Pfennig v The Queen

The account of the facts which follows is largely taken from the reasons for judgment of Mohr J. in the Court of Criminal Appeal. Such departures as there are from his Honour's statement of the facts are occasioned by reference to the evidence and to our understanding of it.

Michael left his home in Murray Bridge at about 1 p.m. on that day to go fishing. He rode his bicycle. He took with him a fishing rod and canvas bag. He was wearing a red and blue striped cotton knit shirt, shorts, underpants and blue thongs. He did not take swim shorts with him. It is possible, though unlikely, that he was wearing swim shorts as underpants. He had no towel. He was accompanied by the family dog. He arrived at Sturt Reserve between 1 p.m. and 1.30 p.m., probably close to 1.30 p.m.

Sturt Reserve is on the same side of the Murray River as the town. The Reserve consists of an area of lawn, running down to the bank of the river. There is a large rowing shed, shelter sheds, a wharf, swings, barbecue facilities and a kiosk which serves refreshments.

Michael was seen by various people at Sturt Reserve. No witness spoke of seeing a boy who could have been Michael at the Reserve later than just before 3 p.m. on 18 January. The evidence of the witnesses as to time was not precise, a matter to be expected in light of the fact that they were giving evidence at the trial of events which had happened more than three years earlier.

Mr. Haines, who knew Michael by sight, saw a boy, probably Michael, at 2.25 p.m. This seems to have been the last occasion on which Michael was seen on 18 January. Some witnesses spoke of seeing Michael a few minutes before 3 p.m., but on their own admission, and in the opinion of the trial judge, their recollection as to the precise time was vague.

The next evidence of Michael's whereabouts relates to Thiele Reserve which is up river and on the other side of the river. Later in the afternoon of 18 January, Michael's bicycle, fishing rod, bag and thongs were found neatly stacked against perma-pine railings at Thiele Reserve. His shirt was found in willows growing on the bank of the river at the upstream end of the Reserve. Nearby was a piece of twine tied to a tree stump. It is possible that Michael's dog had been tied with it. Michael was not seen by anyone at Thiele Reserve.

Evidence was given by Byron Drechsler, who was about fifteen years old at the time of Michael's disappearance, to the effect that he had at one stage believed that he had seen Michael riding across the bridge away from Sturt Reserve, with his dog and fishing rod, at about 3.10 p.m. on the afternoon of 18 January. However, other evidence clearly established that any sighting of Michael by Dreschler must have been prior to 1.22 p.m. on that day and Dreschler, in the course of his evidence, expressly conceded that his original belief as to time was mistaken.

After Michael's disappearance had been notified, an extensive search was made. Police divers carried out an underwater search of the river bottom and banks and found nothing. The divers and the officer in charge testified that there was only a remote possibility that Michael's body was in the river but undiscovered. The evidence was that in a period of twenty-four years there had been 137 operations conducted to recover bodies from inland waters in South Australia and on only one occasion had there been a failure to discover a body and that was in flood conditions.

The evidence was significant because it meant that the jury was entitled to find that Michael's body was not in the river. There was evidence from Dr. Manock, a pathologist, that, if a person drowned, the body would sink to the bottom after death and would remain there for two or three days. It would then rise to the surface and float.

As well as the river search, an extensive land search was made without success. Apart from the official search, unofficial searches of the river and its banks were made for some time. Some searching continued for a period of three years without success.

At the trial, counsel for the appellant conceded that there were only two possible explanations for Michael's disappearance: either that he drowned or that he was abducted. The presence at Thiele Reserve of the bicycle and the other articles associated with Michael indicates that, if Michael drowned, it was at Thiele Reserve. Although his mother gave evidence that he told her that he was going swimming, the evidence points to the inference that he did not swim at the Reserve. He had earlier declined an opportunity to go swimming that day; he disliked swimming alone and he did not swim that day at Sturt Reserve which was one of his favourite swimming spots. There were people at the Reserve, including two boys whom he knew, at the time when he would have arrived there, if he arrived at all. The two boys were fishing at the Reserve yet they did not see him, let alone speak to him. He took no towel; it seems that he did not take his swim shorts and his father testified that Michael did not like Thiele Reserve. He had been forbidden by his parents to ride his bicycle across the bridge. There was evidence that he was a competent swimmer.

The appellant's movements on 17 and 18 January 1989

The evidence led by the Crown as to the appellant's movements on 17 and 18 January was not in dispute. According to Mrs. Pfennig, who was then living with the appellant but has since been divorced, he left home on 17 January driving his white Kombi van and returned on 20 January. The purpose of the trip was to study physics for his next teaching position. The van was equipped with a television set and radio and sleeping and cooking facilities.

One witness, Mr. Smart, gave evidence that he spoke to the appellant at about 4 p.m. on 17 January. He asked Mr. Smart if he knew of any place where it was possible to swim in the nude. Mr. Smart gave him directions to Thiele Reserve. Shortly afterwards, the appellant joined two children who were swimming from the wharf at Sturt Reserve. He asked them if there were any interesting places to visit and invited them to join him in the van if they wished to accompany him. They declined his invitation.

On the next day, the appellant was seen at Sturt Reserve. Mrs. Rooke observed him sitting in his van, watching her and two young children with her. Mr. Haines, who saw Michael Black at Sturt Reserve at about 1.55 p.m., saw a man speaking to a boy, in all probability Michael, at about 2.25 p.m. That was the last occasion on which Michael was seen at Sturt Reserve. The description given of the man by Mr. Haines was not an accurate description of the appellant and may not have been an accurate description of the clothes he was wearing. Mr. Haines said he could not recall seeing any facial hair and he described the man's hair as being a bit longer than short back and sides. The appellant had a moustache and had just had a haircut with the result that, according to his daughter, his hairstyle would have been described as short back and sides. The appellant claimed that he left Sturt Reserve by 2.30 p.m. though he conceded that he left by 2.45 p.m. at the latest.

At about 2.45 p.m., Mr. Toogood, a council employee, saw an old white Kombi van leaving the Reserve at an unsafe speed travelling towards the main street of Murray Bridge. Continuing in that direction the van would have crossed a bridge to the south of the river where Thiele Reserve is located. Mr. Toogood said that he had earlier seen the same or a similar vehicle travelling in the direction of Sturt Reserve. The trial judge regarded Mr. Toogood's evidence as somewhat unsatisfactory and instructed the jury that it would be unsafe to rely on his evidence except to the extent that he testified to seeing an old white Kombi van, whose engine made a typical VW noise, and that the van could have been the appellant's van.

Mrs. Gould, who then lived in a house overlooking Thiele Reserve, heard a commotion below her house at about 3 p.m. to 3.30 p.m. The noise was caused by a vehicle with a loud engine with its wheels spinning in gravel. She heard a dog barking in an excited manner. She gave evidence of an experiment in which she was able to identify more than twelve months later the noise from the appellant's van as being similar to the noise which she heard on the afternoon of 18 January 1989. Michael's dog was seen in the vicinity not long after 3.20 p.m. Michael's belongings had been placed against the perma-pine railings before that time.

Mr. Jones gave evidence that he went to Thiele Reserve waterskiing on 18 January at 4 p.m. or thereabouts. He saw a Kombi van on the opposite side of the road that runs virtually parallel to the river bank. It was parked off the road opposite the boat ramp that he was using, in the vicinity of the place where Michael's bicycle and gear were found. Mr. Jones' description of the vehicle fitted the appellant's van. Mr. Jones originally gave the date as 17 January. Some question arises about the date on which he went on to Thiele Reserve. If the evidence of the Haebich children and Mr. Smart that the appellant was at Sturt Reserve in the afternoon of 17 January is accepted, it seems that Mr. Jones visited the Reserve on 18 January, that being a day on which he considered that he was there, if he was mistaken about 17 January. Another witness, Mr. Neindorf, saw a white Kombi van in the car park area of Thiele Reserve on 18 January. But no witness saw a vehicle entering or leaving that Reserve which was likely to have been used by an abductor.

Michael was not seen at Sturt Reserve at any time after Mr. Toogood saw the white van leave at about 2.45 p.m. This fact is significant as there were people at Sturt Reserve who knew him. Mrs. Regnier and Miss Giles, who closed the canteen at 3 p.m. or 3.15 p.m. and left the Reserve then, did not see him at that time.

There was evidence from the appellant's daughter Petra Pfennig, Nicola Davis and Dianne Davis about the appellant's statements to them about his dealings with Michael Black at Sturt Reserve on 18 January. He admitted to them that he had spoken to Michael at the Reserve. He also said that he had patted Michael's dog. Petra Pfennig said that the appellant had told her that he saw Michael twice that day. She said that the appellant told her that Michael was trying to scale a fish with a thong so the appellant lent him a knife to scale the fish and when Michael brought the knife back the appellant asked him to put it in the van. The appellant said that his daughter must have misunderstood him. On the other hand, Nicola Davis said that the appellant told her that he lent Michael a knife to do something with a tangle in his fishing line.

It follows that, on the prosecution case, there was evidence that the appellant was at Sturt Reserve when Michael was present, that he was in conversation with Michael who, on one occasion at least, was close to the appellant's van, that he left Sturt Reserve at or about the time when Michael was last seen at the Reserve. And there was evidence from Mrs. Gould from which it could be inferred that the appellant's van was at Thiele Reserve after Michael's belongings were placed near the perma-pine railings. The appellant's account of his movements after he left Sturt Reserve was not confirmed by any other evidence.

The appellant pleaded guilty in 1990 to having abducted and raped H., a thirteen-year-old boy, at Port Noarlunga. The appellant denied that he abducted H. for the purpose of raping him and there is a question whether the appellant would have released the boy eventually.

H. gave evidence that he was riding a bicycle past the appellant's white Kombi van at Port Noarlunga at 1.30 p.m. on 30 December 1989, that the appellant inveigled him into entering the van and then closed the sliding door, refusing to allow H. to leave. The appellant then brought the bicycle into the van but later left it at the top of a cliff, first wiping it with a cloth. The appellant bound, gagged and blindfolded H. and held him as a prisoner in the van and later in the appellant's house. The appellant sexually molested the boy in a variety of ways and subjected him to various indignities. These incidents occurred during the afternoon of the day on which the boy was abducted, the following night and morning. When the appellant was absent at lunchtime that day, the boy managed to escape and telephone his father and the police. The appellant was arrested when he arrived home.

The appellant's former wife, Sandra Pfennig, gave evidence that he telephoned her twice after his arrest and said that he wanted a solicitor. In the first conversation at about 1 p.m. or a little later, he told her that he had been arrested for kidnapping and sexual assault. In the second conversation at about 5 p.m. after she asked why he had done it, he said that he was lonely and that he had been thinking of "it" on and off for the past twelve months. He said that he had just driven around, found somebody and threw "them" into the van.

Detective Hirlam gave evidence of the following conversation with the appellant when he was arrested. Hirlam said, "We have had an allegation from a young boy that you brought him here against his will and tied him up, is that correct?"

Hirlam: He managed to escape while you were gone. Why did you do it?

Appellant: Loneliness. If he is in there, can you release him?

Appellant: Just fondling. I just wanted someone to hug.

The appellant gave a somewhat different version of this conversation in his evidence.

The trial judge (Cox J.) conducted a lengthy voir dire in advance of the presentation of the Crown case with a view to determining the admissibility of the H. evidence. Ultimately it was agreed by counsel that a decision on that evidence should not be given before the commencement of the trial. Counsel for the appellant indicated that he would be able to cross-examine Crown witnesses without alluding to the H. incident. The trial then commenced; the Crown did not open on the similar fact evidence.

The trial judge, after hearing the evidence of almost 100 witnesses, heard further submissions on the admissibility of evidence of the H. incident. His Honour, while acknowledging the existence of inconsistencies in the evidence of the witnesses, made certain specific findings:

Michael Black probably rode across the bridge before going to Sturt Reserve on 18 January. It is very unlikely he went as far as Thiele Reserve. He was at Sturt Reserve by 1.30 p.m. He did not ride his bicycle to Thiele Reserve after that.

Toogood saw Michael going to Sturt Reserve. He also saw a white Kombi van on two occasions that afternoon. It was probably the appellant's van that he saw on each occasion.

Haines saw Michael at Sturt Reserve about 1.55 p.m. and 2.25 p.m. The man to whom the boy was speaking at 2.25 p.m. was in all likelihood the appellant.

