white horse, we must prove it precisely by the same description of witnesses and evidence as we must prove the act of robbery itself. So if we ransack the history of a man's life, whatever actions we bring forward, whether criminal in themselves or inferences of criminality, these must all be proved by the same sort of evidence. If we are to prove that he issued a murderous mandate, like King Tarquin, by cutting off the head of a poppy, we must prove that act by precisely the same evidence by which we should have proved that he cut off the head of a man: perhaps it might require some additional facts or circumstances to explain it, or to shew that it was the cause of the criminal effect.

What I have advanced so far upon circumstantial evidence is this, viz. That facts, from which guilt is to be inferred, must be established by the same species of evidence as the immediate or principal acts of criminality; but the inference to be collected from those facts must be left, in every court, to the judgment and consciences of those, whose province it is to pronounce upon the guilt or innocence of the party accused. Mr. Baron Mounteney, in summing up the evidence in the trial between James Annesley, Esq. and the Earl of Anglesea, makes this observation: "I remember to have heard it laid down, by one of the greatest men who ever sat in a court of judicature, viz. That circumstances

were

were in many cases of greater force, and more to be depended upon, than the testimony of living witnesses."" Witnesses, Gentlemen, may either be mistaken themselves, or wickedly intend to deceive others: God knows, we have seen too much of this in the present cause, on both sides. But circumstances, Gentlemen, naturally and necessarily arising out of a given fact, cannot lie*."

We hear this observation everywhere echoed; "Circumstantial evidence is the best; for circumstances cannot lie." But if we would give ourselves the trouble to bestow a little consideration upon the subject, I think we shall be convinced that circumstantial evidence is not the best, and that circumstances can lie. There are circumstances which cannot lie, where the conclusion or inference is necessary and unavoidable; but where the conclusion or inference is contingent, circumstances may lie, that is, we may draw an erroneous conclusion from the given facts. The learned Matthæus clearly describes this distinction: Argumentum porrò necessarium vel contingens est: necessarium, cujus consequentia necessaria est, veluti coivisse eam quæ peperit: contingens, cujus consequentia probabilis est, veluti cædem fecisse, qui cruentatus est; Atalantam virginem non esse, quod

Har. St. Tr. vol. XI. 426.

сит

cum adolescentibus spatietur sola per sylvas. In the first case, one fact is a certain demonstration of the other; but in the second, the circumstances must frequently lie, when they charge with murder a person stained with blood, or Atalanta, from such companions and conduct, with a want of chastity. But he proceeds to observe; Contingentia verò quan quam singula fidem non faciant, plura tamen conjuncta crimen manifestare possunt. Rem uno atque altero exemplo declarabimus. Occisus est Kalendis Mævius: Titius perempti inimicus fuit; eidem sæpius non solum interminatus, sed et insidiatus est. Cum deprehenderetur iisdem Kalendis in loco cædis cruentatus, cum gladio cruento, ad mensuram vulneris facto, toto vultu expalluit, interrogatus nil respondit, trepidè fugit. Hic singula quidem argumenta infirmiora sunt, universa tamen cadis auctorem Titium evidenter designant, rectèque Duarenus dixit, non dubitaturum se hunc reum carnifici jugulandum dare. Tit. xv. c. 6. Yet Duarenus might have condemned and executed an innocent man. Every one of these circumstances must be proved by positive witnesses, who may be either wicked or mistaken; but even if they are pure and correct, the conclusion we draw from the facts disclosed may be erroneous.

So that, in circumstantial evidence, there must, of necessity, be more chances for error than in

positive

positive evidence. If any number of witnesses should swear they saw the prisoner draw a reeking sword from the side of a dead man, we have not the same degree of certainty that he either murdered or killed him, as if the same witnesses had sworn they had seen him run it through his body. It affords a violent presumption, but still it might have been the friendly act of an innocent man, who had accidentally passed that way after the murder was committed: or even if it was the prisoner's own sword, it might have been snatched from his side and plunged into the body of the deceased by some one who had escaped; or the deceased might have borrowed it, and have fallen upon it himself. All human testimony is nothing more than a high probability; and it is true that circumstantial evidence in one case may produce a higher degree of it, or may more nearly approach to certainty, than direct and positive evidence in another*. Human testimony

* That both positive and circumstantial evidence may fail, will appear from the following cases. The first is in the Chronicle of the Gentleman's Magazine for Oct. 1772. The other is from the fifth volume of Causes Célèbres, p. 438, where several more trials of the same nature are related.

Sept. 14, 1772. Came on at the sessions in the Old Bailey, the trial of one Male, a barber's apprentice, for robbing Mrs. Ryan, of Portland-street, on the highway, on the 17th of June last. The witnesses swore positively to the identity of the lad, and the whole court imagined him guilty. He said nothing in

testimony is so far distinct from certainty, that it admits of all the degrees of probability, and by

some

his defence, but that he was innocent, and his evidences would prove it. His evidences were the books of the court; to which reference being made, it appeared that on the day and hour when the robbery was sworn to be committed, the lad was on his trial at the bar where he then stood, for another robbery, in which he was likewise unfortunate enough to be mistaken for the person who committed it; on which he was honourably acquitted.

Voici un autre fait, dont j'ignore l'époque, et qui m'a été transmis par la tradition. Avant qu'on eût rebâti cette longue suite de maisons qui bordent la place Saint Michel à Paris, en fàce de la rue Sainte Hyacinthe, une marchande veuve et âgée occupoit, au même endroit, une petite boutique, avec une arrière boutique, où elle couchoit. Elle passoit, dans le quartier, pour avoir beaucoup d'argent amassé. Un seul garçon composoit, depuis long-tems, tout son domestique. Il couchoit à un quatrième étage, dont l'escalier n'avoit point de communication avec l'habitation de sa maîtresse; il étoit obligé, pour s'y rendre, de sortir dans la rue; et lorsqu'il s'alloit coucher, il fermoit la porte extérieure de la boutique, et emportoit la clef, dont il étoit seul dépositaire.

On voit, un matin, la porte ouverte plutôt qu'à l'ordinaire, sans qu'on remarquât aucun mouvement qui annonçât que la marchande, ou son garçon fussent levés. Cette inaction donna de l'inquiétude aux voisins. Cependant on ne remarque aucune fracture à la porte: mais on trouve un couteau ensanglanté, jetté au milieu de la boutique, et la marchande assassinée dans son lit, à coups de couteau. Le cadavre tenoit, dans une main, une poignée de cheveux; et dans l'autre une cravate Auprès du lit, étoit un coffre qui avoit été forcé.

On