AbstractThe trial of Jeanne d'Arc took place from 9 January to May 1431. During the sessions registrars recorded the interrogations, which, in an increasingly abbreviated form and always in French, were registered in a book now lost. This document provided the basis for the wording of the authentic Latin instrument formulated by the Sorbonist Thomas de Courcelles and the registrar Manchon. It is generally accepted that five authentic copies were made from the Latin instrument. Only three of them have survived. Each of the registrars of the trial, Guillaume Manchon, Guillaume Colles and Nicolas Taquel, signed the authentic copies on which the seals of Pierre Cauchon, bishop of Beauvais, and of Jean Lemaistre, vicar of the inquisitor, can also be seen. In 1897 Henri Denifle and Emile Chatelain published an article entitled Le Procès de Jeanne d'Arc et l' Universite de Paris, in which, after pointing out the errors in the names and degrees attributed to the assessors, members or former members of the University present at Rouen in 1431, they claimed that the Latin text of the trial was a late work dating from 1435 at the earliest, and that Manchon and especially Courcelles were guilty of trusting to their memory to fill up the gaps in the minutes written down during the trial. For three quarters of a century these conclusions by Denifle and Chatelain have been unanimously accepted; P. Champion and P. Tisset, the two most recent editors of the trial, adopted them without hesitation. A more thorough examination of the article by Denifle and Chatelain, however, has revealed its weaknesses and leads us to reject their conclusions. Amongst the one hundred and thirteen persons who attended the trial of Jeanne in 1431, twenty four were members of the University of Paris. Denifle and Chatelain claimed to have discovered errors in the degrees or the names of sixteen of them. But a close investigation reduces the number of errors to five, of which four are insignificant. So there is only one error which makes a claim on our attention: that concerning the Christian name of the abbot of Mortemer, whom the text of the trial calls Guillelmus although in reality, according to the documents of the University of Paris, his name ought to be Nicolaus. It is not very much... Denifle and Chatelain fail by omission. They forget to state - although they know it perfectly well - that on the authentic copies of the trial Cauchon must have set his seal as Bishop of Beauvais on 8 August 1432 at the latest, because on that day he took possession of the episcopal see of Lisieux. Not only do they omit this fact, which alone destroys their whole argument, but Denifle and Chatelain have also tried to strengthen their case by calling to their aid the declarations made by the registrars Manchon and Taquel during the rehabilitation trial, which took place about twenty five years after the death of Jeanne d'Arc. Manchon does indeed say that the Latin version of the trial had been made longe post mortem of Jeanne, and Taquel fails to add more precision by declaring per magnum temporis spatium post mortem. But these late and vague declarations - must we understand by longe four years, as Denifle and Chatelain would like us to believe, or thirty five days as we read in the same rehabilitation trial - lose all their demonstrative value when we know that they are answers produced to please the accuser who had already used similar words to invalidate the trial of condemnation: longo tempore post mortem dictae Johanae, dictus processus fuerat confectus. Finally, documents unknown to Denifle and Chatelain - they are of a financial nature - inform us that Cauchon's involvement with the trial of Jeanne ended on 30 November 1431. On that day - longe (five months) after the death of Jeanne - Cauchon, we suggest, set his seal as Bishop of Beauvais on the copies of the trial.