Bernardi saw the bicycle leaning against the railing near the toilet block at Thiele Reserve before 3.26 p.m. Probably the fishing gear was also there then, and both the bicycle and the gear remained there afterwards.

Drowning was not a reasonably possible explanation for Michael's disappearance. The theory encounters a number of improbabilities - that he went to Thiele Reserve despite his known dislike for the place, that he did not observe or speak to the children there whom he knew, that he was not noticed by any of the people there, that he went swimming alone though he had not intended to swim when he set out, and that no body was discovered after a search of the river.

The only rational alternative to drowning was abduction, a matter which was conceded by counsel for the appellant. Any abduction was likely to have been for a sexual purpose. And, if he was abducted and sexually assaulted, it must be inferred that he was murdered by the person or persons who abducted him.

The prosecution case was presented on the footing that abduction might have occurred as a result of Michael being lured into the clutches of his abductor. His Honour's references to abduction are to be understood as including such a situation.

His Honour considered that Michael might have been abducted at any one of a number of places but concluded that it was likely that he left Sturt Reserve in a vehicle driven by his abductor, the vehicle being commodious enough to take Michael's bicycle, gear and dog. This conclusion, coupled with the evidence concerning the appellant's presence in Sturt Reserve at the relevant time, including his admissions that he spoke to Michael, and the evidence relating to his van and its movements, provided the foundation for the reception of evidence of the H. incident. In other words, the appellant had the opportunity to abduct and murder Michael. The H. evidence revealed the appellant's propensity to abduct a young boy for sexual purposes and by means which were likely to have been adopted by the abductor in the present case. On this footing, the trial judge considered that the H. evidence was admissible.

In my opinion, the evidence of the accused's presence and behaviour in the area on January 17 and 18 including his dealings with Michael Black on the afternoon of January 18, the opportunity his van gave him to carry out an abduction, the vehicle and dog noises heard coming from Thiele Reserve, and the very unusual nature of the particular crime in question, combine to render the evidence of the [H.] abduction admissible in proof of the identity of Michael Black's abductor. The nature of the crime in this case is of great importance. The more unusual the type of crime, the more difficult it may be to accept mere coincidence as a reasonable explanation. Less additional evidence may be needed in such a case to prove a connexion between the propensity evidence and the alleged instant crime and so establish overall the requisite high degree of proof. His Honour later said: It would, in my opinion, be an affront to common sense to postulate two persons in Michael Black's vicinity at Murray Bridge, and both almost certainly at Sturt Reserve, about the same time that afternoon, each with a propensity to kidnap and sexually assault young boys and each having the physical means that afternoon of doing so, one of them befriending the boy and lending him a fishing knife and the other within a fairly short space of time but quite independently engaging, presumably, in some kind of pre-abduction dealing with him, however brief, and both leaving Sturt Reserve in separate vehicles at much the same time.

His Honour pointed to the similarities in the pattern of conduct which must have been engaged in by Michael's abductor and by the appellant in the H. incident, particularly leaving the victim's belongings in a way that was calculated to lay a false trail and suggest accidental drowning - compare the disposal of H.'s bicycle at Port Noarlunga. However, his Honour noted that the fact that there was no murder in the case of H. was an important dissimilarity which would normally tell against the admission of the H. evidence. But the inconclusiveness of the evidence as to the appellant's intentions with respect to the ultimate fate of the boy militated against that treatment of it.

In the result, his Honour considered that it was a case about improbabilities, sufficiently strong to compel a factual conclusion adverse to the appellant. Further, his Honour thought that the basic similarity of the two crimes, including the appellant's behaviour at Sturt Reserve and the laying of a false trail with the bicycle, gave the two incidents an "underlying unity". Although the evidence was very prejudicial, its probative force was so strong that it should be admitted.

The test of admissibility for propensity evidence is exceptionally high. It has been said that the trial judge "must apply the same test as a jury must apply in dealing with circumstantial evidence, and ask whether there is a rational view of the evidence that is inconsistent with the guilt of the accused." [1] Although his Honour seems to have regarded the question of admissibility as one governed by the principle applicable to circumstantial evidence, he also applied the principles applicable to propensity or similar fact evidence. Thus, his Honour held that the probative value of the H. evidence transcended its prejudicial effect and, to put it another way, its probative force was so strong that it would not be unfair to admit it. 1. Hoch v The Queen (1988), 165 C.L.R. 292, at p. 296.

Although the admissibility of the H. evidence fell to be determined before the appellant's case was presented, it is convenient to state the effect of the appellant's evidence at the trial as it relates to Michael. The appellant stated that he had driven his Kombi van to Sturt Reserve on 17 January. He said that he first saw the boy, who, he acknowledges, must have been Michael, fishing near a willow, when he was sitting or leaning against a small retaining wall. The appellant stated that he saw Michael trying to beat a fish to death with a thong. The appellant yelled out, "Hey, haven't you got a knife?". The appellant went to his van, obtained a knife, went over to Michael, showed him how to dispatch the fish and handed over the knife. There was a conversation about bait and the hook size. The conversation, which lasted a matter of minutes, took place at the water's edge where Michael had his gear. The appellant saw a dog there which he identified from photographs as Michael's dog and asked Michael to bring the knife back when he was finished.

The appellant then returned to a position in front of the van which was 10-15 m. away from the place where the first conversation had taken place. Subsequently, Michael returned the knife within ten minutes and the appellant said, "Just put it back in the van." When Michael came back from the van, he said, "You have got a TV set in there." He went back to the van to see whether the cricket was on television in response to a question from the appellant who asked whether the cricket was on. Shortly after, he returned to the willow where he had been fishing.

The appellant said that shortly afterwards he packed up and went into town in the van. The trial judge summarized the appellant's evidence in response to the question whether it was 2.45 p.m. when he left as follows:

That would be close to the latest time. I believe that I left earlier than that. I now think it was 2.15 to 2.30. I would have left before 2.30. It could have been 2.45, but it is unlikely to have been as late as that. It is unlikely that I left via Sturt Reserve Road but I can't exclude that. He denied that he went to Thiele Reserve on 18 January.

The trial judge's summing-up with respect to the H. evidence

Cox J. clearly instructed the jury on a number of occasions that the H. evidence was relevant, at the most, to the question of the abductor's identity and that it could not be used to resolve any doubts they may have had on the question whether there was an abduction at all. His Honour told the jury that, if the Murray Bridge evidence did not satisfy them that Michael Black was abducted and murdered, they must find the appellant not guilty. If, however, they rejected drowning as a reasonable possibility and were satisfied that Michael was abducted and murdered, it would be open for them to take into account the H. evidence in deciding whether the Crown had proved that the appellant had committed the abduction and murder. Earlier, the trial judge told the jury that, if Michael was abducted, the most likely motive was a sexual one and, if he was killed afterwards, the likely explanation might be a wish to silence a potential witness.

The trial judge stated that the Crown pointed to the unusual type of crime which had been committed - the abduction of a ten-year-old boy for sexual purposes - and relied on the H. evidence as showing that the appellant had a disposition to abduct and sexually assault young boys. This, along with other circumstantial evidence, was relied on to prove that the appellant was the abductor and the murderer. His Honour went on to say that, if he was not, there were two persons in Michael's vicinity at Murray Bridge, both almost certainly at Sturt Reserve, each with a propensity to kidnap and sexually assault young boys and having the physical means that afternoon to do so. According to the Crown, it was an affront to common sense to postulate such a possibility.

His Honour referred to the Crown argument that the jury could infer from the H. evidence that the appellant intended to kill H. because he knew too much and went on to remark that the jury might think that the evidence was inconclusive on that question. His Honour observed that, in that event, there was possibly a very important difference between the H. incident and what happened at Murray Bridge and this would require the jury to think very carefully before reaching a conclusion on the question of identity adverse to the appellant.

The appellant's basic challenges to the reception of the H. evidence are that propensity reasoning is an inadequate foundation for its admission and that, even if admission on that footing might be legitimate in some circumstances, no adequate factual foundation was established in the present case. The appellant submits that the trial judge's directions transgressed the general principle that it is not competent for the prosecution to adduce evidence tending to show that the accused has been guilty of criminal acts other than those covered by the indictment for the purpose of leading to the conclusion that the accused is a person likely from his criminal conduct or character to have committed the offence for which he is being tried [2] . The appellant further submits that there was lacking in the present case that underlying unity between the offences charged and the conduct disclosed by the H. evidence which is essential before similar fact evidence amounting to propensity evidence can be admitted [3] . In other words, there was an absence of that striking similarity to which Gibbs C.J. referred in Sutton v The Queen [4] when speaking of similar fact evidence adduced to establish identity.

1. Makin v Attorney-General (N.S.W.) [1894] AC. 57, at p. 65. 2. Moorov v H.M. Advocate [1930] JC. 68. 3. (1984) 152 C.L.R. 528, at p. 535.

Contemporary discussion of the problems attending the reception of similar fact and propensity evidence has its origins in the statements of principles by Lord Herschell L.C. in Makin v Attorney-General (N.S.W.) . Lord Herschell enunciated two relevant principles which have had continuing influence. The first was that the prosecution cannot "adduce evidence tending to shew that the accused has been guilty of criminal acts other than those covered by the indictment, for the purpose of leading to the conclusion that the accused is a person likely from his criminal conduct or character to have committed the offence for which he is being tried" [2] principle was later described as a "fundamental" principle in the law of evidence [5] .

1. Makin v Attorney-General (N.S.W.) [1894] AC. 57, at p. 65. 2. Maxwell v Director of Public Prosecutions [1935] AC. 309, at pp. 317, 320, and see also Burrows v The King (1937), 58 C.L.R. 249, at p. 253, per Latham C.J.

The second principle was that "the mere fact that the evidence adduced tends to shew the commission of other crimes does not render it inadmissible if it be relevant to an issue before the jury, and it may be so relevant if it bears upon the question whether the acts alleged to constitute the crime charged in the indictment were designed or accidental, or to rebut a defence which would otherwise be open to the accused" [6] . It is evident that there was an element of tension between the two principles as thus stated. That tension was partly due to the ambiguity inherent in the use of the word "relevant". The second principle, as expressed by Lord Herschell, seemed to imply that propensity evidence was not as such relevant to the determination of the crime charged, rather that it was relevant to that determination but inadmissible for some overriding policy reason, i.e., that in many cases its prejudicial effect would outweigh its probative force.

Subsequent development of the Makin principles in the English cases

The judgments in succeeding cases suggested that the second principle was directed to justifying the admissibility of similar fact or propensity evidence when it tended to disprove a denial or defence which was otherwise available to the accused [7] , e.g., an alibi [7] or the absence of guilty intention [8] . In Noor Mohamed v The King [9] , the Judicial Committee of the Privy Council specifically rejected the view which Lord Goddard L.C.J. had expressed in R. v Sims [10] . His Lordship had said [11] :

If one starts with the assumption that all evidence tending to show a disposition towards a particular crime must be excluded unless justified, then the justification of evidence of this kind is that it tends to rebut a defence otherwise open to the accused; but if one starts with the general proposition that all evidence that is logically probative is admissible unless excluded, then evidence of this kind does not have to seek a justification but is admissible irrespective of the issues raised by the defence, and this we think is the correct view. It is plainly the sensible view. 1. Thompson v The King [1918] AC. 221. 2. Thompson v The King [1918] AC. 221. 3. Noor Mohamed v The King [1949] AC. 182, at pp. 191-192. 4. ibid., at pp. 191-192, 194. 5. [1946] K.B. 531. 6. ibid., at p. 539.

In Noor Mohamed [12] , Lord du Parcq, speaking for the Judicial Committee, criticized this statement on the ground that "logically probative" might be understood to include much evidence which is held to be "irrelevant". Just what Lord du Parcq meant by "irrelevant" is not altogether clear. It has been suggested that his Lordship was referring to categories of inadmissible evidence including hearsay and secondary evidence of documents [13] .

1. [1949] A.C., at p. 194. 2. Reg. v Boardman [1975] AC. 421, at p. 449, per Lord Hailsham of St. Marylebone.

