The Ballinderry Harp — A Reassessment

by Michael Billinge

Page Menu


Review of early records and some published sources:

The Crannog of Ballinderry

Further observations and comments:





In the National Museum of Ireland, Kildare Street, Dublin, are the surviving metal remains of an ancient harp. These are on display, at the time of writing, mounted onto a modern hypothetical wooden model of the instrument. They consist of the copper–alloy fittings which were attached to the harp’s neck (the curved bands, an end cap, and a strap affixed to the neck and forepillar), along with many of its tuning pins. These pieces were said to have been discovered in the Crannog of Ballinderry, hence they are generally referred to as the Ballinderry Fragments or the Ballinderry Harp Fittings/Mountings. More detailed comment will follow.

Review of early records and some published sources

Royal Irish Academy:

Before they came into the keeping of the National Museum, these fittings were displayed for a time in the Museum of the Royal Irish Academy, who obtained them in 1866. [1] The RIA’s earliest records relating to the fragments indicate that they had previously formed part of a collection belonging to a Mr. Murray of Mullingar, noted in the their Treasure Trove Register as having been found in Ballinderry Crannog. [2]

McIntyre North:

Possibly the earliest published reference to the existence of these mountings, albeit a brief and obscure one, is contained in Charles Niven McIntyre North’s 1880 work, The Book of the Club of True Highlanders. It is regrettable that North chose not to provide any details regarding them in his text, because his observations on other harps are informative. But it is clear that he knew of their existence, for in a footnote he cites two examples of neck–to–forepillar straps that "we have seen", and says:

The Lamont, and with the mountings of a harp in the Museum of the Royal Irish Academy. [3]


In 1886 W. G. Wood–Martin published Lake Dwellings of Ireland, a major work on the subject of crannogs. This volume provided a description, as well as an illustration by W. F. Wakeman:

In the accompanying plate XXVIII., is figured (1) an Irish harp as restored: it stands at present 37 inches in height, all the metal portions were discovered in the crannog of Ballinderry, county Westmeath, and the proportions and form of the modern woodwork were regulated by the size and form of the original metal work; the thirty–five pins show the number of strings that were formerly attached. Although the style of decoration is of a very early character, yet as the letters I.H.S., surmounted by a cross, appear engraved on a brass plate in front of the instrument, the work cannot be older than the sixteenth century. Nos. 1, 2, 3, 4, and 5, convey a clear idea of the kind of ornamentation on the metal fittings, and No.7 (from Drumdarragh crannog) represents a harp pin of the usual kind. [4]

Drawing of Wakeman

Figure 1: Plate XXVIII from W. G. Wood-Martin, Lake Dwellings of Ireland, 1886, facing page 125.


Robert Bruce Armstrong’s seminal 1904 work The Irish and The Highland Harps contains a more detailed account, in which he gives the following description, along with two high–quality photographs and sketches.

Harp Mountings Found At Ballinderry

In the National Museum, Dublin, may be seen in a singularly fine state of preservation the brass mountings for an Irish Harp from the Crannog of Ballinderry near Moat, King’s County. [sic]

For the termination of the harmonic curve there is an enrichment, the front end of which is triangular in form (see illustration). Upon the front is the monogram I.H.S. surmounted by a cross, and beneath an interlaced cross enclosed in a circle. This triangular front is 6⅞ inches long, the lower side of the triangle being 2⅜ inches broad. Attached to the edges of this triangular face by five nails with ornamented heads there is an elaborately wrought border or frame. The sides of this termination without the border are 1⅛ inches broad, [5] each side having differently engraved patterns. Upon the right side there is a brass support for strengthening and retaining in position the harmonic curve and the fore–pillar (see illustration). This support, which is 8¼ inches long by 2⅞ at the widest part, is pierced in the centre in the form of a cross. The upper termination of the support is a dragonesque head; the lower portion divides and curves outwardly; the termination is somewhat shorter, to allow for the commencement of the T formation. The bands for strengthening the harmonic curve upon either side, pierced for thirty–six tuning–pegs, form single curves and are ornamented, as are the tuning–pegs. The measurement from the first tuning–peg in the treble to the triangular face is 19 inches.

These brass mountings (the property of the Royal Irish Academy), probably late sixteenth century, were either attached to or intended for a Harp of a large size, perhaps five feet in height. [6] At present they are placed upon a model, and the tuning–pegs, almost all of which are extant, are inserted in the respective holes. These mountings are exceptionally fine, finer than any other known specimens. The border or frame attached to the termination is deserving of special notice. [7]

Plate from Robert Bruce Armstrong, Musical Instruments Part I, The Irish and The Highland Harps, 1904, facing page 62.

Figure 2: Plate from Robert Bruce Armstrong, Musical Instruments Part I, The Irish and The Highland Harps, 1904, facing page 62.
[You may click on the picture to view a larger–sized image]

Illustration from R. B. Armstrong, page 64.

Figure 3: Illustration from R. B. Armstrong, page 64.


