The development of the singularity approach of fracture mechanics is at its dead end because it is not possible to describe real failure at the crack boundary and to replace the real failure criteria by general energy conditions and the method remains empirical. Therefore the theoretical approach based on the elliptical flat crack has to be followed, leading to the possibility to derive and explain the empirical mixed "mode 1-11" interaction equation. Because it is shown that the singularity approach does not apply for wood, the theory is based on the flat elliptical crack. This book examines a new fracture mechanics theory of wood. Further discussed: the derivation of the power-law; the energy method of notched beams and of joints loaded perpendicular to the grain; the necessary rejection of the applied crack growth models and fictitious crack models and the Weibull size effect in fracture mechanics.