Endoscope image showing the materials and construction details visible inside the body cavity of horse. (XVII.9)
Endoscope image showing the materials and construction details visible inside the body cavity of horse. (XVII.9)
Carved wooden hand from the Line of Kings. British, (XVII.113)
Carved wooden hand from the Line of Kings. British, (XVII.114)
Endoscopic photo inside a carved wooden horse (XVII.7)
Carved wooden head of Henry VIII. English, about 1689-91 (XVII.1)
Carved wooden hand from the Line of Kings
Carved wooden head of Charles I by Grinling Gibbons. English, 1686-7 (XVII.2)
Carved wooden head of Edward III. English, about 1688-90 (XVII.41)
Carved wooden head of Henry V. English, about 1688-91 (XVII.44)
Carved wooden head of Henry VI. English, about 1688-91 (XVII.43)
Carved wooden head of Henry VII. English, about 1689-91 (XVII.40)
Carved wooden head of James I. English, about 1688-90 (XVII.47)
Carved wooden head of William I. English, about 1688-90 (XVII.777)
Carved wooden head of William III probably by Nicholas Alcock. English, 1702 (XVII.45)
Carved wooden head of Charles II by Grinling Gibbons. English, 1685-6 (XVII.3)
Head of a carved wooden horse from the Line of Kings. (XVII.7)
Head of a carved wooden horse from the Line of Kings. (XVII.10)
Head of a carved wooden horse from the Line of Kings. English, 1685-90 (XVII.11)
Head of a carved wooden horse from the Line of Kings. English, 1685-90 (XVII.12)
Head of a carved wooden horse from the Line of Kings. English, 1685-90 (XVII.14)
Head of a carved wooden horse from the Line of Kings. English, 1685-90 (XVII.16)
Head of a carved wooden horse from the Line of Kings. English, 1865-90 (XVII.17)
Head of a carved wooden horse from the Line of Kings. English, 1685-90 (XVII.18)
Head of a carved wooden horse from the Line of Kings. British, 19th century (XVII.20)
Head of a carved wooden horse from the Line of Kings. English, 1685-90 (XVII.30)
Head of a carved wooden horse from the Line of Kings. English, 1685-90 (XVII.8)
Paint analysis of a carved wooden horse from the Line of Kings.( XVII.7)
When the Line of Kings was being re-displayed in the 1990s the horses and heads were examined for dendrochronological study and dating. While some of the horses were undergoing conservation in the mid 1990s oak samples were taken from one of them and the cores were analysed by Dr Daniel Miles. The cores came from horse XVII.17 and some of this oak was identified as having grown in the Eastern Baltic region. The tree rings from certain of the timber baulks were correlated with the region’s sequence for dates, suggesting the felling of this oak between about 1675 and 1691.
In February 2012, as preparations began for a re-display of the Line in 2013, Ian Tyers made an initial examination of the twelve horses and concluded that in their present condition, with their exteriors fully painted, it would be inappropriate to attempt further tree ring analysis on them. As no areas of exposed timber are available, the use of non-destructive methods to obtain patterns of tree rings which could be measured and matched to established sequences would not be possible. However, although tree-ring dating of the horses could not be undertaken, it was possible to draw some fresh conclusions about the types of materials used in the construction of the horses and it was noted that while most of the twelve horses on display in the White Tower in 2012 appear to be constructed mainly of oak, there is one horse which was mainly made using a softwood while another was made from elm. The significance of the different choices was noted, both for the carving of the statues and for dendrochronological examination of them in the future.
The choice of timber would have been important to the carvers because the type and quality of the wood used would have affected the results that could be achieved and their durability. The selection of oak would have provided the carvers with a material well suited to the production of finely detailed statues, whereas softwoods would be less suitable for fine carving but adequate otherwise. As the horses were to be fully painted and no areas of bare wood would be seen, it would have been irrelevant to choose one wood rather than another for reasons of its appearance and selection must have been on the basis of its physical qualities, availability and cost. Slower growing oak would have been a more expensive wood to buy than softwoods, especially if it had been imported from the Eastern Baltic region rather than obtained from trees grown in England and transported only a short distance to London.
