The following and last experiment that I tried on this occasion was the most crucial. I requested Mr Zancig to go out with me on to the landing outside the door of the flat. I did not previously inform Madame Zancig nor Mr Zancig of the nature of the test that I was about to put. Madame Zancig remained in the room with my wife. The door was closed, but not completely. When we were on the landing I suddenly drew my cheque-book out of my pocket, tore out a cheque, and handed it to Mr Zancig, requesting him to transmit the number. ... He then remained silent while he looked at the cheque. My wife then came out on to the landing, and handed me a slate upon which Madame Zancig had during the experiment written the words, "In the year 1875." Mr Zancig then said aloud, "This is not what we want; it is the number." My wife returned into the room with the slate, and the door was closed, but not completely. It was impossible, however, for Madame Zancig to see her husband. The suspicion arose in my mind that the number on the cheque might have been communicated to Madame Zancig by the words that Mr Zancig had spoken aloud. I therefore took the cheque that he had in his hand and substituted another one with a different number that I tore from the bottom of my cheque-book. Mr Zancig remained absolutely silent during the whole time that this second experiment lasted. My wife again came out of the room with the slate, upon which Madame Zancig had written quite correctly, in their proper order, four of the five numbers of the second cheque, with the exception of the last figure, which was wanting, but just as we were returning to the room Madame Zancig said, "There was another figure; it was four" - which was correct.
This impressed me as a good test, with regard to the three last numbers of this cheque, which were different from the corresponding ones of the first cheque. Madame Zancig could not see her husband, and he remained absolutely silent while the experiment was being carried out.