quadrilateral castle is sited on the S bank of the River Lochy at the
strategically important S entrance to the Great Glen - the main passage
through the Scottish Highlands. The castle seems architecturally
related to Castle Sween to the W, Coull to the NE and Bothwell and Dirleton to the SE.
Inverlochy castle is said on Wikipedia
and elsewhere to have been built circa 1270-1280 by John ‘the
Black' Comyn, lord of Badenoch and Lochaber (d.1303). It is even
claimed that he died there in 1300! However this statement
appears to be in error and no near primary source shows his death at
anywhere other than Lochindorb castle, not that its easy to see how Lowchyndorb
can be mistaken for Inverlochy. Lochindorb, incidently, is like
Inverlochy in being a quadrilateral fortress with a round tower at each
The early history of Inverlochy castle is certainly obscure, but more
can be said of it than could ever be gleaned from Wikipedia. The
John Comyn who died in 1303 was the father of the John ‘the Red'
Comyn murdered by Robert Bruce in Dumfries cathedral in 1306.
With this act it seems likely that Inverlochy castle passed to the
lords of the Isles in the form of Alexander of Lorne, who was married
to a daughter of the elder John Comyn (d.1303). What we do know
for certain is that Alexander fought a battle against 2 great galleys
in the loch near to the castle of John Comyn in Lochaber probably in
early June 1297. This was after he had been asked to arrest
malcontents in the area by King Edward I of England
in March. It would seem likely that the castle passed into
Alexander's hands soon after this. He had been married to John's
third of four daughters since before 1290.
With the death of King Edward I in July 1307, Robert Bruce marched from Dumbarton, through the Great Glen, supposedly taking the castles of Inverlochy by negotiation, Urquhart
through it having an insufficient garrison and finally Inverness.
The culmination of this was the forcing of Earl John Comyn of Buchan to
a truce. This John was second cousin to the John murdered by
Bruce - so the idea that he was holding Inverlochy castle in 1307 is
unlikely. In any case, after 2 battles, both Earl John and his
brother Alexander were dead by December 1308 and possibly Inverlochy
castle was ruined as a part of Bruce's policy of not allowing
fortresses in his lands that could be used against him. This of
course does not prove who was holding the castle and it is possible
that the castle had passed into the hands of the lords of the Isles
before that as is hinted in the later correspondence between Alexander
MacDougall, the lord of the Isles, and King Edward II (1307-27).
While defending Loch Linnhe in the spring of 1309, Alexander stated
that he had three castles to defend the inland sea. Presumably
these wereDunstaffnage, Inverlochy and Coeffin. Dunstaffnage castle
was subsequently besieged by Bruce and fell before 16 June 1309.
It is possible that Inverlochy fell at the same time, although it is
also possible that the agreement reached with the garrison when Bruce
campaigned in the Great Glen had earlier led to its ruination.
Regardless of the castle's fate, the lords of the Isles were
subsequently to fight for the Great Glen again and again in the
following hundred years.
Whoever had lost the fortress to Bruce, the defeat of King Edward II
and the surviving Comyns at the battle of Bannockburn in 1314 probably
sealed the castle's fate. Considering Bruce's well-known policy
of slighting fortresses, it could have been only after his death that
the castle was re-occupied, but certainly nothing is heard of it, while
Urquhart, just 40 miles up the Glen, was
mentioned again and again in the 14th century. It is possible
that the castle was re-occupied by the MacDonalds in the 1350s when the
great-grandson of Alexander reacquired the district. This Lord
John of the Isles died between 1371 and 1377, without a legitimate
son. Despite this, many of his lands passed with the lordship of
Lorne to his son-in-law, John Stewart of Innermeath (d.1421).
However it would seem likely that Lochaber, presumably with Inverlochy
castle, stayed with his distant cousins, the MacDonald's of Islay by
the grant of Robert Bruce to Angus Og. Certainly Angus' son, John
(bef.1315-86), made grants within Lochaber lordship to his son,
Reginald, that show that the barony and therefore the castle of
Inverlochy must have belonged to him. Inverlochy castle should
have been a useful base from which John's son, Lord Donald MacDonald of
the Isles, launched his assaults on Urquhart and Inverness in the late
1390s as are described under Urquhart castle.