However, the higher price would have been partly offset by the fact that fine Baltic oak could be used with less wastage than most English hard- or softwood as its straight grain and absence of knots would have been ideal for carving into shape. Twisted and knotted pieces of cheaper timber would have been acceptable for use in the interior structures of the horses but would not have been appropriate for use where carving needed to achieve an unblemished surface effect of muscles and veins. As the archival evidence shows that all the wooden horses purchased in the late seventeenth century were carved at fixed prices, each carver’s workshop would have had to balance the availability and price of materials against the labour costs and the need to achieve a particular standard in the end result.
Carving freestanding, life-sized horses in wood was an unusual commission and there is no evidence that any of the workshops which were awarded contracts had executed a similar job before. Consequently, in the absence of a well-established method for producing a realistic-looking wooden figure of a horse, each workshop would have had to draw on its workers’ knowledge and experience but also to use their initiative and to innovate both with the materials and methods employed. As a result of the discussions following Tyers’ examination of the horses it was confirmed that some of the carvings are composite, as Dr Alan Borg had noted in the case of the softwood and oak horse which was taken apart and rebuilt in the 1970s – made of more than one kind of timber rather than simply oak, elm, spruce or pine. While it is possible in several cases to detect the character of the wood grain through the paint on the exterior surfaces of the horses, it was agreed that only by the use of an endoscope would it be possible to identify whether the oak of the outer cladding was mounted on an inner structure incorporating other woods. This thinking helped guide the endoscopy carried out subsequently in the cavities inside each of the horses.
The choice of timber would need to be matched by appropriate decisions about the ways in which blocks and planks of wood would be used together to produce the required figure. The durability of the horses would be affected not only by the fact that in general oak is a stronger wood than spruce or pine for withstanding knocks and that it is also less vulnerable to attack by woodworm, certainly if the younger sapwood is used facing inwards leaving the tougher heartwood facing out. Another factor is the skill of the carver in ensuring that correct choices are made about the direction of the grain in blocks to be carved. Physical damage and woodworm infestation are certainly factors which must be borne in mind when considering why the number of horses still surviving from the late seventeenth century is thought to be about 60% of the total commissioned.
By examining the unpainted wood surfaces, which are visible just inside the sockets into which the horses’ tails fit, Tyers suggested that a considerable amount of the oak appears to have been worked whilst it was still relatively green. This is indicated by the size and shape of the marks made by adzes, gouges and similar tools used to remove excess wood, as on the interior surfaces these marks have not been rubbed smooth by use of glass-paper or similar, as they have on the exterior. Use of oak in this softer state would have made it much easier to work than after it had dried out and the practice may have been much more usual than is indicated by the common belief that timber has to be well seasoned before use.
While the drying of the timber after its use may have caused some shrinkage and consequent cracks in the horses, this would usually have been minimised by using the wood in such a way that the plane of greatest shrinkage would not be placed in an important position in relation to the exterior of the carving. In addition, as the horses were to be painted, any cracks or overhangs caused by movements in the pieces of wood could be concealed by fillers which would then be covered by layers of paint. External examination of the horses in detail revealed lines and round marks visible through the layers of paint that indicate some of the joints between different pieces of timber and in-filled holes. Some of the latter seem, because of their locations, to be holes for dowels which were used when the horses were carved in sections and assembled to ensure that different parts fitted well and were strongly held together. The positions of some of the other holes suggested that in the past fixings may have been attached to the horses’ bodies for struts to give added stability. After he had re-arranged the line in 1826-7 Dr Samuel Meyrick reported that he had removed wooden posts or props, which had previously been used to help support the horses in the Line. It is possible that some of the metal rods which today help support the horses were added at this time, and that others have been added, or replaced, more recently.
While it was disappointing that tree ring sequences could not presently be taken, it was recognised that the potential of this dating technique should be utilised when conservation work on one or more of the horses next gives opportunities to access unpainted internal surfaces of oak suitable for scanning and analysis.