Joan Rimmer’s 1964 article in The Galpin Society Journal provided further information:

A harp identical in superstructure with that given by Praetorius is implied by the Ballinderry fragments, which are preserved in the National Museum of Ireland. These bronze harp fittings comprise two neck bands with thirty–six pin holes; thirty–one pegs of which nineteen, probably original, are forged and have arrowtail decoration, and twelve, probably replacements, are cast and have simple linear decoration (Plate VIIb); the neck finial boss and the large neck–to–forepillar brace. From these fragments the neck and the top of the forepillar can be reconstructed exactly. The neck must have been of great depth and strength with a deep curve at the treble end, while the forepillar’s T–formation must have started not less than 11.5 cm. down from the joint with the neck. It is impossible to arrive at an exact date for these fragments, or to be certain of the shape of the soundbox of the harp to which they belonged. Dr Raftery, of the National Museum of Ireland, considers them, on decorative grounds, to be post–fifteenth century, and their resemblance to the fittings on Praetorius’s Irish harp suggests a late sixteenth–, or early seventeenth– century date and a box deepening from bass to treble. [8]

In her 1969 publication The Irish Harp, Rimmer also reproduced, courtesy of the National Museum of Ireland, two photographs of the fragments, in Plate 17 (a view of the right side), and Plate 18 (a front view of the finial boss). [9] But otherwise the book contains no extra information to add to that already given in her earlier article.


The harpmaker Robert Evans also published an article in The Galpin Society Journal in 1997, which includes an interesting piece of information relating to a fragment of wire string:

Whilst examining the Ballinderry fragments at the National Museum of Ireland, I noticed a short length of wire in the drilled hole of the tuning pin then in the 10th position from the bass. This wire was fused to the pin by a layer of corrosion. This was a section of drawn wire 4.78 mm long and 0.7 mm in diameter. I was able to have this sample analysed using energy dispersive X–ray spectrometry (EDS) by Dr Peter Northover, Department of Materials, University of Oxford. This analysis showed its composition to be a simple alloy of copper and zinc, with probably no more than 10% zinc. [10]

The Crannog of Ballinderry

Of the crannog itself there is little now to be seen, and one would need good directions to even find the site. As a result of various drainage works in the area, the lake these days is much smaller than in former times, and the current shoreline has retreated some two hundred metres from where the crannog once stood. [11] But no study of the Ballinderry harp would be complete without examining the archaeological background of the site in which the remains were said to have been discovered.

It has been written elsewhere that these harp fittings were found by archaeologists at a crannog, near Moat, King’s county, Ireland, in the 19th century. [12] This comment indicates a general misunderstanding and perhaps some confusion on the part of its author. Not only is the town of Moate actually in County Westmeath, but the statement that they were found by archaeologists — despite being presented as fact — is merely an assumption based on a misinterpretation of previous sources, none of which cites archaeologists as having made this discovery. This inference is misleading, for if these fittings had truly been found by 19th century archaeologists, it would establish a firm provenance for them. As it stands, a more detailed study reveals a completely different story.

One of the earliest accounts of the Crannog of Ballinderry can be found in Wood–Martin’s Lake Dwellings of Ireland:

Ballinderry, in the parish of Kilcumreragh, barony of Clonlonan, is situated not far from Moate. When, as the result of drainage, the water of this lake fell, it was discovered that it had formerly contained a large crannog surrounded by a stockade of oak piles, around and on which was an immense quantity of the antlers of red deer, and fractured bones of deer, oxen, sheep, and other mammalia, all afterwards sold as manure. Many objects of archaeological interest found here were obtained by various collectors — some are in the Museum, R. I. A., and others have been figured in the Journal of the Royal Historical and Archaeological Association of Ireland. The first notice of the crannog occurred in 1844, when Mr. Hayes forwarded to Sir W. Wilde a description of it, together with a plan and map of the locality. Two single–piece canoes were disinterred from this site, and a portion of an ancient harp of wood. The pendent amulet of stone, figured p. 115, was found here. [13]

Despite any impressions to the contrary, it was not until 1942 that any proper archaeological investigation actually took place, when the Harvard Archaeological Mission (led by the noted American archaeologist Hugh O’Neill Hencken) undertook a thorough excavation of the site. This painstaking examination revealed evidence of two distinct eras of occupation, one in the late Bronze Age and another in early mediaeval times; but — significantly — there was none for any later period. Thus the crannog had gone out of use many centuries before the harp could even have been made.

In the introduction to his excavation report, published in the Proceedings of the Royal Irish Academy, Hencken made the following interesting comments:

In the National Museum of Ireland is a collection of antiquities said to have come in the past from this crannog. But the collection includes such diverse objects as a neolithic antler axe–holder of Swiss type and a harp which according to Wood–Martin is not older than the 16th century. A considerable number are also forgeries. Of the remaining it is impossible to tell which really do come from this site, and the information in the old registers of the National Museum is seldom precise. Probably objects found anywhere in the neighbouring bogs and loosely called from Ballinderry have been connected with it. Indeed that point of view survived locally at the time of the excavation, for the writer was shown two bronze spear–heads, which were alleged to have come from the crannog itself, but which had actually been found in a bog close by. Furthermore, a drainage ditch had long ago been dug across Ballinderry Crannog No. 1, and some objects must have come to light there. It was also found in sorting out the old material in the National Museum from the Lagore Crannog that Wood–Martin was careless in attributing localities to the finds which he published, so that the illustrations of so–called Ballinderry objects are not good evidence. Hence it seemed better to omit all consideration of objects said to have been found previously at the site, and to confine this report to an account of what was actually found in the excavation. [14]

It is pertinent that the harp was one of the two items Hencken specifically singled out as examples which clearly could not have come from this site. Another telling observation is his statement that a considerable number of the objects said to have been from the crannog were also forgeries. It is an indication of how the growing thirst for antiquities amongst 19th century collectors had led to the manufacture of such items by less–than–scrupulous dealers; and the Crannog of Ballinderry conveniently provided them with a suitably noteworthy site where they could claim the artefact was discovered. It is now generally recognised by modern archaeologists that not all of the objects once said to have been found at this crannog can genuinely be attributed to it, as the following two comments made by Aidan O’Sullivan in his Ph.D. thesis bear out:

Various other items of late Bronze Age date including tools, weaponry and ornaments have been reported from the site in the NMI files, although some of these may have been deliberately misprovenanced so as to increase their value.


