart family. During the 18th and 19th centuries the house was extensively rebuilt and renovated and became a large manor house with landscaped grounds and gardens. It was in this final state that the house was bought by Glasgow City Council in 1936. It was then used as a children's home until the early 1960's, after which it fell into disrepair. The main house was then demolished in 1969 leaving the centre stump of the tower house which still survives today.
The initial trench situated outside the south-west corner of the tower failed to uncover any material associated with the earlier phase of the house. It appears that this whole area was severely truncated when the main house was demolished in 1969. A second small trench excavated round the north-west corner uncovered the wall of the 1851 extension but the presence here of a sub-basement rendered further investigation unnecessary within the terms of the remit.
Two small trial trenches were then excavated in the interior of the toweer. These showed that any occupation levels had been bulldozed and that the surviving ground level is in fact the bottom of the original foundations. These were later strengthened after the demolition of the main tower house by a concrete apron which two sides of the tower.
A further two trial trenches were excavated to the west of the tower beside the existing asphalt playground. Both of these picked up a defensive ditch, as might be expected given that the eastern approach is the least naturally defended. In the southernmost trench the ditch appears to curve round to the east. Due to lack of time and for safety reasons the ditch was not bottomed but would appear to be excess of 2.00m deep. In the northern trench the ditch had been filled with midden material which produced pottery, bone, bottle glass and a clay pipe bowl that would date the deposit to the 18th century if not earlier. A mortared stone stone structure with a dressed stone face was also uncovered running EW across the line of the ditch and has been interpreted as a possible bridge.
The southern trial trench which has the ditch turning to the east, also contained a stone built drain and a possible stone wall. These features run parallel to one another and do not respect the line of the ditch. While their function remains unclear they do confirm that activity was taking place beyond the ditch in the area currently occupied by the children's playground and that any buildings in this area were demolished in or by 1851 when the house was extensively remodelled and this area landscaped. Two more trial trenches were excavated but only uncovered re-deposited natural and were abandoned at a depth of 1.20m. Sponsor: Glasgow City Council
R Will 1991

NS 609 596 The survey area comprised the portion of the undeveloped grounds of the former Castlemilk House, including the remains of Castlemilk Tower and a possible motte. A desk-based study and field survey traced the development of the designed landscape, identifying and describing the surviving elements, and relating them to their historical context. A surprising number of minor elements have survived, including bridges, paths and drives, dams and plantations, enabling the landscape phases to be reconstructed. The survey findings were also placed within the context of the archaeology of the immediate area.
Sponsor: City of Glasgow Council Parks & Recreation.
D Topen 1997

Castlemilk Tower dated probably from the late 15th century (Tranter 1965,124-5). The House incorporated the Tower and was largely of eighteenth century date, with substantial alterations and additions in 1841. Both the House and the Tower were demolished in 1969. There was a small trial excavation in 1991 on the site of Castlemilk Tower, which is believed to date from the 15th century. (Will, R and Leslie A GUARD 52, 1992) The excavation was in advance of redevelopment work by Glasgow District Council which involved constructing a childrens' playground. The conclusion records that'... despite the high small find recovery rate from such a limited archaeological exercise, no closely datable medieval deposits were located'.
Topen, D., 'An archaeological survey of Castlemilk Glen, City of Glasgow', ACFA (1997)

Early medieval (pre-Tower House)
There is a possibility that an earth and timber castle existed in the Castlemilk area, but how strong that possibility is, must remain a matter of interpretation and judgement. The suggestion arose with an article by Eric Talbot, a medieval archaeologist on the staff of Glasgow University in the early 1970s. Talbot's pioneering article reviewed the potential evidence for earth and timber castles in the west of Scotland (Talbot, 1974). He raised the possibility that the large circular earthwork on Cathkin Braes Golf Course might have been a 12th century earth and timber castle relating to Henry, son of Anselm who is recorded in a charter of Paisley Abbey as having granted land at Carmwmock to the Abbey.

An alternative point of view was put forward by T C Welsh in an unpublished letter to The Royal Commission, that the earthwork might in fact be the remains, not of an early medieval castle, but of the large prehistoric burial mound known as 'Queen Mary's Cairn' which the Rev Ure (History of Rutherglen, 1793) described when it was being destroyed in the late 18th century. This unpublished interpretation was recorded on an Ordnance Survey record card in the National Monuments Record as being 'not widely supported'. However it was given added weight when the ACFA survey of the Cathkin Braes Country Park in 1995 revealed, on an aerial photograph, a hitherto unknown secondary burial contiguous to the larger feature. (ACFA, 1995, Cathkin Braes Country Park). This re­interpretation removed the original reason for postulating an early castle at that site. However it is still within the bounds of possibility that an early castle may have existed elsewhere on the lands of Carmwmock. The early 15th century reference to the 'two vills' (or touns) of Carmwmock is perhaps best interpreted as referring to a settlement at Carmunnock village and at what was known from at least the early 15th century onwards as 'Castleton of Carmwmock', in its various spellings. (ACFA, 1995)

The appearance of the place name 'Castleton' is the first strong indication of the actual presence of a castle. The name can be interpreted as referring to the 'home farm' or 'demesne' attached to the castle to support it and it would normally be an indication of the most productive and desirable part of the landholding. This is where we would expect the proprietor to be resident. The farm of Castleton is some distance from the later House site and possible castle site, but it is not too far away to have functioned as the 'home farm'. The aerial photographs examined as part of the Cathkin survey did not reveal any soil features that would suggest earthworks or enclosures in the vicinity of Castleton.

