suppose that a direct assertion, not shown to be false, leaves the CHAP.VII. jury no choice, and takes all responsibility out of their hands. Such cases are in fact precisely those, which impose the heaviest responsibility, and in which the judge ought to do his utmost to assist the jury, and to arouse their imaginations-generally torpid and sluggish-to a sense of the fact, that it is by no means a light thing to believe the oath of a man whom they have never seen before, so fully as to punish another unknown person on the strength of it.

conceals

the duty

Another great practical objection to the phrase under The phrase consideration is, that it is a contrivance by which the fact the true that juries ought to make conjectures in criminal trials is nature of concealed. I have already given my reasons for thinking of the jury. that this is part of the duty of juries, and perhaps the most important part. The phrase, “Circumstantial evidence," timidly and plausibly conceals this fact. It proposes a sham canon of proof, and leads jurymen to believe that they are deciding on a particular kind, and a highly scientific and ingenious kind, of evidence when, in fact, they are making a conjecture. This is effected by the introduction of the word "reasonable" into the canon of proof quoted above. "The "facts must be absolutely incompatible with the innocence. "of the prisoner, and incapable of explanation upon any "other reasonable hypothesis than that of his guilt." If for "reasonable," "possible" were substituted, this rule would have a distinct meaning. It would mean that the jury ought to assume the existence of any possible state of facts, by which the hypothesis of the prisoner's innocence may be reconciled with the evidence. If it does not mean this, it can have no meaning short of the one which I have already assigned to the phrase, "Reasonable doubt;" that is, that the jury are to fill up some gaps by conjecture-an unpleasant, but I believe a true, conclusion. Assuming that this is not what is meant, let us see how the proposed rule would work. Particular cases are the best test.

tion from

In Palmer's case,* would the following theory have been Illustraunreasonable ?-that Palmer meant to murder Cook, and Palmer's bought what he supposed to be strychnine, and administered case.

CHAP.VII. it with that intention. That, by a coincidence, Cook died soon after the supposed poison was administered, of some uncommon convulsive disease not known to medical men, and that, in truth, the powder purchased was not strychnine at all, but some harmless drug substituted for it by the chemist accidentally? Strychnine may be mistaken for other things, and, conversely, other things might be mistaken for strychnine. Forms of convulsive disease may exist which would produce symptoms like those of poisoning by strychnine; and there was no reason why such symptoms should not come on twenty minutes after the administration of the supposed poison, as well as at any other time. In all this there is no impossibility. It is consistent with every fact in the case; yet no jury would or ought to entertain the suggestion for an instant. If it is their duty to give the prisoner the benefit of any possible construction of the facts, they ought to entertain such a suggestion. If their duty is to say whether, under the circumstances, it is reasonable to guess, they clearly ought not, for the guess that Palmer did poison Cook with the strychnine is strongly suggested by the evidence.

Conse

the test

The question whether any possible hypothesis can make quences of the evidence consistent with the prisoner's innocence, must proposed always be answered in the affirmative; for it is always a if strictly applied. possible hypothesis that enough of the evidence may be either mistaken, perjured, or exaggerated, to leave room for a reasonable doubt, and, if this is the case, the prisoner, according to the canon in question, must be acquitted. No reason can be given for distinguishing between the possibility that evidence may be false, and the possibility that states of fact, of which there is no evidence, may be true. This shows that the only intelligible question which can be left to the jury is, "What is the state of your minds in relation to this matter? As a fact, do you think he did it?" This may be expressed by such phrases as, "Is there any reasonable doubt ?"

any solid doubt?" "such doubts as reasonable men allow to operate on their minds in important affairs of their own?" or the like, and there is not only no harm, but great convenience in these phrases so long as they are used in the popular sense, and

are not supposed to convey any precise meaning other than CHAP. VII. that of recommending a sober and discreet conjecture. The objection to the theory of circumstantial evidence is that it attempts to turn such phrases as these to a purpose for which they are not fitted, and so to conceal the true nature of the functions of a jury, not merely from a particular jury at a given moment, but from those who have to charge them.

