“Never mind that,” said he.  “A sound thinker gives equal consideration to the probable and the improbable.”

Thus encouraged, I proceeded to set forth the theory of the crime as it had occurred to me on my way home in the fog, and I was gratified to observe the close attention with which Thorndyke listened, and his little nods of approval at each point that I made.

When I had finished, he remained silent for some time, looking thoughtfully into the fire and evidently considering how my theory and the new facts on which it was based would fit in with the rest of the data.  At length he spoke, without, however, removing his eyes from the red embers—­

“This theory of yours, Jervis, does great credit to your ingenuity.  We may disregard the improbability, seeing that the alternative theories are almost equally improbable, and the fact that emerges, and that gratifies me more than I can tell you, is that you are gifted with enough scientific imagination to construct a possible train of events.  Indeed, the improbability—­combined, of course, with possibility—­really adds to the achievement, for the dullest mind can perceive the obvious—­as, for instance, the importance of a finger-print.  You have really done a great thing, and I congratulate you; for you have emancipated yourself, at least to some extent, from the great finger-print obsession, which has possessed the legal mind ever since Galton published his epoch-making monograph.  In that work I remember he states that a finger-print affords evidence requiring no corroboration—­a most dangerous and misleading statement which has been fastened upon eagerly by the police, who have naturally been delighted at obtaining a sort of magic touchstone by which they are saved the labour of investigation.  But there is no such thing as a single fact that ‘affords evidence requiring no corroboration.’  As well might one expect to make a syllogism with a single premise.”  “I suppose they would hardly go so far as that,” I said, laughing.

“No,” he admitted.  “But the kind of syllogism that they do make is this—­

“’The crime was committed by the person who made this finger-print.

“’But John Smith is the person who made the finger-print.

“‘Therefore the crime was committed by John Smith.’”

“Well, that is a perfectly good syllogism, isn’t it?” I asked.

“Perfectly,” he replied.  “But, you see, it begs the whole question, which is, ’Was the crime committed by the person who made this finger-print?’ That is where the corroboration is required.”

“That practically leaves the case to be investigated without reference to the finger-print, which thus becomes of no importance.”

“Not at all,” rejoined Thorndyke; “the finger-print is a most valuable clue as long as its evidential value is not exaggerated.  Take our present case, for instance.  Without the thumb-print, the robbery might have been committed by anybody; there is no clue whatever.  But the existence of the thumb-print narrows the inquiry down to Reuben or some person having access to his finger-prints.”