On panel making and the post-1600 Holbein panels

Ian Tyers, Dendrochronologist

Making Art in Tudor Britain

Abstracts from Academic Workshops (2007-8)

At this stage in the Making Art in Tudor Britain project we have examined sixty-six panels with a view to identifying the types of timber used, and where appropriate to attempt to establish the date and source of the panel.

The summary results are as follows:

  Panels Boards
Not oak 2 2
Oak, not analysed 8 37
Oak analysed, not dated 5 13
Oak dated, English 8 (+1 Mixed) 9
Oak dated, East Baltic 41  73
Oak, dated, German 2 3
Total 66 137

It should be noted that it is possible that some of the five currently undated oak panels will be successfully dated as the project proceeds, and that some of the eight un-analysed oak panels may be analysed using different methods. 

Twenty-eight of the panels looked at had been examined in the 1970s and 1980s as part of a pioneering project on tree-ring dating for panel paintings. Their re-analysis has led to both an update of the earlier findings and the correction of some errors. It has also helped to put the results in the context of modern scholarship.

The above table demonstrates the significance of imported oak for the boards used in the construction of the National Portrait Gallery’s collection of English paintings of this period.  Most of these boards were derived from the Eastern Baltic area of Europe (modern Poland and countries to the east and north-east of there). This is an importation pattern that parallels work on panel paintings in the Netherlands, other English panel paintings, and non art-historical groups of boards used in England in the same period. We have also identified in the NPG collections several examples of the use of locally sourced oak boards, some usage of boards from what is now Germany, and at least three examples of the re-use of earlier oak boards. The NPG collection of panels is therefore diverse in a way that is typical of Tudor panel paintings. This talk attempted to summarise the methods of panel construction used then, as we currently understand them.

NPG 5268, reverse of the panel in raking light, with two sawn boards and one riven board, © National Portrait Gallery

NPG 5268
Reverse of the panel in raking light with two sawn boards and one riven board

Just as there are a wide variety of ways of using pigments, a diversity in the use and types of underdrawing, materials used in ground layers and methods of handling paints, so there is a similar diversity of ways in which someone could make a panel support in this period.  From a combination of analytical work on panels, and evidence from documentary research, archaeological and building timbers, it seems reasonable to assume that boards for panel making were selected on the basis of their availability, price, suitability for the size of panel, and conventions in terms of suppliers and use of materials. Most Tudor panels use riven or quarter sawn oak boards, which were then trimmed of sapwood, to make a board 250-300mm wide. These were then usually joined edge to edge, sometimes with pegs, to make panels wide enough for the desired image. Some very large panels have six or ten boards (for example the Holbein’s ‘The Ambassadors’ in the National Gallery has ten), although all the NPG panels so far have had one, two, three, or four boards per panel. The commonest size of portrait, 350-450 mm wide, typically used one wide and one narrow board, it is common for these to be derived from a single board; the riven Baltic boards being slightly wedge shaped can be sawn width-wise to produce two latter sections, one wider than the other.

It is unknown if it was a desirable feature, or is merely co-incidental, that the joint typically lies off the face of the sitter in this format. There is a tendency for the English boards to be wider than the Baltic boards, in fact the majority of the English boards identified so far are used in single board panels, and are hence without board joints (e.g. NPG 6855 Lady Dacre and Baron Dacre, NPG 4912 Southwell, NPG 2918 Clinton, NPG 5387 Foxe). It is difficult to interpret this observation; it might suggest these panels were made by people less confident in their ability to make permanent joints, it may mean English oaks were larger than typical Baltic traded oak boards, or it may imply that the selection process of local oaks for panel making tended to exploit very large oaks.

Whilst in timber terms the Baltic material is much the better raw material for easily producing flat panels it is not reasonable to suppose the same value judgements were applied at the time they were made. Although many of the best panels used Baltic oak there are examples of high class panels using unusually wide English boards such as Hans Eworth’s double portrait of the Dacres, the board for which is 500 mm wide running horizontally (NPG 6855).

Queen Elizabeth I
by Unknown English artist
circa 1588
NPG 541

There are also plenty of examples of poor copies of portraits that use Baltic boards. It is evident therefore that the patterns of board availability are complex and they are not necessarily the same throughout the study period and area. The NPG’s ‘Armada’ portrait of Elizabeth I, for example, uses both English and Baltic boards (NPG 541), a very rarely observed combination. It may be relevant that this was constructed at a period when Baltic boarding was perhaps more difficult to obtain due to the fact that it was a strategic supply subject to Spanish blockading.

A major topic of this workshop related to the copies of Holbein’s images, many of which are well documented and of exceptional quality. The dating evidence for these has been discussed in the other workshop contributions (see Plender, Foister and Cooper). The analysis of a number of copies of Holbein images in the NPG collection has shown that many are from the later sixteenth century and early seventeenth centuries (NPG 2809 Butts, NPG 5245 Kratzer, NPG 4912 Southwell, NPG 2094 Warham, NPG 1727 Cromwell).