However, the criticism of the dicta in Sims did not extend to the decision itself which has been generally regarded as correct. In Sims , where a person was charged in one indictment with several offences of sodomy and gross indecency with several men and the evidence of each man was that the accused invited him into his house and there committed the acts charged, the acts in each case bearing a striking similarity to each other, it was held that the evidence of the other acts was admissible in relation to each charge to show the nature of the act done by the accused. That evidence was held to be admissible, not only because the acts bore "a striking similarity" to each other [14] , but also because "[t]he probative force of all the acts together is much greater than one alone"; they showed that the visits to the accused's house were for a guilty not for an innocent purpose [15] . The correctness of the decision in Sims on that footing was accepted in Reg. v Kilbourne [16] and Reg. v Boardman [17] .

1. [1946] K.B., at pp. 539-540. 2. ibid., at p. 540. 3. [1973] A.C. 729. 4. [1975] A.C., esp. at p. 444, per Lord Wilberforce.

However, before Boardman was decided, the received doctrine was that mere propensity evidence was inadmissible; to be admitted the evidence must go to something other than disposition. Boardman changed that received doctrine by discarding the earlier approach to admission of similar fact evidence based on identifiable categories. Instead, in Boardman , the House of Lords adopted as the guiding principle to determine the admissibility of similar fact evidence the test whether the prejudice to the accused is outweighed by the probative force of the evidence. In that case, the headmaster of a boarding school for boys was charged with buggery with S., a pupil aged sixteen, and inciting H., a pupil aged seventeen, to commit buggery on him. It was held that the evidence of S. on the count concerning him was admissible as corroborative evidence in relation to the count concerning H. and vice versa. The trial judge, it was held, was entitled to decide that the probative force of the similar fact evidence outweighed its prejudicial effect because there was a striking similarity or underlying unity between the similar fact evidence and the evidence of the acts relevantly charged [18] .

Lord Cross of Chelsea, who reflected the majority view in Boardman , observed that the reason for the general rule of exclusion in relation to propensity evidence is [19] :

not that the law regards such evidence as inherently irrelevant but that it is believed that if it were generally admitted jurors would in many cases think that it was more relevant than it was, so that its prejudicial effect would outweigh its probative value. Circumstances, however, may arise in which such evidence is so very relevant that to exclude it would be an affront to common sense. His Lordship went on to say [20] : The question must always be whether the similar fact evidence taken together with the other evidence would do no more than raise or strengthen a suspicion that the accused committed the offence with which he is charged or would point so strongly to his guilt that only an ultra-cautious jury, if they accepted it as true, would acquit in face of it. In the end - although the admissibility of such evidence is a question of law, not of discretion - the question as I see it must be one of degree. 1. [1975] A.C., at p. 456. 2. ibid., at p. 457.

As will appear later in these reasons, the passage just quoted is important and has significance for the question to be decided in the present case.

Subsequently, in Director of Public Prosecutions v P. [21] , the House of Lords rejected the proposition that "striking similarity" was an essential prerequisite of admissibility of similar fact evidence in all cases, holding that the essential feature of the evidence to be admitted is that its probative force is sufficiently great to make it just to admit despite its prejudicial effect [22] . In Director of Public Prosecutions v P. , the accused was charged with rape and incest against each of his two daughters. The trial judge refused an application that the counts relating to each girl should be tried separately and admitted evidence of an offence against one victim in connexion with an alleged offence against another. The House of Lords held that the evidence was properly admitted on the ground that its probative force was so great as to make it just to admit it notwithstanding that it was prejudicial to the accused.

1. [1991] 2 A.C. 447. 2. ibid., at pp. 460-461.

Lord Mackay of Clashfern L.C. (with whom the other Law Lords agreed) rejected the notion that "striking similarity" is an essential element in every case in allowing evidence of an offence against one victim to be heard in connexion with an allegation against another, though his Lordship acknowledged that, in cases of identity, "evidence of a character sufficiently special reasonably to identify the perpetrator is required" [23] .

[T]he essential feature of evidence which is to be admitted is that its probative force in support of the allegation that an accused person committed a crime is sufficiently great to make it just to admit the evidence, notwithstanding that it is prejudicial to the accused in tending to show that he was guilty of another crime. Such probative force may be derived from striking similarities in the evidence about the manner in which the crime was committed But restricting the circumstances in which there is sufficient probative force to overcome prejudice of evidence relating to another crime to cases in which there is some striking similarity between them is to restrict the operation of the principle in a way which gives too much effect to a particular manner of stating it and is not justified in principle. The Lord Chancellor went on to say [24] : Whether the evidence has sufficient probative value to outweigh its prejudicial effect must in each case be a question of degree. 1. [1991] 2 A.C., at p. 460. 2. ibid., at p. 461.

The Lord Chancellor regarded the relationship between the evidence relating to one victim and the evidence relating to another victim as critical. In this respect, his Lordship said [25] :

This relationship, from which support is derived, may take many forms and while these forms may include "striking similarity" in the manner in which the crime is committed, consisting of unusual characteristics in its execution the necessary relationship is by no means confined to such circumstances. Relationships in time and circumstances other than these may well be important relationships in this connexion. Where the identity of the perpetrator is in issue, and evidence of this kind is important in that connexion, obviously something in the nature of what has been called in the course of the argument a signature or other special feature will be necessary. To transpose this requirement to other situations where the question is whether a crime has been committed, rather than who did commit it, is to impose an unnecessary and improper restriction upon the application of the principle. 1. ibid., at p. 462.

The exposition of the principles in Director of Public Prosecutions v P. represents an authoritative statement of the relevant law as it presently stands in England.

The approach adopted in Director of Public Prosecutions v P. has been followed by the New Zealand Court of Appeal [26] .

1. Reg. v Accused (1991), 7 C.R.N.Z. 604; Reg. v McIntosh (1991), 8 C.R.N.Z. 514.

In Reg. v Morin [27] , the Supreme Court of Canada adhered to the view that, although evidence of propensity was relevant to establish the commission of a crime, it was inadmissible for that purpose unless it was relevant to an issue apart from its tendency to show propensity and its probative value exceeded its prejudicial effect. But, in Reg. v B. (C.R.) [28] , the Supreme Court recognized that the general exclusionary rule against the reception of evidence adduced solely to show that the accused is the sort of person likely to have committed an offence is not an absolute. According to McLachlin J. (with whom Dickson C.J., Wilson, L'Heureux-Dubé and Gonthier JJ. agreed) [29] :

Whether the evidence in question constitutes an exception to this general rule depends on whether the probative value of the proposed evidence outweighs its prejudicial effect where the similar fact evidence sought to be adduced is prosecution evidence of a morally repugnant act committed by the accused, the potential prejudice is great and the probative value of the evidence must be high indeed to permit its reception. The judge must consider such factors as the degree of distinctiveness or uniqueness between the similar fact evidence and the offences alleged against the accused, as well as the connexion, if any, of the evidence to issues other than propensity, to the end of determining whether the probative value of the evidence outweighs its potential prejudice and justifies its reception. 1. [1988] 2 S.C.R. 345, at pp. 369-371; (1988) 44 C.C.C. (3d) 193, at pp. 217-218. 2. [1990] 1 S.C.R. 717; (1990) 55 C.C.C. (3d) 1. 3. ibid., at p. 735; p. 25.

The law in Canada, as thus stated [30] , closely resembles the law as expressed subsequently by Lord Mackay of Clashfern L.C. in Director of Public Prosecutions v P. and as applied in New Zealand.

1. See also Reg. v C. (M.H.) [1991] 1 SC.R. 763, at p. 771; (1991) 63 C.C.C. (3d) 385, at p. 392; Reg. v B. (F.F.) [1993] 1 SC.R. 697; (1993) 79 C.C.C. (3d) 112.

In this Court, in conformity with earlier English authorities, it was accepted that propensity evidence is not admissible if it shows only that the accused has a propensity or disposition to commit a crime or that he or she was the sort of person likely to commit the crime charged. But it was accepted that it is admissible if it is relevant in some other way, that is, if it tends to show that the accused is guilty of the offence charged for some reason other than that he or she has committed crimes in the past or has a criminal disposition [31] . It was also accepted that, in order to be admissible, propensity evidence must possess "a strong degree of probative force" [32] or the probative force of the evidence must clearly transcend the prejudicial effect of mere criminality or propensity [33] . Very often, propensity evidence is received when there is a striking similarity between different offences or between the evidence of different witnesses [34] . In particular, it was recognized that the existence of such striking similarity is necessary in cases such as Sutton where the prosecution seeks to lead the evidence on the basis that the similarity between different offences founds a conclusion that they must have been committed by the one person with the consequence that evidence which would be admissible to show that an accused committed one of the offences is admissible to prove that he or she committed another or the others of them.

1. Markby v The Queen (1978), 140 C.L.R. 108, at p. 116; Perry v The Queen (1982), 150 C.L.R. 580, at p. 609; Sutton v The Queen (1984), 152 C.L.R., at pp. 533, 545-546, 556-557, 562-563. 2. Markby (1978), 140 C.L.R., at p. 117; Perry (1982), 150 C.L.R., at pp. 586, 589, 604; Sutton (1984), 152 C.L.R., at p. 533. 3. Perry (1982), 150 C.L.R., at p. 609; Sutton (1984), 152 C.L.R., at pp. 548-549, 559-560, 565; Harriman v The Queen (1989), 167 C.L.R. 590, at p. 633. 4. Markby (1978), 140 C.L.R., at p. 117; Perry (1982), 150 C.L.R., at pp. 603, 607, 610; Sutton (1984), 152 C.L.R., at pp. 535, 549, 559, 566-567.

The insistence in some of the judgments of this Court on the need to show that propensity evidence was relevant to "some other issue" as one of the prerequisites of its admissibility so as to prove the commission of the offences charged contributed to a misunderstanding of the Makin principles and to statements of principles which lacked a clear and coherent theoretical foundation. So much was recognized by Mason C.J., Wilson and Gaudron JJ. in Hoch v The Queen [35] where their Honours stated that the basis for the admission of similar fact evidence lies in its possessing a particular probative value or cogency such that, if accepted, it bears no reasonable explanation other than the inculpation of the accused in the offence charged. In other words, for propensity or similar fact evidence to be admissible, the objective improbability of its having some innocent explanation is such that there is no reasonable view of it other than as supporting an inference that the accused is guilty of the offence charged [36] . Mason C.J., Wilson and Gaudron JJ. said [37] :

Assuming similar fact evidence to be relevant to some issue in the trial, the criterion of its admissibility is the strength of its probative force That strength lies in the fact that the evidence reveals "striking similarities", "unusual features", "underlying unity", "system" or "pattern" such that it raises, as a matter of common sense and experience, the objective improbability of some event having occurred other than as alleged by the prosecution. This passage should not be understood as asserting that "striking similarities" or the other characteristics mentioned in relation to propensity or similar fact evidence are essential prerequisites of its admissibility in every case. 1. (1988) 165 C.L.R. 292, at p. 294. 2. (1988) 165 C.L.R., at p. 295. 3. ibid., at pp. 294-295. This passage was subsequently cited by Dawson J. in Harriman (1989), 167 C.L.R., at p. 600.

An important distinction is to be drawn between cases such as the present case in which the "similar facts" are not in dispute and cases in which such facts are in dispute. Thus, their Honours said (59):

Where the happening of the matters said to constitute similar facts is not in dispute and there is evidence to connect the accused person with one or more of the happenings evidence of those similar facts may render it objectively improbable that a person other than the accused committed the act in question, that the relevant act was unintended, or that it occurred innocently or fortuitously. The similar fact evidence is then admissible as evidence relevant to that issue.

Where the propensity or similar fact evidence is in dispute, it is still relevant to prove the commission of the acts charged [38] . The probative value of the evidence lies in the improbability of witnesses giving accounts of happenings having the degree of similarity unless the events occurred. Obviously the probative value of disputed similar facts is less than the probative value those facts would have if they were not disputed. But the prejudicial effect of those facts may not be significantly reduced because the prejudicial effect that the law is concerned to guard against is the possibility that the jury will treat the similar facts as establishing an inference of guilt where neither logic nor experience would necessitate the conclusion that it clearly points to the guilt of the accused. Because propensity evidence is a special class of circumstantial evidence, its probative force is to be gauged in the light of its character as such. But because it has a prejudicial capacity of a high order, the trial judge must apply the same test as a jury must apply in dealing with circumstantial evidence and ask whether there is a rational view of the evidence that is consistent with the innocence of the accused [39] . Here "rational" must be taken to mean "reasonable" [40] and the trial judge must ask himself or herself the question in the context of the prosecution case; that is to say, he or she must regard the evidence as a step in the proof of that case. Only if there is no such view can one safely conclude that the probative force of the evidence outweighs its prejudicial effect. And, unless the tension between probative force and prejudicial effect is governed by such a principle, striking the balance will continue to resemble the exercise of a discretion rather than the application of a principle.

