HORSE SKULLS AT BAY FARM COTTAGE, CARNLOUGH

On the 27^th September 1983, Mr B Doherty informed the Department of Archaeology at Queen’s University, Belfast, that workmen had uncovered a group of animal skulls while renovating his cottage at Bay Farm, Carnlough. On the next day we visited the cottage and discovered that the skulls had already been dug up and that the floor from which they had been obtained had been cemented over. In front of the cottage was a spoil heap formed from the soil underlying the floor and after digging through it we were able to obtain 10 horse skulls and the foreleg (radius) of a cow. A description of the discovery of the bones by the workmen permits us to reconstruct the location of the skulls.
The skulls were apparently discovered under the north sitting room of the cottage which faces on to the sea. The floor of the room was of wood set on four-inch joists which in turn rested on a layer of brown soil and beach cobbles from the raised beach near-by. One of the workmen claimed that the horse skulls were uncovered at a depth of at least one foot into the fill. He also indicated that most of the skulls were found towards the middle of the room rather than near the walls. One of the workmen who dug up the skulls estimated their total number at twenty, however we were able to recover only ten skulls from the spoil heap and it is unlikely that the number of skulls much exceeded this figure.
The discovery of a series of horse skulls beneath a house floor naturally raises a number of interesting questions, the most important of which include:

When were the skulls deposited?
How were so many skulls collected and buried?
Why were the skulls buried?

Date of Burial
Since the skulls had already been removed from the context before our arrival on the site and the floor sealed with concrete our assessment of the date of deposition must be fairly conjectural. We do know that a substantial structure stood at Bay Farm at least as early as 1669 when it served as the residence of James Donaldson. As J Irvine observes (The Glynns Vol Xl, p.19) remains of this house survived at Bay Farm into the present century. While it is possible that the horse skulls were deposited as long ago as the mid-seventeenth century, there is some evidence to suggest that the deposition may have been more recent.
The main objection to a mid or late seventeenth century date stems from the presence of seventeenth/early eighteenth century rubbish within the same section of the spoil heap as the horse skulls.  In possible association with the soil from which the skulls were removed there were also recovered several sherds of decorated glazed pottery known as sgraffito ware. This pottery is generally dated to the seventeenth and early eighteenth century and would be characteristic of the fine wares that one might expect in a fine residence such as that erected by James Donaldson. While this may help to confirm the historical evidence concerning the Donaldson family, it does little to support a very early date for the deposition of the horse skulls. A very plausible explanation of the broken sherds is that they were rubbish associated with the layer of soil into which the horse skulls were later deposited. The finding of the foreleg of a cow, again typical kitchen rubbish, in the same soil as the skulls would further support the theory that the skulls were deposited into an earlier layer. Furthermore we observed that several of the skulls appeared to have been stained with the impression of wood. At first we suspected that this indicated that the horse skulls had been placed directly beneath the floor-boards of the house but on further enquiry the workmen insisted that the skulls were at least one foot to a foot and a half deep in the soil and beach cobble layer. It is therefore possible that the wood impressions might have been derived from wood remains associated with the soil underlying the present Bay Farm cottage. All of this line of reasoning, of course, rests on the assumption that the seventeenth/eighteenth century rubbish dug up by the workmen while renovating the Bay Farm cottage was associated with the house remains into which the skulls were later buried. It is entirely possible, however, that the rubbish accumulated above the floor and after the skulls had been deposited and that they were subsequently mixed together during more recent modifications to the house. Naturally, a determination of whether the skulls or the seventeenth/.eighteenth century rubbish were deposited first could only have been accomplished by a controlled archaeological excavation of the site. In any event, the evidence indicates that the horse skulls ma have been deposited as early as James Donaldson’s erection of a house by the mid-seventeenth century although one line of evidence suggests that it might have been somewhat later.
If one chooses to exclude a seventeenth/eighteenth century date for a deposition, then what is the next probable date for the burial of the skulls? A glance at the first edition of the Ordnance Survey Map indicates a rectangular structure typical of the period and area standing at Bay Farm by 1834. Here we may recall James Boyle’s depiction of the typical farmhouse in this parish as being two roomed (sleeping apartment and kitchen). Subsequent maps indicate extensions of the structure and it is clear that most of the additions associated with the present cottage have involved extensions to the west. How much earlier this structure was built we cannot say for certain. Mr Jimmy Irvine has informed me that Lendrick’s map of 1810 marks a house on the site but gives it no name 3which may indicate that it had already fallen into decline. Whether this house is the same as that on the Ordnance Survey Map or the declining remains if the Donaldson family’s earlier residence, we cannot say.
One more item of dating evidence must be taken into consideration since it will affect our interpretation of the horse skulls. It is clear from the workmen’s account that the skulls were located well below the present ground surface and not between the joists of a wooden floor. Consequently, it is unlikely that they represent a deposition associated with more recent remodelling of the cottage with a wooden floor but should refer back to a time when the cottage had an earthen floor. Unfortunately, this does not provide us with a clearly defined date since the wooden floor may not have been laid until the twentieth century.
In conclusion, there would seem to be clear evidence for a house at Bay Farm from at least the seventeenth century. It is possible that the horse skulls were buried under the original house floor or possibly under a somewhat later renovation house occupying the same position. On the evidence of other horse skulls burials from Ulster (see below) we might expect that the deposition was made before the mid-nineteenth century.

