with its association of 'theft,' would trouble both men.
"You must have seen, I think, that the little speech I made before giving the test was not merely what it pretended to be. That speech was an excuse for me to couple together and lay particular emphasis upon the natural associations of certain words. So I coupled and emphasized the natural association of 'safe' with 'combination,' 'scraps' with waste-basket,' 'dollars' with 'ten thousand,' and so on. In no case did I attempt by my speech to supplant in anyone's mind his normal association with any one of these words. Obviously, to all your clerks the associations I suggested must be the most common, the most impressive; and I took care thus to make them, finally, the most recent. Then I could be sure that if any one of them refused those normal associations upon any considerable number of the words, that person must have 'suspicious' connection with the crime as the reason for changing his associations. I did not care even whether he suspected the purpose of my test. To refuse to write it would be a confession of his guilt. And I was confident that if he did write it he could not refrain from changing enough of these associations to betray himself.
"Now, the first thing which struck me with Ford's paper was that he had obviously erased his first words for 'reship' and 'ethics' and substituted others. Everyone else treated them easily, not knowing them to be the combination words. Ford, however, wrote something which didn't satisfy him as being 'innocent' enough, and wrote again. There were no 'nor-