Intriguingly, there were also some modern forgeries from the site, inspired by antiquarian interest in it in the nineteenth century... [15]

Further observations and comments:

One can hardly view these metal remains without certain obvious questions coming to mind, e.g. what might the original harp have looked like, how big was it, what was its musical range, and would it be possible to reconstruct it?

Unfortunately, because so little survives of the instrument, there are too many variables to provide answers with any degree of confidence. A reasonable deduction of the form of the neck–to–forepillar joint can be made, but the length and curvature of the forepillar, the position at which the neck and soundbox met, and the form of the soundbox itself can only be guesswork. [16] Therefore, efforts to assess the lengths, angles, and spacings of the strings would likewise be pure conjecture. Any attempts at a reconstruction need to be based upon a given premise, and unless a plausible case can be made for choosing a particular one as a starting point, the matter must remain open.

However, though it may be impossible to definitively answer any of the above questions, a study of the remnants can reveal further information to add to that already published. But before examining the surviving metal components themselves, I feel that some brief comments relating to the modern model are helpful.

The modern wooden model:

It may have seemed sensible to the curator of a 19th century museum to attempt to recreate some sort of visual interpretation of the harp’s missing timber parts, in the hope of giving an idea of what the original might have looked like. But since such a model can only be conjectural, it is a side issue to the main study of the fragments. However, for anyone who is interested, information regarding its construction is included in W. F. Wakeman’s unpublished catalogue of the Royal Irish Academy Museum, in which he wrote:

Head fittings of brass or bronze, mounted on a modern harp model, made for their reception by Simon Byrne, a servant of the Royal Irish Academy, under the direction of the late Edward Clibborn, Curator of the Museum.... [17]

It would appear that Edward Clibborn ceased being the curator in 1872. [18] If so, Wakeman’s statement indicates that this construction would have been undertaken sometime between 1866 and 1872.

It seems likely that this is the same model as the one upon which the mountings are currently displayed, for a comparison of the present woodwork (examined in 2014) with the detailed photograph reproduced in Armstrong’s 1904 volume shows the various dents, marks, and grain pattern to be identical. [19] However, Wood–Martin’s statement that it stood 37 inches in height is somewhat confusing, as this measurement neither agrees with the model as seen today, nor fits in with the scale and proportion of the illustration provided in his book. Indeed, given the size of the surviving metalwork, a height of only 37 inches would produce a disproportionately squat–looking harp, and the figure as printed is probably an error.

Neck bands:

The form of these carefully constructed copper–alloy neck bands, or harmonic curves, bears a clear similarity to those found on three other early instruments: the Trinity College harp, the Queen Mary harp, and the Lamont harp, though the ones on the Ballinderry are larger, and pierced to take thirty–six tuning pins. This is a greater number than on any of these examples, and comparing the shapes of their respective curves tends to suggest that the extra strings on the Ballinderry probably extended the range in both the treble and the bass, though this cannot be definitively proven.

It is generally thought that the reason for fitting such bands was to give additional reinforcement to the neck. Although attaching metalwork will clearly add strength to the timber, this was not their only — nor necessarily even their primary — purpose. It is more likely that their main role was to resist the wear of the tuning pin and provide a solid surface at either end to hold it more securely. This would help to prevent the neck holes from becoming excessively or irregularly worn, and also maintain the quality of string tone. (Perhaps one way to visualise this is to think of the tuning pin as an axle and the brass plates as bearings at either end). [20]

The holes in the bands would have been enlarged to the diameter necessary for the tapered tuning pins to fit correctly, and this was probably done by rotating a similarly–tapered strip of hardened metal within the hole. But whatever the tool used for this purpose, it does not appear to have been sharp enough to cut the metal cleanly (unlike a modern reamer), because the process seems to have raised a slight ridge, or burr, around the edges of the holes. [21] An examination of the distance between the centres of the holes reveals a variance ranging from approximately 13 mm. in the treble to 15 mm. in the bass. This seems to indicate that the strings of the Ballinderry harp were spaced wider apart than those found on the three aforementioned instruments.

The neck bands themselves, which are about ⅞ inch in width, have distinctly chamfered edges, which appear to be a unique feature (no other surviving harp has bands formed in such a manner). This indicates that the bands were intended to sit flat against the surface of the neck, rather than being located in a shallow recess, as they are on some of the other instruments. Therefore, these bands were only held in position by the tuning pins passing through them, plus a number of small nails which were driven through holes in the lug–like extensions neatly incorporated into the chamfer. The position of these nail–holes alternates between the upper and lower edges. [22] Armstrong said that the measurement from the first tuning–pin in the treble to the triangular face [at the far end of the band] is 19 inches. In order to gain a better idea of the overall length of the bands, one needs to add an extra inch or inch–and–a–quarter to this distance. [23]

Detail view of end of neck band showing the decoration, chamfered edges, nail-hole lugs, and burrs around edges of tuning pin holes.

Figure 4: Detail view of end of neck band showing the decoration, chamfered edges, nail-hole lugs, and burrs around edges of tuning pin holes.

