His wife declared that she had been asleep in her bed, and hearing his cry had rushed out to find him lying on the stairs; but this was immediately questioned. In the first place, it was proved that from her room she could not have heard the struggle on the stairs, owing to the thickness of the walls and the length of the intervening passage; then it was evident that she had not been in bed and asleep, since she was dressed when she roused the house, and her bed had not been slept in. Moreover, the door at the bottom of the stairs was ajar, and it was noticed by the chaplain (an observant man) that the dress she wore was stained with blood about the knees, and that there were traces of small blood-stained hands low down on the staircase walls, so that it was conjectured that she had really been at the postern-door when her husband fell and, feeling her way up to him in the darkness on her hands and knees, had been stained by his blood dripping down on her. Of course it was argued on the other side that the blood-marks on her dress might have been caused by her kneeling down by her husband when she rushed out of her room; but there was the open door below, and the fact that the finger-marks in the staircase all pointed upward. The accused held to her statement for the first two days, in spite of its improbability; but on the third day word was brought to her that Hervé de Lanrivain, a young nobleman of the neighbourhood, had been arrested for complicity in the crime. Two or three witnesses thereupon came forward to say that it was known throughout the country that Lanrivain had formerly been on good terms with the lady of Cornault; but that he had been absent from Brittany for over a year, and people had ceased to associate their names. The witnesses who made this statement were not of a very reputable sort. One was an old herb-gatherer suspected of witchcraft, another a drunken clerk from a neighbouring parish, the third a half-witted shepherd who could be made to say anything; and it was clear that the prosecution was not satisfied with its case, and would have liked to find more definite proof of Lanrivain’s complicity than the statement of the herb-gatherer, who swore to having seen him climbing the wall of the park on the night of the murder. One way of patching out incomplete proofs in those days was to put some sort of pressure, moral or physical, on the accused person. It is not clear what pressure was put on Anne de Cornault; but on the third day, when she was brought in court, she “appeared weak and wandering,” and after being encouraged to collect herself and speak the truth, on her honour and the wounds of her Blessed Redeemer, she confessed that she had in fact gone down the stairs to speak with Hervé de Lanrivain (who denied everything), and had been surprised there by the sound of her husband’s fall. That was better; and the prosecution rubbed its hands with satisfaction. The satisfaction increased when