Site Report

WoSASPIN 7105

(NS 3692 9030) Castle (NR) (Ruins of)
OS 6" map (1923)

The castle on Inchgalbraith is ascribed to the period between 1542 and 1700. The only remains are an ivy-covered tower.
D MacGibbon and T Ross 1892

The remains of the castle consist of the freestone masonry walls of a rectangular tower between 1.5 and 7.0m in height, and measuring 10.5m east to west by 12.7m north to south. The island is covered by quantities of building material.
Visited by OS (IA) 22 March 1973

NS 369 903 The fragmentary remains of this castle stand on an island in Loch Lomond. They are possibly of 16th-century date.
RCAHMS 1978
D MacGibbon and T Ross 1892

This small island measuring no more than c.25m in diameter is roughly circular tapering to a point at the N end though the shape will no doubt change when the loch is lower.
The island is considered to be a crannog and while it is clear it is constructed of stones it is not possible to determine if these are glacially deposited or introduced by human activity, the whole of the island is now covered with the collapsed stones from the castle itself. The island will be further investigated during the proposed Crannog Survey.
The island and castle belonged to the Galbraiths of Bandry. The Galbraiths once owned Bannachra Castle in Glen Fruin as well as Inchconnachan and Inchlonaig prior to the 16th century. The island is called Yl na Castel on the first map of Loch Lomond drawn up by Pont in the 17th century.
The castle takes up most of the space on the island and there is no room for any further structures. MacGibbon and Ross (1892) considered this castle to be late in date and of the period 1542 to 1700. However, the prominence of the Galbraiths of Bandry in the earlier records of the Lennox suggests an earlier date should perhaps be favoured. The castle is described as ruined and overgrown in the early 18th century by Walter MacFarlane.
The castle is a square tower constructed of quarried sandstone boulders and blocks and it retains a fine entrance and portcullis in its E wall. The S and W walls, which are exposed to the prevailing weather, are very denuded and stand to no higher than c.1.5m. The N and E walls are preserved to the base of the second storey. Occasional schist/phylitte and conglomerate sandstone was also seen in the tumble.
It is interesting to note that there are no firing slots visible in the tower's remains.
Restoration work has been undertaken at the castle in the past, perhaps in the late 19th century when such projects were a favoured past to time. Gravelly concrete has been used to consolidate the castle in a number of places and the window / doorway in the N wall has been entirely reconstructed by the early conservationists.
The castle is the only feature on the island. Although the island is probably a crannog nothing of its construction could be determined due to tumbled masonry covering any earlier constructional details.
In the centre of the N wall underneath the first floor door/window is a blocked arch with a triangular keystone, probably a postern gate. Immediately to the W of this four stones project from the wall and the broken stub of a fifth projecting stone is also apparent. These stones may be associated with an external stair which led to the first floor opening or some other external addition to the tower. It appears as though the area around these projecting stones may also have been blocked or altered at some time although this may reflect an original constructional detail.
Traces of original harling may be visible above first floor level.
Above the first floor door or window opening it is clear that some alteration has taken place and it is possible a chimney was present, perhaps added at a later date. Although the wall is badly damaged and cracked above this opening the wall does appear to show signs of alteration.
The N wall survives to the greatest height and survives in part to second floor level.
At the W end of the N wall of the tower is a 2.1m stretch of wall that projects from the tower. This wall has a rounded butt end and may be associated with a boat landing on the W side of the tower. The relationship of this wall to the tower is a little obscure due to fallen masonry and reconstruction works and it is not possible to determine whether it is abutting the tower or integrated into it. Some clearance work would be required to determine the relationship with certainty.
The W wall is overgrown and denuded and stands no more than c.1.5m high. At the N end of the W wall is the projecting wall described above which runs to the west. At the S end of the W wall is another wall projecting from the main tower for 2.4m. This wall at the SW corner is only 0.7m wide. It has been restored in the past but is now eroding away. The face of the W wall is visible in places but in the main is obscured by tumble and vegetation.
In the centre of the W wall is the possible stub of a wall that projected to the W. This wall may be no more than a linear arrangement of collapsed masonry but its linear nature suggests it may in fact be a collapsed wall. If so, it would form a small enclosed bay or landing place at the N end of the W wall in conjunction with the westward extension of the N wall. There is no trace of a wall, on the island, to the W and thereby no evidence to suggest an ancillary building may have stood here.
