R v Faure [1999] VSCA 166

23. At about 4.10 a.m. the applicant was examined by Dr Holmes, a part-time police medical officer, who pronounced him fit to be interviewed. The applicant said that he and the deceased had been playing Russian roulette and that it was all an accident. He said he had been drinking heavily that night and taken medication for back pain and other reasons. He said that he routinely took Rohypnol, Endone and Panadeine Forte but that he was unable to specify which drugs he had taken over the previous 24 hours. He claimed that he could not remember any gun going off and that it was all a blank. According to Dr Holmes the applicant's level of alertness and consciousness was perfectly normal and there was no evidence of any brain dysfunction which could be associated with drug or alcohol use and no evidence of thought disorder. The applicant was talking rationally, although somewhat inconsistently. A blood sample taken by the doctor at 4.55 a.m. showed a blood alcohol level of 0.154%. Other evidence showed that the deceased had died from a shot fired at her head from a distance of about 5 mm. The applicant took part in a lengthy interview with members of the Homicide Squad and also returned with them to the vicinity of Seaview Parade, where a further unsuccessful search for the firearm was made. It was in fact not found until two days later, when a householder in McCulloch Street, Dromana found a revolver on his front lawn under some clumps of grass which had apparently been placed on top of it. This revolver contained a single fired cartridge case in one chamber, all the other chambers being empty.

26. The recorded interview conducted with the applicant began at 2.17 a.m. It was broken off more than once, and these interruptions included the medical examination and the second unsuccessful expedition in search of the revolver. A number of times during the interview the applicant said he had been drunk or half drunk at the time of the shooting. He had begun drinking at about 11.30 a.m. on the day of the shooting and had drunk a great deal. (I continue to give parts of his account.) He had met the deceased at about 2.30 that afternoon and had been with her from then on. That night they had ended up at her bungalow. There he had unprotected vaginal intercourse with her and she also performed oral sex on him. Then they decided to play Russian roulette with a six-shot revolver he had borrowed. (I do not draw attention to any of the numerous inconsistencies appearing in the record.) He discussed with her the dangers of playing that game, saying, "You can die here". She replied, "You gotta die sometime". She, too, had been drinking. They agreed that they would fire no more than two shots each. (By this, he no doubt meant to refer to the pulling of the trigger.) He placed a single cartridge in the cylinder. The cylinder was spun each time before the trigger was pulled. When he pulled the trigger he aimed to skim her head. First she pulled the trigger; then he did; then she did it again; finally, he pulled the trigger and a shot was fired. They were lying on the bed as they played the game. He must have been too drunk and pointed the revolver the wrong way. The game of Russian roulette was not preceded by any argument or raised voices. The alcohol and drugs had affected his judgment and made him stupid. He had smoked marijuana at a friend's place that afternoon. After the shooting he panicked. She was still alive: she had a pulse. He went to the Dromana police station but no one was there. He was on his way back to the bungalow when he was apprehended. He knew that the deceased had hepatitis C. He also knew that she did not have AIDS because she had showed him the result of blood tests she had had.

29. According to the deceased's mother, when the deceased and the applicant visited her at about 8.30 on the night of the shooting her daughter asked her, "Mum, have I got AIDS?" The mother replied, "No. You have got hepatitis C, not AIDS". The daughter then said to the applicant, "I told you, Les, that I didn't have it", and added that she would have a test that week to prove to him that she did not have AIDS. The Crown put forward as a possible motive for the shooting the applicant's concern and anger that he had been having intercourse for some time with a woman who might have AIDS.

32. There was expert evidence that the blood alcohol concentration of 0.154% when the sample was taken from the applicant represented one of about 0.22% at the time of the shooting. There was a good deal of evidence from persons in whose company the applicant had been during the afternoon and night of the day of the shooting about how much liquor he had had to drink and how much he appeared to be affected by drink. There was also evidence that he had smoked marijuana during that afternoon. I need not summarise that evidence or the expert evidence given about the effects of alcohol and marijuana, especially the former. In addition to evidence dealing specifically with how much the applicant had had to drink and specifically with whether he appeared to be affected by drink, the jury had, of course, evidence of his behaviour after the shooting including his conversations with civilian and police witnesses and the whole of his answers to questions during the recorded interview. None of this considerable additional body of evidence need be summarised. It is enough to say that, putting the matter in terms of a positive finding (an approach too unfavourable to the applicant), it was, to say the least, well open to the jury to consider that the applicant had had a great deal to drink by the time of the shooting and was considerably affected by it. The Crown prosecutor addressed on the basis that the applicant was in a state of some intoxication.

