and upper floor.
The Tempul Clogas or Belfry Church on Iniscloran has
the same arrangement. It is a Romanesque church much
altered and the tower is considered to be an addition, and
to be not earlier than the I2th century or even to be post-
Norman. 1 In the three Mayo churches I take the tower
to be a part of the original plan. They all show Gothic
work and I take Attyrickard to be the earliest, and Kinlough
to be probably later than Illaunnaglashy. The ground
plans of the western parts of these two differ
Illaunnaglashy presents the very unusual feature of two
very small narrow windows in S. wall, one similar window
in the W. wall, one similar window in the N. wall near
the W. wall, and a ruinous opening, of either a door or a
window in the N. wall of the ground floor room, which
was cut off from the church by a cross wall which reduced
the length of the church to the E. to 29' 6". But I am not
quite sure that this cross wall is original. The church walls
to E. of it are only about 3 ft. high and covered with
1 R. S.A.I., 1900, xxx. pp. 81, 168, 257.
166 DIOCESE OF TUAM
rubbish. Nor am I sure that the gap in the N. wall is a
doorway and not a broken-down window.
Above the joist holes in the S. wall are the openings
of two windows, of which the western is over the western
window of the ground floor and the eastern more to the
E. than the eastern ground floor window. The heads of
those windows are gone, but what remains assures me that
they were lancet windows, and certainly much wider and
higher than the lower windows which were flat headed 26" x 4".
The western window in the upper room is arched and is not
over that of the lower room but more to the N.
In the absence of the eastern part of the walls it is im-
possible to make out the arrangements with certainty but
the church is remarkably narrow for its length, suggesting
that a very early church has been lengthened to the W.
The upper windows are very large for such an upper room,
and their distance apart suggests that the cross wall did not
run up so high ; the whole appearance suggests that they
were church windows. Yet the only thing to explain such
an arrangement is the very early St. Columcille's House at
Kells. 1 It measures 19' x 15' 5" inside, and had three stories.
The first floor was a chapel to which there was access by a
door 8 feet from the ground. It had a barrel vault and
a loft above under a pointed stone roof. Under the chapel
was a crypt, without door or window, accessible by a hole
in the chapel floor.
This does look as if the upper room might have been
the church with a crypt under it.
The Ullard church had a crypt under the chancel, lighted
by a narrow slit. 2
These three churches measured inside
Attyrickard, 40' x 19' church, 16' x 19' tower.
Illaunnaglashy, 58' 6" x 14' 6".
Kinlough, 65' x 22' 4".
1 Dunraven, Notes on Irish Architecture, ii. p. 50.
3 Ibid., ii. pp. 86, 87.
CHURCH ARCHITECTURE 167
These churches certainly belong to the Gothic period,
but are treated here on account of their towers.
A small square tower is sometimes attached to the side
of the church as in Cormac's Chapel at Cashel and at Mungret,
which seem to have grown out of the round tower springing
from a square base. 1
The church of Inishmaine has a square building on each
side. The larger on the N. side has good windows and
is plainly intended for ceremonial or domestic uses. That
of the S. side at junction of nave and chancel has no
opening on the ground floor. It seems to be the butt cf a
The churches of the 4th Type, the Gothic, are divided
sharply into the abbey and the parish churches.
The former are on a quite different scale, and after the
death of Cathal Crobderg and the conquest of Connaught
lost all distinctive Romanesque characteristics, preserving
only reminiscence of the past in the mason's methods of
making small windows and the like, but developing certain
peculiarities of their own. Sometimes a very archaic little
window has been utilised as in the Errew Abbey.
Some of the large parish churches may be classed rather
with the abbey churches, but on the whole it may be said
I think that the parish churches of the I3th and I4th cen-
turies are on the plan of Romanesque churches with door
and window frames of the new fashion, which was materially
modified from the English style. This modification was very
much in the direction of using very narrow slits as windows,
I suppose to keep out rain and wind in the absence of glass,
and very few windows indeed.
