The experimental investigations that first showed excess of sound absorption in gases above the classical value are described, and the theoretical ideas that lead to their explanation by relaxation processes are discussed. A survey is then given of later progress, which is closely connected with the rapid development of electronics, and of the parallel refinement of the theory. The work in liquids started somewhat later. The theoretical situation is more complex, but, for that very reason, ultrasonic measurements in liquids give information about problems absent in gases. The consideration of viscous liquids led to a connection with the behavior in glasses, in accordance with Simon's theory. In metals, the starting point, the methods, and the important frequency range for loss measurements were originally different; but here, too, the investigations and theoretical viewpoints have been multiplied and diversified.