That witnesses should be tortured, in order to obtain or confirm their testimony, is an abuse which, repulsive as it may seem to us, has been, with more or less disguise, a practice wherever torture has been used. It is true that the Roman law prohibited that one who had admitted his own guilt should be examined as to that of another, and this principle, adopted in the False Decretals, became a part of the early canon law.[33] The Inquisition, however, regarded the conviction of a heretic as only the preliminary to forcing him to denounce his associates; the earliest papal utterance, in 1252, authorizing its use of torture, prescribed the employment of this means to discover accomplices and finally Paul IV and Pius V decreed that all who were convicted and confessed should, at the discretion of the inquisitors, be tortured for this purpose.[34] The _question préalable_ or _définitive_, in which the convict was tortured to make him reveal his associates, became, through the influence of the Inquisition, a part of the criminal jurisprudence of all lands in which torture was employed. It was, in reality, the torture of witnesses, for the criminal's fate had been decided, and he was thus used only to give testimony against others. The Spanish Inquisition was, therefore, only following a general practice when it tortured, _in caput alienum_, those who had confessed their guilt. No confession was accepted as complete unless it revealed the names of those whom the penitent knew to be guilty of heretical acts, if there was reason to suspect that he was not fully discharging his conscience in this respect, torture was the natural resort. Even the impenitent or the relapsed, who was doomed to relaxation, was thus to be tortured and was to be given clearly to understand that it was as a witness and not as a party, and that his endurance of torture would not save him from the stake. The Instructions of 1561, however, warn inquisitors that in these cases much consideration should be exercised and torture _in caput alienum_ was rather the exception in Spain, than the rule as in Rome.[35] In the case of the _negativo_, against whom conclusive evidence was had, and who thus was to be condemned without torture, the device of torturing him against his presumable accomplices afforded an opportunity of endeavoring to secure his own confession and conversion. We have seen this fail, in 1596, in the Mexican case of Manuel Diaz, nor was it more successful in Lima, in 1639, with Enrique de Paz y Mello, although the final outcome was different. He persistently denied through five successive publications of evidence, as testimony against him accumulated in the trials of his associates. He was sentenced to relaxation and torture _in caput alienum_; it was administered with great severity without overcoming his fortitude, and he persisted through five other publications as fresh evidence was gathered. Yet at midnight before the auto de fe, in which he was to be