The S wall does not survive to any higher than c.1.5m. Two entrances are visible at ground floor level leading directly onto the shore and both must be considered as postern gates. It is interesting to note that these doors are located on the S side into the prevailing weather which suggests the N side of the castle may have been reserved for formal use and the S side was the area of domestic activity. The westernmost of these gates is hardly visible through the tumble and vegetation and only the E jamb of sandstone ashlar can be clearly discerned. The SW corner of the castle is very overgrown and damaged.
The easternmost of these gates retains several of its original details. Only the W jamb survives intact and it is constructed of sandstone ashlar blocks. The doorway tapers towards the interior where the sill and interior jamb survives. The holes for an iron yett are clearly visible in this interior jamb. The E jamb is missing and fallen masonry blocks the doorway.
The external width of this doorway is 1.20m and the internal width is 0.7m. The wall is 1.10m thick at the doorway and the S wall appears to be 1.05m thick on average.
The E wall is in very bad condition and the SE corner has completely collapsed taking with it the S part of the fine portcullis located at the S end of the E wall. The N side of this portcullis survives intact and is a very fine example of this type of entrance gate. The wall has been penetrated by tree roots from an established tree in the first floor window and a very serious crack is present. The wall is bowed and bulging and in danger of collapse. At least two doors were present. The outer door consisted of a gate that was lowered from second floor level. The slot in the wall where the gate would have been held within the wall is clearly visible. Between the outer portcullis gate and the interior gate the springers of an arch are visible in both the rubble core of the eroded wall and also in the remaining dressed sandstone. The jamb of the interior gate survives and a drawbar slot is also present. It is possible there may have been yet another gate as small holes suggesting some kind of fitting are present in the jamb to the interior side of the drawbar slot.
The dressed sandstone facing stones on the external side of the portcullis are chamfered c.1m above ground level and it appears that there was some function to the design of the external face of the doorway. Also of note is a vertical slot cut into the outer edge of the portcullis slot in the third block up from ground level, the function of which is not clear.
Three steps lead up into the castle but these are not original, or at least not in their original position. The two lower steps are fine dressed sandstone slabs and the third step is of sandstone blocks bonded in concrete.
At the NE angle three sets of initials have been carved onto one of the quoins.
The E wall has been repointed in the past.
The interior of the castle is overgrown and it was not possible to determine any internal features in plan. However, there is a noticeable ridge in the ground surface with the N half of the castle being some 0.3m or so higher than the S side. Although the interior is filled with collapsed masonry it is possible this ridge represents an internal division and perhaps the presence of a vault on the N side.
The N wall is dominated by a large vertical opening representing either a window or a door at first floor level. This opening has been heavily restored in the past with concrete clearly visible and it is also apparent that the stones have not been restored in their original position. This arched opening has projecting jambs into the interior. Any mouldings have been removed and the lack of facing stones on either side of the opening suggests they were gouged out.
The facing stones from the interior of the wall have been removed and the wall core is exposed.
Above the first floor opening the W jamb of another opening can be seen at second floor level, perhaps for a fireplace. The castle stands to its highest point here. The N wall is 1.5m thick.
Nothing could be determined about this wall from the interior due to the thick vegetation.
The easternmost doorway can be seen from the interior but the wall is so overgrown that very little could be determined.
The SE corner, S of the portcullis has completely collapsed. The N part of the wall survives to second floor height.
Towards the N end of the wall is a large vertical opening extending from first to second floor level. Very little of the original jambs survive and it is clear that the mouldings have been robbed. The jambs are angled tapering from an internal opening of 1.65m to the exterior opening of 0.9m. The E wall is 1.60m thick.
FIRAT 1995; NMRS MS 993/2

The island is part of a glacially deposited mound. The castle on the mound is the only evidence of artificial construction and the mound itself does not appear to have been artificially raised which is not surprising as bedrock is exposed above the water level adjacent to the castle. In the shallows around the castle it appears that the natural stones have been cleared to the underlying gravel to create placesas where boats could be pulled up. The site was considered a crannog by the Islay Archaeological Survey Group.
STUA and FIRAT 1997; NMRS MS 993/5, no.13