The Abbey Church of Errew is not dated but I class it
as probably the earliest. It may I think be taken as certain
that it is earlier than the year 1210 when the comarb lands
were transferred to the bishops in this province. At the
suppression it owned only the land given it by Robert Barrett
in 1413. The extensive see lands about it I take to have
been its endowment transferred to the bishop. The archi-
tecture is very coarse and rough. Though the windows
generally are pointed the}' are very few and very small for
a church of this class. In the N. wall close to E. end is
1 Stokes, Early Christian Archi. in Ireland, pp. 62, 63, 71.
168 DIOCESE OF TUAM
a small window with a round top hollowed out of a stone,
looking very early, as if it might have been taken from an
earlier church. A similar narrow slit is opposite in the
S. wall but with a pointed top. The cloisters were very
low and very dark, lighted only by a few narrow slits, and
might more properly be called vaults. I suppose it to have
been built by the O'Dowdas or O'Lachtnas sometime in
the i2th century for the old abbot and convent transformed
into canons of St. Augustine with a house suited for the new
practices and ideas, built by Irish architects not yet familiar
with the style.
Knockmoy Abbey founded in 1189 is entirely Gothic, but
at that time Gothic architects were abundant and it was
a Cistercian house.
Ballintubber Abbey founded in 1216 is somewhat
composite. The windows of the E. end have a decidedly
Norman aspect, but all the rest of the church is Gothic.
Norman influence appears also in the conventual buildings.
I know of no other abbey in these dioceses that can be
dated earlier than the Anglo-Norman occupation in 1237.
They are all distinctively Gothic, are unmistakable, and
are dated within at least a few years.
What may be called the great parish churches are a
small group, only three known to me, which from identity of
plan and size seem to have been built at the same time.
Shrule, 91' io"x24' 4". Burriscarra, 9i'x23' 10".
Holyrood at Ballinrobe, the ancient Roba in Carra, 102' 6"
x 24' 6", which has certainly been lengthened towards
the W., and seems to have been the same as the other two.
Allowing for my measurements taken with a tape being not
absolutely exact I think it may be taken that they were laid
out to be identical in area. They have two or three lofty
narrow pointed windows in the east end, a few similar
windows in the side walls, two doors opposite each other
in the N. and S. walls near the W. end, and at Shrule
and Holyrood a small door in the S. wall near the E.
end, probably for the use of the clergy.
Annaghdown Abbey nave is the same length.
They are certainly of about the same date as Kinlough,
but this last has a Romanesque connection in the western
CHURCH ARCHITECTURE 169
dwelling ; the others seem to have been intended for some-
thing more than the ordinary parish uses, and made pro-
vision for priests' dwellings elsewhere. It is I think safe to
take them to be the earliest of the Gothic parish churches
and to assign them to the I3th century, and to attribute
them to the Irish lords of the time of Cathal Crobderg
rather than to the first Anglo-Norman lords, who set up
monasteries and used parish church rectories to aggrandise
The rest of the parish churches usually show the later
ogival ornament and mouldings where any are left. But on
the other hand they show rather the proportions and the
arrangements of the Romanesque churches, as if there was
a reversion to Celtic uses and requirements in parish church
practices concurrently with the adoption of Irish social
customs and laws. I infer that as the Anglo-Norman families
threw out branches those branches rebuilt or reconstructed
ancient parish churches. Of course the plan would remain
the same when the " restoration " consisted of insertion
of more fashionable door and window frames in existing
walls, which is a very common case. But there are
instances where the new church was built on a new site,
and the architect was free to design what was thought
best. A very good example of this is Tempul na Lecca
at Cuslough. The old parish church is in ruins on Inish-
robe. The new one showing ogival ornament is on the
This church needs only to be stripped of ivy and shrubs,
to be roofed, and to be plastered all over to restore it to its
original condition. It is in plan typical of most of the parish
churches of its period, whether restored Romanesque or
original. They differ a little in proportion of length and
breadth but the arrangements of door and windows are
in substance the same. There is an E. window generally
very narrow, but sometimes larger and even double as in
Islandeady. In the S. wall is another narrow splayed
window close to the E. wall in order to light the altar.
In Tempul na Lecca it is so close to the E. wall that there
is but 4" of splay on that side. This is sometimes larger,
as in Kilmolara where there is a mullion. A door is in the
S. wall near the W. end. If the church is long a small
DIOCESE OF TUAM
slit may be found between the door and the W. wall, or
even two, as in Islandeady. The E. window and the
window in S. wall near the E. wall are sometimes mere slits,
as in Easky.
The following list shows the general run of dimensions
52' 2" x 20' 9"
42' 3" x 17'
A B Moyne ..
