Dendrochronology in Dating Timber Framed Buildings and Structures

Jonathan Taylor

 

Tree rings: each ring represents one year's growth. A good year for growth will be recorded by a wider ring in all trees affected, whatever their age, like an annual date stamp. The outermost ring records the year that the tree was felled.

Dendrochronology, or 'tree ring dating' as it is often known, can provide an invaluable insight into the history of a building by revealing the year in which the timbers used in its construction were felled.

It was discovered early in the 20th century that trees of the same species in the same region displayed remarkably similar ring patterns across the tree trunk and in the end grain of timber beams. Each year a tree gains another ring as it grows; the thickness of which depends on the amount of growth. In a year with ideal growing conditions, trees will produce a wider ring than in a year with poor conditions, and all the trees in the same region are likely to display the same general chronological growth pattern, despite any local ecological variations. By plotting the relative thickness of these rings in a newly felled oak of say 200 years old, a clearly identifiable sequence of variations will emerge like a date stamp for each period. By comparing variations in the first 100 years growth (ie the innermost 100 rings) with those of the last 100 years growth (ie the outermost 100 rings) of similar timber felled locally 100 years ago, the match should be immediately apparent. If the older timber retains its bark, the year that it was felled will be recorded by the outermost ring, the ring which was grown in the year that the tree was felled.

Tree ring data for most areas of the country are now documented by master chronologies spanning hundreds of years, based on timbers of the same tree species, from the same region, with overlapping periods of growth. Oak is the best documented species because it was the one most widely used for the construction of timber-framed buildings in the past. By cross-matching the tree rings of historic timbers from existing buildings with the master chronology, dendrochronology laboratories are able to determine when the timbers were felled.

The appeal of dendrochronology as a dating tool is that it is objective and entirely independent of other evidence such as datable design features and documented information. Furthermore, where analysis results in a clear match with the master chronology the results are completely accurate and reliable.

However, not all buildings can be dated by dendrochronology. A project to examine the medieval timber-framed buildings of Kent (1) which was established in 1986 examined 74 buildings across the county and firm results were obtained for 53 of them. Nevertheless, data from the study of these buildings established clear dates for certain features such as joint details and mouldings which are known to change chronologically, from which it was possible to gain a much clearer picture of the development of all the buildings in the study.

For tree ring analysis to produce an accurate result, it is necessary to have samples of timber which retain their bark, so that it is clear which ring was outermost when it was felled. If only some sapwood remains (this is the outer layer of timber which lies beneath the bark and transports sap), the year in which the tree was felled can be estimated, probably to within 15 to 20 years. This is because the number of rings in the sapwood varies widely, with some estimates suggesting that the range may be from 15 to 50 rings in the sapwood of mature oak trees in 95 per cent of the cases considered in the UK. In Kent, 18 samples of oak taken from different medieval timbers indicated a smaller range of 15 to 35 sapwood rings.

TAKING SAMPLES

In practice, samples are prepared by the dendrochronologist either from cores drilled out of the timber or, if the timber is to be replaced as a result of a repair, by taking complete slices through the whole timber. Core samples leave a hole of up to 15mm diameter. The hole may be filled with a dowel to disguise it, but nevertheless some damage is done to the timber. Slices on the other hand are far more destructive but give the most clear picture of tree ring sequences.

Samples should include the bark, and finding suitable timbers for examination may require dismantling part of the structure. In some cases it may be possible to carry out some analysis from exposed beam ends in situ.

It becomes progressively easier to date timbers the more rings there are and the more samples taken. English Heritage recommends (2) that generally a minimum of 50 rings should be present in each sample and that eight to ten timbers should be sampled per building or per phase of the building's development, with no more than two core samples taken from the same timber to avoid unnecessary damage.

The samples are first polished. Then the dimensions of each ring are measured under a microscope and the results recorded on both a graph and on a computer for statistical analysis. All the samples are then cross-checked with each other to identify any possible measurement errors and abnormalities before a master curve is prepared based on average tree ring sequences.

LIMITATIONS

• Not all timbers used in timber framed buildings and roofs are of oak. Elm and other species were also used. If the only master chronology available for the region is oak, cross-matching with timbers of these other species cannot be relied on.
  
• Where trees were felled at a relatively young age there may not be enough rings to cross-match accurately.
  
• Sapwood is highly susceptible to decay particularly by beetle larvae. As a result all sapwood may have been removed from the accessible surfaces of timbers during building repairs and conservation work, making it impossible to determine when the timber was felled.
  
• The date the timber was felled may not necessarily be the date that the building was constructed. A surprisingly early date may suggest the use of salvaged timbers.
  
• Occasionally no cross-match will be identifiable. This may be for a number of reasons such as the use of timber imported from another area or unusual growth conditions caused by pollarding for example.
  
  

Notes

(1) Sarah Pearson, The Medieval Houses of Kent, RCHME, HMSO, London, 1994 (the book describes a thorough dendrochronology programme and, in some detail, the results obtained)

(2) English Heritage, Dendrochronology: Guidelines on producing and interpreting dendrochronological dates, English Heritage Publications, 1998