though, if they are of exemplary character, their evidence may justify some moderate penalty.[266] It is probable that, for awhile, practice was not uniform in all tribunals. In that of Valladolid, in 1621 and 1622, there are several cases in which arrest was voted on the evidence of a single witness and these votes were confirmed by the Suprema.[267] On the other hand, about 1640, an inquisitor tells us that, when the accused denies, conviction requires the evidence of three witnesses whom he has been unable to disable for enmity, low rank of life, or doubtful repute. Some authors, he adds, insist that four are necessary, but he admits that, when there are two whose characters stand thorough investigation and there are supporting indications, conviction may follow.[268] It is impossible not to recognize the charitable motives that prompted this reluctance to punish. The requirement thus established of two independent denunciations threw serious impediments in the way of suppressing a crime in which it was so notoriously difficult to find accusers. The routine gradually established was, when a denunciation was received, to search the records for a previous one. If none were found, letters were addressed to all the other tribunals requesting a similar examination of their registers and, if this was unsuccessful, the denunciation was filed away to await the chances of another accuser presenting herself, thus giving the accused, if guilty, the opportunity of continuing his profligate career, and leading the woman to believe that the case was too trivial to deserve the attention of the Inquisition. These long intervals of impunity illustrate the difficulty of obtaining denunciations, and the preponderant chances of escape, when prosecution was thus obstructed. [Sidenote: _TWO DENUNCIATIONS REQUIRED_]