During the last fifteen years the field of the investigation of glasses has experienced a period of extremely rapid growth, both in the development of new theoretical approaches and in the application of new experimental techniques. After these years of intensive experimental and theoretical work our understanding of the structure of glasses and their intrinsic properties has greatly improved. In glasses we are con­ fronted with the full complexity of a disordered medium. The glassy state is characterised not only by the absence of any long-range order; in addition, a glass is in a non-equilibrium state and relaxation processes occur on widely different time scales even at low temperatures. Therefore it is not surprising that these complex and novel physical properties have provided a strong stimulus for work on glasses and amorphous systems. The strikingly different properties of glasses and of crystalline solids, e. g. the low­ temperature behaviour of the heat capacity and the thermal conductivity, are based on characteristic degrees of freedom described by the so-called two-level systems. The random potential of an amorphous solid can be represented by an ensemble of asymmetric double minimum potentials. This ensemble gives rise to a new class of low-lying excitations unique to glasses. These low-energy modes arise from tunneling through a potential barrier of an atom or molecule between the two minima of a double-well.