EVIDENCE 743 the jury altogether. They are strongly objected to by Beutham (Rationale of Judicial Evidence, Introduction, c. 22) on this very ground. " On trial for a criminal offence, amongst others murder, in this and that case the law pre sumes malice. Of the presumption in this case, what is the plain English ? That fearing that by a jury the man would be acquitted, the determination of the judge is that he shall be convicted." If the presumption, however, is the safest conclusion to act on in the circumstances, there would seem to be no harm in saving the jury the trouble of drawing the inference for themselves. Besides these two classes of presumptions, and along with them, legal writers often discuss the presumptions which are said to be within the province of the jury itself. These are neither more nor less than various degrees of probability, in cases of circumstantial evidence ; thus the leading text book on criminal practice (Archbold), following Coke and lllackstone, states that these presumptions are of three kinds violent, probable, and light or rash. A case of violent presumption, generally given as an illustration, is where a person is found in a house run through the body, and a man is seen running away with a drawn sword in his hand, no other person being found about the premises. The conclusion that this man is the murderer is irresisti ble. The other cases are simply inferior degrees ot proba bility established by circumstantial evidence, the lowest degree being described as such that it ought to have no weight with us at all. The distinctions are of no value, and are probably retained in text books because they are described by the same name as the two classes of legal pre sumptions above described those, namely, which Sir J. Stephen distinguishes as " conclusive proof " and " pre sumption" respectively. Presumptions of the second class abound in every branch of the law, and are to be explained with reference to its peculiar principles. Of the more general presumptions a number of examples have been collected in Stephen s ])i<jest, part iii. c. 14. One of the most common is the pre sumption of death after seven years absence which has been a good deal debated in the courts, but may now be considered to be settled. A person who has not bsen heard of for seven years is presumed to be dead, unless the circumstances are such as to account for his absence otherwise. But there is no presumption as to his having been dead at any particular time, e.g., if a person was last heard of in I860, the court in 1870 presumes that he is dead, but not that he was dead in 1867. The question of survivorship, where two or more persons are shown to have perished by the same catas trophe, as in cases of shipwreck, has been much discussed. It was at one time thought that there might be a presump tion of survivorship in favour of the younger as against the older, of the male as against the female, &c. But it is now clear that there is no such presumption. Another common case is the presumption of legitimacy in favour of persons born during the continuance of lawful wedlock. The presumption of regularity in official proceedings (oinnia e^e rite ado) is also one of frequent occurrence. The effect of presumptions may be compared with that of estoppel. The former establishes against a party a conclusion which stands unless and until he positively disproves it. By estoppel a party is prevented from dis proving a fact which he has actually or constructively asserted. For examples see article ESTOPPEL. With few exceptions all witnesses are now competent to testify iu courts of justice. The following are the chief exceptions : (1) persons incapacitated by extreme youth, or mental disease ; and (2) in criminal cases the wife or husband of a prisoner, except when the prosecution is for injury or violence to such wife or husband. The old rules of exclusion have been noticed supra. Certain classes of facts are protected from disclosure on various grounds. Thus, no person can be compelled to disclose communications made to him by his wife during marriage, and servants of the state cannot be compelled to give evidence in official matters without the consent of the bead of the department to which they belong. But perhaps the most important case is that of communications between lawyer and client. The lawyer is not allowed to disclose such communications without the client s assent, nor can the client be compelled to disclose such communi cations himself. The rule, however, will not extend to communications in furtherance of any crime or fraud. Medical men and clergymen have no such privilege. There is, however, a general consensus of opinion in favour of protecting confessions made by prisoners to their spiritual advisers ; and judges have from time to time expressed their reluctance to compel disclosure in such cases. To this class also belongs the rule that no person can be compelled to answer a question tending to criminate himself, although the fact that the answer might expose him to a civil action will be no protection. In some few cases the evidence of more than one witness is required. Thus, in trials for treason, there must be at least two witnesses testifying to the same act or to different acts of the same treason, except when the treason consists iu an attempt on the life or person of the queen. So in perjury, one witness, unless corroborated by circumstances, will not be sufficient to convict the prisoner. In actions for breach of promise of marriage, in affiliation cases, and in prosecutions when the only witness is an accomplice, such corroboration is also necessary. Otherwise in the law of England the testimony of one witness is sufficient to prove any fact. The general rule is that testimony must be given on oath, but an oath is binding if administered iu any form which the witness declares to be binding. By recent enactments, however, a person objecting on grounds of religious belief to the taking of any oath may be permitted to make a solemn affirmation instead ; and any person who objects to take an oath, whether on religious grounds or not, or is objected to as incompetent to take an oath, may 1 " solemnly promise and declare." In all cases the punishment of per jury attaches. At the trial a witness is first of all examined by the party producing him (examination-in-chief); he is then cross- examined by the opposite party, and re-examined by bis own party. The re-examination must refer to matters arising out of the cross-examination. There are certain questions which may be asked in cross-examination only. Thus, in the examiuatiou-in-chief, leading questions (i.e., questions suggesting their own answer) are not allowed; in cross-examination they are. So also in cross-examination a witness may be asked any question tending to test his accuracy or credibility, or to destroy his credit by in juring his character, and he must answer them, however disgraceful may be the imputation they convey. 1 No 1 The unlimited licence of cross-examination to character is the one flagrant abuse of the existing law of evidence ; and but for the restraint imposed upon counsel, partly by public partly by professional opinion, would be a much more serious evil than it is. The illustration in Stephen s Digest is a notorious but perfectly fair example. "The question is whether A. committed perjury in swearing that he was R.T. B. depones that he made tattoo marks on the arm of R. T., which at the time of the trial were not and never had been on the arm of A. B. may be asked aud compelled to answer the question, whether many years after the alleged tattooing, and many years before the occasion on which he was examined, he committed adultery with the wife of one of his friends." The Indian Evidence Act restricts the licence of cross-examination by the following provisions: (1) Such questions are proper if they are of such a nature that the truth of the imputa tion would seriously affect the opinion of the court as to the credi
bility of the witness on the matter tQ which he testifies ; (2) Such