correct in

The very phrase "circumstantial evidence" is admirably Phrase is contrived to serve as the foundation for this sort of evasion. itself. It has no distinct meaning, but is capable of being used in several different senses. Each particular item of evidence is (adopting that classification) direct and not circumstantial. The production of a footmark is direct evidence that a foot has been pressed on that piece of earth. The assertion that the prisoner had blood on his hands is direct evidence that he really had blood on his hands. In what sense, then, are such items of evidence "circumstantial" evidence? Because (it is said) the facts to which they testify are part of a set of circumstances from which when collected together guilt is to be inferred. This admits that the guilt is to be inferred, not from the evidence, but from the inferences drawn from the evidence. The evidence-the words spoken, or things producedleads the jury to infer that a certain set of facts existed, but the existence of the facts is matter of inference, not matter of evidence, and thus the true meaning of "circumstantial evidence" is the process of combining and arguing upon the inferences drawn from evidence. This is not an accurate way of using language.

relative

of direct

tial evi

Waiving all objections to the use of the term, and assuming No differthat there is a real distinction between circumstantial and ence in the direct evidence, is there any important difference correspond- strength ing to the distinction? There is no sort of difference between and cirthe cogency of the different kinds of evidence, whether the cumstancomparison is made between weak cases, or strong ones. dence. Compare two strong cases. How is it possible to say whether the evidence of several credible witnesses, who say they saw a man put his hand into another man's pocket, and take out his purse and run away, is stronger or weaker than that of the same number of equally respectable witnesses who prove

T

CHAP.VII. that the purse was taken, and that immediately afterwards the prisoner was seen running away, and on being stopped was found to have the purse in a secret pocket, no explanation being given. Or take two weak cases. A man swears that he was robbed on a dark night, and that the prisoner is the man who robbed him. The light by which he saw him was the reflection of a furnace a long way off, which would cast a light at once strong and unsteady; and the robber was exposed to it only for a moment. A sack is stolen, and is found three months afterwards, apparently concealed, in the house of a marine-store dealer. He says something on the subject which may be, and probably is, a lie. Other people had access to the place where the sack was found. Which of these cases is the stronger of the two? Their relative strength cannot be shown to depend in anyway on the properties of either direct or circumstantial evidence as such.

The phrase

useful only for sophistical pur.

poses.

The only real purpose which the phrase ever serves is that of supplying prisoner's counsel with a convenient sophism. Instead of saying, this evidence is incomplete, because such a fact is not proved; or it is inconclusive, because such an explanation of it may be suggested; they say it is circumstantial, and all circumstantial evidence as such is inconclusive or incomplete. They go on to quote cases of incomplete and inconclusive evidence, like the famous story of the wicked uncle quoted by Lord Hale from Coke,* and then argue or insinuate that, as in each case the evidence is circumstantial, so in each case it must be inconclusive. It is, in short, a word useful only for the sake of puzzling juries, and providing them with a loophole for avoiding a painful, but most important, duty. Such cases as those related by Lord Hale are, in fact, no precedent at all. Indeed there are, and can be, no such things as precedents for verdicts. The use of a precedent is to establish a general rule by showing that it was recognised in deciding upon particular facts; but in weighing evidence, and drawing inferences from it, there neither are, nor can be, general rules. The whole matter is an affair of individual shrewdness and common sense. It is a case in

* 2 Hale, P. C. 290, note.

which men argue, and ought to argue, as for the most part CHAP.VII. they do in the other affairs of common life, directly from particulars to particulars, without conscious intermediate generalisations.

of French

Procedure.

The whole subject is admirably summed up in the declara- Provision tion, which, by a somewhat theatrical provision,* the foreman Code of of a French jury is bound to read to his colleagues before Criminal they begin to deliberate, and of which a copy must be posted up in the room. "The law does not require the jury to "account for the means by which they are convinced. It "does not prescribe to them rules by which they are to "test specifically the fulness and sufficiency of a proof; it "orders them to ask themselves in silence and retirement, "in conscientious sincerity, what impression the proofs pro"duced against the accused, and the points raised in his favour, have made on their reason. The law does not say "to them, 'You shall take as true every fact attested by so "many witnesses;' neither does it say, 'You shall not regard "as sufficiently established any conclusion not supported by “such a procés-verbal, such documents, so many witnesses, or so many circumstances.' It asks only one question, "which comprehends the whole sphere of their duty, “Are "you satisfied in your own minds?" (Avez-vouz une conviction

intime?)

* Code d'Instn. Criminelle, Art. 342.