In dealing with less intelligent people, with whom this construction cannot be performed, one must be satisfied with general rules. By demanding complete accuracy and insisting, in any event, on the ratio sciendi, one may generally succeed in turning a perception, uncertain with regard to any individual, into a trustworthy one with regard to the confessor. It happens comparatively seldom that untrue confessions are discovered, but once this does occur, and the trouble is taken to subject the given evidence to a critical comparison, the manner of adaptation of the evidence to the confession may easily be discovered. The witnesses were altogether unwilling to tell any falsehood and the judge was equally eager to establish the truth, nevertheless the issue must have received considerable perversion in order to fix the guilt on the confessor. Such examinations are so instructive that the opportunity to make them should never be missed. All the testimony presents a typical picture. The evidence is consistent with the theory that the real confessor was guilty, but it is also consistent with the theory that the real criminal was guilty, but some details must be altered, often very many. If there is an opportunity to hear the same witnesses again, the procedure becomes still more instructive. The witnesses (supposing they want honestly to tell the truth) naturally confirm the evidence as it points to the second, more real criminal, and if an explanation is asked for the statements that pointed to the “confessor,” the answers make it indubitably evident, that their incorrectness came as without intention; the circumstance that a confession had been made acted as a suggestion.[63]