Photograph on OS Cards
Lumsden correspondance with Harvey Brown. Further information from John Mitchell, Nature Conservancy Council - see NMRS letters:
'DUNBARTONSHIRE: INCHGALBRAITH CASTLE
Thank you for your informative letter of 30th October. As you surmised, I am gathering material on the former osprey site on the castle ruins. I was unaware of Gilpin's description, and will follow this one up. The earliest possible date for the castle of 1542 is interesting, for when Pont surveyed the area c1590s he described the Inchgalbraith castle as 'old' and the island as overgrown (ref: Macfarlanes geographical collections Vol II, Mitchell 1907). It would appear therefore that occupation of the building was very brief, and even raises the question of whether or not it was ever completed. I have a copy of the brief description by Fraser (1869) :- 'The north wall or tower remain, covered with ivy and surrounded by a thicket of trees. A portcullis of falling gate-door hung over the gate for greater security. The groove in which the portcullis was raised and let down is still to be seen.'

The third and final phase of fieldwork of the Loch Lomond Islands Survey project took place in September 1997. The following is among sites searched where no crannogs were located:.
NS 369 903 Inchgalbraith.
Sponsors: Historic Scotland, Friends of Loch Lomond, Dunbartonshire Enterprise, Loch Lomond Park Authority/SNH.
F Baker and Dixon 1998
Report date for above text from NMRS 21/10/2003

The island was investigated during the FOLL 1997 crannog survey and found to be a natural glacial moraine and not a crannog. The exceptionally low loch level (c.7m) during the 2007 survey indicated the island to be more oval in shape with a tapering tail at the North end. The low loch level also revealed wall lines amongst the tumbled masonry on the South and West sides of the island and amongst the tumbled masonry a considerable number of dressed architectural fragments were observed. The low loch level exposing more of the island and the expert eye of Tom Addyman revealed that the building we see on Inchgalbraith is actually only the central core, probably enclosing a courtyard, of what was a much larger castle with ranges on the North, South and West sides. The re-appraisal of this castle leads to a complete revision of the interpretation given in 1995 and the interpretations given in the 2007 survey notes below supersede the 1995 interpretation given above. The 2007 survey also confirmed the original survey’s suspicions that Inchgalbraith is an early castle and construction details were dated to the 13th century by Tom Addyman.

No structural elements were observed on the North end of the island although tusking in the North wall indicates there were structural elements on the North side but silting up over this point may have obscured structural remains and excavation to the North of the standing building may reveal archaeological features here.

Some ivy roots have been cut at the castle but the ivy has not been removed from the walls. No real change to the interior of the castle. The window in the North wall has been reset and the lintel reset in the top of the window is bowed and twisted. However the window seat itself appears to be an original medieval feature. The mouldings of the window have been removed and cement repointing introduced at the exterior opening. The NE corner of the window has also been rebuilt / repointed with cement at the base.

The low loch level exposed structural elements on the South and West sides of the castle indicating that Inchgalbraith was in fact a much larger, older and more complex castle than originally thought. The surviving tower is in fact the interior of the tower and was probably an interior courtyard or open space and the ‘doors’ previously described in the South wall are in fact windows looking on to this interior courtyard. It seems likely that the original castle was in fact larger than the surviving island on which its remaining element survives. The original castle probably consisted of three ranges of buildings around a courtyard.

This revelation that Inchgalbraith was considerably larger than it is today indicates the severe problem of erosion over centuries.

The line of two E-W aligned walls could be traced to the South of the standing tower and fragments of dressed stone including quoins, mouldings and architectural fragments including some with medieval tooling marks were identified in the collapse. One E-W stone alignment located 1.6m S of the projecting N-S wall at the SW corner indicates a c. 4m long NE-SW wall line, which may be the remains of a wall or the edge of a building platform. On the West side of the standing tower a distinct linear stone setting aligned N-S and measuring 4.8m and 1.5m wide and 0.40m high was found. This stone setting of apparently vertically set stones may represent the collapse of a wall, i.e. four or five courses of a collapsed wall lying on its side.

The sandstone slab foundations observed towards the West end of the North wall were identified as a 13th century building technique by Tom Addyman. This contradicts the late 1542-1700 date given to the castle by MacGibbon and Ross but is more in keeping with the presence of the portcullis gate and the documentary evidence referring to Yl na Castel dating back to 1574. Further documentary research may shed further light on its apparently early date although information on the Galbraith clan’s activities in the area was scarce during the initial 1995-98 research although the desk based research undertaken at that time did not fully explore the documentary sources.

