**1. Introduction**

Metallic glasses (MGs) constitute an amazing example of a man-made non-crystalline state, which is not observed in nature. These materials are very promising in the sense of technological applications that results in the growing interest to the investigation of their physical properties. At first glance, since MGs do not have directional chemical bonding, one can expect that the understanding of their structure and properties should be a simpler task than for other types of glasses. However, any commonly accepted theory describing their formation and main structural features has been absent thus far, and any general theory of non-crystalline substances is still lacking as well. This largely constrains the development of new type MGs with the physical properties predicted in advance. Since MGs are prepared by melt quenching while the melt is obtained by fusion of the crystalline state, one should naturally expect a relationship between these states, including a connection between the properties of the glass and those of the maternal crystal (whose melt is used for the glass production). These interdependencies should be taken into account by a theoretical model of glass. On the other hand, since it is commonly accepted that glass is a frozen liquid, this model should also imply certain relation to the melting mechanism and formation of the liquid state. By that, it should include as a major ingredient the notions on structural defects, which are intrinsically related to the whole glass prehistory, i.e., maternal crystal→melt→glass. To date, quite a few theoretical models describing the structural features, defects and different properties of MGs have been suggested [1–12]. In our opinion, however, the above general requirements are satisfied by another approach—the Interstitialcy theory (IT) of condensed matter states suggested by Granato [13,14]. It was demonstrated in recent years that the IT provides a powerful tool for the understanding and predicting different relaxation phenomena in MGs and unambiguously shows a genetic relationship of the glass with the maternal crystal. Therefore, this brief overview is firstly devoted to an analysis of the major hypotheses of the IT and main experiments related to its verification. Since certain issues related to the IT have already been discussed [15,16], this work is largely devoted to the new experiments and their interpretations while earlier information is mentioned briefly. Finally, a relation of the IT to other models of the metallic glass structure and its defects is discussed.