If a pre-Tower House, earth and timber castle did exist at Castlemilk then there are two possibilities worth bearing in mind. The first is that some sort of defensive enclosure may have existed on the Castlemilk ridge, perhaps as some sort of ringwork, which would suit the topography. Local parallels can give some useful indications of the form such a hypothetical structure might have taken. The best preserved local example is at Crookston Castle, where the later stone 'tower sits within an earlier circular ditch, which would have been protected on the inside by a timber palisade with a substantial gate, enclosing timber buildings. These timber buildings would include living accommodation for the proprietor and his family. Some sort of 'hall house' would be appropriate, probably of one story, with separate areas for living, sleeping and cooking. Sometimes the cooking was done in a detached building because of the fire hazard. Occasionally, hall houses can be of stone, even at an early date e.g. Skipness Castle has a two story stone example dating to the 13th century. Other buildings would be required, such as cottages, stables, barns and sheds, which could be of quite flimsy construction, of timber, or stone and turf, with thatched roofs. There would also be tracks, drains and middens.

Actual evidence for a pre-Tower House phase on the ridge is not strong. The GUARD excavation of 1992 recorded' despite the high small find recovery rate from such a limited archaeological exercise, no closely dateable medieval deposits were located.' (Will and Leslie, 1992, pI6). The excavators concluded that demolition work in the 1960s would have destroyed any contextual evidence in and around the tower house. However the excavation did reveal a substantial ditch running north west to south east about ten metres from the remains of the tower house. This ditch did produce some late medieval pottery (not closely dateable) but 'due to lack of time, and also for reasons of safety, the ditch was not bottomed and the full extent of the feature was left undetermined' (op cit, p 13) It seemed to have been in use until the end of the 18th century or early 19th century. 'The discovery of a robbed out wall at the bottom of the ditch in trench 3 indicates an earlier phase of activity, details of which have been lost in the subsequent landscaping' (op cit p 16). So although occupation before the tower house cannot be definitely established, it cannot be ruled out either

The other possibility is that the mound close to the east end of Croftfoot Drive might be a motte, which it does resemble. This site is discussed in the text and it is recommended that any development in that area should involve specialist archaeological input.

The Tower House phase
The tower appears to have been erected in the mid or late 15th century for the Stewart family. Early engravings such as that in the Reverend Ure's History of Rutherglen and East Kilbride, 1793, and in the Glasgow Scrapbook show the tower before the remodelling of 1841. It was an typical rectangular tower of four stories with a stair emerging from a square cap house on to a corbelled parapet walk without angle turrets. As for the original arrangements, internally we could reasonably expect the ground floor to be vaulted in stone and used for storage, and perhaps for kitchens. The first floor would have housed the Great Hall, or principal apartment, perhaps with a small retiring room or closet and a private stair to the wine cellar. The floors above would house bed chambers. All floors would be accessed from a turnpike stair rising in one corner.

The tower house would not have stood alone. The occupants required support services, typically accommodation for servants, stables, barns, sheds and also gardens or orchards e.g. Crookston Castle in 1511 had 'edificia' (buildings), ortos (gardens), terras dominicales (demesne lands or 'home farm') as well as 'parcam et clausuram' (a park and enclosures). Sometimes the buildings around the tower were gathered inside a barmkin wall and should really be considered as the predecessors of the 'court of offices' of later centuries.

Post Tower Housel Designed landscape
After around 1600 there was a general movement away from the earnest defensiveness of the preceding century and a half. Houses of the gentry gradually became less military and more domestic. Curiously, there was in many cases a late flowering of a profusion of gunloops and battlements, but as consciously anachronistic features, as a statement of baronial pretensions. The 1600s also saw a growing interest in the recreational use of the immediate vicinity of the house, with the creation of parterres or flower gardens, knot gardens and the erection of fountains, sundials etc. After the Civil War there was a reaction against the militaristic trappings of preceding decades and a movement towards more horizontal, classically-inspired styles took hold and became the norm by the early eighteenth century.

As the century progressed, the ideal country house came to be seen increasingly as standing alone, 'like a Classical villa in an Italianate landscape'.(Buxbaum, 1989, pI7). This is what we see already in place on Richardson's map of 1795 and Forrest's map of 1816.

