M. Dupin then begins his now-famous method of ratiocination. He maintains that one should not ask "what has occurred," but, instead, "what has occurred that has never occurred before." He maintains that the solution of the mystery is in direct ratio to its apparent insolubility, according to the police. He announces to his friend, the narrator, that he is waiting on confirmation of his solution; he expects a person to arrive momentarily to confirm his theory.
M. Dupin then points out to the narrator some of the obvious things that the police have overlooked. Among the witnesses who heard the two voices were an Italian, an Englishman, a Spaniard, a Hollander, and a Frenchman. Each one thought that the shrill voice they all heard was the voice of a foreigner, but none agreed on the nationality; furthermore, the Englishman thought it belonged to a German, but he does not understand German, the Spaniard thought it to be English, but he does not understand English, the Italian believed it to be Russian, but he does not understand Russian, and so on in every case. No person can identify the nationality of the shrill voice. And whereas they all agree that the deep French voice uttered discernible words, such as mon Dieu (my God) and sacre and diable, the shrill voice uttered no discernible words — only sounds.
As to the matter of egress from the room being impossible, the police reject the notion because of its impossibility. M. Dupin, however, says that he will show that "these apparent 'impossibilities' are, in reality," possible. Using this logic, he discovers that the locked windows have a spring in them that, once pressed, can be opened. Furthermore, since the police abandoned further examination of the windows after they saw that they were nailed down, M. Dupin decided to examine the nails. He found a nail in one window to be broken off just at the shaft so that it only appeared to be nailed shut; the nail separated when the window was open. Thus someone could have entered by the open window and closed it upon leaving, thereby springing the spring closed and making it appear as though it were nailed shut since the two parts of the nail met again after the window was closed.
