) wood, and their orientation measured in relation to the longitudinal axis of tracheids in order to obtain information on the orientation of the cellulose microfibrils in wood cell walls. All observations supported the view that cavities are aligned along the microfibrils. Observations at low magnification of a large number of tracheids in radial section or on a large number of tracheids from the same annual ring suggested that most cavities were oriented at a similar angle to the tracheid axis. More detailed observations at higher magnification revealed, however, that there was a considerable variation within the S2 layer, particularly in earlywood tracheids. Cavities in cell corners of earlywood tracheids were often oriented parallel to the longitudinal axis of the tracheids. The majority of the cavities in the S2 spiralled in a Z-helical arrangement, but S-helices were also occasionally observed. Cavities observed in pit free areas of tracheids were often oriented at different angles to the tracheid axis and sometimes even crossed one another. Observations suggest that there is a large variation in microfibril orientation even within one tracheid. The orientation of cavities was also used for measuring microfibril orientation in two cross-sections cut from different heights of a pine tree. The mean values for the microfibril angle of earlywood tracheids at both heights were smaller in the tangential than radial walls. The microfibril angle decreased from pith to the bark in both of the cross-sections.