1. Boardman [1975] AC., at pp. 452, 458-459; Sutton (1984), 152 C.L.R., at pp. 556-557; Hoch (1988), 165 C.L.R., at p. 295. 2. Hoch (1988), 165 C.L.R., at p. 296 (where Mason C.J., Wilson and Gaudron JJ. expressed agreement with the remarks of Dawson J. in Sutton (1984), 152 C.L.R., at p. 564). See also Harriman (1989), 167 C.L.R., at p. 602. 3. See Peacock v The King (1911), 13 C.L.R. 619, at p. 634; Plomp v The Queen (1963), 110 C.L.R. 234, at p. 252.

In our view, the principles stated above which derive from Hoch correctly state the law with respect to the admissibility of similar fact evidence. Those principles have not been disavowed by any subsequent decision of this Court and they were accepted and applied by the trial judge in this very case. The discussion in Hoch was expressed in terms of evidence of similar facts rather than propensity evidence. That was because the evidence in that case lent itself to that classification though, in the light of the possibility of concoction, it was held to be inadmissible.

There has been a tendency to treat evidence of similar facts, past criminal conduct and propensity as if they each raise the same considerations in terms of admission into evidence. The difficulty is that their probative value varies not only as between themselves but also in relation to the circumstances of particular cases. Thus, evidence of mere propensity, like evidence of a general criminal disposition having no identifiable hallmark, lacks cogency yet is prejudicial. On the other hand, evidence of a particular distinctive propensity demonstrated by acts constituting particular manifestations or exemplifications of it will have greater cogency, so long as it has some specific connexion with or relation to the issues for decision in the subject case. That evidence, as has been said, will be admissible only if its probative value exceeds its prejudicial effect. But that statement, it seems to us, is of little assistance unless it is understood that the evidence sought to be admitted is circumstantial and as such raises the objective improbability of some event having occurred other than that asserted by the prosecution; in other words, that there is no reasonable view of the evidence consistent with the innocence of the accused. In stating the question in that way, we point out, as Lord Cross of Chelsea suggested in Boardman [41] , that the purpose of the propensity evidence is to establish a step in the proof of the prosecution case, namely, that it is to be inferred, according to the criminal standard of proof, that the accused is guilty of the offence charged. Accordingly, the admissibility of the evidence depends upon the improbability of its having some innocent explanation in the sense discussed.

Acceptance of the statement of principles stated above means that striking similarity, underlying unity and other like descriptions of similar facts are not essential to the admission of such evidence, though usually the evidence will lack the requisite probative force if the evidence does not possess such characteristics. What is more, that approach conforms with the approach that now exists in the United Kingdom, Canada and New Zealand.

In Harriman v The Queen , Dawson J. pointed out [42] :

In the past, evidence of a criminal propensity to commit crime in general, or a particular kind of crime, appears to have been regarded as inadmissible because it was thought to be purely prejudicial, and therefore irrelevant, rather than relevant but excluded because of its prejudicial nature. Upon this basis it was said that it became admissible only if some relevance could be shown beyond the propensity itself. His Honour observed that, although this view had drawn some support from Lord Herschell's statement in Makin , that view had given way in Reg. v Boardman to the opinion that Lord Herschell was pointing [43] "to the high degree of relevance required to render propensity evidence admissible rather than to the requirement of relevance of a different kind". That is how Lord Goddard L.C.J. had stated the law in Sims when his Lordship said [44] : "Evidence is not to be excluded merely because it tends to show the accused to be of a bad disposition, but only if it shows nothing more." What his Lordship was insisting on was that the evidence of bad disposition should also have some "specific connexion" with the commission of the offence alleged. That is because, as a matter of policy, the courts have taken the view that propensity evidence if it does no more is likely to have a very prejudicial effect and should not be received unless its probative force exceeds that prejudicial effect. So the evidence of propensity needs to have a specific connexion with the commission of the offence charged, a connexion which may arise from the evidence giving significant cogency to the prosecution case or some aspect or aspects of it. However, as we have already said, the criterion of probative force as against prejudicial effect and thus of admissibility is that deriving from Hoch . 1. (1989) 167 C.L.R., at pp. 597-598. 2. ibid., at pp. 598-599. 3. [1946] K.B., at p. 537.

The role of the trial judge in admitting propensity evidence

Once that criterion of admissibility is accepted, it is apparent that the trial judge is required to discharge an important responsibility. That point was made by the Supreme Court of Canada in Reg. v B. (C.R.) [45] where it was accepted that the process of balancing the probative value of the evidence against its prejudicial effect was a delicate one. But the trial judge, in making that judgment, must recognize that propensity evidence is circumstantial evidence and that, as such, it should not be used to draw an inference adverse to the accused unless it is the only reasonable inference in the circumstances. More than that, the evidence ought not to be admitted if the trial judge concludes that, viewed in the context of the prosecution case, there is a reasonable view of it which is consistent with innocence [46] .

1. [1990] 1 S.C.R. 717; (1990) 55 C.C.C. (3d) 1. 2. Sutton (1984), 152 C.L.R., at p. 564; Hoch (1988), 165 C.L.R., at p. 296; Harriman (1989), 167 C.L.R., at p. 602.

It is of first importance in the present case to recognize that the trial judge decided to receive the H. evidence after conducting a voir dire which led him to conclude that drowning was not a reasonably possible explanation for Michael's disappearance, that the only rational alternative was abduction for a sexual purpose and that, if he was abducted and sexually assaulted, the inference must be that Michael was murdered by his abductor. Later the trial judge carefully instructed the jury that the H. evidence was relevant, at the most, to the question of the abductor's identity and that it could not be used to resolve any doubts on the question whether there was an abduction at all. It was only if, on the Murray Bridge evidence, the jury rejected drowning as a reasonable possibility and were satisfied that Michael was abducted and murdered, they could take into account the H. evidence in deciding whether the appellant had committed the abduction and murder. His Honour's directions on this point were clear and there was no possibility that the jury could misunderstand them.

The correctness of the trial judge's conclusions on the voir dire

The challenge to the trial judge's conclusions on the voir dire rests very largely on the possibility that Michael drowned, though his body was not recovered, and the further possibility that he was abducted or induced to leave the Murray Bridge area for a purpose other than sexual gratification. The prosecution case, which it was clearly open to the jury to accept, was that drowning at Thiele Reserve must be rejected as a reasonable possibility. On the evidence, Michael would not have gone voluntarily to Thiele Reserve; he was forbidden to go there and he did not like the Reserve. Further, if he had gone to Thiele Reserve, he would not have been swimming there. He disliked swimming alone, and Sturt Reserve was one of his favourite swimming spots yet he did not swim there that day. It is very unlikely he would have had a swim in his shorts and it seems he did not take his swim shorts with him. If he did have a swim at Thiele Reserve, he would have joined the children he knew there. If, contrary to all the indications, he went swimming, drowning would be unlikely because he was a competent swimmer. And, apart from the presence of Michael's bicycle and belongings, neatly stacked, contrary to his practice, there is no evidence that he was at that Reserve. No one, including the boys who were fishing and knew him, saw him there. On top of all that, there was the failure to discover his body in the area, despite thorough and extensive searches.

Once accidental death is put aside, the evidence points inexorably to abduction as the cause of Michael's disappearance. The presence of the bicycle and the neatly stacked belongings at Thiele Reserve strongly suggest that the bicycle and the belongings were placed there with the intention of laying a false trail in order to create the impression that Michael drowned at Thiele Reserve. In addition, Mrs. Gould's evidence of hearing the commotion is quite unexplained except on the hypothesis that Michael was abducted and the commotion was associated with his abduction.

The evidence does not suggest that abduction would have taken place for any purpose other than sexual purposes. Kidnapping for the purpose of seeking ransom was not suggested; nor was mindless killing suggested and it exists only as a mere theoretical possibility. Abduction for sexual purposes was the most likely reason for any abduction.

The connexion between the H. evidence and the offence charged

The prosecution case against the appellant based on the Murray Bridge evidence therefore pointed to abduction for sexual purposes. That required the presence in the area of a person of the requisite disposition equipped with the means of effecting an abduction. The appellant was in the area at the relevant time; on the H. evidence, he was of the requisite disposition; and he had a van which, on the H. evidence (which was not in dispute), was used for just such a purpose on that occasion. More than that, there was evidence that the appellant had spoken to Michael on two occasions, patted his dog at Sturt Reserve, lent him his knife and asked him to put it in the van at Sturt Reserve. And, on the previous day, the appellant had invited two other children to accompany him in the van to visit other places in the vicinity. Furthermore, there was the evidence of his statement to his former wife after he had been arrested by the police in connexion with the H. abduction that he was lonely and had been thinking of "it" on and off for the past twelve months. In that statement, he said that he had just driven around and picked up somebody and threw "them" into the van. In the context in which that statement was made, the reference to "it" must be understood as a reference to an abduction of a boy for sexual purposes, the abduction being effected by means of the van. The reference to the past twelve months extended to the time when the offence charged was committed. The former wife's evidence, if accepted, established that, at the time when Michael disappeared, the appellant was thinking about abducting a boy for sexual purposes. The H. evidence indicated that the appellant was prepared to carry these thoughts into effect when an opportunity arose. The Murray Bridge evidence, if accepted, established that such an opportunity arose on 18 January 1989.

Because propensity evidence may well have a prejudicial effect which is disproportionate to the probative force of that evidence, it is necessary to maintain an insistence on that evidence having a high level or degree of cogency in the circumstances of the particular case. In this context, the reference to prejudicial effect is a reference to the undue impact, adverse to an accused, that the evidence may have on the mind of the jury over and above the impact that it might be expected to have if consideration were confined to its probative force.

Propensity evidence (including evidence of bad disposition and prior criminality) has always been treated as evidence which has or is likely to have a prejudicial effect in the sense explained. That is because the ordinary person naturally (a) thinks that a person who has an established propensity whenever opportunity arises has therefore yielded to the propensity in the circumstances of the particular case and (b) may ignore the possibility that persons of like propensity may have done the act complained of. Hence, the necessity to find something in the evidence or in its connexion with the events giving rise to the offences charged which endows it with a high level or degree of cogency.

Often that high level or degree of cogency is found in the striking similarity, underlying unity or "signature" pattern common to the incidents disclosed by the totality of the evidence. So, in the present case, had the prosecution case been based on direct evidence of abduction of Michael for sexual purposes by means of inveigling him into a van, there would have been, in our view, no doubt about the admissibility of the H. evidence. The pattern of similarity, underlying unity or "signature" common to both incidents would have resulted in such a degree of cogency that the probative force of the H. evidence would have outweighed its prejudicial effect, notwithstanding that there was but one other incident of the kind alleged, that it occurred virtually twelve months later and that it did not establish that the appellant intended to kill the boy H.

Here, however, because the cause of Michael's disappearance cannot be established by direct evidence, the Crown case lacks direct proof of the details which might otherwise constitute a detailed pattern of striking similarity or underlying unity. What we have is a case of circumstantial evidence, based on the Murray Bridge evidence, leading to an inference that Michael was abducted for sexual purposes and, in the circumstances, that meant that a van of the kind which the appellant drove was an ideal vehicle for the execution of such an enterprise. If that be accepted, as we think it should be accepted, then there is a similarity and unity between the two incidents of abduction for sexual purposes involving the use of a van. In addition, according to the undisputed evidence of each incident, the appellant was present on each occasion, initiating conversation with the victim in close proximity to his van. The fact that the Murray Bridge evidence as to abduction is circumstantial and inferential rather than direct means that the H. evidence lacks some of the cogency that it would have had if the Murray Bridge evidence on the point were direct.

Granted the circumstantial evidence in this case falls short of that level of precision, there are other factors present which indicate that the H. evidence has very considerable cogency. First, there is the Murray Bridge evidence as to the appellant's presence with his van in Sturt Reserve and his contact with Michael; in other words, there is convincing evidence of opportunity. Then there is Mr. Toogood's evidence of a van which could have been the appellant's van travelling at an unsafe speed in the direction of Thiele Reserve at about 2.45 p.m. and the evidence of a commotion at Thiele Reserve. These two pieces of evidence suggest that the opportunity was availed of by someone with a vehicle which could have been the appellant's vehicle, there being no evidence that another vehicle with similar characteristics was in the vicinity at that time. The evidence therefore points to the appellant taking advantage of the opportunity which presented itself.

