This view that the differential features of emotions are not to be traced to the viscera is in accord with the experimental results of Sherrington,[222] who has demonstrated that emotional responses occur in dogs in which practically all the main viscera and the great bulk of skeletal muscle have been removed from subjection to and from influence upon the brain, by severance of the vagus nerves and the spinal cord. In these animals no alteration whatever was noticed in the occurrence, under appropriate circumstances, of characteristic expressions of voice and features, indicating anger, delight or fear. The argument that these expressions may have been previously established by afferent impulses from excited viscera was met by noting that a puppy only nine weeks old also continued to exhibit the signs of emotional excitement after the brain was disconnected from all the body except the head and shoulders. Evidence from uniformity of visceral response and evidence from exclusion of the viscera are harmonious, therefore, in minimizing visceral factors as the source of differences in emotional states.[*] *[Footnote: The paucity of afferent fibres in the autonomic system, and the probability of an extremely low degree of sensitiveness in the viscera (for evidence, see Cannon: The Mechanical Factors of Digestion, London, 1911, p. 202), likewise support this conclusion.] If these differences are due to other than visceral changes, why is it not always possible by voluntary innervations to produce emotions? We can laugh and cry and tremble. But forced laughter does not bring happiness, nor forced sobbing sorrow, and the trembling from cold rouses neither anger nor fear. The muscle positions and tensions are there, but the experiencing of such bodily changes does not seem even approximately to rouse an emotion in us. Voluntary assumption of an attitude seems to leave out the “feeling.” It is probable, however, that no attitude which we can assume has all the elements in it which appear in the complete response to a stirring situation. But is not this because the natural response is a _pattern reaction_, like inborn reflexes of low order, such as sneezing, in which impulses flash through peculiarly coöperating neurone groups of the central system, suddenly, unexpectedly, and in a manner not exactly reproducible by volition, and thus they throw the skeletal muscles into peculiar attitudes and, if sufficiently intense, rush out in diffuse discharges that cause tremors and visceral perturbations? The typical facial and bodily expressions, automatically assumed in different emotions, indicate the discharge of peculiar groupings of neurones in the several affective states. That these responses occur instantly and spontaneously when the appropriate “situation,” actual or vividly imagined, is present, shows that they are ingrained in the nervous organization. At least one such pattern, that of anger, persists after removal of