An example of the regional type defined by the plan is found in Co. Limerick (Donnelly, 1994). Combining their state of preservation and their plan, he isolated a number of types found in the county. Of these, the commonest were his type 1A, with the plan of the rectangular tower divided into two unequal
Figure 127 Walterstown, Co. Meath: the tower-house, showing the vault and floor levels
Spaces: in the upper floors, the larger space housed the main chambers, with the lesser ones and the stairs in the small block. The entry was in the smaller of the two, the door leading into a lobby from which opened a subsidiary chamber on the one side and the spiral stairs on the other. This, and the closely allied tower-houses of his types 1B-C, were all to be found within an area about 18 miles by 8 (28x13 km), the central part of the county, south and west of Limerick city (Donnelly, 1994, 104-5). Within this area these are the predominant form of the tower-houses, while they are not to be found in the neighbouring counties, especially Tipperary. Given that many Limerick tower-houses have been destroyed, of course one cannot be dogmatic about it, but this does look very much like a definite local type of plan.
Jope (1966, 120-1) pointed out the occurrence of a particular plan type in south-east Co. Down, based on two projecting turrets on one side. They are linked by an arch at the level of the battlements, which provided a wide machicolation commanding the door which led into one of the turrets below (Fig. 128). They are found neither in the rest of the county, nor in Antrim to the north, nor in Louth to the south. In the latter, interestingly, the ‘Courthouse’ in Ardee has two turrets projecting on one side, and the entry door leads into the stairs within one of them. However, the plan seems almost deliberately to eschew the Co. Down scheme by placing the door in the outer wall of the turret concerned (Buckley and Sweetman, 1991, 343-5). Killagh in Co. Meath does the same with the door on the side opposite from the two turrets (Abraham, 1991, Fig. 36).
The tower-houses of Louth, Meath and the Pale in general are notable for their use of subsidiary turrets, though not on the one side as in Co. Down. Where there are two, they tend to be at diagonally opposed corners, like the Scottish Z-plan. Unlike it, however, the door is not placed in the angle so as to be protected from one of the flanking turrets (Fig. 129). There are also tower-houses with turrets at each of the four corners. The type has not been defined more tightly, in terms either of the diagnostic traits or of its boundaries. In Wexford, the so-called Coolhull type of castle has been linked to tower-houses (Jordan, 1991, 36-7). It combines a hall with a narrow, higher chamber turret at one end; the western core of Fethard castle is one such, although it is here included with enclosure castles because it has a gate tower as well. In general, the Wexford tower-houses lack the turrets of the Pale (Jordan, 1991, 36). In the south-west, in Clare and Tipperary (but not Limerick), are found a number of tower-houses which are round in plan (Fig. 130) but otherwise share the same internal and external features of the other tower-houses (Craig, 1982, figs 65 and 67). Similarly, the two rectangular blocks with angle turrets at Delvin, Westmeath and Dunmoe in Meath should be linked to tower-houses although they are formally different in plan.
The size and plan would clearly have been the patron’s business for they would have dictated how the building would function and how much it cost. In the contract quoted earlier the presence of a vault was specified; this is a feature which seems to reflect some regional variation as well as a possible chronological pattern. The tower-houses of Munster as a whole are more elaborate than those of the east and north. In Limerick and Tipperary, a substantial minority of tower-houses have a vault over more than one floor. This is unknown in Ulster or in Louth and there is only one example in Meath (Newcastle—Abraham, 1991, 611). The amount of decoration would also be a matter for the patron, and does seem to vary regionally. Loops in the arris of the tower (see fig. 118) are found quite frequently in the south but not at all in
Figure 128 Audley’s castle, Co. Down: the door of the tower-house protected by the two turrets and a machicolation arch
Eastern Ulster or in Louth, and only once in Meath, at Summerhill (Abraham, 1991, plate 76). Likewise, the elaborate corbelled-out angle turrets found in the south and west are only found (and then in rather different form) in Ulster in seventeenth-century Scottish Plantation contexts (Jope, 1951) and are absent from the Louth and Meath.
Figure 129 Roodstown, Co. Louth: the flanking turret does not protect the door of the tower-house
A curious, and unexplained, procedure can be seen in the construction of some of the tower-houses of the west. Here their plans are rectangular, with the smaller rooms collected at one end, as is common elsewhere. What is odd is that it is clear, from the breaks visible in the masonry, that the section with the smaller rooms was built before the rest. It is difficult to believe that the building of a tower house was such a formidable undertaking that it was split into two
Figure 130 Ballynahow castle: a round tower-house
Parts, to allow the patron to collect his resources in two stages. Even if it was, it seems equally strange to envisage a period of time when only the lesser rooms were used, until the main ones were built.
Details which may owe more to the craftsman than the patron include those studied by Duggan, looking at the timberwork in a selection of tower-houses in the south, and McKenna, who compared the remains of timberwork (floors and roofs) in Counties Tipperary and Down, to investigate these traditions (Duggan, 1982; McKenna, 1984a). Duggan distinguished between floors constructed with the joists supported by a number of beams with their ends sunk into wall sockets (his ‘double floors’), and floors where the ends of the joists rested on two wall-plates (one on each opposing wall) supported by corbels and with their ends sunk into sockets in the end walls. He tried to distinguish between them in terms of date, arguing that the double floors were earlier. As an argument this has not stood up to testing. With one exception (Audley’s castle, where the floor joists were supported on offsets of the walls), all the floors in the Co. Down tower-houses were built in what McKenna called her type 2, Duggan’s double-floor system. This system used more timber and was more difficult to repair than the one normally used in Co. Tipperary, McKenna’s type 3, Duggan’s wall-plate system. While this system was used in all the Tipperary tower-houses McKenna looked at, in two of them the Co. Down system was used on one of the floors as well (McKenna, 1984a, 24-5), which seems to dispose of any chronological distinction between the two systems. In Limerick, Donnelly (1994, 140-1) also found both systems in use in different tower-houses, with no hint of a chronological difference involved. The Pale tower-houses also use both systems, sometimes in the same tower-house, like Glaspistol, Co. Louth (Buckley and Sweetman, 1991, 313-15).
The question of roofs is more complex. The first oddity, as noted by Craig (1982, 107-9), is that only one apparently original wooden roof has survived, at Dunsoghley. For the rest, we must try to reconstruct the arrangements of the wood from the sockets in the stone, left when the original timber was removed. McKenna pointed out (1984b, 42-3) that the form of the roof was intimately linked to the use of the attic space and the arrangements of the parapet and wall-walk. In Co. Down the attics were very simple with a floor carried apparently on the tie beams of the roof trusses: any windows in the gables at this level were very small and there were no other domestic features. In Co. Tipperary, the windows at attic level were larger and better made, indicating that the attics were seen as spaces for more than just storage or low-grade accommodation and their floors were carried on wall plates like those below them. In order not to block the attic space, the roof trusses must have been based on collar beams rather than tie beams: at Ballindoney, for example, the gable has sockets for collar purlins.
World History