Neck end cap, boss or enrichment:

This highly ornamented end cap would have made a striking and beautiful addition to the instrument, but it also had an important practical purpose which went beyond that of a mere decorative embellishment: to protect the termination of the neck, where the end–grain of the timber was otherwise extremely vulnerable to damage. Therefore it is not surprising to find similar caps extant on some of the other early harps too. [24] Although there are several examples of these, this one is by far the largest. Its capacious size indicates that the neck would have had a profile far deeper than that of any other known instrument, and probably appeared somewhat bulky in comparison. [25]

The cap seems to have been assembled from three basic components: a band of brass wrapped around the end of the neck, forming a collar or cuff. Affixed to this (probably using a soldering or brazing process) was a flat face–plate, placed against which was what Armstrong described as the border or frame. The assembled unit would have been secured to the end of the neck by the five large and skilfully decorated nails, which were driven through the holes in the border and plate. This manner of attachment suggests that the inside face of the cap fitted flush against the end face of the neck, with no gap between the two. There also appear to be four small nail holes, two on each side, at the rear edge of the collar, which were probably added to secure these edges and stop the thin brass from accidentally being pulled away; or perhaps to prevent any buzzing or rattling which might otherwise have occurred. [26]

There is good reason to believe that the cap was specifically tailored to fit the individual shape of the neck–end, and not made independently of the harp. This becomes evident from observing the slightly misshapen and asymmetrical nature of the end plate, for when viewed from the front, this can be seen to bulge to the left of the border; whereas the right side aligns more consistently. Had the end cap been made separately, to a given set of dimensions, it is unlikely that what appears to be an idiosyncratic characteristic of the neck would have been so accurately reproduced.

Detail view of part of neck end cap showing the quality of some of the decoration.

Figure 5: Detail view of part of neck end cap showing the quality of some of the decoration.
[You may click on the picture to view a larger–sized image]

Illustration from R. B. Armstrong, page 64.

Figure 6: Exploded diagram showing how the main component parts of the neck end cap (the collar, plate, and border) were assembled, then fitted and attached over the end of the neck with decorative nails.

Neck–to–forepillar strap:

Armstrong refers to this as a brass support for strengthening and retaining in position the harmonic curve and fore–pillar, a statement which is open to over–interpretation. For the sake of clarity, it is helpful to take a closer look at how the harmonic curve (i.e. neck) and forepillar were joined, and undertake a more thorough assessment of the purpose of this strap.

Illustration from R. B. Armstrong, page 64.

Figure 7: Frontal view of neck end cap showing its asymmetrical nature.

The necks and forepillars of the earlier harps were fitted to each other with a carefully–constructed mortice and tenon, which was not only drawn tightly together by the tension of the strings, but was also pegged by the tuning pins which passed through it, effectively locking the whole thing into place. This formed a substantial joint which was quite capable of fulfilling the structural function by itself, without any requirement for an external strengthening and retaining support. As a result, early harps such as the aforementioned Queen Mary harp and Trinity College harp, were made without such straps.

However, at a later period it seems to have been deemed desirable to attach this type of strapping across the joints of certain harps, for it was subsequently added to the Trinity College. [27] Evidence of two other similar examples can be noted as well. [28]

This strapping was not really necessary for reinforcing the joint or holding the parts together: its primary purpose was to help counter the effect of the twist — or torque — of the neck which, after a period of time, could sometimes open a slight gap on the non–string side of the joint. This in itself does not represent a major structural problem, as the tension of the strings would prevent it from coming apart. But the small visual separation can look unsightly, and attempting to rectify this was probably more of a cosmetic exercise than a physical necessity. (The accompanying diagram should help to explain this effect, and how the strap would have prevented any gap appearing.)

cross-sections through a typical neck-to-forepillar joint

Figure 8: Three diagrams showing cross-sections through a typical neck–to–forepillar joint, which illustrate:

(Left) How the snugly–fitting mortice and tenon joint is pulled together by the string tension and pegged securely by the tuning pins passing through it.

(Centre) How, with time, a combination of worn tuning pin holes and the torque on the neck can contort the mortice slightly, allowing the joint to creep and a small gap to open at the side.

(Right) Adding the neck-to-forepillar strap prevents any gap from developing.

A closer examination makes it apparent that this strap was solely intended to constrain the relatively small component of force which acted to open the joint; for apart from the single tuning pin passing through it, the fitting was only attached to the harp by four small nails, one into the neck and three into the pillar. It is clear from the lightness of this method of fixing that this construction was never meant to resist any great force. Further evidence is presented by the fact that the lower part of the strap forms the outline of a Greek cross. Decorative as this design may be, its open nature makes it structurally weaker. Had strength been considered a priority, a solid shape would have been used.

Figure 9: Detailed views of parts of neck-to-forepillar strap and neck end cap, showing similarity of scalloped (and other) decorative patterns.

Figure 9: Detailed views of parts of neck-to-forepillar strap and neck end cap, showing similarity of scalloped (and other) decorative patterns.
[You may click on the picture to view a larger–sized image]

A few other observations relating to the strap are also worthy of comment:

Like the neck bands, the strap also was formed with bevelled edges, which indicates that it too sat flat against the face of the wood and was not recessed into the side of the forepillar. The metalwork at each end is finished with zoomorphic embellishments, which on the upper one takes the form of a beautifully executed animal head, described by Armstrong as dragonesque. This figure certainly appears more reptilian than mammalian, but exactly what creature it’s meant to represent is uncertain. At the lower end, extending down from the bottom edge of the cross and facing away from its counterpart, are the heads and necks of two birds. These extensions also served to take the two lowest of the small nails used to attach the fitting. Whether the appearance of these animal heads was purely decorative or intended to communicate some deeper symbolism must for the moment remain conjecture. Overall, the general characteristic of the metalwork on the strap is very much in keeping with that found on the other brass fittings. Although it is not conclusive proof, this tends to reinforce the likelihood that these were all produced at the same time, and by the same craftsman.

