On the basis of data in the literature and findings in the present study, it is clear that emotion may be recognized by patients as an intrinsic part of their attacks, and not as a reaction to the occurrence of the attack itself. The emotion is stereotyped in that it is constant in both quality and time of occurrence during the attack. The quality of the emotions varies from patient to patient, and occasionally sequential alterations of emotion occur in the same individual. Fear or sensations of anxiety are reported most frequently. Ill-defined unpleasant sensations, in many instances akin to fear, also are common. Depressive symptoms are relatively rare. On the contrary, a minority of patients reported pleasurable experiences.Evidence appears overwhelming that in the majority of patients the discharges associated with these seizures arise in the temporal region. In many instances there was convincing evidence of origin in the depths of the temporal lobe. Uncinate hallucinations and vegetative phenomena occurred either singly or in varying combinations in 38, or almost 75%, of the patients. Both these epileptic manifestations are indicative of discharges deep in the temporal region. Masticatory attacks, for which Magnus and co-workers(17) have postulated that the amygdaloid nucleus is the efferent mechanism, were also relatively common. This impression was further supported by the anatomic location of the verified vascular anomalies, atrophic lesions and the tumors. Thus, evidence favors the fact that the deeper temporal regions, areas either intimately associated with or an actual part of the rhinencephalon, are involved in these seizures.It will be recalled that in addition to the temporal area, certain supracallosal structures in the medial parasagittal regions also are parts of the so-called limbic system. Examples of affect associated with lesions in this area are rare. Erickson's(25) patient is one instance. As was previously mentioned, one patient in my series had an oligodendroglioma located on the mesial surface of the frontoparietal parasagittal region. This patient experienced attacks which would begin with a sensation of fear and a rising epigastric feeling. Therefore, it seems likely that the so-called limbic system is involved in ictal affect.These conclusions are in accord with experimental evidence concerning the functions of this region. In 1933, Herrick(30) suggested that the rhinencephalon might have activities other than those related directly to olfaction. He wrote thatAn important function of the olfactory cortex, in addition to participation in its own specific way in cortical associations, is to serve as a non-specific activator for all cortical activities.This type of non-specific activity is one of the major functions of the olfactory cortex.. . . Having no localization pattern of its own, it may act in two ways: first upon other exteroceptive systems whose localized mechanisms are adapted to execute adjustments where external orientation is demanded, and, second, upon the internal apparatus of general bodily attitude, disposition and affective tone.In 1937, Papez(31) proposed a concept of the anatomic substrate of emotion based on purely theoretic grounds. He suggested that the hypothalamus, the anterior thalamic nuclei, the gyrus cinguli, the hippocampus and their interconnections constitute a harmonious mechanism that may elaborate the functions of central emotion, as well as participate in emotional expression.Almost simultaneous experimental confirmation of these general concepts was made by Klü and Bucy(32), who studied the effect of resection of both temporal lobes in monkeys. These authors noted profound behavioral changes in the monkeys, including pronounced placidity and a "psychic-blindness" that was characterized in part by inability to recognize fearsome stimuli. Great changes in the sexual behavior of the animals also were present.Spiegel and co-workers(33) reported states of rage in cats with experimentally produced lesions in the region of the olfactory tubercle and the amygdala. Bard and Mountcastle(34), in an elaborate study of cats observed over a prolonged period, reported that extensive ablation of the isocortex was not associated with any striking emotional alteration; however, subsequent destruction of either the region of the amygdala or the cingular zone was associated with the development of "sham-rage" reactions. If the isocortex was left undisturbed, then only lesions in the region of the amygdala were effective in the production of ragelike states.These observations have been directly opposed by the reports of Schreiner and Kling (35), who described profound placidity combined with hypersexuality in carnivores after destructive lesions were produced in the region of the amygdaloid nucleus. The cause of these discrepancies is not clear and further investigation of this problem is indicated. In any event, regardless of the character of the emotional disturbances, all authors agree that the affective behavior and responses of animals to external stimuli are greatly altered by lesions in this region.Considerable attention has been devoted recently to elaboration of these animal studies by methods of chemical and electric stimulation. MacLean(36) has observed "fearful alerting" after stimulation of the hippocampal area in animals. Shealy and Peele(37) report that in cats, stimulation of the amygdaloid nucleus produces either frightened behavior, undirected rage or an "alert" state. In man, stimulation of this area has caused fear, anxiety or "weird feelings"(38). On the basis of an extensive review of the literature, Pribram and Kruger (39) have proposed subdividing the rhinencephalic structures into 3 systems, and have formulated hypotheses relating to their functions. Further experimental studies will be necessary for evaluation of these concepts.The massive visceral accompaniments of affective states have led some(28) to equate rhinencephalic functions with visceral function and to use the term "visceral brain." The implication exists that emotion is basically visceral in origin. Such a conclusion is more readily justified on the basis of animal studies than on the basis of the experiences of human beings with seizures. In man, there is plentiful evidence that during minor seizures extensive autonomic changes may occur entirely without emotional accompaniments. As was previously pointed out, at times there is a discrepancy between the quality of the emotion and the character of the vegetative events. Thus, while emotion and vegetative states may proceed pari passu, they are not in one-to-one correspondence. Vegetative activity is not the skeleton of emotional sensation. Pribram(40), on different grounds, has expressed a similar opinion and propounded a broader concept in which autonomic function plays a varying role as one of several variable but interrelated factors producing emotion.Williams(9) has suggested that there is a certain topographic localization of emotion in the temporal region. This concept has the appeal of symmetry, but on theoretic grounds it is hard for me to accept the concept of "centers" for fear or for pleasure in such a phylogenetically ancient system. An alternate possibility is that the patterns of excitation in this general area, with associated widespread alterations in physiologic levels of activity of both the central nervous system and the internal milieu of the body, may bring about the state interpreted as "emotion."It is obvious that any attempt to develop a theory of the physiology of emotion, in which the behavior of experimental animals is used for analogies, founders on the absence of subjective reports of the emotional states in the animal. Furthermore, one must be cautious in attempting to extrapolate on the basis of epileptic events. Certainly, in the majority of patients, ictal emotions are meager and crude. They lack the profusion and subtlety of normal emotions. The patient is not deceived: these are experiences happening to him. They are as involuntary as the jerking of a hand during a focal motor seizure. In other focal seizures there is a derangement or fragmentation of normal cortical activity, so that the epileptic phenomena constitute a distortion of the normal function of that area. Just as the "clotted mass of movement" in a jacksonian seizure may give no clue as to the complexity of motor patterns latent in the cortex, so also may the emotion in an affective seizure only hint at the richness of function in these areas.In view of such reservations, what do these observations tell us concerning emotion? The frequent concomitant occurrence of affect, vegetative phenomena and olfactory and gustatory sensations underlines their importance in activities essential to the preservation of the individual and to the maintenance of the species. I tender the suggestion that emotion exerts a selective function in emphasizing and coloring incoming stimuli. In a sense, it filters present experience in the light of the past—encouraging certain actions, denying others. By its relatively long duration, it possesses an executive capacity, providing continuity to the activity of the individual and orienting future acts in terms of the present. It is the archer of time's arrow.