In the Opening Remarks Dr. Kudritzki asked what tricks were used in calculating stellar structure. Though many fine models and discussions have been presented, answer has not been given yet to this question. The theory of stellar structure consists of two very much different kinds of building blocks. One is the local physics such as equation of state, opacity, nuclear reaction rate, and so on. They are determined only locally when the values of temperature and density at a point are given, i.e., they are irrelevant to the conditions whether they are considered in the stellar interior or not. Another is the physics of self-gravitating systems in which the global spatial structures are discussed. Characteristics of astrophysics are often said to lie in the fact that it covers much wider parameter ranges than in laboratories. This, however, grasps only one side of the problem, i.e., only local physics. Those which are not encountered in laboratory physics lie mainly in the global physics. It shows characteristics out of common sense; examples are negative specific heat and associated gravothermal catastrophe which are related with the gravitational contraction of the stars, formation of core halo structures which is observed not only in red giant stars but also many celestial objects, and tendencies dividing a system into two sub-systems one with high energy and/or high entropy and the other with low energy in the deep potential well and/or low entropy as seen in supernova explosion of type II and formation of jet in various objects. They behave contrary to the general trend, i.e., against equipartition and thermal equilibration.