Tuning pins (or pegs):

At the end of the 19th century, thirty–five tuning pins appear to have been fitted into the display model. [29] Yet an examination of the harp today reveals only twenty–nine pins in place. This same number, twenty–nine, can be counted in the National Museum of Ireland’s photograph, published as Plate 17 in Joan Rimmer’s previously cited 1969 work The Irish Harp. However, it is clear from this picture that at least some of these tuning pins are not now in the same position as they were when that image was taken; and it is possible to look even further back and observe that sometime after Armstrong’s 1904 photograph, the lowest pin had also been removed. [30] Moreover, Rimmer stated in her 1964 article The Morphology of the Irish Harp (also referred to above) that there were 31 tuning pins then extant. [31]

Therefore, when Robert Evans commented on the finding of a small length of wire attached to a tuning pin, he quite correctly stated that the pin was then in the 10th position from the bass. His deliberate choice of the word then indicates his awareness that a degree of caution is necessary when assessing this string. This does not so much concern the nature of the wire or the findings on the alloy itself, [32] but mainly addresses any assumptions that the tuning pin and its string fragment were necessarily still in their original position. It should also be noted that in the 19th century, when the fittings were mounted on the newly–made wooden model, Byrne and Clibborn would have needed to remove all the pins from the neck bands. There is no guarantee that these were subsequently reinserted into their previous locations.

A further note of caution should also be raised in relation to the pins themselves: Rimmer appears to have divided them into two groups, saying that nineteen, probably original, are forged and have arrowtail decoration, and twelve, probably replacements, are cast and have simple linear decoration. Yet a closer examination suggests that there may be an even greater variety in the pins than this statement implies. Wakeman noted, length of pins four and one–eighth to three and a quarter inches, which also indicates quite a degree of variance. [33] It certainly does look as though many of the pins now set into the display model may not be original to the instrument. [34] Whether these were added to the harp as replacements during its working life is difficult to say, but it seems unlikely that so many pins should somehow be lost while the instrument was still strung and in use. Without evidence to the contrary, the possibility that some of these could perhaps have been taken from the Royal Irish Academy’s collection of loose tuning pins, and added to the model at the time it first went on display, might also be considered. [35]

The heads of the tuning pins currently in the neck bands, showing the apparent variety of forms.

Figure 10: The heads of the tuning pins currently in the neck bands, showing the apparent variety of forms.
[You may click on the picture to view a larger–sized image]

View along the string side of the neck, indicating the degree of difference in the lengths of the tuning pins.

Figure 11: View along the string side of the neck, indicating the degree of difference in the lengths of the tuning pins.


Although it is unfortunate that only the metalwork remains by which to judge the instrument, it can certainly be argued that the quality of the craftsmanship exhibited on these fragments surpasses that found on the fitments of the other surviving Irish and Highland (as Armstrong called them) harps. But it would require an extensive description to attempt to do justice to this, and the workmanship can probably be better appreciated from the accompanying illustrations and photographs. Nevertheless, there are a few observations worth drawing attention to, as they potentially provide further information relating to the Ballinderry’s background.

There is an undeniably strong Christian theme to certain elements of the decoration, as attested by the I H S and cross engraved prominently on the front plate, as well as the obvious symbol of the Greek Cross predominant in the form of the neck–to–pillar strap. Moreover, the incorporation of this cross–shape into the very structure of the strap was clearly a deliberate — and unconventional — choice at the time of its manufacture. This suggests that the religious theme was a part of the harp from its very conception, and that whoever commissioned the instrument wished to make some conspicuous expression of Christianity. Whether this indicates an ecclesiastical origin is hard to say, but it should be noted that several elements of Christian symbolism have been identified within the designs found on the Queen Mary harp harp too. [36]

This would not be the only link relating to the decorative patterns of these two instruments. In their comments regarding the engravings on the Ballinderry’s end cap, Keith Sanger and Alison Kinnaird have observed:

On one side of the mount, and in the peak of the face, they also show patterns of palmette leaves closely similar to those carved on the Queen Mary Harp. [37]

A close examination does indeed reveal that the patterns are similar to a remarkable degree (see illustration below). This could be significant, for not only does it suggest that the two harps may share a common root, but it also becomes more interesting when viewed in light of a comment made by John Bannerman who, when assessing the provenance of some of the early clàrsachs, wrote:

But the feature that clinches their Scottish provenance is the ornamentation on the upper part of the sound–box and on the fore–pillar of the Queen Mary harp, in particular the spiral arrangement of clusters of split–palmette leaves which is one of the characteristic patterns of late medieval monumental sculpture in the West Highlands and which has no parallel in Irish art of this period. [38]

There is also another pattern, found in the decorative engraving running along the edge of the left–hand (string–side) neck band of the Ballinderry, which resembles that on one of the other harps. The same type of design runs parallel to the edge of the right–hand neck band of the Lamont harp. It is worth noting that of all the surviving instruments, the metalwork of the Lamont, though simpler, is (arguably) the closest to that of the Ballinderry.

