T HE TERM "psychosomatic" is a clinical, descriptive one which refers to the appearance of physical symptoms in relation to emotional states or mental activity. From a dynamic viewpoint it seems grossly inadequate. In the first place, it indicates little about the particular patient before us in analysis. Secondly, it tells nothing about the psycho-somatic relationship, and thirdly it lumps together, indifferently, manifestations of quite different order and dynamic process. Can we equate, for instance, signs of direct body participation in an affective state--e.g., the crying of joy, the tachycardia of rage, or the pallor of humil ia t ion--with such complex symptoms as aia hyster ica l paralysis? Or such symptom-complexes as asthma or tuberculosis with the somatic symptoms arising out of a hypochondriacal conviction of illness? In one case the somatic symptoms may be the physiological component of a healthy, affective state of simple order; in another the affect may be compulsive or irrational. Or anxiety may be involved. Generally, this is considered an affect, yet resulting from the interplay between "simpler" emotions, so that although a large volume of experimental work may give factual evidence of particular somatic changes being "produced" by particular affects, it does not help too much in understanding the patient, whose emotional states are more complex, or unconscious, or cannot be easily identified either by patient or observer. Still other somatization reactions, such as the "conversion" symptoms, may not only involve affects but conceptualideational complexes in conflict. And even more complex clinical syndromes, such as asthma or tuberculosis, have been related to personality types or "profiles" or groups of character traits. In these cases not only is the psychosomatic relationship often most strained, but the origin and nature of the psychological state is difficult to explain. Most attempts to explain these symptoms have focused on the nature or type of affective-ideational factor, and on the relationship of such particular condition to a type of somatic response. They assume an interaction between more or less definable psychic and somatic events. These theories have been well reviewed in a recent article 1 and it is unnecessary to detail them here. Suffice it to say that the degree of relatedness has been seen to vary from a close, precise specificity (affect --> symptom, personality --> syndrome) to a limited specificity (non-specific anxiety --~ symp-