Summary From the foregoing data, certain suggestive and still tentative interrelationships seem apparent. It appears clearly established that an emotional component was operative in all of the twelve cases seen in this series. The evidence for such a component was based on the observed or historically inferred correlation between the appearance of spontaneous or experimentally induced angry or aggressive emotions and an immediate activation of a hypermotile colonic response leading to tenesmus or diarrhea. In all but one case, a precipitating event or combination of circumstances was demonstrated in relation to the onset of symptoms, occasioning simultaneously the activation of previously repressed, wholly unacceptable negative emotions and a sweeping loss of emotional support. In the group of sixteen children reported here, an apparently inverse ratio existed between personality strength or integrative capacity and the rapidity and severity of onset. In general, a more direct relationship appeared to exist between such integrative capacity and the intensity of stress required to precipitate a somatic response. Other factors, such as gastrointestinal or parenteral infection, allergy, and dietary alterations, appeared to play a predominant role in occasional exacerbations of symptomatology. During the course of psychotherapeutic play and verbal interviews, a reciprocal relationship has been suggestively demonstrated between therapeutic release of pathogenic angry or hostile emotions and the intensity of the colonic hypermotile response. The sources of the pathogenicity of such emotions for children with such dependent, rigid, immature personalities were discussed in relation to early experiences within the family constellation. A genetic component in the predisposition to disturbances of gastrointestinal function has been tentatively implicated as well, possibly involving the inheritance of specific patterns of autonomic response to emotional stress.