Figure 12: Three illustrations from R. B. Armstrong

Figure 12: Three illustrations from R. B. Armstrong showing the strong similarity in style of the decorative patterns found on

(A) the left side of the Ballinderry neck cap;

(B) the front face-plate of the Ballinderry neck cap; and

(C) the top of the forepillar of the Queen Mary harp.

Comparative drawings

Figure 13: Three illustrations extracted from lithographs reproduced in John Stuart, Sculptured Stones of Scotland, Vol. 2, 1867, showing similar patterns of decoration on West Highland Cross shafts:

(A) Plate XLIX — Keils in Morvern, Argyll–shire [i.e. Keil Church, Lochaline, Morvern, Argyllshire]

(B) Plate XXXV — Keil, Islay [i.e. Keills, near Portaskaig, Islay]

(C) Plate LVI — Kilchousland, Cantire [i.e. Kilchousland, near Campbeltown, Kintyre]


It probably goes without saying that had the woodwork survived, this harp would very likely have presented one of the finest — if not the finest — examples of its genre; but with so little of the instrument remaining, some questions will always remain unanswered. However, despite there being so little to go on, the metalwork does contain some clues that may help with understanding its origins.

Although nothing is known of the harp’s history prior to its becoming part of Murray’s antiquities collection in the 19th century, the prominent Christian theme in the decoration may hint at some ecclesiastical source. Regarding a date for its manufacture, both Wood–Martin and Dr. Raftery remarked that, on decorative grounds, it would not have pre–dated the 16th century. The strong similarity to the artwork found on the Queen Mary harp (an instrument thought to date from around 1500), the form of the metalwork in general, and the possible religious connection, all tend to suggest a date in the first half of the 16th century. [39]

To turn to the elephant in the room, i.e. the claim that these fragments were found at Ballinderry Crannog: this site is now known to have been derelict for many centuries before the harp was even made. Therefore it could not have been there during its occupation, and in later periods the crannog was a deserted island; so the belief that the harp was discovered in that location no longer holds water.

The origin of the Ballinderry connection appears to lie with the entry in the Treasure Trove Register of the Royal Irish Academy. This could simply have been a mistake or misunderstanding, with the fittings carelessly being listed along with other parts of the collection which may genuinely have been discovered at the crannog. However there is another possibility which cannot be discounted.

How Murray actually obtained this metalwork is uncertain, but it seems likely to have been through one of the dealers in antiquities who made it their business to supply such artefacts to interested parties. This was not always the most honest trade in those days, given the fact that they dealt in rarities and supply was not always adequate to demand. Murray was from Mullingar, and the proximity of the Crannog of Ballinderry would have provided a convenient known archaeological site to use in claiming a new provenance for an existing relic, by any dealer who was keen to sell it to him. We will never know whether the Treasure Trove entry was deliberate or some sort of misunderstanding, but it would likely have been in Murray’s own interest to have these fittings registered as found at a recognised site such as this. Establishing this provenance for the items could be to the owner’s advantage.

Since the purported connection with the Ballinderry Crannog can no longer be considered credible, any assumption that this harp was made by a craftsman working in the midlands of Ireland is now open to question. In fact, what evidence there is suggests an origin much further to the northeast. The strong similarity of elements in the decoration between these metal fragments and those on the Queen Mary (and what Bannerman called the West Highland School of carving in general), as well as the parallels between the neck fittings on the Lamont and Ballinderry, tip the balance of probability towards the west of Scotland as the place of construction.


In an intriguing coincidence, on 8th July 1863, three years before the Royal Irish Academy acquired the Ballinderry fittings, a photograph was presented at the quarterly meeting of the Kilkenny and South–East of Ireland Archaeological Society, along with a brief account concerning some curious remains of an ancient Irish harp. Their journal notes:

By Arthur Gerald Geoghegan, Esq., Londonderry: a photograph of five characteristic ancient bronze fibulae, and another of eight bronze spearheads; also photographs of the front, back, and side views of a human skull; and photographs of a portion of an ancient Irish harp, and two ancient bells. All the photographs were executed by Mr. Ayton, of Londonderry. With regards to the skulls, harp, and bells, Mr. Geoghegan forwarded the following notices:–

... ANCIENT IRISH HARP. — Found at the bottom of a bog at Taughboyne, county of Donegal. The wood was lying beside, crumbled to dust. There is a tradition in the neighbourhood that a battle was fought in former times in the locality where the harp was found. The workmanship is of a superior description. As the bards joined in the onset of battle, it is not improbable that this is the iron framework of the clairseach. This interesting relic is now in my possession. [40]

If, as reported, the remains were found in a bog, and the decay of the timber was such that it crumbled to dust, it is difficult to equate the statement that the workmanship was of a superior description with the mention of the iron frame (by which Geoghegan probably meant the neck bands). There are a few surviving harps known to have neck bands made of iron, but these all appear quite crude in construction and probably date to the 18th century. [41] Earlier harps — and those of any quality — invariably had neck bands of copper–alloy. Given the circumstances in which this relic was stated to have been found, it is hard to see how any ironwork could have remained sufficiently free from corrosion for Geoghegan to be able to judge the quality of its workmanship as superior. It seems likely that he was either careless or over–generalising, and that his use of the word iron is incorrect. It would be difficult to distinguish the exact type of metal using only the naked eye, on a specimen which was discoloured from having lain underground for so long.