ABC Kilmainemore 52' 6" x 21' 3"
Killedan . . . 48'x2o'
Kilkinure . . 53'6"xi8'
C Tempul na Lecca . 41' x 18' 6"
C Tempul an Machaire 37' x 17' 6"
C Kilmolara . . 57'4"xi8'6"
C Islandeady . . . 52'xi8'6"
Ballyovey . . . 45' x 19'
A denotes Romanesque original altered to B or C. B early
Gothic. C Later Gothic showing ogival forms.
Tempul Som at Knockatample in Kildacommoge parish is
an exceptional church. It measures 26' 6" x 14' and had a
western loft. The walls are remarkably high for the size, in
order to allow such a loft. None of the openings have been
left in their original state, but the church seems to be early
Romanesque, or even earlier, judging from these indications.
Into these walls an E. window has been fitted consisting of
a wide rectangular limestone frame with a mullion ; a similarly
wide rectangular window in the W. gable over the pointed
door ; another in the S. wall near the W. wall. It seems
to be the latest mediaeval restoration of all, applied to the
earliest existing altered structure.
ANGLO-NORMAN ABBEY CHURCHES
They are in two divisions. The first shows a long rect-
angle with a chapel at the W. end opening into the nave
N. or S. wall, and conventual buildings on the opposite
side, such as Ballinrobe, Burriscarra, Ballyhaunis, Urlare.
This may be called the I3th century type.
The second comprises churches which are divided into
choir and nave by two arches supporting a nearly square
central tower, with sometimes transept and aisle. The
tower is less than the full width of the church and is there-
fore elegant and slender in appearance. In one case, Bur-
rishoole, the tower is the full width of the church. The
CHURCH ARCHITECTURE 171
choir is sometimes less than the full width of the nave. Such
are Claregalway, Rosserrilly, Rosserk, Court Abbey.
The date of foundation is known or the style indicates
the period of most of the abbey churches. But that of
Kilnamanagh is exceptionally difficult to date. It is men-
tioned in an ancient tract on the Muintir Murcada. 1 It is
the church of the parish of Struthir in Muntercuda (Muintir
[Mur]cada) of the Taxation. The parish merged in that
of Donaghpatrick and the rectory of the whole belonged to
this monastery at the suppression. The Four Masters record
the death of the Abbot of Kilnamanagh in 1438, who seems
to have been a Connaught abbot. A Franciscan house has
no abbot, but I suppose the term was used laxly. It may
be assumed that this small house was not founded before
the great house of Claregalway, reputed to be the first Fran-
ciscan house in Connaught. The Rackets were then in
possession of this country, and were probably the founders.
It measures 94' 10" x 19' 9'. Part of E. wall and a great part
of S. wall are gone. The E. wall stands save a part of the
S.E. angle. A small flat-headed splayed window is not in the
middle of it but nearer the N. waU, the middle of the window
being only 6' 8" from it. A small window in the N. wall
about 30 ft. from E. end has an uneven splay / I
nearly straight on the E. side. Elsewhere I have seen an
uneven splay only when a window in a S. wall is so close
to the E. wall as not to allow a splay. A part of the N. wall
to E. of this window is gone, so that it cannot be said that
there was or was not another window there. The S. wall
is down except at the W. end.
At the W. end are joist holes and a small flat-headed
window just above them in the S. wall, which I guess to
measure in the opening about 15" *6". In the S.W. angle
is a small window N on ground floor.
Near the middle of the N. wall is a bit of much better
masonry like the W. jamb of a door, apparent inside. Out-
side, about opposite, seems to be a joint as if the church
had been lengthened, and some appearance as if part of the
1 ff. W.C. t p. 368.
i 7 2 DIOCESE OF TUAM
E. side of a doorway was carried on to the W. Mounds
adjoining in the graveyard to the S. seem to be ruins of
buildings in connection with the church.
It seems that a Romanesque church with a loft has been
lengthened to the E. to fit it for a monastic church. The
work is done in a very rough way. I class it on the whole
as a late reconstruction.
Killeenbrenan or Murgagagh Abbey also presents the
feature of a mediaeval monastic church constructed on the
site of an earlier Irish church. In the E. wall is a small
piece of very fine walling of pick-dressed stones with very
fine joints, which seems to be a fragment of the E. end of
a very much older church. Unfortunately the upper part
of the E. wall is gone. The character of the rest of the build-
ing agrees with the date of foundation, 1428, given in Arch-
dall's Monasticon. The S. wall began to fall out and was
reinforced by a thickening outside which went so high as
to block a considerable amount of the square windows
high up in the wall. A huge buttress supports this wall at
the eastern end.