Manner of Burial
An examination of the horse skulls permits is to say a few things about their probable manner of collection and deposition. To begin with, there were no lower jaws accompanying the skulls which would suggest that they had been deposited after the flesh had decayed from them rather than soon after death. This is further supported by the absence of nose (nasal) bones on eight of the ten skulls. In short, all evidence points to the conclusion that the burial involved that of horse skulls and not horse heads.
As to the manner of death, only two skulls clearly indicate pole axing. Concerning the age of the horses at death, three of them were old individuals, three were mature and three were between the ages of 2½ and 4½ years. The tenth individual was represented by a fragment which could not be aged. It is likely, then, that most of the horses died of old age/disease and there is no uniformity of slaughter pattern that might suggest that the horses were deliberately dispatched immediately before their burial.
Finally, we have the problem of their post-mortem treatment. In most cases we suspect that the flesh simply decayed from the animal over a period of time. This is supported by the evidence of a dog gnawing on the face of some of the individuals and the general absence of nasal bones. However, several of the skulls show evidence of knife marks consistent with either the butchering or skinning of the horse. Consequently, we seem to have a situation in which the horses died over a protracted period and their heads were exposed to both the elements and occasionally dogs which reduced them to their bare skulls. Eventually, they were collected together for their final deposition beneath the floor of the cottage. This leads us to our final problem – why were they buried?

Function of Burial

The deposition of horse skulls beneath the floor of a house is reasonably wide-spread in Ireland and it has been the subject of two interpretations – foundation sacrifices and acoustic enhancement.
In 1945 Sean O’Suilleabhain published an article – ‘Foundation Sacrifices’ in the Journal of the Royal Society of Antiquities of Ireland (Vol75, part 1, pp.45-52). In it he related the results of an inquiry carried out by the Irish Folklore Commission to a number of informants scattered throughout the Republic. Eleven people responded that they were acquainted with the practice of burying one or more horse skulls beneath the floor of a house. With the exception of one correspondent from Co. Galway who suggested that the burial was made to bring luck to the house, all other correspondents associated the burials of the skulls with @making an echo’, especially for dancing. O’Suilleabhain, however, argued that this acoustic explanation was merely a practical justification for continuing an ancient pagan practice of sacrificing an animal and burying it on the erection of a building. This was done in order to secure the protection of the victim and bring luck to the householders as well as to ensure the stability of the structure. O’Suilleabhain cited examples from around the world and suggested that such practices may have gone back to the Bronze Age in Ireland.
The second hypothesis, acoustic enhancement, takes the explanations of the practice made by O’Suilleabhain’s informants at face value. i.e., it explains the burial of skulls as an attempt to provide a hollow area below clay stamped floor or stone flagging so that it would increase the sound of musicians and dancers in a room. This use of horse skulls as sounding boxes is by no means limited to the Republic and has been discovered on a number of occasions in Ulster. Dr Alan Gailey describes in the Yearbook of the Ulster Folklore and Transport Museum (1969-70: 13-14) the discovery of nine horse skulls beneath a house floor near Crossgar, Co. Down and another from an 18century house in Crumlin, Co. Antrim. Dr Gailey points out that these finds as well as others reported tend to associate with ceilidhing houses. Consequently, the location o the skulls in the middle of rooms (and not under wall foundations or hearths) the explanation offered by the informants themselves, and the association of the practice with structures used for music and dancing tends to support the ‘echo; theory.
In evaluating these two theories we should keep in mind a tendency among folklorists to seek ‘ultimate explanations’ no matter how non-functional they may be or whether they are the motives of the people carrying out a custom at a particular time. We have already seen O’Suilleabhain’s willingness to discard what his informants told him as merely a practical justification for a much earlier custom of foundation sacrifice. The same must also be said of some of the examples of the sounding box theory. For example, the horse skulls found in the house near Crossgar were no below an earthen floor but rather between the joints supporting a wooden floor. Acoustically they would have been of no value whatsoever since the hollows formed by the joists would have provided the echo effect. Dr Gailey explains this retention of a custom for enhancing the sound of a room even when it has become meaningless as a typical example of people retaining old customs even after their rationale has ceased to exist. Whether one accepts this as an explanation or nor, the evidence from Bay Farm Cottage does tend to support the sounding box interpretation.
In conclusion, from Bay Farm we certainly have no evidence of horses deliberately sacrificed for their ritual deposition. Rather the heads seemed to have been decayed over a period of time and were then buried together. The skulls were found within soil and beach cobbles underlying the house rather than between the wooden joists and so they were likely to have been buried beneath a stamped floor rather than a wooden floor. We cannot say precisely how long ago they were buried although they may have been deposited as far back as the mid seventeenth century. All of this suggests that the north sitting room in the Bay Farm cottage may have served not only as part of the farm house but was the scene of ceilidhs among locals in the northern part of the parish of Tickmacrevan. This may have been some time before 1835 when James Boyle in his notes to the Ordnance Survey Memoirs of the parish related ‘the people do not seem much given to amusement. Dancing is the only one and it is now greatly given up’.
The authors would like to thank J Irvine, E Mc Donnell, C Bell, and N Brannon for their assistance and advice in writing this paper.