In any case, it is doubtful that he would have commissioned a costly (as they were in those days) photograph of something not well enough preserved to justify his statements. Attempts to locate this picture have thus far proved unsuccessful, and it is possible that it was returned to Geoghegan after the meeting.

Unfortunately, nothing is known of these remnants today, at least under such an identity. Geoghegan’s account seems to be the only record of any harp claimed to have been found at Taughboyne (about six miles southwest of Derry city). [42] If the workmanship truly was of a superior description, as Geoghegan says, then it seems unlikely that he (or his estate) would have casually disposed of this interesting relic without either selling it to the antiquities trade or presenting it to an institution. Yet no trace has ever been found of any Taughboyne harp fragments. Items of such apparent value don’t normally just disappear.


I would like to thank the National Museum of Ireland, and particularly Mary Cahill for her assistance with the Royal Irish Academy records.

[1] Royal Irish Academy Museum Register of Antiquities 1859–1886.

[2] Royal Irish Academy Treasure Trove Register 1861–1872.

[3] C. N. McIntyre North, The Book of the Club of True Highlanders, 1880, p. 37, Note r.

[4] W. G. Wood–Martin, Lake Dwellings of Ireland, Dublin and London, 1886, pp. 125, 126. The number of strings formerly attached, given here as thirty–five, is misleading, as the harp would originally have taken thirty–six.

Note on Plate XXVIII (facing p. 125): While some of these figures do convey a clear idea, they are not totally accurate. This is especially true of the representation of a harp pin, reproduced as No. 7, which is poorly drawn and can give a misleading impression. (There is also a brief mention of the instrument on p. 206 but it adds no information).

[5] At this point Armstrong adds in a footnote, The border rises ⅜ inch; the heads of the nails are ⅜ inch above the border.

[6] It seems most unlikely that a harp of this period, and with this form of harmonic curve, would have been anywhere near five feet in height. Even the larger, slightly later, instruments known from the early 17th century were only around four feet tall. The closest comparative example, in terms of number of strings and likely age period, would probably be the Otway harp. Since this measures approximately 41 inches, a more appropriate suggestion for the height of the original instrument seems to be in the region of three and a half feet.

[7] Robert Bruce Armstrong, Musical Instruments Part I, The Irish and The Highland Harps, Edinburgh, 1904, pp. 63, 64.

[8] Joan Rimmer, The Morphology of the Irish Harp, Galpin Society Journal, Vol. 17, 1964, p. 42, and Plate VII b.

At the time of writing her article, Rimmer seemed unwilling to entertain the idea that the harp Praetorius described was a chromatically–tuned instrument, thus atypical of the normal genre. A note of caution is therefore advisable regarding any comparisons that might be made or implied between these two harps. They are two distinct forms of instrument, with differently shaped harmonic curves and numbers of strings. To suggest that these were identical in superstructure is technically incorrect.

[9] Joan Rimmer, The Irish Harp, Cork, 1969, p. 44 (Plate 17) and p. 45 (Plate 18).

[10] Robert Evans, A Copy of the Downhill Harp, Galpin Society Journal, Vol. 50, March 1997, p. 124.

[11] For a brief description and more precise details of the site’s location, please view the National Monuments Service Record Details on http:/ Ref. OF001–001 (Archaeological Survey of Ireland, Record Details), posted: 23 May 2011.

[12] (accessed 22nd August 2015).

[13] W. G. Wood–Martin op. cit. pp. 205, 206.

[14] Hugh O’Neill Hencken, Ballinderry Crannog No. 2, Proceedings of the Royal Irish Academy, Vol. 47C, 1942, p. 5.

[15] Aidan O’Sullivan, The Social and Ideological Role of Crannogs in Early Medieval Ireland, Ph.D. Thesis, NUI Maynooth, Vol. 2, 2004, pp. 166 & 168.

[16] It is not just the soundbox’s size that would be in question, but its shape too. An examination of what seem to be the four oldest harps still extant reveals that two of them, the Trinity College harp and the Lamont harp, have soundboards which were probably originally formed with a flat face; whereas the two others, the Queen Mary harp and the Otway harp, appear to have had their boards deliberately carved into a curved profile. This is important because the shape of the soundboard also affects the instruments’ string lengths.

[17] W. F. Wakeman, Royal Irish Academy Museum. Catalogue, Vol. 2, unpublished, 1894, p. 81.

[18] Edward Clibborn was Curator of the Museum for many years, but seems to have resigned most of his positions at the RIA in 1872. He died April 1880. See: http:/

[19] In producing this model, Clibborn appears to have been influenced by elements of the Sirr harp and a plaster cast of the Trinity College harp, both of which were in the possession of the Royal Irish Academy Museum at that time.

[20] An observation by James Talbot reinforces the point that the primary purpose of these metal neck bands was to act as bearing surfaces for the tuning pins. In the 17th century he noted in a manuscript (Christ Church Library Music MS., 1187): Brass Cheeks of each side to prevent pins turning by draught of strings and Brass cheeks on each side to which the pins are fitted to prevent their turning round by the draught of the strings.

Joan Rimmer, James Talbot’s Manuscript VI. Harps, Galpin Society Journal, Vol. 16, 1963, p. 67.

[21] Similar burrs can also be observed on the neck bands of some of the other early harps.

[22] There are thirteen nail holes clearly visible (six along the upper edge and seven along the lower) but it seems probable that there could be two more hidden under the neck cap, to secure the end of the band. The possible total of nails would then be fifteen for each side.