35. The applicant stood mute and called no evidence. Since he stood mute, the only evidence of the playing of Russian roulette was that contained in the out-of-court statements of the applicant himself. The Crown's primary case was one of wilful murder, it being said that he had put the revolver to the woman's head in anger in the course of an argument which may have concerned whether she was suffering from AIDS. The game of Russian roulette was, the Crown contended, an invention. And so on the Crown's primary case there was no game but simply the putting of the pistol to the woman's head in anger and the pulling of the trigger with the intention of shooting and killing her. But the Crown put also an alternative case of reckless murder, and this seems clearly enough to have been tied to the game of Russian roulette, the suggestion being that the applicant must have known that death or serious injury was likely to result from what he did. The judge left the case to the jury as one in which the Crown suggested that the proper verdict was one of murder, based either on murderous intent or on recklessness, and the defence did not suggest that a verdict of manslaughter, based on unlawful and dangerous act, was not appropriate and indeed went so far as to invite such a verdict.

38. No exception was taken to the charge. Nor was it suggested on behalf of the applicant at any other stage of the trial that the jury should be directed about the use they might make of intoxication. Now it is put by him that the verdict cannot stand because of the judge's failure to give such a direction. What is said by Mr Tehan to be fatal is the omission of the judge to give the jury any directions about the relevancy of evidence of intoxication to the matters of murderous intent and, especially, the state of mind necessary for reckless murder. He makes no complaint about the failure to direct the jury about intoxication and "conscious, voluntary act", it having been announced to the judge by the applicant's counsel in the absence of the jury that there was no dispute that the act had been a conscious, voluntary one. This was said in answer to the Crown prosecutor's inquiring whether it was necessary for him to address the jury on "conscious, voluntary act". The applicant's counsel went on to suggest to the judge, in effect, that what was important was his client's "appreciation of the risk".

41. Unfortunately, the judge's directions on the law were of a general kind. There was little done by way of modifying or expanding general directions so as to make them more appropriate to the particular case and little was done by way of relating the law as given to the jury to the facts of the particular case.

44. His Honour distinguished wilful murder from reckless murder - understandably, those expressions were not used - by telling the jury, by way of modification of what might be called the usual definition of murder given to juries, that there were two different intents, either of which would, in conjunction with the other elements of the crime, constitute the crime of murder. The first was the intent either to kill or to do really serious injury and the second was the knowledge that it was probable that death or really serious injury would result from the act done. His Honour then gave a direction about inferring state of mind that was in perfectly general terms. The judge went on to tell the jury, turning to the particular case, that "there is no argument from the defence that the acts that caused the death were conscious, voluntary and deliberate and were unlawful and dangerous" and that the defence contended for a verdict of guilty of manslaughter. His Honour added that the real issue was "whether the prosecution has proved beyond reasonable doubt that the accused acted with one of those two intents which are the essential elements of murder". But unfortunately the judge did not go further at this point. Although he did, in the course of summarising the arguments of counsel, deal with the question whether the accused had been, or might have been, playing Russian roulette, he did not, before turning to the evidence and then to the arguments of counsel, say to the jury that, in considering what he had rightly described as the real issue, it would be important for them to consider whether they were satisfied that, as the Crown alleged in its primary case (of wilful murder), he simply aimed the revolver at the deceased and pulled the trigger as opposed to taking part with her in a game in which they took turns at pointing the revolver at each other and pulling the trigger. For, as one would expect and as one is able to ascertain from the charge where it summarised the arguments of counsel, the Crown case of wilful murder was put on the hypothesis that no game was being played. It also appears from the charge, in so far as it deals with the arguments of counsel, that the Crown case of reckless murder was put on the hypothesis that a game was being played. It would have been helpful to tell the jury, clearly and distinctly, and immediately or shortly after defining the crime of murder to them, that the Crown put two alternative views of the facts, its first and primary case being one of wilful murder, no game having been played, and its secondary case being one of reckless murder, on the basis that a game was being played. It is, as I have said, possible to get this from his Honour's summary of the case for the Crown and the case for the defence (although it does not appear as clearly as one would wish), but it would have been much preferable to deal with this clearly and distinctly and at an earlier stage, when relating the law to the facts of the case.

47. Had the Crown's two alternative cases been highlighted in this way, the jury would have readily appreciated - and indeed they would have been told by the judge in terms - that as regards the drawing of the inference of murderous intent they were being invited by the Crown to draw that inference in a case in which, as the Crown submitted, the applicant, in the course of a heated argument, put a revolver to the woman's head and pulled the trigger. The jury could then have been told that the Crown submitted that the only inference open in those circumstances was one of murderous intent. The jury would also have been told that the reckless murder state of mind which they were asked to infer was an inference which the Crown said was the only proper one in all the circumstances at the time when, in the course of the playing of the game, the applicant discharged the firearm. Had the matter been put to the jury in this way, it may be that the desirability of giving the jury some direction about the relevance of intoxication to the mental element of the crime would have become more apparent. It might have been more obvious that it was desirable to tell the jury, as regards wilful murder, that, in considering whether to draw the inference of murderous intent if they were satisfied that Russian roulette was not being played, the jury should consider the evidence that the man who put the revolver to the woman's head and pulled the trigger was not sober but much affected by drink. Similarly, it might have become more apparent that it was desirable to tell the jury that, in considering whether the man who pulled the trigger in the game of Russian roulette knew that the likely consequence was death or really serious injury, they should take into account the evidence of intoxication.