Unlike the carved horses, which are of hollow construction, the heads were made from solid wood, presumably because they were much smaller and lighter. The savings in materials used and the need to reduce weight were therefore not important factors in making the heads, which were carved either from single pieces of timber or from two or more pieces fixed together. Most of the ‘kings’ heads’ are still painted but at some time a few have been stripped back to bare wood. For those where suitable areas of wood are visible it is possible to identify the timber used and on some pieces of oak to measure the tree rings for correlation with established dendrochronological records. Originally the carved heads would have been joined to the rest of the figures but the bodies from the late seventeenth century seem to have been destroyed or disposed of and surviving mannequins are later. The heads themselves are in several instances significantly altered as they have been re-used over the centuries, sometimes cut down to fit inside helmets of a particular shape and size, and pieces have been removed or lost. The necks of these heads end in flat surfaces which were not intended to be visible. These bases provide access to the wood, even for the heads which are otherwise covered in paint.
Ian Tyers examined the carved heads and selected several made of oak and with their tree rings accessible on the base. From these he attempted to take measurements of tree ring data. Unfortunately some proved unsuitable for providing ring sequences for dating purposes as in the past the bases had been repeatedly struck hard with a hammer, which had distorted their ring patterns. Tyers suggested that these heads might in future be submitted for a computerised tomography (or CT) scan, which would produce a print-out of the tree ring patterns within the object, from which dating might be possible. This non-destructive technique could also be usefully applied to the oak heads from which ring sequences had been obtained. Their bases only reveal a part of the full sequence of rings in the wood from which the head is made. The CT scanner would capture the pattern of rings that do not show up on the base, potentially doubling the length of the ring sequence and increasing the likelihood of achieving a match with an established and dated sequence.
From the samples successfully taken from heads, one match was quickly achieved between the oak of one of the carved and painted heads and oak from horse XVII.17 which was examined in the 1990s. The high probability score of the match indicates that timber from the same tree was used in the making of both these items, which may well have been carved in the same workshop. The wood matches the Eastern Baltic oak sequence for the late 17th century, but as neither sample includes bark or wood from the outer part of the tree, the rings do not give a precise date for its felling. The tree ring scans from the other three heads were not possible to match with published ring sequences at present. However, as ring sequences are still being built up for various potential regions of origin, such as Poland, Lithuania, Latvia and Estonia, coverage is currently far from complete. The fact that the oak ring samples have not produced matches with datable sequences should certainly not be regarded as evidence that they are not contemporary with the oak which successfully gave a good match. Future research, including the taking of longer sequences from some of the carved horses when suitable opportunities arise, may well lead not only to links with dated sequences but even further matches between the oak used in different horses and/or heads. The identification of the subject of XVII.62, the head newly restored to the ‘royal’ group, poses a challenge as none of its features are particularly striking. What was reassuring, however, was that having obtained a positive dendrochronological identification, the head closely matched the characteristics of the group recognised by other means – weight, dimensions, materials, construction methods and paint treatments. By approaching the objects from different starting points we have now established a far better understanding of them than we had previously.
In the course of the dendrochronological tests it was noted that, despite their apparent simplicity, even the one-piece heads differ in the way in which the wood is used. In most the grain ran in the direction of the axis of the head, nose to back or vice versa, but in at least one instance the grain ran at 45° to this axis. It is unclear at present whether this is an indication of differences between carvers, or an experiment within a workshop responsible for producing multiple heads. It may or may not prove significant that one of the carved horses was made differently to the others, mounting the four blocks of wood at a 45° to the axis of the body and thus working in a manner somewhat similar to the head. The initial findings of the examination of the royal heads alerted us to the possibility that there could be other heads originally made for the 1685-90 Line of Kings but which have been overlooked and dated much later. A systematic examination of all the other heads for possible 17th century examples may increase further the number of items from the Stuart Line of Kings that we can identify.