[23] W. F. Wakeman’s Catalogue (op.cit.) states that the length of the curved straps is twenty–two inches. At first sight this seems to contradict Armstrong’s information, but it is likely that Wakeman measured around the curve rather than in a straight line.

[24] Examples of similar metal finials can be found on the Trinity College harp, the Lamont harp, and the Otway harp. To emphasise the functional importance of these end caps, it is worth noting that the neck of the Queen Mary harp, an instrument of similar form and construction, did not originally have one. (Although there is evidence that some sort of metal fitting was attached at a later date, it was subsequently removed.) The absence of such protection means that at some stage, the Queen Mary suffered damage at the apex of the neck, which has broken away.

[25] Armstrong measured the length of the end cap at 6⅞ inches. By way of comparison, the equivalent dimensions for some similar harps are:
Lamont = 5¾
Trinity College = 4½
Otway = 4¾
Queen Mary = 3½ (NB: This is for the end of the neck only, as this harp currently has no actual end cap).

[26] The construction and fitting of this end cap is in fact very similar to that used on the Lamont harp, though the Lamont’s, being smaller and less decoratively complex, utilises fewer nails to attach it. In comparison, the Otway harp’s brass finial seems almost crude, consisting merely of a collar and plate, with very little in the way of decoration. The cap on the Trinity is of a slightly different form, and is made of silver, originally set with two cabochon crystals, with no frontal nails or border to secure it.

[27] When the harp now known as the Trinity College, first came to public notice in the 18th century, there were two short metal straps placed across the neck–to–forepillar joint on the non–string side. Later — probably in the 1840s when the harp underwent study — these were removed and not replaced. Once they had been taken off, it would have become quite apparent that the straps were attached over the top of (and partially obscuring) some of the harp’s earlier decoration, and that they were therefore clearly a later addition. Today, traces of the small holes left by the nails used to affix these straps can just be discerned in the neck and forepillar.

[28] Apart from the example mentioned above, a pair of brass straps were also added to the Lamont harp, and the illustration of an Irlandische Harff, reproduced in Plate XVIII of Michael Praetorius’ 1619 publication Syntagma Musicum II, De Organographia, clearly shows two finely–wrought metalwork straps attached across the neck–to–forepillar joint.

[29] This is the implication of Wood–Martin’s statement (op. cit.) that the thirty–five pins show the number of strings that were formerly attached. This figure was apparently confirmed a few years later by a comment made by W. F. Wakeman in his unpublished catalogue (op. cit.) that the neck bands are traversed by thirty–five string pins of brass or bronze.

[30] It is not possible to say exactly how many pins were in situ when R. B. Armstrong saw it, for although he states that almost all were inserted, he unfortunately gives no specific figure for these.

[31] A possible explanation for the apparent numerical mismatch evident in Rimmer’s information may be found in the fact that at some stage, two of the tuning pins had been removed for detailed photography; and it seems likely that these were not replaced in the model afterwards, but stored separately. Photographs of these two pins are reproduced in Plate VII b of Rimmer’s 1964 article (op. cit.).

[32] The results of the metallic analysis for this piece of wire seem perfectly in keeping with what may be expected for a string of the period. For a further overview of the type of wire used for harp strings prior to 1700 see:
Michael Billinge and Keith Sanger, Early Pre 1700 References to wire harp string composition
and also Historic Records of Wire Harp Strings

[33] W. F. Wakeman, op. cit.

[34] It seems strange that a harp whose quality of metalwork was otherwise of the very highest order would intentionally incorporate tuning pins of such an irregular, even haphazard, construction. By way of comparison, it may be of interest to note that all the surviving pins found on the Dalway harp (1621) were made to the same shape, size and decorative pattern.

[35] In William Wilde’s catalogue of the antiquities of the Royal Irish Academy, 1856, a collection of twenty–two metal harp pins is noted: the majority were obtained from crannoges. By the time the fragments were mounted for exhibition, it seems likely that the Academy would have acquired several more.
William Robert Wilde, A Descriptive Catalogue of the Antiquities in the Museum of the Royal Irish Academy, 1856, p. 599.

[36] For an overview of some of the Christian themes found in the decoration on the Queen Mary harp see: Keith Sanger and Alison Kinnaird, Tree of Strings — A history of the harp in Scotland, 1992, pp. 59–61.

[37] Ibid. p. 57.

[38] John Bannerman, The Clàrsach and the Clàrsair, Scottish Studies, Vol. 30, 1991, p. 9.

[39] Rimmer’s suggestion that the fragments may belong to the late sixteenth–, or early seventeenth– century was based on an inappropriate assumption, and is unlikely (see note 8 above).

[40] Journal of the Kilkenny and South–East of Ireland Archaeological Society, New Series, Vol. 4, No. 2, 1863, pp. 343, 345.

[41] Some examples of Irish harps with neck bands made of iron, rather than copper–alloy, are: the Hollybrook harp, the Royal Irish Academy No.2 harp (also known as the Carolan harp) and the William Kearney No.1 harp.

[42] Since Geoghegan did not say how the remains came into his possession, the possibility that he obtained them from an antiquities dealer is a distinct one. As such, it would be wise to treat the statement that they were found at the bottom of a bog at Taughboyne with a certain amount of caution, for he may only be relating a story told to him by the vendor.

Submitted by Michael Billinge. August 2015.
Note: Except where otherwise stated, all photographs and drawings are ©2015 Michael Billinge

Except where otherwise noted, content on this site is licensed under a Creative Commons License.
Permissions beyond the scope of this license may be available by contacting us at

Creative Commons License