In order to understand what part this indecision of the ear performs in the feats of the ventriloquist, let the reader suppose two men placed before him in the open air, at the distance of one hundred feet, and standing close together. If they speak in succession, and if he does not know their voices, or see their lips move, he will be unable to tell which of them it is that speaks. If a man and a child are now placed so near the auditor that he can distinguish, without looking at them, the direction of the sounds which they utter, that is, whether the sound comes from the right or the left hand person, let the man be supposed capable of speaking in the voice of a child. When the man speaks in the language and the accents of the child, the auditor will suppose that the child is the speaker, although his ear could distinguish, under ordinary circumstances, that the sound came from the man. The knowledge conveyed to him by his ear is, in this case, made to yield to the more forcible conviction that the language and accents of a child could come only from the child; this conviction would be still further increased if the child should use gestures, or accommodate his features to the childish accents uttered by the man. If the man were to speak in his own character and his own voice, while the child exhibited the gestures and assumed the features which correspond with the words uttered, the auditor might be a little puzzled; but we are persuaded that the exhibition made to the eye would overpower his other sources of knowledge, and that he would believe the accents of the man to be uttered by the child: we suppose, of course, that the auditor is not allowed to observe the _features_ of the person who speaks.