The relationship of the heads and horses and the timber used in their construction could be significant but is not straightforward. The archival sources make it clear that the nineteen wooden horses made between 1685 and 1690 were each supplied with a figure with a carved face, with one possible exception. For a £20 fee a carver had to provide a ‘horse and figure’ set, with a face apparently carved to resemble a particular king. It seems that the price included the carvers providing the timber that they used, unlike certain other occasions when the Board of Ordnance records refer to supplying a quantity of wood to a craftsman for the construction of a figure. Gibbons’ orders specify that he was to supply the carved faces of Charles I and II but unfortunately none of the orders to Emmett, Morgan, Nost I, Quellin or Townson make clear which carvers were to portray which monarchs. It has formerly been argued that Nost I supplied the head of Henry VIII amongst others but the only post-Gibbons royal head with specific documentation linking a carver and a subject is the head of William III by ‘Alcock’ – identified as Nicholas Alcock – commissioned to be added to the Line in 1702. As at least 18 faces were supplied but only 14 kings seem initially to have been in the Line in the 1690s, it poses the question whether some carvers supplied generalised, ‘non-royal’ faces, intended for use on equestrian figures displaying contemporary cavalry arms and equipment. The carved head which provided the tree ring data matching one of the horses leaves us with the challenge of identification. Was it carved as the head of a king – and if so which? Or was it a non-specific face for a figure to represent a cavalryman? However, there is another possibility to consider, which is that some monarchs may have been represented by two or more carved heads. This is likely for Henry VIII, as 18th-century accounts mention a standing figure as well as the one on horseback in the Line. Head XVII.1 might have been used on either of these two figures but we may no longer be able to determine which. It would appear from written accounts that at times there may also have been two figures of King Henry VII.
The fact that some heads are made from oak and others from elm or a soft wood may not prove particularly helpful in connecting the heads with particular horses and carvers. Some heads were carved from single blocks, others were built up of two or three smaller pieces. However, this may also not help in making attributions to carvers. The parts of composite heads were sometimes of the same wood but in other cases were of different woods. The composite wooden heads indicate that even with the smaller carvings, oak was sometimes chosen only for the face while the back of the head was made of softwood. The size of the pieces needed for making a head were small enough to have made use of off-cuts, as available. On the other hand the size and quantity of the blocks needed for constructing horses would have been great enough to have required the workshops either to draw upon the stock of blocks that they carried or to order new materials specifically for the job.
The carvers worked on a range of commissions, often using limewood, fruitwood and other timber particularly favoured for the deep carving and undercutting in fashion at the period for the decoration of doorcases, overmantels, picture frames and other architectural ornaments such as friezes. It is possible that supplies of oak blocks would have been only a small part of their regular stock in trade but timber-yards beside the Thames would have been an easy source for London carvers. Analysis of the wood used for the heads offers two possibilities for establishing associations. If two or more heads can be shown to have been made from the same piece of timber it would strongly support the case that they were carved in the same workshop, as long as there is no evidence that the wood is part of a repair or replacement. In addition, it might prove possible to identify wood from the same tree used in other heads and horses, linking them as likely products of the same workshop – if at some future date it proves possible to examine more timber from the horses. However, it is also a possibility that the different carvers bought their wood from the same timber dealers, so imported oak baulks from the same tree and in a single shipment could have found their way into more than one of the workshops. Caution is need in the interpretation of results unless other sources of evidence are available to confirm the conclusions. Knowledge about the importation of Baltic oak into England at this period is very scarce, which is a further reason for being hesitant until further research into the scale and nature of the trade is undertaken.
Finally, it was noted that a completely different, denser and darker type of wood was used for the carving of the head of the ‘blackman’ and that this might be of interest to the timber research department at Kew Gardens. Experts there may be able to identify the wood, its place of origin and the dates between which it was imported into Britain.
A representative sample of the collection of hands was examined but no examples were found which were made from oak and potentially datable by tree-ring analysis. The hands seem to have been carved from light and very light woods which were more easily broken than hardwoods like oak. The hands may thus have been treated as more disposable than the heads and horses, and it seems likely that many of the surviving hands date from the 19th and 20th centuries. At present there is no evidence to suggest that any of the extant hands date from the late 17th century.