50. It is unfortunate that the jury were not told, before the evidence and arguments were summarised, not only that the only element of the crime of murder put in issue by the defence was the mental one - this they were told in substance - but also that the principal question of fact for them to consider as regards that element was whether the Crown had satisfied them that this was simply a shooting in the sense that Russian roulette was not played.

53. The Crown submits now that when regard is had to the judge's summary of counsel's final addresses - these have not themselves been transcribed - the jury can have been in no doubt that the evidence of intoxication was relevant for their consideration in both respects. But I do not consider that the charge was adequate to apprise the jury of the relevancy and indeed importance of the evidence of intoxication. It may well be that, as Mr Tehan argued, the defect in the charge is of more significance in relation to reckless murder. But even as regards wilful murder, while the nature of the weapon used and the manner of its use - discharging a firearm into the head at point blank range - could be said to provide a firm basis for an inference of murderous intent even in the face of considerable evidence of intoxication, I doubt very much whether this verdict could have been upheld by the application of the proviso to s.568(1) of the Crimes Act even if the Crown case had been left to the jury as one of wilful murder only, on the basis that the game of Russian roulette was an invention. It is very surprising that in the present case the judge gave no direction to the jury about the significance of intoxication in relation to murderous intent and the state of mind required for reckless murder, and it is surprising that the very experienced counsel appearing to prosecute and for the accused did not draw his attention to the omission. Apart from a fleeting and adventitious reference to intoxication in the introductory part of the charge in which the credibility of witnesses is dealt with in the usual general way, there is no reference to intoxication in that part of the charge which gives the jury the law. There is in the part which summarises the evidence a summary of the evidence given about how much drink the applicant had taken and its apparent effect on him and of the general evidence of an expert witness about the effect of liquor, and there is a summary of that part of the argument of the Crown prosecutor and counsel for the accused in which reference is made to the evidence of intoxication. Counsel for the accused seems to have relied on this evidence as accounting for what might appear to be guilty behaviour after the shooting, as explaining inconsistencies between accounts given by the accused and, as regards the mental elements of the crime charged, as throwing doubt on the existence of that element as regards reckless rather than wilful murder.

56. Perhaps one could imagine a case in which, notwithstanding the authorities I shall mention in due course, a charge could be treated as adequate where on a trial for wilful or reckless murder the evidence sufficiently raised an issue as to intoxication and the judge, in giving the law to the jury and relating it to the facts, failed to explain the significance of intoxication in relation to the mental element of the crime. Possibly the arguments of counsel and the judge's treatment of those arguments could be such that one could say, notwithstanding the distinction often drawn between what is said with the authority of the judge's office and what is not, that the jury were adequately apprised of the relevance and possible importance of the question of sobriety, no exception having been taken to the charge in this regard. Compare the view taken by Begg, J., in dissent, on the circumstances existing in R. v. Peterkin (1982) 6 A.Crim.R. 351, a very much better case than the present in this respect from the Crown's point of view, for the judge in Peterkin did give the jury some directions about intoxication. But the charge given in the applicant's trial must be regarded as inadequate.

59. Directions about the significance of evidence of intoxication in relation to specific intent in murder have been commonplace in this State for decades. They have certainly been given for the last 30 years, and I have little doubt that they have been given for much longer than that; compare what was said by Starke, J. in R. v. O'Connor [1980] VicRp 60; [1980] V.R. 635 at 647. Until 1975 it was usual to direct juries in Victoria that intoxication itself is not a defence to a criminal charge; that the mere fact that drink has given the accused Dutch courage or stirred his passions will not assist him; and that the only way in which intoxication can provide a defence is if it is such as to make the accused incapable of committing a conscious, voluntary act or incapable of forming the specific intent. The last of these three directions stems from what was said by Lord Birkenhead, L.C. in Director of Public Prosecutions v. Beard [1920] A.C. 479 at 499-502 and the Judicial Committee in Broadhurst v. R. [1964] A.C. 441 at 461-3. The direction is, it is now well established, erroneous: Viro v. R. [1978] HCA 9; (1978) 141 C.L.R. 88 at 111-2 per Gibbs, J. (the judgment on this branch of the case being concurred in by Stephen, J. at 128-9, Jacobs, J. at 147 and Aickin, J. at 171-2). But even before Viro the practice in Victoria had changed, as a result of R. v. Sheehan (1975) 1 W.L.R. 739, and from 1975 or thereabouts on Victorian judges directed juries that intoxication is to be taken into account in considering whether murderous intent in fact existed.

62. In Viro Gibbs, J., at 111, rejected as contrary to fundamental principle the view that evidence of intoxication was relevant to specific intent only where it was evidence showing, or raising as a possibility, incapacity to form that intent. His Honour then dealt, at 112, with what the jury should be told about intoxication and specific intent:

"In a case where there is evidence fit to be considered by a jury that the accused was intoxicated as a result of the consumption of drink or drugs, it is not enough to tell the jury that the Crown must prove beyond reasonable doubt that the accused had in fact formed the requisite special intent. They should also be told that the fact that the accused was intoxicated, whether by drink or drugs or by a combination of both, may be regarded for the purpose of ascertaining whether the special intent in fact existed. It should be explained that evidence that the accused was intoxicated will not in itself entitle him to an acquittal, because a person when intoxicated may form the necessary intent, and one who has formed the intent does not escape responsibility because his intoxication has diminished his power to resist the temptation to carry it out. However, the jury should be told that if, because of the evidence as to the effect of the intoxication or otherwise, they are not satisfied that the accused did in fact have the necessary intent, they must acquit of the crime which involves that intent."