Viewed against that background provided by the Murray Bridge evidence, the H. evidence, including the statement made by the appellant to his wife (which was not disputed) after he had been arrested in connexion with the H. incident, is cogent, circumstantial evidence pointing to the appellant's guilt. The modus operandi disclosed by the H. evidence, namely, the inveigling of a child into the van, has its parallel with the two children swimming from the wharf at Thiele Reserve on 17 January and to a lesser extent with the initiation of the conversation and contact with Michael at Sturt Reserve on 18 January. That conversation and contact is given a particular character by the appellant's statement made to his wife after his arrest in connexion with the H. incident.

The sense of that statement, in terms of its bearing upon the offence charged, cannot be fully appreciated unless it is accompanied and explained by the evidence of the H. incident. In one sense, that is by the way because it points to an independent ground for receiving that evidence. But it highlights the significance of the H. evidence and indicates that it would be an affront to commonsense to reject the statement made to the appellant's wife and the evidence of the incident on the basis that it was not admissible.

In the context of the Murray Bridge evidence, the H. evidence demonstrates not only propensity and criminality but also established the appellant's modus operandi in abducting a young boy for sexual purposes and his acknowledgment that he was thinking about indulging his propensity by recourse to the same modus operandi. The H. evidence ties in with the prosecution evidence relating to the appellant's activities in the Murray Bridge area on 17 and 18 January and indicated, if accepted, that he took the initial steps that the H. evidence instructs us were the steps that he engaged in when he gave effect to that propensity. Once that evidence is taken into account and once drowning is excluded (as the jury clearly excluded it) as a reasonable possibility, there is no reasonable hypothesis that explains Michael's death in a way other than that contended for by the prosecution. In other words, there is no reasonable view of the evidence which is consistent with the appellant's innocence.

In reaching that conclusion, we have taken account of three matters which have been, or could be, relied upon to sustain a contrary conclusion. First, there is the argument that, as the H. incident involved abduction and sexual abuse but not murder, it cannot fall into the category of acceptable propensity or similar fact evidence. The H. evidence is equivocal on the score of propensity to murder. That evidence does not deny the existence of such a propensity. The dropping of the bicycle at the clifftop, after first wiping it with a cloth, may indicate an intention to fabricate an accident. But, in any event, for the reasons already given, the H. evidence is cogent and its probative force outweighs its prejudicial effect.

The second matter is that the H. incident occurred after Michael's disappearance. But it does not matter whether similar facts occurred earlier or later [47] . The third matter is the possibility that, despite the trial judge's instruction to the jury against using the H. evidence to decide whether Michael drowned or was abducted for sexual purposes, the jury may nonetheless have had regard to that evidence in relation to that issue. For our part, we do not see what ground there is for concluding that the jury failed to give effect to the clear instruction given to them by the trial judge. It may be that the jury was not precluded from having regard to the evidence generally once it was admitted into evidence. But there is no occasion to consider that question.

The appellant was convicted of the murder of Michael Black, aged ten, on about 18 January 1989. His appeal to this Court is from the judgment of the Court of Criminal Appeal of South Australia dismissing his appeal against that conviction.

The appeal turns on the admissibility at his trial of evidence of events which took place in December 1989, that is, nearly one year after the date of the alleged murder. Those events, involving a boy of thirteen, led to the appellant's conviction, following pleas of guilty, for the offences of false imprisonment, rape, indecent assault and abduction. For those offences he was sentenced to imprisonment for fifteen years, with a non-parole period of thirteen years. The events were referred to in the courts below by reference to the name of the boy involved. To preserve his anonymity, I shall refer to those events as the H. incident and the evidence relating thereto as the H. evidence.

To understand how and on what basis the H. evidence was admitted on the trial of the appellant, it is necessary to refer in some detail to the events of 18 January 1989.

Michael Black lived at Murray Bridge, a country town south east of Adelaide. On the day in question he left home at about 1 p.m. to go fishing. He rode his bicycle, taking with him a fishing rod and canvas bag. He was wearing a red and blue striped cotton knit shirt, shorts and blue thongs. He did not carry bathers. He may have been wearing bathers instead of underpants but this is unlikely. He had no towel. He was accompanied by the family dog, a red heeler.

At about 1.30 p.m. he arrived at Sturt Reserve. The reserve is on the town side of the Murray River. It is an area of lawn running down to the river bank. There is a large rowing boat shed, shelter sheds, a wharf for pleasure boats, swings, barbecue facilities and a mobile kiosk where light refreshments are sold. Michael was seen by a number of people at Sturt Reserve up until nearly 3 p.m. He was not seen again after 3 p.m. that day; his body has never been found.

The story then moves to Thiele Reserve which is up river from Sturt Reserve and on the other side of the river. Later on the afternoon of 18 January Michael's bicycle, fishing rod, bag and thongs were found neatly stacked against some railings at Thiele Reserve. His shirt was found in some willows growing on the bank of the river at the up-stream end of the Reserve. Nearby was a piece of twine tied to a tree stump. In his judgment in the Court of Criminal Appeal Mohr J. said: "It was assumed that Black's dog had been tied with it." The basis for that assumption is that Michael's father found twine still attached to the dog's collar at about the length the dog would have chewed it through.

When Michael did not return home, his father reported the matter to the police. Police divers carried out an extensive underwater search of the river bottom at Thiele Reserve and among the willows growing on the bank at either end of the Reserve. A visual search was not possible; the divers operated by touch. The divers gave evidence that there was only a remote possibility they would have failed to find a body. As well, there was evidence that over twenty-four years there had been 137 police operations to recover dead bodies from inland waters in South Australia and, the present case apart, there was only one occasion when they had failed to recover the body. That occasion was during flood conditions. The jury were entitled to conclude that there was no body to be found in the river and such a conclusion was not seriously challenged. An extensive land search was undertaken and unofficial searches were made of the river and its banks. All to no avail.

At his trial the appellant's counsel conceded that there were only two explanations for Michael's disappearance. Either he drowned or he was abducted. Michael's mother, who was separated from his father, said that she spoke to the boy by telephone on the morning of the day he disappeared and that he said he was going swimming. The presence of the bicycle and other items, in particular the shirt in the willows, it was said, pointed to the possibility that Michael had gone for a swim among the willows although a short distance downstream there was a sandy beach. However, there were factors pointing against the likelihood that he went for a swim at Thiele Reserve. No one saw him at the reserve, including two boys known to Michael who were fishing there. Michael's father gave evidence that the boy disliked swimming in underpants although he said, in cross-examination, that if Michael wanted to swim he would if necessary swim in shorts and underpants. If he had been wearing bathers, one would have expected to find his shorts and underpants at the scene. Mr. Black also said that his son did not like Thiele Reserve and was unlikely to have gone there.

There was a shrimp pot suspended from the willows near where the shirt was found. The defence advanced a theory that Michael may have tried to recover the pot and fallen into the river. The theory did not explain why Michael would have taken off his shirt to recover the pot. More importantly, it did not explain why his body was not found.

In these circumstances the Crown invited the jury to conclude that Michael did not drown and that he did not leave Sturt Reserve of his own volition. Such a conclusion would inevitably require a finding that the items found at Thiele Reserve were placed there by Michael's abductor to lay a false trail by creating the impression that Michael had gone to the reserve and had drowned there.

Against this background it is necessary to look at the movements of Michael and the appellant at relevant times. The appellant's former wife said that he left home on 17 January, driving his white Kombi van. The van was fitted out with radio and television and had sleeping and cooking facilities. Mr. Smart gave evidence that he spoke to the appellant at Sturt Reserve at about 4 p.m. on 17 January. He identified the appellant from photographs and in court. The appellant asked him if there was somewhere he could swim in the nude as he had no bathers. Mr. Smart directed him to Thiele Reserve. Shortly after, the appellant joined two children, Christian and Chelsea Haebich, who were swimming from the wharf at Sturt Reserve. He asked them if they knew of any places of interest he could visit. He suggested that if they knew of such places they could accompany him in his van. They did not go to the van. The children identified the man they had spoken to as the appellant in court.

On the following day, the day of Michael's disappearance, the appellant was observed at Sturt Reserve by Mrs. Rooke who saw him sitting in his van watching her and the two children in her care. By his own account he went to the kiosk where he spoke to some children who were short of money. He offered to make up the shortfall for their purchases. Early that afternoon Michael was served at the kiosk several times by Miss Giles who knew him. The first time was about 2 p.m. She estimated his last visit as close to 3 p.m. But, to some extent, that estimate was based on the first time she saw Michael, which she later said may have been earlier than 2 p.m. Miss Giles said that when she finished work shortly after 3 p.m. she noticed that Michael was no longer at the river. Mrs. Regnier was also working at the kiosk that day. She served Michael a packet of crisps at about 2 p.m. At about 2.25 p.m. Mr. Haines saw a man speaking to Michael. This was the last firm sighting of Michael [48] . Mr. Haines described the man as about 5′ 10″, 5′ 11″, of light build, clear complexion, dark brown hair and about forty years of age. He could not recall any facial hair. This description generally fitted the appellant except that he had a neat moustache. Also, it must be said, Mr. Haines described the man as neatly dressed, in "normal business trousers and a shirt". Other evidence suggested that the appellant was wearing a dark tracksuit with matching top or jeans. The man seen was not known to Mr. Haines who had lived in Murray Bridge for twenty-five years. At about 2.45 p.m. Mr. Toogood, a council employee, saw a white van leaving Sturt Reserve at an unsafe speed and travelling towards the main street of the town. This would have taken the van to a position where it could have crossed a bridge to the south of the river where Thiele Reserve is located.

1. There was evidence from Byron Dreschler, who was fifteen years of age at the time. At one stage he thought that he had seen Michael riding across the bridge from Sturt Reserve at about 2.10 p.m. on 18 January. There was however other evidence that any such sighting by Dreschler could not have been later than 1.22 p.m. on that day and, in evidence, Dreschler conceded that his earlier belief as to the time he saw Michael must have been mistaken.

Is there any evidence which places the appellant at Thiele Reserve on the afternoon of 18 January? Mrs. Gould lived in a house overlooking the Reserve. Between 3 p.m. and 3.30 p.m. she heard a "commotion" coming from the Reserve, what sounded like a vehicle with a loud engine with its wheels spinning in gravel, and the sound of a dog barking excitedly. Much later Mrs. Gould took part in a police test in which, out of her sight, eight vehicles, including the appellant's, were driven past. The appellant's vehicle was driven past twice. The other vehicles were driven past once. Mrs. Gould identified the sound of the appellant's vehicle as similar to that she heard on 18 January. The admissibility of that evidence was challenged but it was admitted by the trial judge. However, he said to the jury:

I think you might be best advised to put the sound test itself to one side in so far as it purports to identify the sound as very similar to the accused's van. If you think you can rely on Mrs. Gould's memory and description, confine yourselves simply to her description of what she heard on the Wednesday itself, a noisy engine, a sort of rattling chugging noise, with an uneven beat. Mr. Jones gave evidence that he went to Thiele Reserve to go water skiing on 18 January at about 4 p.m. He saw a Kombi van matching the description of the appellant's van on the opposite side of a road running almost parallel to the river bank. It was in the near vicinity of where Michael's bicycle and gear were found. However, the trial judge commented to the jury that his evidence was confused and they might think it better to put the evidence to one side. Another witness, Mr. Martin, saw a man at Thiele Reserve at about 4 p.m. on the day in question. He appeared to be a "loner" and was neatly dressed, wearing trousers and a shirt.

The appellant's daughter Petra testified that on one occasion the appellant told her he had seen Michael twice on that day and on another occasion he told her that he had loaned the boy a knife to scale a fish. Michael brought back the knife and put it in the van. Dianne Davis, a friend of the appellant, and her sister Nicola gave evidence that as part of a general conversation whilst they were watching a news item on the disappearance of Michael Black, the appellant had mentioned that he had seen and talked to Michael, patted his dog and loaned him a fishing knife to untangle a line.

Thus there was evidence that the appellant was at Sturt Reserve at a time when Michael was there, that he spoke to Michael and that he left the reserve at about the time Michael was last seen there. A vehicle, which the jury were entitled to find matched the sound of the appellant's van, was heard at Thiele Reserve shortly after the time when the so-called false trail was laid. The appellant's movements are thereafter unaccounted for by witnesses until he arrived home on 20 January.

