In 1867 G.H. Meyer published a paper called "Die Architektur der Spongiosa" in which he presented a line drawing of the cancellous bone structure he had observed in the proximal part of the human femur. Stimulated by this drawing C. Culmann constructed the principal stress trajectories in a crane-like curved bar loaded as the human femur. J. Wolff convinced that Culmann's crane exactly represented the cancellous structure of the human femur, made it the basis of his trajectory hypothesis and his "law of bone transformation". A reconstruction of Culmann's crane shows, however, that it is most probably based on a straight cantilever having a parabolic shearing stress distribution on its free end. To make the geometry more consistent with that of the proximal end of the femur, some curvature was added to the straight cantilever beam without an accompaning alteration in the stress distribution. In 1880 W. Roux developed the principle of functional adaptation and tried to combine it with Wolff's trajectory hypothesis to obtain a coherent theory of bone modeling. Although his theory is not entirely convincing, Roux's papers contain at least two basic ideas important in later theories:​ Apposition and resorption of bone as a biological control process and the dependence of this process on the local state of stress.