Murgagach is Irish for cracked, having a crack or chink,
and is a descriptive name. But it might have been applied
to the far older church close by, called the Killeen, as in
that case the chancel was built simply against the E. wall
of an older church. It is impossible to fix its date, but the
dimensions 61' x 19' whereof 19' 6" is length of chancel marks
it as a comparatively late reconstruction. It suggests to
me that this Killeenbrenan is the old parish church, and
that the abbey was formed on another disused ancient church.
The Killeen is in Moorgagagh Townland, .the abbey in that
of Kill. The Killeen was once a very important establish-
ment ; the land N. and W. and S. of it is covered with
foundations of walls and buildings marking a large settle-
HOLY WELLS, BULLAUNS, LONG STONES
CLOSELY associated with the ancient churches are Holy
Wells. Bullauns, and Long Stones.
Wells were objects of worship by the Irish and by the
other nations of western Europe. But how they were
worshipped and for what reason is now obscure. People
went to them to pray for what they wanted and to leave
offerings as they do to this day. Of the views held about
them we have an indication in one direction in Tirechan's
account of St. Patrick's proceedings at the well called Slan,
from which we learn that the well was honoured and that
offerings were made to it as a god, and that the people be-
lieved that a dead prophet had been placed in a coffin in
the well under the stone cover. This suggests a belief that
burial in such a holy place would give a good start in the
next life. The well worship was made tolerable in Irish
Christianity by dedication of the well to a saint because
it could not be eradicated, but it seems to have gained no
more than toleration and so has retained its pagan features.
In only a few cases have chapels been built at or over wells.
Such a chapel is seen at the Holy Well near the Round
Tower and old church of Balla.
The mysterious bullauns are intimately connected with
the holy wells, and in some instances are themselves used
as holy wells. The bullauns seem to be a pagan survival.
They are found not only as wells or bowls for water but on
upright and sloping stones where they could not hold water.
Those which are used as wells are on a stone sunk in the
ground. In the parish of Killedan there are three. One
called Gloonpatrick is at Oxford by the side of a stream
where a large bullaun is in a stone sunk in the earth with
i 7 4 DIOCESE OF TUAM
a few stones built round to keep out rubbish. One called
Patrick's Well is in the demesne of Ballinamore which is a
similar bullaun in a stone sunk in a low mound of stones
and grass, by the side of which are the foundations of a
small rectangular building. The third is a little east of
Ballinamore House, and is but a small hollow which may
be natural, but it is accepted as a bullaun. Local tradition
says that St. Patrick knelt in prayer at these three places,
the hollows being worn by his knees, and there is an old
saying that the part between these three stones will always
be safe from wars and destruction. The first two are under
old ash trees.
Patrick's Well is most likely the place in Mag Foimsen
where St. Patrick left Conan. The church has disappeared
but in the circumstances it is I think fairly certain that
there once was a church in Lisnacrus, or at least a chapel
at the well. The mound and the foundations show clearly
such an arrangement as exists at Patrick's Well in Kilcorkey
parish of Co. Roscommon. This well is not a spring but
a large stone with a large and small bullaun sunk in the
ground. A sort of alcove has been built over it, and the
alcove is approached by a narrow passage about 9 ft. long,
sloping downwards so that at the opening of the alcove
it is below the level of the bullauns. The passage is open
above. Stones are piled all round so that the alcove is
in the middle of a small cairn about 3 ft. high.
Adjoining the cairn on the north are remains of a small
rectangular building, of which enough of the west end
remains to show that it was built with very large stones.
It is like the cell or house sometimes seen in similar close
relation with a holy well. Stations are still made here.
The high road separates the cairn from the foundations
of a small church. 1
A little to the south of the Ballinamore Patrick's Well
are the remains of a rath, and the country people say that
stations used to be held, marked by little wooden crosses,
starting from the well and round the rath back to the well.
It is still called Lisnacrus.