The appellant gave evidence at his trial. He admitted speaking to Mr. Smart at Sturt Reserve on 17 January, though he denied asking about nude swimming or being given directions to Thiele Reserve. He admitted swimming with the Haebich children. He said he spent the night at Sturt Reserve though around midnight there was a disturbance which prompted him to cross to the other side of the river for a time.

As to 18 January, the appellant said that he believed he saw Michael Black at Sturt Reserve though he could not be exact about the time. Michael had a large fish which he was beating with a thong. He gave Michael a knife from the van, telling him how to kill the fish. Michael returned the knife. The appellant said: "Just put it back in the van." There was some conversation between them as to whether the television broadcast of cricket had started. He left Sturt Reserve at a time "probably after 2 but before 2.30". He said that he did not go to Thiele Reserve that day or on any other day until an Easter canoe trip with friends, though it was put to him in cross-examination that his former wife gave evidence that they went to Thiele Reserve about a week after 18 January. After he left Sturt Reserve, the appellant said, he stopped at a butcher's shop in the main street. He later drove down a dead end road where he spoke to a farmer. The farmer was not identified or called as a witness. Thereafter he drove to several towns, camping both nights in his van, before returning home on 20 January.

The Crown case involved the proposition that whoever abducted Michael Black drove to Thiele Reserve with Michael's belongings (and presumably Michael). Michael's dog was later found running free with twine around its neck. On the case as advanced, the dog may have been taken to Thiele Reserve where it later broke free or the twine found there may have been intended to suggest that the dog had been taken there by Michael. In any event, the proposition necessitated the abductor placing the bicycle, the fishing rod, the haversack and thongs against a wooden fence near the toilet block and taking the boy's shirt (and perhaps the dog) to the river bank.

This scenario required a conclusion that Michael did not go voluntarily to Thiele Reserve on the day in question. No one saw him there. He had been told not to go there by his parents and he did not like the place. The last time he was seen was at Sturt Reserve, no later than 3 p.m. The only explanation for his disappearance was drowning or abduction. The evidence was strongly against drowning. If he was abducted, it is most likely to have been from Sturt Reserve. In that event the placing of his bicycle, clothing and other items at Thiele Reserve must have been by his abductor and have been done to lay a false trail, pointing to the boy's disappearance from Thiele Reserve.

Against this background the Crown sought to call the boy H. to give evidence of those events in December 1989 which led to the appellant's conviction on four serious charges. The application was opposed by the defence but was acceded to by the trial judge after a lengthy voir dire hearing.

The evidence of H. was that the appellant, using the same Kombi van, abducted him at Port Noarlunga South. H., who had been riding his bicycle, was persuaded by the appellant to get into the van to look for some keys. The appellant pushed him to a seat, threatened him not to "try anything" and retrieved the bicycle, placing it in the van. The appellant tied up H. with tape, blindfolded him and placed a sock in his mouth. He then drove the van some distance from the place of abduction. H. had mentioned that the bicycle belonged to a friend and he wanted him to have it back, adding: "just put it somewhere safe." After driving some distance the appellant stopped the van, wiped the bicycle for fingerprints and left it in the bush, near a path at the top of a cliff. The appellant eventually drove to his home where he was living alone and kept H. overnight, during which time he sexually assaulted him more than once. The boy escaped the next day.

The appellant's former wife testified that when she asked him why he abducted H., the appellant said: "I was lonely. I'd been thinking about it on and off for the past twelve months" or words to that effect. This conversation took place when the appellant telephoned her from the Christies Beach police station after his arrest. He also said: "I just drove around and found somebody and threw them in the van." In answer to his wife's question: "Did you tie them up?", he said: "Yes." The reference to "the past twelve months" has an obvious significance, though the appellant offered an explanation in cross-examination that he was referring to the period he and his wife had been separated.

Although the H. incident occurred nearly twelve months after 18 January 1989, that of itself is no bar to admissibility of evidence if the evidence is otherwise admissible in accordance with rules relating to similar fact evidence or the like [49] . The appellant did not argue otherwise.

1. Thompson v The Queen (1989), 169 C.L.R. 1; Reg. v Morris (1969), 54 Cr. App. R. 69.

In order to determine the admissibility of the H. incident, it is first necessary to consider where the Crown case stood, absent that evidence. It is of particular importance in the present case because one of the appellant's complaints, perhaps the principal complaint, is that the evidence pointing to abduction and murder was at best equivocal but that the H. incident might readily lead the jury to conclude that there had been an abduction, not just to conclude that the appellant was the abductor in an otherwise proven case of abduction.

Any discussion of this aspect of the appeal begins with a concession made by counsel at trial that "abduction is the only rational alternative to drowning". As Legoe J. said in the course of his judgment in the Court of Criminal Appeal: "On the evidence I would think that such a concession was inevitable." There was more than sufficient evidence for the jury to conclude that Michael was abducted from Sturt Reserve. And, at the time of the appellant's trial in May 1992, the jury were certainly entitled to conclude that he had been murdered. Indeed, once they concluded that there had been an abduction, a finding that the boy had been murdered was inevitable.

There can be no doubt as to how his Honour saw the relevance of the H. evidence. In the course of ruling that the evidence was admissible, he said:

The [H.] evidence shows not merely that the accused was of homosexual orientation but that he had a disposition for young boys and was willing to pursue it by the extreme means of kidnapping and rape. He had with him at Murray Bridge on January 18 the van that he used a year later when he abducted [H.]. He had the opportunity to abduct Michael Black. The relevance of his criminal propensity to the question whether he was Michael Black's abductor is obvious. Later in his ruling his Honour said: I do not think there is a rational view of all of this evidence that is inconsistent with the guilt of the accused. It would, in my opinion, be an affront to common sense to postulate two persons in Michael Black's vicinity at Murray Bridge, and both almost certainly at Sturt Reserve, about the same time that afternoon, each with a propensity to kidnap and sexually assault young boys and each having the physical means that afternoon of doing so, one of them befriending the boy and lending him a fishing knife and the other within a fairly short space of time but quite independently engaging, presumably, in some kind of pre-abduction dealing with him, however brief, and both leaving Sturt Reserve in separate vehicles at much the same time.

The trial judge saw the situation as "fundamentally a case about improbabilities, sufficiently strong to compel a factual conclusion adverse to the accused". He said that "the very unusual type of crime in this case itself provides, exceptionally, a sufficient "similar fact" which, coupled with the other linking factors to which I have referred, justifies the admission of the propensity evidence".

While the ruling that the H. evidence was admissible was at the forefront of the appellant's attack, it is also necessary to look at the use the jury were invited to make of the evidence. The appellant argued that the evidence must inevitably have led to a greater readiness on the part of the jury to conclude that Michael was abducted and murdered. And this, notwithstanding that the trial judge more than once cautioned the jury against using the evidence for that purpose. For instance, his Honour said in the course of summing up:

So the Crown's case moves, as it were, from a rejection of drowning to positive evidence of an abduction. That, in outline, is the Crown's evidence - as I understand it - on the primary and essential question whether it has been proved to your satisfaction that Michael Black was abducted. In forming any judgment about that matter, you must confine your attention to the Murray Bridge evidence. That means that you cannot have any regard to the evidence relating to [H.] on that issue.

The [H.] evidence is relevant, at the most, to the question of any abductor's identity. It may not be used to resolve any doubts you might have on the question whether there was an abduction at all. For that, I say again, you must have regard only to the Murray Bridge evidence and put the [H.] matter completely out of your minds. I cannot emphasize that too strongly. If the Murray Bridge evidence does not satisfy you that Michael Black was abducted and murdered, you must find the accused not guilty.

Later in his directions the trial judge again emphasized that the jury could not use the H. evidence unless they were first satisfied "that Michael Black was, in fact, abducted and murdered by some one". Still later he asked the jury to remember "the way in which you may and the way in which you may not use the [H.] evidence, and I shan't go over all that again".

The argument as to the admissibility of the H. evidence wears two faces. If the evidence was inadmissible on any count, the appeal must be upheld because the evidence undoubtedly played an important, if not crucial, part in the conviction of the appellant. On the other hand, the evidence may have been admissible for a particular purpose; in that event a further question arises as to whether the trial judge's directions in regard to the use the jury might make of the evidence were accurate and comprehensive. However the two aspects are not entirely discrete; they impinge on each other.

Reference has been made already to the trial judge's directions to the jury as to how they must not use the H. evidence. But what did his Honour say as to the use the jury might make of it? He began by saying:

If you are satisfied that there was an abduction and a murder, you may take the [H.] evidence into account with the other evidence on the question whether the accused had been proved to have been the person responsible. His Honour then put to the jury that the Crown had pointed to "the unusual type of crime that was committed in this case - the abduction and murder of a ten-year-old boy, presumably for sexual purposes". He continued: What the Crown does say, however, is that the [H.] evidence shows that the accused has a disposition to abduct and sexually assault young boys and that this and all the other circumstantial evidence proves that he was Michael Black's abductor.

While not telling the jury whether they should accept or reject the Crown's submission, the trial judge put to them the Crown case that if the appellant was not the abductor there must have been two persons at Murray Bridge about the same time on the afternoon Michael disappeared, "each with a propensity to kidnap and sexually assault young boys and each having the physical means that afternoon of doing so". He added: "The Crown says that it is an affront to commonsense to postulate such a possibility. It stretches coincidence to breaking point." Expressed as they were, these remarks may have tended to suggest that the Crown's argument had his Honour's endorsement. But later he said:

You are entitled to use your own general knowledge about human behaviour, as well as your assessment of the accused himself, in drawing any conclusion as to whether the [H.] evidence establishes that the accused had such a propensity towards young boys and with a willingness to abduct and assault; and also that he had the same propensity twelve months earlier, because that is the critical time, so far as this case is concerned.

If the evidence was admissible, I do not think that the appellant has a legitimate complaint that the jury may have been misled into using the evidence to reach a conclusion that Michael was abducted and murdered. In the light of the trial judge's directions, unless it be assumed that the jury were unable to comprehend the warning about using the H. evidence or perversely ignored it, this aspect of the appellant's argument must fail. There is no reason to conclude that the jury lacked understanding or were perverse.

Admissibility of propensity or similar fact evidence

It is appropriate now to look at the principles which govern the admissibility of evidence such as that of the H. incident. Counsel used the terms "similar fact" and "propensity" when referring to the evidence and the use made of it by the trial judge. These terms tend to obscure exactly what is at issue. Since most of the argument in this Court was in terms of the admissibility of the H. evidence to establish a propensity in the appellant to abduct young boys, it is convenient to speak at this stage of propensity evidence. Later it will become necessary to refer to similar fact evidence and their relationship.

A convenient starting point is a passage from the judgment of Dawson and Gaudron JJ. in B. v The Queen [50] :

Evidence that an accused has committed offences other than those with which he is charged ought not be admitted if it tends to show only that the accused has a propensity or disposition to commit criminal offences of a particular type or generally. If a propensity or disposition of that kind is all that the evidence tends to show, then its prejudicial nature must be greater than any relevance which it might have. To admit such evidence would be to invite the jury to proceed upon prejudice or suspicion rather than proof. But if the evidence of other offences goes beyond showing a mere propensity or disposition to commit crime or a particular type of crime and points in some other way to the commission of the offences in question, then it will be admissible if its probative value for that purpose outweighs its prejudicial effect. That is not to say that it may not still be evidence of propensity or disposition, but it will then be evidence of propensity or disposition of a particular kind which in the circumstances has a degree of relevance justifying its admission. 1. (1992) 175 C.L.R. 599, at pp. 617-618.

Expressed in that way, the principle accepts that evidence of other offences is not irrelevant. Indeed, in a passage in his judgment in Harriman v The Queen [51] cited by Dawson and Gaudron JJ. in B. v The Queen , Dawson J. said [52] :

When a person is charged with a criminal offence, evidence is ordinarily inadmissible that he has on other occasions been guilty of behaviour indicating a criminal disposition. This is not because the evidence is irrelevant. On the contrary, it is excluded because a jury is likely to regard it as proving too much and is for that reason likely to proceed upon prejudice rather than proof. 1. (1989) 167 C.L.R. 590. 2. ibid., at p. 597.

