The hypothesis is examined that short-term creep response of earlywood and latewood tissues of some coniferous species, stressed in tension parallel to the grain, is a function of microfibril angle of the S 2 layer of and relative degree of crystallinity in the tracheid cell wall, along with specific gravity of that wood tissue and its extractives content. A new technique was developed to measure the total creep that occurred over a 60-minute period of time for small specimens (nominally 0.010 in, thick) of Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco) (normal and compression wood), Sitka spruce (Picea sitchensis (Bong.) Carr.) and western hemlock (Tsuga heterophylla (Raf.) Sarg.), taken from earlywood and latewood zones of the same increment. Total creep was determined at two initial deformation levels, 3,000 microin.per in. (strain level No. 1) and 6,000 microin. per in. (strain level No. 2). Microfibril angle was measured by a modified mercury impregnation method, while cell wall crystallinity was determined on small, unoriented pellets by the X-ray diffraction technique. Air-dry specific gravity (oven-dry weight) and alcohol-benzene plus hot water extractives were also determined by conventional methods. Multiple regression analyses were carried out and prediction equations, based on the experimental results, have been constructed. It is shown that the variability in total creep response can best be explained by using the prediction equation which contains microfibril angle of the S2 layer, specific gravity and extractives content. The multiple coefficients of determination (R²) using this subset of variables are 0.7680 and 0.8550 for initial strain Numbers 1 and 2, respectively. Cell wall crystallinity was eliminated from the prediction equations as the least important variable due to its high inverse correlation with the microfibril angle of the S2 layer (r=0.9284). Two possible reasons are suggested to explain this correlation. First, in the case of a small angle, the scatter around the mean microfibril angle is smaller and the microfibrils probably lie almost parallel to each other. As a result, the relative degree of amorphous material required to fill the micro-spaces between microfibrils would be smaller. Considering the case of a large microfibril angle, the microfibrils are probably not parallel to each other; consequently, relatively large micro-spaces would be occupied by the amorphous material. A second possible reason for this relationship may be that cellulose chain molecules, in the case of a small microfibril angle, will have a better chance for increased frequency of cross links (bonding between neighbouring chains) along their unit length. Consequently, a tendency of improved geometric order should be observed with better chain coherence in the resulting cellulose as compared to situations associated with tracheids characterized by larger microfibril angle. It must be indicated that reasons for this high degree of correlation, as noted above, remain conjectural. Among the structural features studied, microfibril angle was shown to control creep response to the greatest extent. As it increases, total creep increases, the reason being that with a small angle, microfibrils are in a position to bear most of the applied load and therefore their relative movement towards a smaller angle would be less. This results in a small plastic deformation. In the case of a large angle, there is a possibility that the microfibrils have a large tendency to move to a smaller angle causing a large creep response. Wood samples of low specific gravity creep more than those with high specific gravity. This behavior is explained by the higher relative per cent of the S2 layer in the latter. Extractives are shown to contribute significantly to the variation in total creep. They probably act as plasticizers causing a reduction in the primary and secondary bonding between microfibrils. This would facilitate the movement of the stiff inextensible microfibrils to accommodate the creep-inducing stresses. Results obtained in this study were compatible with the proposed hypothesis. / Forestry, Faculty of / Graduate