1. The receptor organs of the acoustico-lateralis system in fish respond in various ways to pressures and pressure gradients and provide the fish with information about external sources of vibration. 2. A fish's movements will set up pressures and pressure gradients and this poses three questions. (i) Can a fish obtain useful information from self-generated pressures and pressure gradients? (ii) To what extent do self-generated pressures mask signals from external sources? (iii) Can interactions between external and self-generated pressures and gradients in the acoustico-lateralis system give patterns of activity from the receptor organs which have special significance? 3. In herring (Clupea harengus L.) and sprat (Sprattus sprattus (L.)) measurements have been made of dimensions of various parts of the acoustico-lateralis system particularly of the subcerebral perilymph canal which crosses the head between the lateral lines. 4. Self-generated pressures produced by lateral movements of the head are antisymmetric, i.e. equal and opposite in sign on the left and right sides of the head. They oppose the accelerations of the head that produce them. In contrast, external sources give pressures that are largely symmetric. Any pressure gradients they give will accelerate the fish and the surrounding water together and any net pressure gradients will be small and so will any flows through the subcerebral perilymph canal. 5. Flows of liquid between the lateral lines across the lateral-recess membranes have been measured at various frequencies for pressure gradients applied across the head. Between 5 and 2000 Hz the velocity of flow per unit pressure does not vary by more than than a factor of 2. At low frequencies the absolute values of flow are very much larger (more than 50 times) than those found for equally large symmetrically applied pressures (as from an external source) due to flow into the elastic gas containing bullae. 6. It is calculated that a net pressure difference (at optimum frequency) across the head of only 0.008 Pa will reach threshold for the lateral line neuromast nearest the lateral recess and one of 0.02 Pa for that under the eye. The responses of these neuromasts are expected to saturate and provide little information when the pressure differences across the head exceed 6 to 18 Pa. The pressures given by the swimming fish are discussed in the light of a theory advanced by Lighthill in the paper that follows this paper. With such antisymmetric pressures the direction of flow in the lateral-line canals will be towards the lateral recess on one side of the fish and away on the other and so differ from the situation found with an external source when flow at any instant will be either towards or away from the lateral recess on both sides of the head. 7. Antisymmetric pressures can produce large flows past the utricular maculae. However, at low frequencies flows across the maculae, on which their stimulation depends, will be small. We do not know the direction of these latter flows though they will be in opposite sense on the two sides of the head, again unlike the situation with an external source. 8. Calculations of impedances below 30 Hz show that the observed flows across the head are consistent with the dimensions and properties of the known structures. 9. There are major and systematic differences in the patterns of receptor organ stimulation between those expected from external sources and from a fish's own movements. 10. Experiments on the red mullet (Mullus surmuletus L.) showed that it too has a transverse channel connecting the right and left lateral-line systems. At low frequencies its properties resemble those of the subcerebral perilymph canal of the clupeid.