In ordinary cases, they require two witnesses to prove a fact; and therefore they held, "that, if there be but one witness, and no probable grounds of presumption of some kind (_nulla argumenta_), that one witness is by no means to be heard"; and it is not inelegantly said in that case, _Non jus deficit, sed probatio_, "The failure is not in the law, but in the proof." But if other grounds of presumption appear, one witness is to be heard: "for it is not necessary that one crime should be established by one sort of proof only, as by witnesses, or by documents, or by presumptions; all the modes of evidence may be so conjoined, that, where none of them alone would affect the prisoner, all the various concurrent proofs should overpower him like a storm of hail." This is held particularly true in cases where crimes are secret, and detection difficult. The necessity of detecting and punishing such crimes superseded, in the soundest authors, this theoretic aim at perfection, and obliged technical science to submit to practical expedience. "_In re criminali_," said the rigorists, "_probationes debent esse evidentes et luce meridiana clariores_": and so undoubtedly it is in offences which admit such proof. But reflection taught them that even their favorite rules of incompetence must give way to the exigencies of distributive justice. One of the best modern writers on the Imperial Criminal Law, particularly as practised in Saxony, (Carpzovius,) says,--"This alone I think it proper to remark, that even incompetent witnesses are sometimes admitted, if otherwise the truth cannot be got at; and this particularly in facts and crimes which are of difficult proof"; and for this doctrine he cites Farinacius, Mascardus, and other eminent Civilians who had written on Evidence. He proceeds afterwards,--"However, this is to be taken with a caution, that the impossibility of otherwise discovering the truth is not construed from hence, that other witnesses were not actually concerned, but that, from the nature of the crime, or from regard had to the place and time, other witnesses could not be present." Many other passages from the same authority, and from others to a similar effect, might be added; we shall only remark shortly, that Gaill, a writer on the practice of that law the most frequently cited in our own courts, gives the rule more in the form of a maxim,--"that the law is contented with such proof as _can_ be made, if the subject _in its nature_ is difficult of proof."[48] And the same writer, in another passage, refers to another still more general maxim, (and a sound maxim it is,) that the power and means of proof ought not to be narrowed, but enlarged, that the truth may not be concealed: "_Probationum facultas non angustari, sed ampliari debeat, ne veritas occultetur._"[49]