The exceptionally high loch levels (c.10.2m) of December 2006 had come up to about 1m high on the walls indicated by rubbish lines along a high water mark. Mortar has been washed out of the base of all walls at the foundation course level by the high water levels.

East Wall
The East wall of the tower is, remarkably, still standing but the major structural crack in this wall has increased in size and is up to 70mm wide. Note that this wall was repointed in Victorian times and this visible movement of the wall has occurred in the last 100years. This bowed wall could collapse at any time and its recording and repair is urgent. There is approximately a 0.40m overhang caused by the bowing of this wall which is caused by erosion and softening of the foundations which is causing rotating masonry. The re-set steps at the portcullis entrance have been eroded and undercut by the high water levels. The North side of the portcullis, the better preserved side, shows further erosion of the mortar at the abutting interior E-W wall undermining the abutment and leading to free standing masonry. The ivy on the East wall has increased considerably and the ash tree in the window has matured increasing the disturbance it would cause if it fell. The East wall appears to be the only wall that appears as it would have done originally with the ranges attached to the South, West and North walls. Flood defences would be required to protect this wall along with underpinning of the wall in order to preserve it.

South Wall
Further water erosion has taken place and is still active particularly along the most exposed South wall of the standing tower. Further masonry collapse has occurred at the SW corner. In the East window of the South wall increased water erosion is apparent at the sill and under the W jamb. The East jamb of the East window is in better condition but the mortar and stones are being pulled apart by ivy roots. New graffiti has been carved on the West side of the East window.

The West window in the South wall and the projecting wall of the range that was originally attached to the South side of the standing remains is completely overgrown and is riddled through with roots, some of the trees here are dead and when they collapse will disrupt the masonry and cause further collapse.
The East end of the South wall / SE corner of the standing tower has suffered severe erosion and recent collapse is evident, the rubble wall core is exposed and the impressions in the mortar indicate that the facing stones have only recently become dislodged, presumably during the December 2006 high water levels. A corner stone at the SE corner is loose and dislodged and needs immediate repointing to save it from collapse. A rowan tree growing out of the SE corner is further pulling the masonry apart and should be removed immediately.

North Wall
Generally in similar condition to 1995 but there is erosion of mortar at the base of the wall / foundation level, particularly at the sandstone slab foundation at the West end of the North wall. The ivy roots have penetrated under the window opening and into the East end of the wall. The saplings at the base of the wall have increased in size and their roots will penetrate into the foundation courses as they increase in size. The chimney head appears more unstable than in 1995 – 97 but the cracks in the chimney are concealed under ivy growth making the stability hard to assess. The North wall is bowed. The foundations at the NE corner are in relatively good condition. Trees growing out of the masonry at the West end of the North wall are pulling the masonry apart. Tusking projecting from the North wall indicates a wall joining the North side of the castle of which no other trace is visible. The large sandstone slab foundations are indicative of an early 13th – 14th century build date.

West Wall
The collapsed West wall is badly overgrown and the trees, saplings and ivy roots are causing damage to the surviving mortar and masonry and further damage by tree roots has been caused since the original survey. Dead trees are present which will pull the masonry apart even further when they fall. There is evidence of further collapse of the wall. The SW corner of the West wall where there is a projecting wall to the South and the stub of a probable wall projecting to the West has suffered recent water erosion and collapse and newly exposed rubble core is present. The foundation footings are visible on the West side of this wall but the East side has been eroded to the rubble core. A possible sub circular stone setting and further N-S alignments of stones were observed towards the South end of the West wall. It is unclear if these stones represent original features and wall lines or if they are later attempts at consolidation and protection of the castle from erosion.

