On the hypothesis of fraud, therefore, one of two things must have happened.
Either Slade must have contrived, during the sitting, to possess himself of the box, open it, abstract the coin, close the box again, and return it to the table; or else he must have substituted for the box, which at the beginning of the sitting contained the coin, another (empty) box, previously prepared to resemble the original.
I do not think the former method to be at all likely.
One cannot unstick a length of glued paper and stick it up again in a few seconds unobserved.
On the other hand everything lends itself to the supposition that the second method was actually adopted.
In the first place we know that the box was prepared some six months previous to the experiment.
It is true that Zöllner is a trifle hazy as to dates, saying at the outset that Slade's first visit to Leipsic was in December 1877, and, later, that the first and second visits were in November and December 1877.
But this is comparatively immaterial, the point being that Slade had presumably had ample time and opportunity for finding out all about these boxes and for preparing substitutes. I say "presumably" because in the absence of definite evidence to the contrary, we have no reason to suppose that these boxes were kept in an inaccessible place or that Zöllner had never mentioned his intentions with regard to them to Slade himself or to anyone else. I consider then that so far as the records go, we are perfectly entitled to suppose that Slade was able to prepare, and, in fact, actually did prepare, an empty counterfeit box, externally similar to that prepared by Zöllner. The second, and almost incredible, point to be noticed is that apparently no steps of any sort were taken by Zöllner to identify either the box or the coin after the sitting with those originally prepared by him.
In fact, he definitely says that he had completely forgotten, indeed had never so much as observed, the value or dates of the coins used!