The respondent relied on this approach to support the argument that evidence which discloses a propensity to offend is prima facie admissible unless the prejudicial effect of the evidence outweighs its probative effect. However, an approach which assumes the admissibility of evidence of other offences but measures the probative force of the evidence against the prejudice it is likely to generate tends to by-pass the basic question of whether and why the evidence is admissible at all. Propensity evidence necessarily involves the use of character against the appellant. As Wigmore has observed [53] :

The evidentiary use of character for or against a defendant in a criminal case cannot be understood without separating the principles of relevancy and of undue prejudice. Evidence of other offences is circumstantial [54] ; the first inquiry must be whether it is relevant. If it is not, the evidence is simply inadmissible. If it is relevant, it may still be rejected by reference to some principle of exclusion, generally by reason of undue prejudice to the defendant. 1. Wigmore on Evidence, Tillers Revision (1983), vol. 1A, s55. 2. Hoch v The Queen (1988), 165 C.L.R. 292, at p. 296.

A well-known example of a situation where evidence of propensity was admitted because it was relevant to an issue, without relying upon any special test, is Makin v Attorney-General (N.S.W.) [55] . The appellants were convicted of the wilful murder of a child they had received from the child's mother on representations as to their willingness to adopt it and upon payment of a sum inadequate for its support for more than a very limited period. The body of the child was found buried in the garden of a house occupied by the appellants. Evidence that several other infants had been received by the appellants on like representations and on like terms and that their bodies had been found buried in a similar manner in the gardens of several houses occupied by them was held admissible.

The judgment of the Privy Council was delivered by Lord Herschell L.C., who said [56] that the evidence was not "irrelevant to the issue to be tried by the jury". The judgment does not elaborate on the point but earlier [57] there appears this passage:

On the other hand, the mere fact that the evidence adduced tends to shew the commission of other crimes does not render it inadmissible if it be relevant to an issue before the jury, and it may be so relevant if it bears upon the question whether the acts alleged to constitute the crime charged in the indictment were designed or accidental, or to rebut a defence which would otherwise be open to the accused. In those situations the evidence is admissible simply because it is relevant to an issue to be tried. Its relevance derives from "the statistical improbability that a number of children which the Makins have at various times had in their care would all have died of natural causes. From this it followed that they were likely to have been murdered" [58] . 1. [1894] A.C., at p. 68. 2. ibid., at p. 65. 3. Hoffman, Similar Facts After Boardman, Law Quarterly Review, vol. 91 (1975) 193, at p. 199.

What if the evidence does no more than show a propensity to commit criminal offences or criminal offences of a particular character or, it may be, to engage in conduct that is blameworthy but falls short of a criminal offence? The principle, as stated in Makin , begins with a rule of exclusion [57] :

It is undoubtedly not competent for the prosecution to adduce evidence tending to show that the accused has been guilty of criminal acts other than those covered by the indictment, for the purpose of leading to the conclusion that the accused is a person likely from his criminal conduct or character to have committed the offence for which he is being tried. But this exclusionary rule is followed by the passage quoted earlier in this judgment which expresses an inclusionary rule, based on relevance. This has the effect of negating the exclusionary rule because, if the evidence is relevant, it is admissible. 1. ibid., at p. 65.

Zuckerman [59] argues, with great force, that there is a fallacy involved in the inclusionary exclusionary distinction, illustrated by Thompson v The King [60] . The accused was charged with acts of indecency on two boys. The boys testified to the acts and also to an arrangement to meet them again later. The accused turned up at the appointed time but, suspecting that police were present, gave the boys money and told them to go away. He was arrested after trying to escape. At his trial he denied any prior acquaintance with the boys. The prosecution was allowed to adduce evidence of articles found in his possession and, on appeal to the House of Lords, the evidence was held admissible, even though it showed a disposition towards homosexuality on the part of the accused, because it was relevant to an issue in the case. As Zuckerman observes, the issue was variously identified in the judgments as the credibility of the boys, identity and intention, and the accused's "identifying mark". Yet, in truth, it was the accused's disposition which was regarded as throwing light on the issue, however that issue was described.

1. Zuckerman, The Principles of Criminal Evidence (1989), pp. 224-225. 2. [1918] A.C. 221.

In Reg. v Boardman [61] the accused, who was the headmaster of a boarding school for boys, was charged with sexual offences against two pupils. The House of Lords upheld the trial judge's ruling that the evidence of each boy was admissible as corroborative evidence in relation to the count against the other. There was a "striking similarity" and "underlying unity" in the evidence of both boys which gave it the necessary probative force to justify its admission. Lord Hailsham of St Marylebone said that two theories had been advanced to explain Lord Herschell's general rule of exclusion and that "both have respectable judicial support" [62] . The first theory is that the evidence is simply irrelevant, on the basis that no number of similar offences can connect a particular person with a particular crime. The second is that the prejudice created by the admission of such evidence outweighs any probative value it may have. Lord Hailsham's view was that "both theories are correct". But the second theory can only stand with the first on the footing that, in a particular case, the evidence may have some probative force. The question then arises: in what circumstances will such evidence have probative force? It will be remembered that Lord Herschell said that the mere fact that the evidence adduced tends to show the commission of other crimes does not by itself render the evidence inadmissible "if it is relevant to an issue before the jury", for instance where there is a question whether the acts alleged to constitute the crime were designed or accidental. Lord Hailsham regarded these instances as an "independent proposition". But if it is an independent proposition, not an exception grafted on to the initial rule, it is hard to know in what other circumstances the evidence will have probative force.

In Director of Public Prosecutions v P. [63] the accused was charged with rape and incest against each of his two daughters. The trial judge admitted evidence of an offence against one daughter in relation to the counts against the other. The admissibility of the evidence was affirmed by the House of Lords in a judgment delivered by Lord Mackay of Clashfern L.C. with which the other Law Lords agreed. Referring to Boardman his Lordship said [64] that "it is not appropriate to single out "striking similarity" as an essential element in every case in allowing evidence of an offence against one victim to be heard in connexion with an allegation against another". He went on [65] :

I consider that the judge must first decide whether there is material upon which the jury would be entitled to conclude that the evidence of one victim, about what occurred to that victim, is so related to the evidence given by another victim, about what happened to that other victim, that the evidence of the first victim provides strong enough support for the evidence of the second victim to make it just to admit it notwithstanding the prejudicial effect of admitting the evidence. This relationship, from which support is derived, may take many forms and while these forms may include "striking similarity" in the manner in which the crime is committed, consisting of unusual characteristics in its execution the necessary relationship is by no means confined to such circumstances. 1. [1991] 2 A.C. 447. 2. ibid., at p. 460. 3. ibid., at p. 462.

In the present case evidence of the commission of other offences was not led to rebut a defence of accident or the like. There was a complete denial by the appellant of any unlawful conduct in relation to Michael Black. If the evidence was admissible, its relevance could only be because it served to identify the appellant as the person who, on other evidence, the jury could reasonably conclude abducted the boy and murdered him. One thing is clear [66] :

[T]he fact that the accused is the sort of person who would be likely to have committed the offence, though relevant, is not admissible. As a result, evidence adduced solely for the purpose of proving disposition is itself inadmissible, or, to put it otherwise, evidence the sole relevancy of which to the crime committed is through proof of disposition, is inadmissible. 1. Morris v The Queen [1983] 2 SC.R. 190, at pp. 200-201; (1983) 1 D.L.R. (4th) 385, at p. 394, per Lamer J.

What then is the relationship of propensity to similar facts? It is in this regard that some of the uncertainty that has bedevilled this area of the law has arisen. Cross on Evidence [67] discusses "Relevance via disposition" in the chapter "Similar Fact Evidence". And that approach is supported by authority. In Markby v The Queen [68] Gibbs A.C.J., in a judgment with which Stephen, Jacobs and Aickin JJ. concurred, said:

The first principle, which is fundamental, is that the evidence of similar facts is not admissible if it shows only that the accused had a propensity or disposition to commit crime, or crime of a particular kind, or that he was the sort of person likely to commit the crime charged. Gibbs A.C.J. reinforced this principle by adding: The second principle, which is a corollary of the first, is that the evidence is admissible if it is relevant in some other way, that is, if it tends to show that he is guilty of the crime charged for some reason other than that he has committed crimes in the past or has a criminal disposition. That second principle inevitably invites the question: in what circumstances does such evidence tend to show guilt other than in terms of propensity? Deane J. offered a test in Sutton v The Queen [69] : On the other hand, if the similarities between offences which the accused undoubtedly committed and the offence with which he is charged are such as to warrant, in the context of all the evidence, the conclusion that, in the absence of extraordinary co-incidence, the same person committed all offences the "similar fact evidence" will, in the context of the evidence as a whole, have probative force on the question whether the accused committed the offence with which he is charged which is distinct from its prejudicial tendency as evidence of mere propensity. What was said by Gibbs A.C.J. and Deane J. does place propensity evidence within the area of similar fact evidence. And, in terms of admissibility, that is an appropriate place so long as the focus of inquiry is on the relationship between the evidence in question. As was said earlier, there was a complete denial by the appellant of any unlawful conduct in relation to Michael Black. Since there was only one charge, there was no question of relying upon the H. evidence and the evidence relating to Michael to support each other, as in Boardman . It was the appellant's role in the later incident which was used as evidence of his role in the earlier one. 1. Aust. ed., vol. 1, Ch. 11, par. 21030. 2. (1978) 140 C.L.R. 108, at p. 116. 3. (1984) 152 C.L.R. 528, at p. 557.

To deal with evidence of similar facts (including propensity) on the footing that the evidence is admissible if its probative force outweighs its prejudicial effect is, I think, to beg the question unless it be accepted that such evidence is always relevant. The preferable approach is indicated by the judgment of Mason C.J., Wilson and Gaudron JJ. in Hoch v The Queen [70] :

The basis for the admission of similar fact evidence lies in its possessing a particular probative value or cogency by reason that it reveals a pattern of activity such that, if accepted, it bears no reasonable explanation other than the inculpation of the accused person in the offence charged. Their Honours went on to point out [71] that the strength of the probative force of the evidence: lies in the fact that it reveals "striking similarities", "unusual features", "underlying unity", "system" or "pattern" such that it raises, as a matter of common sense and experience, the objective improbability of some event having occurred other than as alleged by the prosecution. While that passage refers to "striking similarities" and the like, its emphasis is on the relationship between the evidence and on the improbability which follows from that relationship. Although the language used is not identical with that in Director of Public Prosecutions v P. , there is a consistency of approach in the two judgments. 1. (1988) 165 C.L.R. 292, at p. 294. 2. ibid., at pp. 294-295.

It is important, in my view, that the admissibility of evidence of other offences be tested against the criterion identified in Hoch , read in the light of Director of Public Prosecutions v P. An approach in terms only of probative force outweighing prejudicial effect leaves too many questions unanswered. In that connexion it should be said that Harriman v The Queen [72] , which is sometimes referred to in this context, was not truly a decision on similar fact or propensity evidence. The evidence in question was admissible for the light it threw on an association between the accused and another man. This was directly relevant to the primary issue raised by the indictment, whether the accused was "knowingly concerned" in the importation of heroin. Likewise, S. v The Queen [73] , which is sometimes mentioned in the same context, is not truly a case of similar fact or propensity evidence.

1. (1989) 167 C.L.R. 590. 2. (1989) 168 C.L.R. 266.

The judgment of Mason C.J., Wilson and Gaudron JJ. in Hoch identifies the basis upon which evidence of similar facts is admitted. "The criterion of its admissibility is the strength of its probative force." [74] However, it remains with the trial judge to consider the prejudicial effect of the evidence and whether, in the circumstances, it is just to admit it. That is not precisely the same as saying that the evidence is admissible if its probative force exceeds its prejudicial effect. The probative force of the evidence, in accordance with the criterion discussed, is the basis for its admissibility; nevertheless, the trial judge may exclude it. Evidence that an accused has committed other relevant offences must inevitably have a prejudicial effect. But, in the language of Director of Public Prosecutions v P. , it may nevertheless be "just" to admit the evidence. The reference to just aptly conveys the notion that it is not only the interests of the accused that are involved. The legitimate interests of the Crown and of the community cannot be overlooked [75] . The admissibility of the H. incident depends upon it possessing those features identified in Hoch . Unless it does, the evidence is not admissible. If it does, the evidence may be admitted notwithstanding its prejudicial effect if the trial judge considers it just to admit the evidence.

