Let me illustrate this a little more fully. Suppose you desired to tell a Chinaman, who spoke not a word of English, to fetch a certain object from the next room. It would be useless for you to say "watch," because he would not know what the word meant. Probably you would tap your waistcoat pocket, pretend to take out a watch, wind it, look at the hands, etc., in your endeavour to convey to him your meaning. If this was not recognized, for any reason, you would have the utmost difficulty in conveying your meaning to him--and equal difficulty in telling him to fetch the watch from the next room. Now, suppose these antics--or somewhat similar ones--were resorted to by a "spirit" in his attempt to convey the word watch--perhaps to remind the sitter of a particular watch he used to wear. The medium might well proceed as follows: "He taps his stomach, and looks at a spot over his left side.... He seems to wish to convey the impression that he suffered much from his bowels--perhaps a cancer on the left side. Yes, he seems to be taking something away from his body; evidently they removed some growth, and he wishes to convey the idea that something was taken from him.... Now he is examining his hands; he is looking intently. He is doing something with his fingers.... I can't see what it is ... a little movement. Was he connected with machinery in life? Now he is pointing to the door ..." etc. Such an interpretation of the facts, it will be observed, while describing all his actions, is wholly misleading in interpretation; the symbolism has been entirely perverted and misconstrued. And inasmuch as the subject probably never died of cancer, had no bowel trouble, underwent no operation, and was never connected with machinery, it is highly probable that the "message" would be put down wholly to the medium's subliminal, or even to guessing or conscious fraud. Yet, it will be observed, the message was, in its inception, wholly "veridical"--the fault lying in the erroneous symbolic interpretation of the medium. There is evidence to show that other forms of symbolism are adopted also--applying to the auditory as well as to the visual presentation of the messages. _Names_ afford some of the best evidence for this; e.g. in the sitting of Mrs. Verrall with Mrs. Thompson, November 2, 1899 (_Proceedings_, xvii. pp. 240-41), "Nelly," the control, gave the names "Merrifield, Merriman, Merrythought, Merrifield," and later went on: "I am muddled. I will tell you how names come to us. It's like a picture; I see school-children enjoying themselves; you can't say Merrimans, because that's not a name, nor merry people...." (Mrs. Verrall's maiden name was Merrifield.) If I remember correctly, there was similar symbolism with regard to the name Greenfield at another sitting.