66. In R. v. Pearce and Castano (unreported, Court of Criminal Appeal, 19 December 1978) it was argued that the judgment of Gibbs, J. in Viro required that whenever there was plausible evidence of intoxication the jury be directed that if, because of that evidence, they were not satisfied that the accused did in fact have the specific intent, they must acquit of murder. It was held that there was no such rule. Young, C.J. and McInerney, J. said, at 4:

"... it will generally be sufficient to tell a jury that they must be satisfied beyond reasonable doubt that in all the circumstances, including intoxication, the accused had in fact formed the necessary intent."

"... it is in my opinion proper to direct the jury quite simply that, in deciding whether they are satisfied beyond reasonable doubt that the accused had the necessary specific intent, they should consider inter alia the evidence relating to intoxication."

74. The year 1975 and the ensuing three or four years were a period of transition in Victoria as judges came to make the adjustment from charging juries under the influence of Beard (intoxication and incapacity to form the specific intent) to charging juries in the light of Sheehan and then Viro (intoxication and actual intent). In that period the Court of Criminal Appeal had on at least two occasions to consider whether the charge adequately conveyed to the jury that the ultimate question was whether, in the light of the evidence of intoxication and all other relevant circumstances, the intent necessary for murder existed: R. v. Pearce & Castano; R. v. Leone, Makhoul & Jolly (unreported, 26 June 1978).

77. Another illustration of the need for a proper direction about specific intent and intoxication in a case of murder is R. v. Sanftl (unreported, Court of Criminal Appeal, 23 June 1983). There the trial judge had begun his direction about the relevance of intoxication by saying it was no defence to a criminal charge. The Court, while accepting that such an introductory direction was commonly given, said that it was undesirable, but none the less held the charge to have adequately conveyed to the jury that evidence of intoxication was relevant to the issue of specific intent. Not surprisingly (since I was the trial judge in that case), I have always entertained respectful doubts about the correctness of the criticism of the charge there made. In Viro, at 112, Gibbs, J. seems to have accepted that a jury might be told that drunkenness was never, of itself, a defence. In R. v. Tucker (1984) 36 S.A.S.R. 135 at 138 King, C.J., speaking in effect for the Court of Criminal Appeal, described as correct a direction that intoxication is of itself no defence. The same direction was not treated as creating any difficulty in Herbert v. R. [1982] FCA 147; (1982) 62 F.L.R. 302; the direction appears at 309. Indeed, it was said by Hunt, J., speaking for the Court of Criminal Appeal in R. v. Coleman (1990) 19 N.S.W.L.R.467 at 486, that the judge should warn the jury that intoxication does not amount to a defence. I see no reason why a jury should not be told that intoxication is of itself no defence, provided that the judge goes on to make it clear how the matter of intoxication is relevant in the case.

80. In R. v. O'Connor [1980] HCA 17; (1980) 146 C.L.R. 64 at 88 Barwick, C.J. observed that if the evidence about the taking of alcohol or another drug was capable of raising a doubt as to either voluntariness or the existence of an actual intent, the jury should be directed about the relevance of that evidence. To the same effect are the decisions of the Court of Criminal Appeal of New South Wales in R. v. Peterkin (1982) 6 A.Crim.R. 351 (a case of reckless murder), R. v. Coleman (1990) 19 N.S.W.L.R. 467, R. v. Stokes & Difford (1990) 51 A.Crim.R. 25 (where Coleman is described at 32 as showing "the usual direction concerning the relationship of intoxication to the formation of a specific intent") and R. v. Rose (1996) 87 A.Crim.R. 109. In South Australia the Court of Criminal Appeal has several times expressed the same opinion in decisions which include cases of reckless murder: R. v. Tucker [1984] 36 S.A.S.R. 135; R. v. Williamson (1996) 67 S.A.S.R.428 and the unreported decisions there cited at 447-448. The Full Court of the Federal Court endorsed the same approach to the need for a direction in Herbert v. R. [1982] FCA 147; (1982) 62 F.L.R. 302. In R. v. Pemble [1971] HCA 20; (1971) 124 C.L.R. 107 at 120-1 Barwick, C.J. stressed the importance, where reckless murder was alleged, of assisting the jury with regard to the circumstances that bore on whether the necessary foresight of consequences should be inferred; these circumstances included the sobriety or insobriety of the accused. To the same effect are the remarks of Crockett, J. (concurred in by the other members of the Court) in R. v. Allwood (1975) 18 A.Crim.R. 120 at 128-9. Reference may also be made to His Honour Judge Kelly's Book of Directions to Juries in the Criminal Jurisdiction, paras.9.1 and 9.1.1 and Glissan & Tilmouth, Australian Criminal Trial Directions, para.5-1400.

