The conceptual difficulties in the diagnosis of depression, and the study of affect generally, hinder a clear approach to the "organic" side of depressive illness. While autonomic, biochemical and electrophysiological disturbances follow a general pattern in severe retarded depressions, correlation of these with the variable somatic symptomatology is only partial and demarcation from other psychiatric syndromes vague. Various attempts have been made to relate the patterns of autonomic reaction to thought content on the one hand and to central nervous system activity and organization on the other. In recent years, fragments of knowledge about physiological changes accompanying affective experiences in infancy have appeared. Isolated descriptions of what appears to have been the replacement of depression by somatic symptoms with or without somatic pathology have not yet been organized into a coherent whole. In this article the various studies and opinions relevant to the above conclusions have been reviewed and the findings discussed.