The walled garden was normally situated at some distance from the house, beyond the stables. This was the system recommended by John Reid, whose Scots Gard'ner (1683) became the gospel for Scottish gardening, running through several editions. It was convenient to keep the manure smells away from the house. This is what we see at Castlemilk, with the walled garden some distance to the south beyond two detached buildings in the vicinity of the present stable block which probably indicate the pre 19th century stables. Reid recommended a southerly aspect and a sheltered, sloping position to encourage drainage. The Castlemilk gardens are well sited, in an open yet sheltered position. The 1795 map shows the interior divided up into squares which again is very much in line with the prevailing practice; Reid recommended a broad walk around the interior parallel to the wall and went on 'it's also ordinary to make the Garden into four plots, by two walkes crossing from side to side' (p 22) which we see on the 1st Edition OS map. Forrest showed the garden conventionally divided up into squares

By 1795 the pond at Castlemilk had been created by damming the Croftfoot burn. The pond was an essential component of a designed landscape. As such, it fulfilled several functions ­
1 aesthetic and recreational, as an attractive and interesting feature in the landscape, to admire when on walks through the grounds and to enhance the view of the house, especially when approached via the drive from the east lodge.
2 sporting, as a source of fish and ducks. Reid recommended that ponds should be 'large and broad, such being best both for the health of Fish and Fowll, Clean, and most preferrable water for watering plants: squair, Triangle, Circle, Ovall, or what figure suits your ground best; let them be 5 or 6 foot of solid water at least, with Sluces to let it Run in and out at pleasure'.

The Avenue was an essential element which we can see already in place in 1816 and fully planted by 1864. Also by 1816 there were lodges on the east and west, presumably with gates and gate-keepers. The carriage drives and bridges seem to be in place by 1816. Ross Shearer, in his History of Rutherglen, p 71 refers to a 1794 illustration of 'Falstaff', based on a statue on the lawn at Castlemilk. The statue was between seven and eight feet high, and was probably made in two parts. It was well carved, showing Falstaff 'conventionally dressed in doublet and hose, with leathern girdle encircling his portly corporation, top-booted, and armed with shield and rapier'. After extensive enquiries, Shearer was unable to discover any information about its origins or date.

(f) The Victorian Country House, post 1841
The First Edition, Ordnance Survey (Figure 7) shows the House at its fullest extent after the alterations of c. 1841. Likewise, all the components of the 'stately home's park' are in evidence. The north block was added, the old tower was extensively re­modelled. An arched doorway was slapped through the gable and the window openings and the battlements and chimneys have been rebuilt.

One of the most interesting survivals from this phase are the remains of the Ice house. (ACFA, 1996). The ice would be collected from the ponds and carried to the Ice house which had double doors for insulation. These doors were normally made of two skins of planking, running vertically and horizontally, with a leather pad round the rim to ensure a tight fit. They would have been hung on heavy strap hinges. The outer door gave access to a brick-lined passage. This was the working area for loading and unloading. Only the damaged brick floor of this chamber and the bottom couple of courses of masonry, including the door jambs and threshold, still survive. The inner door led to the egg-shaped ice chamber, of which somewhat less than half survives. At the foot of the chamber there would have been an iron grating, above a sump leading through a drain pipe into the stream, passing through another grating at the outflow. The grating above the sump was missing and the outflow could not be located owing to disturbance of the banks of the stream.

When in use, the bottom of the ice chamber would shave been packed with straw for two or three feet, to allow drainage. Above the straw blocks of ice were laid, alternating with layers of straw or sawdust. The chamber would be filled to just below the middle. The contents would be covered with a top layer of straw or rushes, dished slightly towards the centre. Fish could be laid on this top layer.

The walled garden The 1st OS map shows the garden within its wall with two entrances on the north towards the comers and a third entrance in the middle of the south side. Lady Helen (op cit p36) said that the garden had three ornamental iron gates. There is a glasshouse on the middle of the north wall. Trees are shown within the garden, probably for fruit. The garden has a path round the inside of the perimeter on all four sides with paths running north-south and east-west to divide the interior into four squares. By Lady Helen's time, the garden had a large central plot of grass which was subdivided into intricate flower beds which displayed the family emblems of the Buckle Crescent and the Star.

The garden was set within the south east comer of a larger enclosure which was divided into plots, with greenhouses. Outside the garden on its north east comer, beside the approach road from the stable block, here is a small building in the position later occupied by the laundry maids' cottage.

Overall conclusion Although the survey area was small, it contains a great deal of important survivals from the Designed Landscape phase and certain areas, such as the ridge on which the house stood, and the area east of Croftfoot Drive, should be considered as archaeologically sensitive. Topen, D., 'An archaeological survey of Castlemilk Glen, City of Glasgow', ACFA (1997)