This, however, was, like everything else in this nebulous region, open to considerable laxity in application. When Francisco de Tornamira, a boy of eighteen and page of the Duke of Pastrana, was tried in 1592, on the charge of having said that Jews and Moors could be saved if they had faith in their respective beliefs, he denied and was tortured till he confessed, and then the triviality of his offence was admitted by subjecting him only to abjuration _de levi_, to hearing a mass as a penitent in the audience-chamber, and to a reprimand. The same tribunal in 1579, tried Stefano Grillen, an Italian, who, in a discussion with some chance fellow-travellers, maintained that the miracles at the shrines of Our Lady of Atocha and of la Caridad were wrought by the Virgin herself and not by her images. He freely confessed but was tortured--apparently on intention--and was dismissed with the same trivial punishment as Tornamira.[23] Even more suggestive is the case of Juan Pereira, a boy of fifteen, tried, in 1646, for Judaism at Valladolid. The proceedings were dilatory and he gradually became demented; nothing could be done with him and opinions were divided as to the reality of his insanity. The Suprema was applied to and sagely ordered torture to find out. It was administered, April 22, 1648, but the method of diagnosis was not as successful as its ingenuity deserved and, in August, he was sent to a hospital for six months, with instructions to observe him carefully. As his name after this disappears from the records, he probably died in the hospital.[24] It is evident that the Inquisition did not take to heart the warning issued by the Suprema, in 1533, that torture was a very delicate matter.[25]
When we come to inquire as to the character of evidence requiring torture for its elucidation, we find how illusory were all the attempts of the legists to lay down absolute rules, and how it all ended in leaving the matter to the discretion of the tribunal. As confession, though desired, was not essential to conviction, the _negativo_ who was convicted on sufficient evidence was not to be tortured, but was to be relaxed. Even this rule, however, could be set aside at the caprice of the judge, though he was warned, in such cases, to put on record a protest that he did not direct the torture against the matters that had been proved, for the very good reason that endurance of torture might purge them and nullify the proof.[26] It was impossible to reduce to a logical formula that which in its essence was illogical, or to frame an accurate definition of evidence that was insufficient for conviction yet sufficient for torture. It was easy to say that _semiplena_ evidence suffices, but what was semiplena? One authority will tell us that a single witness, even an accomplice, justifies torture, another that three accomplice witnesses are requisite. One impartial and unexceptionable witness, again, is sometimes held to require public fame as an adjuvant, but the records are full of cases in which torture was employed on the unsupported testimony of a single witness. The weight of other more or less confirmatory evidence was also keenly debated, without reaching substantial agreement--whether flight before arrest, or breaking gaol, or vacillation and equivocation when examined, or even pallor, was sufficient justification.[27] It is not surprising, therefore, that, as a practical result, we are told that all these questions must be left to the discretion of the judge, to be decided in each individual case.[28] Under such conditions it would be useless to expect consistency of practice in all tribunals and at all periods. We have seen above that cases were sometimes suspended because evidence had not been ratified, yet the Toledo tribunal, in 1584, tortured Lope el Gordo for that very reason, because the chief witness against him had not ratified his testimony, and it is satisfactory to add that Lope endured the torments and thus earned suspension of his case.[29]