These may be taken to represent a class of artificial
wells. Bullauns are commonly found in large stones and
1 JL R,S.A.L, xxxii. p. 189.
VARIOUS ANTIQUITIES 175
rocks near churches in conditions which afford no ground
for supposing that they had been built over like those two
wells. But they are very frequently used as holy wells. 1
In some cases a holy well which is a natural spring is
found near the church as well as a bullaun stone. It cannot
be said that the bullaun was a substitute for a natural well,
but it is evident that it was such in some cases, and that it
was used in some religious fashion in other cases. The
connection of bullauns with churches and holy wells needs
The ash and the thorn tree are intimately connected with
the holy wells and bullaun wells. One or other is almost
always beside a holy well. The Sacred Trees of antiquity
were called Bile.
Lough Keeraun is a small bog lake, now nearly filled
with water-weeds and the growth of bog, about 400 yards
west of Temple Som or Temple na Lickeen and north of
the road from Bohola to Bellavari in the detached part
of the parish of Kildacommoge, and is remarkable as an
object of unusual reverence like a holy well. Even now
a great concourse of people make stations about it on
Garland Sunday. Its reputation was still greater formerly.
There is some doubt as to the meaning of the name which
might be either Ciaran's Lake or Mountain Ash Lake. It
is probably the former as these objects of pagan worship
were usually Christianised by affixing a Saint's name. It is
the Loughharrow of Wood-Martin's " Traces of the Elder
Faiths of Ireland," ii. 99, with which he mentions a small
tarn in the Co. Cork which is similarly reverenced, which
seems to be also the Loughadrine of pp. 89 and 112.
At a lake called Loch Cill Eascrach half a mile S.W. of
Moylough, there was on Garland Sunday a great assemblage
of people who used to swim horses in the lake to keep disease
from them. 1
Garland Sunday is Crom Duff's Day, and where we find
these patterns at laJces and wells on that day we may safely
assume that the annual ceremony has been taken over from
the worship of Crom Duff. Garland Sunday is so commonly
1 See also Jf. S.A.I., xiil p. 466, xxxii. p. 190. Ulster Jl. of Archaology,
iv. p. 272, and Wilde's Lough Corrib, p. 294, for a few more instances.
3 O.S.L.G., i. p. 232.
176 DIOCESE OF TUAM
the festival day of wells and churches associated with St.
Patrick in these countries that I am inclined to suspect
that he may have often been given for churches the places
where Crom Duff was worshipped. It may be said that
the temples of Crom Duff were given him for churches,
for these objects of reverence were open air objects of nature
and would no more need buildings than the festivals held
at them to-day. By building a church at such a place he
would divert the worship to Christian lines without too
great a break in the associations of the common people
and half-hearted converts.
LONG STONES AND CROSSES
The Long Stone is often found at churches and then
usually bears an inscribed cross. They seem to have been
a pre-Christian form of monument in Ireland and elsewhere,
and it is not unlikely that the church was put near the stone
in some cases because it was a place where the people were
already used to worship as at the wells. In the Doonfeeny
churchyard is a very long and slender stone 21 ft. high, the
longest in Ireland, bearing ancient inscribed crosses.
Groups of three long stones are found in several places
in Ireland. Two only are known to me in these dioceses.
South of the old castle of Moneycrower are two very large
long stones, one north and one south of the high road. Near
the latter lies a third which seems to have been quarried
but not set up. A short way to the east are the remains
of a small ancient church and enclosure called Killeen-
Killocrau a little west of Ballinrobe has some 200 yards
west of it three small pillar stones. St. Patrick's seat at
Duma Selce was among the three inscribed stones. They
are likely to have had some religious significance.
Ogham stones are found in these dioceses at Breastagh
near Rathfran, at Bracklaghboy and at Tullaghan near
Ballyhaunis, and at Ross on Lough Mask where remain but a
few scores. Though the writing cannot be fixed as pre-
Christian they are certainly a very early form of monument
and have in some cases been Christianised by the addition
of a cross.
VARIOUS ANTIQUITIES 177
We have of ancient High Crosses only that of Tuam and
the remains of that of Cong. These appear to have been
put up as memorials and not over graves. The practice of
putting up a memorial cross survived into the I7th century.
By the roadside near Donamona Castle is the pedestal of a
small high cross bearing this inscription " This cross was
made in anno 1633 by David Kelly and Gate Bourke his wife
for the soule of his [father?] Moyler Kelly who died 8 October
1627. F r whom let all men pray " and other Latin in-