83. The judge's failure to direct the jury about the significance of intoxication as regards murderous intent and the state of mind necessary for reckless murder must bring down this verdict. It cannot be saved by the application of the proviso.

86. It goes without saying that in a murder trial it will often be desirable, and sometimes be essential, to direct the jury about the relevance of intoxication to questions or matters additional to those of the voluntary character of the act and the mental state essential to the offence. I forbear from giving examples, for to do so would, I am afraid, nowadays court the risk of attempts to elevate them into propositions about what must always be done in a given class of case.

89. Notwithstanding the failure to take exception to the charge, a new trial must in my view be ordered. For the assistance of the judge presiding at the retrial I shall say something about two other matters. The first concerns the reliance placed by the Crown at the trial last November on certain alleged conduct of the applicant after the shooting, and especially three things: his failure to take any steps to obtain help for the victim, notwithstanding that she was still alive; his pretending to Dunlop that he did not know where the victim was, when he knew she was lying wounded in the bedroom of the bungalow; his pretending to be unable to locate the revolver. These were strong features of the Crown case and, as I read the Crown's opening and the judge's summary of the final addresses, the Crown invited the jury - as it was well entitled to do - to treat the conduct as showing consciousness of guilt. One of the abandoned grounds of appeal complains of the judge's failure to direct about consciousness of guilt. The charge fails to give any directions about this, merely summarising counsel's arguments on the facts, nor does the judge appear ever to have raised with counsel the question whether a direction should be given. In view of the other deficiencies in the charge, we need not determine the effect of what was done at the trial in this regard. But the judge at the retrial should give careful attention to this matter of possible additional directions and also to the desirability of including in those directions remarks about the relevance of evidence of intoxication when the jury consider whether the applicant has by his conduct shown consciousness of guilt of the crime of murder.

92. The second additional point worth mentioning concerns the case of reckless murder. The crime requires knowledge that it is probable that death or grievous bodily harm will result from the act. The requisite state of mind is predicated upon an objective fact - a probability. The existence of this probability - the danger, as it may presently be called - seems to have been not discussed but rather assumed at the trial. Counsel for the applicant addressed on the basis that the question was whether his client had been shown to have knowledge of the (implicitly admitted) danger. There is no reason for doubting the correctness of the tacit concession. If one assumes that the weight of the loaded chamber does not affect the probable outcome, and that the cylinder is spun each time the trigger is pulled, the playing of Russian roulette with a six-shot revolver is like throwing a die in the hope of not getting a six. The applicant told the police that the firearm discharged on the fourth occasion and that he and the deceased had agreed that they would fire no more than two shots each. The probability of getting a six in four throws of a die is 671/1296: Eggleston, Evidence, Proof and Probability, 2nd ed., p.,16. But the case is not to be approached as Pascal would have approached it. Juries are not to be directed in terms of an "odds on" chance: La Fontaine v. R. (1976) 136 C.L.R. 62 at 99 per Jacobs, J.; Boughey v. R. [1986] HCA 29; (1986) 161 C.L.R. 10 at 15 per Gibbs, C.J. and at 19-22 per Mason, Wilson and Deane, JJ.; R. v. Piri [1987] NZCA 6; [1987] 1 N.Z.L.R. 66. The approach of the law has been pragmatic in matters of this kind. It has regarded as "dangerous" for present purposes the pulling of a trigger even though the probability of the discharge of the firearm was mathematically quite low. As long ago as 1839 it was accepted by the distinguished English Royal Commissioners inquiring into the criminal law (Holdsworth, History of English Law, vol.15, p.143) that it would be murder if, knowing that one only of two (or three or four) pistols set before him was loaded, but not knowing which one, the accused, for the pleasure of it, picked up one pistol, put it to another person's head and pulled the trigger with fatal results. It is for this and another reason worth quoting at length from the Commissioners' fourth report, contained in (1839) 19 Parliamentary Papers, pp.23-25:

"We proceed to offer some observations and illustrations, with a view to show that in the ordinary case where death is occasioned by the doing an act attended with manifest peril to life, the conclusion is properly not one of law but of fact, and that it is for the jury to decide whether the offender wilfully exposed life to manifest peril. Upon their affirming or negativing this proposition, the question whether he be guilty or not of murder, ought, we think, to depend. The 'mens mala', - the heart regardless of social duty, - are figurative expressions used to denote the criminal apathy or indifference with which an act is wilfully done which puts human life in peril. Whether such a peril be wilfully occasioned is a question not of law but of fact, depending on a consideration of the nature of the act done, the circumstances under which it was done, the probability that the act under those circumstances would be fatal to life, and the consciousness on the part of the offender that such peril would ensue. If A, knowing that one of two pistols was loaded with ball, but without knowing which, took one, and applying the muzzle to the head of B, pulled the trigger and shot him, having by chance taken the loaded one, this would be murder; for A wilfully placed B's life in peril by his act. Here it was an even chance whether B was killed or not, - but it is evidence that the principle and the law founded upon it must be the same, even although the probability were less; as for instance, though the selection were made from three or four (one only being loaded) instead of two; although no doubt the degree of probability may be an important circumstance in ascertaining the fact of the intention of the offender to put in danger. Neither the nature nor the degree of peril can possibly be the subject of legal definition; as well the knowledge of the offender that such peril exists, as the wilfulness of his act in proceeding notwithstanding that knowledge, are necessarily matters of fact, the existence of which is for the consideration of the jury. It may no doubt frequently appear, from the particular facts found by the jury, that the offender did wilfully expose human life to manifest peril, but still, as in numerous instances the court may be unable to draw the conclusion from the mere facts, it seems to be clear, as a general position, that the question is one of fact for the jury. In particular, it may be noticed that the fact of knowledge on the part of the accused, of the peril he was about to occasion by his act, must very frequently depend upon mere circumstantial evidence, upon the effect of which it is the province of a jury to decide. Malice, in the legal sense of the word, means, according to Mr Justice Foster, the mala mens, the heart bent upon mischief, as evidenced by general symptoms and circumstances. When this definition is applied to an ordinary case of implied malice, as where a man throws a heavy stone over a wall into a street, through which many persons are likely to be passing, malice thus defined assumes a more precise meaning. The offender in such case does not intend the death or hurt of any one in particular, - the act may kill a stranger or may do no hurt at all: there is nothing, therefore, to make it criminal, and so to constitute implied malice, but the peril of destroying life wilfully occasioned. The 'heart bent on mischief' must in this case, and it is apprehended in all others, in the absence of a special definite rule of law which defines what is meant by implied malice, consist simply in the disregard of life manifested by exposing life to peril, although no malice in the popular sense of the term exist against any individual. It is the wilful exposure of life to peril that constitutes the crime. If one, knowing cakes to contain deadly poison, were to scatter them in a public street, and death were the consequence, the act would be murder, for the offender knew that what he did was likely to occasion loss of life; but if one did the same act without knowing or suspecting that the cakes contained poison he would be innocent in law as well as in morals, for, although the act be in truth equally dangerous whether knowledge exist or not, in the one case great peril to life is wilfully occasioned, and in the other not. And so in all cases it is essential to the criminality of the act, both in law and morals, not only that the act should in its own nature under the circumstances be attended with peril to life, but that the offender should be aware of such peril. Where the offender does an act attended with manifest danger to life wilfully, that is with knowledge of the consequences, he may properly be said to have the 'mens mala, or heart bent upon mischief'; but if he does an act, however dangerous it be in its own nature, without any knowledge or suspicion of its tendency, that is, if he does not wilfully place life in peril, he cannot be said to show the mens mala, or heart bent on mischief. In such cases, therefore, it is plain that implied malice in law, according to Mr Justice Foster's notion, means nothing more than the state or disposition of the offender's mind when he wilfully does an act likely to kill or wilfully intends to put life in peril; but this is a conclusion to be drawn from the facts, viz., the probability that death will result from the act under the particular circumstances of the case, and the intention of the offender to occasion the peril regardless of consequences. ... In so atrocious a case as that of scattering poison, above, supposed, no doubt could well be entertained as to the guilt of the offender, dependent as it is on the circumstances of the act being in its own nature attended with manifest peril to life, and of the criminal intention on the part of the offender wilfully to place life in peril. In other instances, although the evidence be not so cogent, it will be found that the difficulties arise not from the necessity of calling in aid the operation of any new principle, or the applicability of those adverted to, but merely from doubt arising upon the matter of fact. The degree of probability that death will ensue from the act is susceptible of every variety from moral certainty to the remotest possibility; - the intention of the actor is also capable of degrees, closely indeed dependent on the greater or less degree of probability of mischief, from a direct intention to destroy life, to perfect innocence. And such varieties, it may be observed, are not merely in proportion to the risk, but depend also on the knowledge and consciousness of the risk to life. These elements are obviously matters of fact, to be decided as facts; they are beyond the reach of definition, and when probability of loss of life from doing the act, the knowledge of that probability on the part of the offender, and his criminal intention to occasion the risk have been determined in fact, the principle of law applies. It is upon these principles that the different degrees of legal guilt, in the absence of a direct intention to kill, depend. If poison instead of being, as in the case above supposed, used in such a manner as to occasion manifest danger to life, and with consciousness and knowledge of such danger, were to be used for the destruction of rats by placing it in a closet locked for greater precaution, the loss of human life would still be a barely possible, although most improbable, result from the act; a thief might by possibility break open the closet, and mistaking the poison for food, eat it and die. Here the fatal result would occur without the least blame, legal or moral, on the part of him who thus laid the poison. But between these two extreme cases lie an infinity of others, in respect of which important legal distinctions are requisite, although they may vary almost imperceptibly from each other, and particularly as to the degree of risk to human life which may be incurred. ... [T]he question is properly one of fact, as dependent on the existence of peril, and the design to expose to peril."