On 25 September 1394, Earl Thomas Dunbar of Moray entered into an
agreement with ‘Alexander of the Isles, lord of Lochaber'.
Presumably this was Alexander MacDonald, the younger brother of Donald
MacDonald of the Isles (d.1423). Inverlochy would have been a
good starting point for their subsequent seizure of Urquhart castle in
Inverlochy may have been occupied in 1431, when clansmen of Alexander
MacDonald, lord of the Isles (1423-49), defeated King James I's larger
army under the earl of Mar in the first battle of Inverlochy. The
earl had taken up position in the hills above Inverlochy when he was
attacked by land and sea (presumably from Loch Linnhe) and
routed. The castle was not mentioned in the fighting.
Over 70 years later in 1505, Inverlochy castle was granted to Alexander
Gordon, 3rd earl of Huntly, who was charged by King James IV with
repairing it for use as a royal garrison to help control the
Highlands. It is possible that the new battlements and N
gatehouse were installed at this time. In 1509 Inverlochy was
designated as the location for holding royal courts in Lochaber and
around he same time, Alexander's brother, William Gordon, the laird of
Gight, became master of Inverlochy. He was slain commanding the
Camerons at Flodden in 1513.
On February 2, 1645, the second battle of Inverlochy was fought, when
the royalist army of James Graham, 1st marquess of Montrose, defeated
an army from the castle consisting of the Campbells under the earl of
Argyll. In the aftermath of the battle, 1,300 men were said to
have been taken out of the castle and massacred. To put the claim
somewhat in context, when a grave pit was opened in Durham of the 1,700
Scots said to have perished there after the battle of Dunbar, no more
than 28 bodies were found. This of course does not rule out later
grave pits, but garrulous stories of massacre should always be treated
with caution. That said, renovations to the W curtain, which
included the building of battlements in the late 19th C, found a
complete skeleton sealed in the N end of the wallwalk. Perhaps
this was a victim of the action. Canon balls were also said to
have been found embedded in the castle walls. In 1654 Inverlochy
castle was abandoned in favour of a new fortification further down the
River Lochy at its mouth where it joins Loch Linnhe. This, in
1690, became the modern Fort William.
With the N side facing the river, the other three sides of the fortress
were protected by a water-filled ditch and unimposing ‘modern'
outer ward with 2 circular turrets. Therefore Inverlochy castle
primarily consisted of a rectangular fort with a substantial, but low
curtain wall and a round tower at the 4 angles. Within the outer
‘defences' was a ditch, separated from the castle by a berm up to
40' across. The main ward is some 102' by 89' across, surrounded
by a wall up to 9' thick and 25' high. There is plentiful
evidence that the wall has been raised 6' in antiquity. The
largest tower to the NW is now known as the Comyn tower and has an
internal diameter of 20' and a wall thickness of 10'. It also has
a third storey the other towers lack. Surprisingly the other
towers are all of differing diameters; the NE tower being 15', SE 13'
and SW 12'. No logical reason for these differences is apparent.
Unfortunately much of the castle facing has been renewed in modern
times and battlements added to the W curtain. Sketches made
before this time suggest that there was an external batter to the walls
and at least a projecting string course capping the most impressive
plinth - that on the donjon. The layout of the smaller towers
appears similar. All 4 towers were entered at internal ground
floor level via the gorge. Each had curving mural stairs leading
up to the battlements and upper floors. This is an unusual
arrangement and is more reminiscent of 12th than 13th century
castles - viz the keeps at Conisbrough, Arundel, Bronllys etc.
The ground floors of the S towers consisted of small circular chambers
equipped with what appear to have been fish tailed loops of a type
common from 1150 to 1250. The upper storeys of the 3 towers all
seem to have had 2 loops set within embrasures on the side where the
mural stairs were not. On the interior sides the S two had loops
covering the courtyard and it is to be presumed that the NE tower had
one too, although this level has been largely destroyed. The NE
tower is also the only one with a windowless basement entered at ground
floor level, the staircase being at 1st floor level - the doorway at
this height obviously having been entered from the destroyed E
range. Excavation has proved the tower had no internal ground
floor other than the natural clay and patches of spilled mortar.
Human bones from at least 4 bodies were found in the soil used to pack
the foundations. This may suggest that part of the soil from the
ditch had earlier been used for burials and was then unknowingly used
to help level the castle site before masonry construction began.
