It is interesting to review the types of energy transfer process shown by different systems. By energy transfer, I mean the coupling mechanism between two parts of the system. I shall not go into the question of energy migration, which involves a knowledge of the relaxation around the excited unit of the crystal. The rare-gas crystal is a very simple molecular crystal, although, as Dr. Knox has shown, it may not be so simple electronically as some organic crystals like anthracene. Inorganic ionic crystals have a range of binding types much wider than organic molecular crystals do, and at the same time there is a much wider variety of energy transfer phenomena. It will be interesting here to present several examples from these different categories and to compare them. A classification scheme which is useful in making these comparisons is to divide the examples into those having overlapping and those having nonoverlapping charge distributions (1). In the second class, we may have spherically connected or not spherically connected charge distributions. By this is meant that a sphere of minimum size can be drawn about each charge distribution, and these spheres may or may not intersect. The case where the charge distributions do not overlap and are not spherically connected is the easiest one to treat mathematically. In this case, the interaction energy reduces to that of a set of classical point multipoles, as long as exchange can be neglected. This state of affairs seems to be a relatively good description of the organic molecular crystal. Davydov splitting can be described in terms of dipoledipole (2) or octupole-octupole coupling (3), and intensity anomalies due to the crystalling state can be similarly accounted for-with the addition of the dipoleoctupole interaction (4) in some cases. Some molecules are also able to be treated as a collection of nearly nonoverlap-