96. The example of playing with pistols was given again by the Commissioners in their seventh report, printed in (1843) 19 Parliamentary Papers, p.24:

"[L]et it be supposed that a person knowing that one of two pistols is loaded, without knowing which, points one of them at the head of another person and draws the trigger, and (the loaded pistol having in fact been taken) shoots that person, the offender (consistently with the foregoing principles) ought to be deemed as fully responsible for the consequences as he would have been had he been aware that he took the loaded pistol, the fatal result being the same, and the mens rea existing in the one case as well as the other, without any such difference as to afford any substantial distinction for legal purposes. It seems to be clear that the application of the same principles would tend to the same conclusion, if instead of taking at hazard one of two pistols, one of three or of four, or of any other definite number, were taken, one only of that number being loaded. The probability of a fatal result would be diminished as the number from which the selection was made was increased, but still there would be a wilful risking of life attended with a fatal result, and as it seems a total absence of any intelligible principle of distinction for penal purposes. The state of the offender's mind in thus exposing life to danger seems clearly to fall within the legal notion of mens rea for all purposes of plenary responsibility."

100. Text writers have adopted the view that Russian roulette is dangerous in the sense now under discussion: Glanville Williams, Criminal Law - The General Part, 2nd ed., pp.59-61; Fisse, Howard's Criminal Law, 5th ed., pp.59-61; Gillies, Criminal Law, 4th ed., pp.632.3; Laws of Australia, Subtitle 10.1, para.68. They give the example of Russian roulette. So did White, J. of the Supreme Court of South Australia in R. v. Marshall (1986) 43 S.A.S.R.448 at 477, although the example given is of firing once with four rounds in a six-round chamber. In my own view it is dangerous in the necessary sense to fire once with one round in a six-round chamber. As a matter of interest, an example of Russian roulette in civil litigation where the player tried to cheat will be found in National and General Insurance Co. Ltd. v. Chick [1984] 2 N.S.W.L.R. 86, while in R. v. Lamb [1967] 2 Q.B. 981 the accused was charged only with manslaughter where he had pointed a revolver at a friend and pulled the trigger as a joke, the cylinder containing two cartridges, neither being opposite the barrel and the accused not knowing that the pulling of the trigger would rotate the cylinder. What Lamb did was of course very different from spinning the cylinder and I mention the case only as a matter of interest. In the criminal law, actual examples of Russian roulette are hard to find except in the United States. A Singaporean case in 1953 is mentioned by Glanville Williams, p.60, but there the charge was not one of murder. In 1946 the Supreme Court of Pennsylvania had before it an appeal by a young man of 17 convicted of the reckless murder of a 13-year-old boy with whom he had played "Russian poker": Commonwealth v. Malone 354 Pa. 180; 47 A. 2d 445 (1946). In Pennsylvania murder was divided into degrees by a statute which did not modify the common law rules defining murder, and in Malone the law was said to require for reckless murder an act of gross recklessness from which the defendant must reasonably anticipate death as likely to result. The youth had placed a single cartridge in one of the five chambers of a revolver - he swore it was the chamber immediately to the right of the firing chamber - and placed the revolver against the boy's side. He had then, without any spinning of the chamber, pulled the trigger three times, the third pull resulting in a fatal wound. The Supreme Court of Pennsylvania had no hesitation in affirming the conviction, observing (at 449) that the act had been done "in reckless and wanton disregard of the consequences which were at least sixty per cent certain from his thrice attempted discharge of a gun known to contain one bullet and aimed at a vital part of Long's body". In a later case - Commonwealth v. Ashburn 459 Pa. 625, (1975) - a hopeless attempt was made by the defendant to treat this remark as defining murder in the second degree as requiring an act at least sixty per cent certain to result in the death of another, and the Supreme Court of Pennsylvania said, at 633, that the application of the principles of law "does not depend on any precise mathematical calculation of the probable consequences of the defendant's acts". This accords with what was said in the passages cited above from La Fontaine and Boughey and in the New Zealand case of Piri. In Ashburn, according to the defendant he had produced a revolver to show to his companion, who then brought up the subject of Russian roulette; he himself had then loaded one chamber, pointed the gun at his companion and pulled the trigger twice; the gun had discharged on the second pull.

103. Ashburn is one of the cases relied on by the Supreme Court of Massachusetts in Commonwealth v. Ward (17 December 1997, Docket No. SJC-07215), where the conviction was reduced to murder in the second degree in a case where a defendant with a blood alcohol content of 0.39 at the time of his arrest had loaded two cartridges into a five-chamber revolver, pointed it at his companion's head and pulled the trigger; he had been playing Russian roulette with others not long before. The Court cited Commonwealth v. Atencio 345 Mass. 627 at 631 (1963): in Russian roulette "the outcome is a certainty if the chamber under the hammer happens to be the one containing the bullet". In People v. Haack 396 Mich. 367; 240 N.W. 2d 704 (1976) the Supreme Court of Michigan held that the judge had rightly accepted a plea of guilty to second degree murder from a defendant who, during the plea-taking colloquy, had claimed that he was carrying a revolver for a friend which had four bullets in a six-chamber cylinder. He placed one of the empty chambers under the hammer and the other immediately to the left of it. In order to frighten a man with whom he was having an argument, so the defendant's story ran, he pointed the gun at him and pulled the trigger and to his surprise the gun discharged, he having been mistaken in his belief about the direction in which the cylinder would rotate. The Court cited authority, at 709, for the view that the mental state necessary for murder could be inferred where the defendant intentionally set in motion a force likely to cause death or grievous bodily harm. In State v. Mackey 687 So. 2d 465 (1996) the Third Circuit Court of Appeal of Louisiana upheld a conviction of second-degree murder where the defendant had decided to play Russian roulette with an unwilling room-mate. He put a revolver with one bullet in the chamber to his own head and pulled the trigger. He then turned the gun on his companion and pulled the trigger, shooting him in the neck.