Excavation also found the interior of the SE tower had a similar
composition, immediately overlain by modern debris. Taken
together this evidence shows that Inverlochy was built as a masonry
castle and had no wooden phase.
All the tower stairs were equipped with narrow external lights and the
upper rooms with fireplaces. The wallwalks used different methods
to pass through or around the towers. The 2 E towers had the
wallwalk pass through the wall in mural passages, while the SW tower,
which was the smallest, had actually to be entered, there apparently
not being sufficient room for a mural passage. The keep was
actually passed around to the S by the wallwalk which then appears to
have ended in a narrow passageway within the N curtain. The main
wallwalk continued from the upper storey of the keep, via steps down
the raised curtain wall. This is a most peculiar
arrangement. Of the battlements on the towers, which appear to
have been reached via another set of 3 curving mural stairs, set in the
thickness of the tower walls opposite the exit from the stairs up from
the ground floor, nothing remains but a part of a solitary merlon on
the S side of the SE tower. These towers bear great similarity in
style to those built by Earl Hubert Burgh in the Trilateral castles
in Gwent between 1219 and 1239, except that his generally did not have
loops on the ground floor. In plan though, the castle perhaps
most resembles Flint castle in Wales - a fortress the Comyns may have
seen in the invasions of Wales in the 1270s and 1280s.
the 3 towers of the castle are peculiar, the keep is even more
aberrant. Firstly it is entered on the ground floor and access is
gained from there to the next floor. In medieval defensive
philosophy this simply was not done for obvious reasons. The
wallwalk entry to the tower was gained via a narrow doorway protected
by a deep drawbar slot. Next to this was a passageway which
currently leads onto the W wallwalk, but originally was a mural
passageway that led to a garderobe overhanging the curtain.
Another smaller latrine seems to have been against the NE tower at this
level using a similar layout, but the walls here are much
damaged. The interior of the keep is awkward at this level.
The round basement interior rises to a semi-circular room with five
straight sides to the N. The central segment of this contains a
round backed fireplace, while those in the other towers are square
backed. Three embrasures contain modern loops, while a small loop
lights the top of the stairs up to the W. There is no access to
the upper storey via a mural stair. Instead access was gained via
the N wallwalk, though it is uncertain how access to this wallwalk was
otherwise gained. The upper chamber of the keep is most ruined,
but once again there were three external embrasures and one facing the
courtyard. There were also 2 mural passageways to the W.
Excavation found a formally buried adult skeleton just 20' from the
keep door. This was probably a member of the last garrison of the
castle. The grave was left untouched, just 2' beneath current
The courtyard was entered to N & S by simple hole-in-the-wall style
gateways. The S gate was the main entrance to the castle being
about 9' wide. It was protected by a portcullis and locked via a
drawbar. The portcullis may have been further protected by a
barbican of which excavation found one side to have been 6' thick and
tied into the curtain wall, suggesting that this was a contemporary
design or that this part of the castle had been totally rebuilt when
the gateway was altered. Whether this was a gatetower or simply
two walls from which a drawbridge was operated over the moat is
unknown. The layout within the gate suggests that this part of
the castle has been much rebuilt. The few surviving gateway jambs
are made of a good quality yellowish sandstone and have the rounded
portcullis groove cut into them. Rounded grooves are supposed to
be 14th century and
later. The portcullis was raised from a chamber set partially
within the curtain and later internal gatetower. This room was
reached via steps down from the wallwalk. Presumably this
rebuilding is of 14th or 15th century date. The segmental arch holding up the upper section of curtain is modern.
The N gate in comparison is only 8' wide, and although it too had a
portcullis, this was of a totally different design. The
fragmentary remains seem to show that it was rectangular - a design of
the 12th-14th century.
Like the S gate this would appear to have been operated from a chamber
above, but this one was wholly set within the curtain, without the
later addition of a projecting internal gatehouse behind. Parts
of the gate arch remain too, which show that this was pointed.
Possibly the S gate was originally similar to this. The N gate
also had an external barbican added, but in this case, contiguous to
the walls. Further this could not have had a drawbridge as there
was no ditch before the river. At a later date a building has
been added that passed right across the rear of the entrance as can be
seen by the floor line.
Paul Martin Remfry