Inchgalbraith clearly indicates the problem of erosion. The further investigation of the site can only be carried out at very low loch levels of 7m or lower or the archaeological remains observed during the 2007 survey will not be visible. Excavation is recommended to reveal as much of the surviving ground plan of the castle as possible. Considering the extent of the erosion in order to prevent or at least slow down further erosion the construction of breakwaters to deflect the water action, which comes with the prevailing wind from the SW, is recommended. Specialist engineering advice will be required to implement this. An article on breakwater protection of a crannog site is included in this report for an indication of the type of works involved. However, the castle is already a shadow of its former self and given the problems of erosion the question must be asked as to whether it is now reached the state where the most viable option is that it is recorded and excavated as far as is safely possible and then left to decay. A full structural survey is required to determine if the castle can in fact be saved and the overall costs of underpinning, repair and repointing works and flood defences.
The trees that surround the island should remain as they provide some protection from wind and rain and break the impact of the waves. An interpretation sign with information on the castle should be provided, advising visitors not to remove stones from the building or shore, to remove rubbish, not to disturb nesting birds and that the site is dangerous.

The following stages of work are recommended. Scheduled Monument Consent would not be required as the castle is only B Listed.:

Stage 1.
Record the remains in detail. The primary purpose of a drawn survey at Inchgalbraith is to provide a series of drawings that will serve as a basis for architectural repair works. However, in light of the importance of the site and the immediate risks it faces some effort will be made to complete a general record for posterity.

Given the extent of vegetation, principally ivy that obscures the standing masonry, it is intended that the record drawings be revised when better access (perhaps during any future repairs) permits.

General ground plan
The castle remains and the limited area of surrounding island should be comprehensively planned. This would be best achieved at the lowest water level possible. As well as a systematic survey of the castle wall remains, particular emphasis will be given to mapping evidence for footings and alignments on the foreshore and the positions of individual ex situ dressings of which there are a considerable number at the site.

The plan will include topographic contouring and the careful marking-up of areas of active erosion, the major individual trees and the extent of the in situ turf cover.

Standing building survey
Standing building survey should include an element of prior vegetation removal where this is practicable and safe - simply to maximise the area of masonry that can be recorded. It is envisioned that outline drawings be made of the 4 exterior (loch-facing) sides of the ruin, with interior elevation drawings of the N and E walls. Important individual details, such as the entrance arrangement, details of moulded stones, etc. will be drawn as appropriate.

The elevation drawings will record the outline of the walls and individual analytically significant details such as phasing breaks, stone dressings, etc. In the absence of better access, particularly to upper wall areas, the drawings will be indicative to an extent (any such limitations to be noted on the individual drawings)

Historical Research
On archaeological grounds the castle is now considered to be c. 300 years older than previous interpretations had indicated. However, no readily available historical information on the castle and its association with the Galbraiths was located during the initial survey. Specialist historical research including primary records is recommended so that a greater understanding of this castle, and its relationship to Ellan Rossdhu and Inchmurrin, with which it is contemporary, can be developed.

Archaeological Evaluation Excavation
The south and west ranges of the castle have been denuded to the extent that they are only visible as tantalising and faint wall lines in amongst tumbled stone. The standing building represents only a small element, perhaps not even half of the original extent of the castle. Excavation would seek to reveal as much of the original plan of the castle as possible and investigate the survival of occupation related deposits. Also to determine if occupation deposits survive how well protected they are by the tumbled stones and if they are suffering active erosion. The collapse deposits would require scale planning before removal and many architectural fragments are present which may relate to collapsed elements of the building. A full project design for excavation would need to be produced before any work commenced. The SW corner of the castle is likely to be the most suitable area for an evaluation excavation as it may remain above loch level for longer but any excavation work should take place at the lowest loch levels. Analyses of the loch levels over preceding years should be undertaken to pinpoint the most suitable time to carry out excavations.

Stage 2.
Consult an engineer about methods of resisting further erosion and loss of fabric. If there is no feasible or affordable method to halt, or at least significantly reduce the erosion from the loch, then a very detailed record is probably as far as this project should be taken. Some funding bodies might be prepared to give money to the recording process but funding bodies would not contribute to consolidation for a monument which is under threat.

Stage 3.
If an appropriate method of preventing damage from the loch waters can be designed, make
architect’s drawings and engineer’s drawings. Quantity Surveyor to carry out a cost report.
Apply for permissions and grants.

Stage 4.
On receiving consent and notification of grant, tender for contractors to scaffold the ruin and complete its consolidation. Standing building recording to be completed from the scaffolding.
Baker, F., FIRAT Archaeological Services Ltd, 'Loch Lomond Islands Survey 2007', (2008)