In Denmark, dendrochronological dating is carried out almost exclusively on oak wood. An analysis is carried out on a cross-section (if necessary a core) of a wood sample, on which radial tracks have been prepared using a specially-mounted razor blade. Measurements are made using a microscope (10-40X magnification), in conjunction with a measuring device to collect the data. The rings in an individual sample are normally measured twice, preferably along two separate radii. The tree-ring curves for each radius are plotted to provide a visual check on the measurements for each sample. The curves are corrected/adjusted as necessary, then averaged to give a single curve which represents the sample. The curves from several samples are the synchronised relative to each other to give one or more site-chronologies. Both single and site-chronologies are then dated with the help of existing regional reference chronologies: master-chronologies. The Dendrochronology Laboratory at the National Museum's Science Research Unit has developed a master chronology complex (several local chronologies) for oak, which extends back from the present day to c. 100 BC. In addition, the laboratory has access to the majority of regional master chronologies for oak in northern Europe. The computerprogramme CATRAS is used for measurements, synchronisation, averaging, plotting, editing etc.
Ideal conditions for the successful execution of a dendrochronological analysys are rarely met. Both the number of samples in an investigation and the number of annual rings in the individual samples are of crucial importance. The main requirements are that each of the analysed samples contains 100-150 annual rings and that a master chronology exists for the area in question. A master chronology is built up on the basis of extensive measurements on living trees and construction timber and the like from the area. The dating of a single sample does not give a reliable date for the whole construction (be it church, house, castle, ship or whatever). There may have been re-use of timber, repairs etc. On the contrary, if there are many samples from the same construction, which the dendrochronological analysis show have contemporary felling dates, it is very likely that the trees were felled ad hoc and used immediately. Furthermore, it is then possible to take account of possible re-used timber, repairs, building phases and the like.
Acknowledgements Prof Dieter Eckstein and Dr Sigrid Wrobel, University of Hamburg, Dr Hanns Hubert Leuschner, University of Göttingen, Dr Burgart Schmidt, University of Cologne, Dr Tomasz Wa ny, Academy of Fine Arts, Warsaw, University of Lund, Prof M.G.L. Baillie, Queens University Belfast and Dr Ian Tyers, Sheffield University have kindly provided The National Museum of Denmark with the master chronologies produced in their laboratories.