After a prolonged interrogation, Israel tangled himself in a web of conflicting statements and confessed that he had killed Father Dahme. When a detective asked what had become of the cartridge of the mortal bullet, Israel said it was in the toilet of his boarding house. The police found it there. With this evidence added to the confession, Israel’s trial and conviction seemed only a formality. Homer S. Cummings, later Attorney General of the United States and then state’s attorney for Fairfield County, at first had no doubt of Israel’s guilt but gradually noticed flaws in the state’s case. First of all, Israel soon repudiated his confession, claiming that when he made it he had been questioned for so long that he was willing to confess anything to get a rest. Then Cummings discovered that after the police had left the boarding house, the landlady had found still another empty cartridge in the toilet. Both shells had been exploded by a dull firing pin. The pin of Israel’s revolver was sharp, and it turned out that a friend of his had sharpened it at his shop several weeks before the murder. Cummings now spent some time with Van Amburgh examining the strip photographs of the bullets. As he testified at the preliminary hearing, Van Amburgh, in order to develop his argument, cut a slit in the photograph of the Dahme bullet. “Then,” said Cummings, “he inserts through that slit the strip picture of the recovered bullet, and pushing this up and down, finally reaches a place where he says the lines coincide. Well, I worked at that for two hours one day, and yesterday, in the presence of Captain Van Amburgh worked on it for more than an hour. It took Captain Van Amburgh fifteen minutes to find the point of juncture himself—well, say five minutes, it will be more conservative. Then a peculiar thing developed. I called his attention to it, and he had absolutely, so far as I could see, no answer to it. I asked him to put the photographs in the position which he claimed demonstrated his point, and he got them into that position. Then I said, ‘Now, lift up the flap and look and see what is under it,’ and I thought, somewhat reluctantly, the flap was lifted and on the under picture we saw a scene totally different from that which we saw on the upper picture. I asked him whether it was not perfectly logical to assume that if the two pictures had been superimposed—that is, if the other picture was transparent and placed over the other—that there ought not to be a continuous similarity in the surface appearance. He admitted that was logical, and I never could get a satisfactory explanation for the discrepancy.”