We propose that the key s tural feature in the conversion of chemical free energy into mechanical work by actomyosin is a myosin-induced change in the length of the actin filament. As reported earlier, there is evidence that helical actin filaments can untwist into ribbons having an increased intersubunit repeat. Regular patterns of actomyosin interactions arise when ribbons are aligned with myosin thick filaments, because the repeat distance of the myosin lattice (429 A) is an integral multiple of the subunit repeat in the ribbon (35.7 A). This commensurability property of the actomyosin lattice leads to a simple mechanism for controlling the sequence of events in chemical-mechanical transduction. A role for tropomyosin in transmitting the forces developed by actomyosin is proposed. In this paper, we describe how these transduction principles provide the basis for a theory of muscle contraction.The central problem of muscle contraction is to account for the mechanism offorce generation between the thick and thin filaments that constitute the sarcomere (1-6). In Huxley's analysis (3,5,6), the proportionality of isometric tension to the overlap of thick and thin filaments, as well as the other mechanical properties of muscle, are neatly explained with so-called independent force generators that are presumed to be distributed uniformly along the overlap zone. In most models, the individual force contributions of these generators are summed up by some relatively inextensible structural element to which each generator has a single point of attachment. In the case of the classical rotating cross-bridge model of force generation, thin filaments themselves serve the role of the inextensible structural elements. In this note, we propose an alternative model in which repetitive length changes in segments of actin filaments, induced by myosin heads, generate forces that are summed and transmitted to the Z disc by tropomyosin.The impetus for this model came from observations that actin in profilin-actin crystals is organized around a ribbon motif (r-actin), which we interpret to be an untwisted, extended form of the helical filamentous form of actin (h-actin) (7). In the transformation of actin from the helical to the ribbon form, the actin subunits undergo a 130 rotation as the segment extends by 8.25 A per subunit. It is presumed that the actin-actin intersubunit contacts are largely maintained during this transformation.The conclusion that r-actin and h-actin are related structures is based on a number of observations. First, growth of the r-actin form in the profilin-actin crystals appears to depend in a precise way on subtly shifting the competition for actin monomers away from h-actin formation toward entry into the growing crystal lattice. This strongly suggests that the same actin-actin bonds are used in both r-actin and h-actin. Second, x-ray diffraction shows clearly that crystals can be transformed into semicrystalline fibrous assemblies by replacing the bathing medium with solutions known to induce...