Thanks to the valuable works by the brain scientist A. R. Damasio (1994, 1999) that underlined the bodily bases for feelings, the correlation between the feeling and the body is receiving much attention today. In this paper I first discussed the significance of his theory as well as Wallon's theory of feelings (1932), which has so far been paid comparatively little attention in the English-speaking sphere. Next I argued that Wallon's theory of feelings is basically a theory of mutually coordinating interaction between center and periphery, and that it is quite different from James-Lange or Cannon-Bard theories. Then I proceeded to consider the various phenomena that I called "not corresponding feelings" and regarded as manifestations of what Wallon discussed in the theory, and classified them under four categories, in the hope that the classification will serve as a theoretical basis for future case-gathering efforts. The first of the categories I called "the feelings toward undefined subjects", under which come Freud's "anxiety", Damasio's "background emotions", and the emotional effects relevant to "the postural functions" that Wallon discussed. The second category was called "remaining and alteration". Anderson's study on the primacy effect (1965), the perseverance effect by Ross, Lepper and Hubbard (1975), and the non-effect of mitigating information by Zillman and Cantor (1976) are the instances of "remaining" I gave, and the suspension-bridge experiment by Dutton and Aron (1974) and the sexual attraction study by Dienstbier, R. A. (1979) come under "alteration". The third was called "strengthening and weakening". "strengthening" was seen, for instance, in a phenomenon when a laughter (or crying) is so compulsive it cannot be stopped, but I could not identify any studies on this subject. An example of the "weakening", on the other hand, was found in the survey by Bowlby (1979) on convalescent processes from the loss of a parent or spouse, which indicates crying is effective in alleviating grief. My argument is that the "weakening" phenomena is contradictory to both the James-Lange theory and the Cannon-Bard theory and can only be explained by the theory of mutually coordinating interaction. The fourth is "deviant expression", which includes such phenomena as tears out of anger or laughter, laughter out of extreme fear, and aggressive behavior as expression of joy, the last of which may be illustrated by some of the behaviors of the C type children in Ainsworth et al.'s SS method. Very few studies have been done on the subject, leaving a great deal of room for future research. I gave some examples of "deviant expression" I observed among children at nursing school. My overall conclusion is that theories on feelings today should be some kinds of revised two-factor theory-they, while recognizing the role of cognition in the formation of emotion emphasized in the two-factor theory by Schachter and Singer (1962), the theories of cognitive appraisal by Roseman et al. (1990), Smith and Lazarus (1993) and Scherer (1984, 1992, 1993), and the theory of secondary emotion by Damasio (1999), should also take the theory of mutually coordinating interaction of Wallon (1932) fully into consideration.