106. A helpful summary of the law of the various American states as regards murder by recklessness will be found in the draft Model Penal Code and Commentaries of the American Law Institute, published in 1980. That Code defines murder as including criminal homicide committed recklessly under circumstances manifesting extreme indifference to the value of human life. Reckless murder at common law is discussed at p.15 of the publication. At p.22 the draft Code is said to reflect the common law in so far as it includes cases of homicide caused by extreme recklessness, and Commonwealth v. Malone is cited as an example of prior decisional law. This provision of the draft Code and this part of the Commentary are referred to in Tison v. Arizona [1987] USSC 61; 481 U.S. 137 (1987): see 157 in the majority opinion and 167 in the dissenting opinion. (The case concerned whether the death sentences imposed were constitutionally permissible having regard to the Eighth Amendment - "cruel and unusual punishments".)

109. I acknowledge my indebtedness to Ormiston, J.A., who drew my attention to a number of the cases from the United States.

112. Earlier I said that the quotation from the Royal Commissioners' fourth report is of interest for two reasons. The first is its discussion of the game with pistols. The second is its discussion of the nature of the questions which arise and the role of the jury in reckless murder cases. The texts cited above give the impression that, as a matter of law, it is probable that death or grievous bodily harm will result from the act of pointing a revolver at another person's body and pulling the trigger after spinning a cylinder containing one cartridge. Similarly, the Commissioners, in their fourth and seventh reports, appear to assert, as a matter of law, that pulling the trigger in the examples which they give of playing with pistols is dangerous to life in the necessary sense so that if the actor has knowledge of the danger he will be a murderer if death results. On the other hand, the Commissioners are at pains to emphasise that, as it is put, for example, at p.24 of the fourth report, "probability of loss of life from doing the act" and "the knowledge of that probability on the part of the offender" are to be "determined in fact".

115. As Barwick, C.J. said in Gammage v. R. [1969] HCA 68; (1969) 122 C.L.R. 444 at 451, "it is for the jury to be satisfied of the elements of the murder charged and not for the presiding judge". When text writers describe a game like Russian roulette as dangerous, they are not gainsaying the role of the jury. They are seeking to convey that a reasonable jury, properly instructed, should find that the necessary element of danger existed in such a case. To make the process of reasoning more explicit, they are asserting a proposition of law, namely, that "probable" as contrasted with "possible but not likely" (Crabbe at 469-70) means a substantial, or real and not remote, chance, whether or not it is more than 50 per cent (Boughey at 21 per Mason, Wilson and Deane, JJ.) and saying that accordingly in the case put a reasonable jury should regard the chance as substantial.

118. The present case is not, according to the applicant, one of a single pulling of the trigger. The only evidence of the playing of a game is that afforded by the admissions of the applicant himself, and that includes admissions that it had been agreed that they should pull the trigger no more than twice each and that the revolver discharged when the trigger was pulled the fourth time. It may be that on the re-trial the danger as opposed to the knowledge of it will again be conceded. There might be some practical, if not theoretical, difficulty, in the accused man's accepting that the jury should find the dangerous act necessary for manslaughter while disputing that the dangerous act necessary for reckless murder should be found. Compare Wilson v. R. [1992] HCA 31; (1992) 174 C.L.R. 313 at 333 per Mason, C.J., Toohey, Gaudron and McHugh, JJ. and at 339-40 per Brennan, Deane and Dawson, JJ. I would not expect a jury to have any difficulty in concluding, on the evidence placed before it last November, that the applicant's act was dangerous in the sense necessary for reckless murder. This is on the assumption that they were not satisfied that the game was an invention.

121. On the re-trial if it is not conceded by the accused that it was probable that death or grievous bodily harm would result from his act, the directions of law given to the jury on this point could be quite short. The judge should direct them in accordance with Crabbe, as explained in Boughey.

126. I agree, for the reasons given by Brooking, J.A., that this application must be allowed and a new trial had.

136. The formal order of the Court will be that the application for leave to appeal against conviction is allowed. The appeal is treated as having been heard simultaneously with that application and it too is allowed. The conviction recorded is quashed. The Court directs that a new trial be had on the charge of murder.

139. The Court will place on its record the notice of abandonment of application for leave to appeal against sentence signed by the applicant today.