Cytoplasmic streaming in characean algae is thought to be driven by interaction between stationary subcortical actin bundles and motile endoplasmic myosin. Implicit in this mechanism is a requirement for some form of coupling to transfer motive force from the moving myosin to the endoplasm. Three models of viscous coupling between myosin and endoplasm are presented here, and the hydrodynamic feasibility of each model is analyzed. The results show that individual myosinlike molecules moving along the actin bundles at reasonable velocities cannot exert enough viscous pull on the endoplasm to account for the observed streaming. Attachment of myosin to small spherical organelles improves viscous coupling to the endoplasm, but results for this model show that streaming can be generated only if the myosin-spheres move along the actin bundles in a virtual solid line at about twice the streaming velocity. In the third model, myosin is incorporated into a fibrous or membranous network or gel extending into the endoplasm. This network is pulled forward as the attached myosin slides along the actin bundles. Using network dimensions estimated from published micrographs of characean endoplasm, the results show that this system can easily generate the observed cytoplasmic streaming.Actin-myosin interactions are thought to be responsible for the generation of many types of biological motility, including muscle contraction, amoeboid movement, cytoplasmic streaming, phagocytosis, cell division, and axonal transport (1). Although it is now generally understood how the interaction between actin and myosin leads ultimately to the motion (contraction) of muscle tissue, in most nonmuscle systems the details of motion generation are not yet understood (1). In particular, detailed mechanisms are lacking for the hypothesized actin-myosin-driven flow of intracellular fluid in amoeboid movement, cytoplasmic streaming, and axonal transport. Implicit in these proposed actin-myosin systems is the assumption that the sliding of myosin molecules past actin bundles is somehow coupled to the intracellular fluid to produce a bulk flow.The purpose of this paper is to examine, from a hydrodynamic viewpoint, the feasibility of various supramolecular structures that might function to provide coupling between actin-myosin sliding and bulk fluid flow. The hydrodynamics presented will be modeled for the geometry of cytoplasmic streaming in characean algae. Various microscopic, ultrastructural, and biochemical observations (2-5) have provided evidence that this streaming is driven by actin-myosin. Although other mechanisms have been proposed (6), most available evidence (3) suggests that characean streaming is driven by interaction between motile endoplasmic myosin and stationary subcortical actin bundles (Fig. l). The highly organized pattern of streaming in characean internodal cells makes this a convenient system for hydrodynamic modeling. Some of the results obtained for this system may then fred applications in other, less organized, nonmus...