M. J. S. Rudwick and others postulate ‘rhythmic-flow’ feeding for the Permian richtho-feniacean bi-achiopods, whereas R. E. Grant claims that they fed by normal ciliary action. Suspension-feeding has two components, current generation and food capture; normal brachiopod lophophores do both, but this is neither universal nor compulsory among animals. Opening and closing the richthofeniid shell generated a ‘tidal-flow’ current precisely analogous to respiratory currents in mammals; this is neither inefficient nor ‘self-defeating’, as Grant claims. Grant's analysis fails because he chose the wrong mechanical analogy (a pump). Morphological features of richthofeniids are better explained on a tidal-flow hypothesis than on a ciliary-flow model, and all the data adduced by Grant in rejecting the former is compatible with it or favorable to it. It explains morphological features that are bizarre mysteries on the ciliaiy-current model, and is therefore superior even though it implies that these Permian brachiopods were radically innovative.