1. (1988) 165 C.L.R., at p. 294. 2. There is in any event a discretion in a trial judge to reject evidence where its prejudicial effect outweighs its probative value, in order to secure a fair trial for the accused: Harriman (1989), 167 C.L.R., at pp. 594, 608, 633; Reg. v Sang [1980] AC. 402, at p. 437.

It is important to say again that if the evidence of the H. incident is admissible, it is on the footing that other evidence justified a conclusion by the jury that Michael Black had been abducted from Sturt Reserve and later murdered. Other evidence did justify that conclusion, for the reasons already given. Consistent with authority, the question then is whether the H. evidence revealed sufficient similarities with the evidence surrounding Michael's abduction so as to raise, as a matter of commonsense and experience, the objective improbability of someone other than the appellant, on the day and about the time in question, abducting Michael from Sturt Reserve.

The respondent points to a number of features of the H. incident. H. was thirteen years and three months when abducted by the appellant. He was lured by a ruse into the appellant's Kombi van. Once he was in the vehicle, the appellant threatened him, retrieved the boy's bicycle, bound and gagged him and drove away. An hour or so later the appellant placed the bicycle on the top of cliffs at Port Noarlunga South and took the boy to his (the appellant's) home that night. H. escaped from the house at about midday on the following day. The purpose of the abduction was the appellant's sexual gratification.

Once the jury concluded (as they were entitled to) that Michael Black had been abducted, it was apparent that the abductor was a person who was at Sturt Reserve around 2.30 p.m. on 18 January 1989. The appellant was such a person. Indeed he gave evidence that he had spoken to Michael just before he left Sturt Reserve some time between 2.15 p.m. and 2.45 p.m. At his request the boy had been in his van to return a knife. He was the last person seen talking to the boy and he was possessed of the means, a van, to effect an abduction at that time and place. The jury were also entitled to conclude that the purpose of the abduction was sexual gratification. No ransom was demanded and there was no suggestion of family discord which might have led to the abduction. It is true that the abduction might have been for non-sexual purposes but that was a hypothesis the jury were entitled to reject in all the circumstances. The appellant was a person who sought sexual gratification by abducting a young boy on another occasion and he had done so with the use of his Kombi van.

The jury were also entitled to conclude that the abductor had deliberately tried to mislead by placing Michael's belongings in a position likely to suggest that he had drowned. The jury were entitled to conclude furthermore that the appellant had tried to mislead in the case of H. by leaving the boy's bicycle in a position where it might be thought that he had suffered an accident. Certainly the appellant gave no real explanation as to why he retrieved H.'s bicycle, put it in the van and then deposited it some distance away.

It is apparent that the H. incident and the circumstances surrounding Michael's disappearance had a number of features in common, such as to warrant in the context of all the evidence a conclusion that the same person was involved in both abductions. There is a particular feature of Michael's abduction that gives relevance and persuasiveness to the H. incident, namely, that if the appellant was not the abductor there must have been someone else at Sturt Reserve around 2.30 p.m. on 18 January 1989, possessing a vehicle or some other means by which the boy could be abducted without any of those present being aware of what was happening. There is no evidence pointing to the presence of another such person. This is not so much an aspect of similar facts as circumstantial evidence which the jury were entitled to take into account (consistently with the principles enunciated in Chamberlain v The Queen [No. 2] [76] and Shepherd v The Queen [1] ) along with the H. evidence in reaching a decision as to the guilt of the appellant. All these considerations justified the trial judge's ruling that the evidence was admissible.

1. (1984) 153 C.L.R. 521. 2. (1990) 170 C.L.R. 573.

A matter of some moment is the trial judge's references to "disposition" and "propensity" in his charge to the jury, because of the emphasis these terms placed on the character of the appellant. At the same time his Honour linked these terms to what was an integral part of the Crown case, namely, the similarities in the two abductions and the improbability that anyone else could have been the abductor of Michael Black. Taking the directions in their entirety, the jury were not likely to have been misled by these references.

In my view the evidence of the H. incident was admissible. No real attack was made by the appellant on the exercise of a residual discretion to reject the evidence, as opposed to the admissibility of the evidence. Once the evidence was admitted, the jury were properly directed as to the use they could make of it.

The question in this appeal is whether, on the trial of the appellant on a charge of murdering Michael Black, a ten-year-old boy, evidence was admissible against him that twelve months after the alleged murder he abducted and sexually assaulted a thirteen-year-old boy. The Crown contends that the proper inference to be drawn from all the evidence in the case was that Michael Black - whose body has not been found - was abducted, sexually assaulted and murdered. The Crown argues that, because the appellant had the opportunity and means to abduct him on the day that he disappeared, evidence concerning the subsequent sexual assault was relevant because it proved that the appellant had a disposition to abduct and sexually assault young boys. Ms. Shaw, counsel for the appellant, denied that evidence of an accused person's propensity to commit crime was admissible to prove a criminal charge. She conceded that in some circumstances evidence of other criminal conduct on the part of an accused was admissible against that person. But she contended that such evidence was only admissible if it proved guilt for a reason other than that the accused was a person who was likely to commit crime or crime of a particular kind.

It becomes necessary in this appeal, therefore, to examine once again the vexed question as to the circumstances in which the prosecution may prove a criminal charge by tendering evidence that the accused has engaged in criminal conduct on occasions other than that which is the subject of the charge before the court.

On 18 January 1989, Michael Black rode his BMX bike to Sturt Reserve, a picnic spot alongside the Murray River at Murray Bridge, South Australia. He was last seen there in the early afternoon of that day. The appellant had arrived at Sturt Reserve the previous day and camped overnight in his Kombi van. On the first day that he was at Sturt Reserve, the appellant spoke to two children in terms that invited them to ride in or enter his van. The appellant admitted in evidence that he had spoken to Michael Black at about 1.30 p.m. on 18 January and lent him a knife to kill a fish. Later that day, searchers found Michael Black's BMX bike, his fishing rod, some of his clothes and his dog at Thiele Reserve, which is not far from Sturt Reserve but is on the opposite side of the Murray River. Some time before 3.30 p.m. one witness saw the bike at the place where it was found. About 3.15 p.m. another witness heard a van, with an engine noise similar to that of the appellant's van, arrive at Thiele Reserve. The driver kept the engine running. The witness heard the doors of the vehicle open and close and the distressed barking of a dog. None of the people who were at Thiele Reserve that afternoon including two persons who knew Michael Black recollect seeing him there. There is no evidence that Michael Black has been seen since he was last seen at Sturt Reserve [2] . A search of the Murray River failed to find his body. Evidence tendered by the Crown suggested that it was extremely unlikely that Michael Black had drowned.

1. Two witnesses gave evidence that they had seen Michael Black riding his bike across a bridge on the Murray River in a direction away from Sturt Reserve. The unaided recollection of these witnesses would have placed the time of this sighting as being after 3 p.m. But an examination of a banking transaction that took place almost immediately after this sighting established that the sighting must have been before 1.22 p.m.

On Saturday, 30 December 1989, H., a thirteen-year-old boy, was riding a BMX bike belonging to a friend along a suburban street in Port Noarlunga, South Australia when the appellant called out to him. The appellant asked H. to help him get some keys out of a compartment in his van. H. got into the van. While he was looking for the keys, the appellant "came in behind me and he closed the door pushed me down onto this seat at the back of the van, and said, "Don't try anything or I'll hurt you" ". The appellant then got out of the van and brought the bike into the van. He tied H.'s hands behind his back and blindfolded him, put a sock in his mouth and taped his mouth and ankles. After driving for about half an hour, the appellant parked the van at a car park. While they were there, H. told the appellant that the bike was not his bike and that he wanted him to put it somewhere safe. H. also told the appellant that he had planned to go swimming that afternoon. Soon after, the appellant drove the van to a beach where he wiped the bike for fingerprints and put it into a bush area. After driving for another two hours, the appellant stopped the van, removed H.'s shorts and sexually assaulted him. They then drove to the appellant's house where the appellant untaped H.'s hands, feet and ankles. The appellant told H. that he would kill him or hurt him unless H. took off his own pants and underpants. The appellant then again sexually assaulted H. The following morning the appellant tied H. to a chair with rope and put a sock into his mouth. While the appellant was absent from the house, H. got his hands free from the rope and escaped. The appellant was arrested when he returned to the house.

After his arrest over the H. incident, the appellant told his wife that he had been "thinking" about doing "it" for twelve months and that he had committed the offence because he was lonely. Subsequently, he pleaded guilty to a number of offences arising out of these events and was charged with the murder of Michael Black.

It will later be necessary to examine the evidence in relation to Michael Black's disappearance and the findings of the trial judge on that evidence. But the above outline of the facts is the foundation of the Crown case that the appellant abducted Michael Black at Sturt Reserve for the purpose of sexually assaulting him, murdered him, and planted a false trail at Thiele Reserve to suggest that Michael Black had drowned while swimming in the Murray River at that reserve. The case against the appellant was purely circumstantial. But, without the evidence concerning the abduction and sexual assault of H., the Crown did not establish a prima facie case against the appellant. Thus, the critical question in the appeal is whether the H. evidence was admissible against the appellant. In my opinion it was. However, my reasons differ from those of the learned trial judge. They also require that the principles of admissibility formulated in Markby v The Queen [3] be modified if propensity evidence is to be admitted in criminal trials.

On the trial of a criminal charge, evidence that reveals that the accused is a person of bad character is not admissible if it proves no more than that he or she has a general disposition or propensity to commit crime or crime of a particular kind [4] . For more than a century, one of the fundamental theses of the common law has been that on a criminal charge guilt is not to be "inferred from the character and tendencies of the accused" [5] . The character and tendencies of a person are relevant in determining whether that person has committed the crime with which he or she is charged [6] . But as a matter of policy the law generally excludes evidence of other incidents that reveals the criminal or discreditable propensities of the accused. Various reasons have been put forward to justify this exclusion. One reason is that it creates undue suspicion against the accused and undermines the presumption of innocence [7] . Another is that tribunals of fact, particularly juries, tend to assume too readily that behavioural patterns are constant and that past behaviour is an accurate guide to contemporary conduct [8] . Similarly, "[c]ommon assumptions about improbability of sequences are often wrong" [9] , and when the accused is associated with a sequence of deaths, injuries or losses, a jury may too readily infer that the association "is unlikely to be innocent" [9] . Another reason for excluding the evidence is that in many cases the facts of the other misconduct may cause a jury to be biased against the accused [10] . In the present case, for example, once the H. evidence was admitted, it would require a superhuman effort by the jury to regard the appellant as other than a person of depraved character whose uncorroborated evidence, whether or not he was guilty, could not be acted upon except where it supported the prosecution case. Functional reasons also play a part in excluding evidence of bad character. Trials would be lengthened and expense incurred, often disproportionately so, in litigating the acts of other misconduct; law enforcement officers might be tempted to rely on a suspect's antecedents rather than investigating the facts of the matter; rehabilitation schemes might be undermined if the accused's criminal record could be used in evidence against him or her [11] . Despite these policy considerations, in some circumstances evidence that discloses a propensity to engage in criminal or discreditable conduct may be admissible against the accused.

1. Makin v Attorney-General (N.S.W.) [1894] AC. 57, at p. 65; Markby (1978), 140 C.L.R., at p. 116, per Gibbs A.C.J. with whose judgment Stephen, Jacobs and Aickin JJ. agreed; Perry v The Queen (1982), 150 C.L.R. 580, at p. 585, per Gibbs C.J.; p. 603, per Wilson J.; p. 609, per Brennan J. 2. Dawson v The Queen (1961), 106 C.L.R. 1, at p. 16, per Dixon C.J. 3. Reg. v Rowton (1865), Leigh & Ca. 520, at pp. 540-541 [169 E.R. 1497, at p. 1506], per Willes J. 4. Perry (1982), 150 C.L.R., at pp. 593-594, per Murphy J.; Reg. v Boardman [1975] AC. 421, at p. 451, per Lord Hailsham; Piragoff, Similar Fact Evidence, (1981), p. 4. 5. Cowen and Carter, Essays on the Law of Evidence (1956), pp. 144-145; Imwinkelried, The Use Of Evidence Of An Accused's Uncharged Misconduct To Prove Mens Rea: The Doctrines which Threaten to Engulf the Character Evidence Prohibition, Ohio State Law Journal, vol. 51 (1990) 575, at pp. 581-582; Ligertwood, Australian Evidence, 2nd ed. (1993), pp. 81-82; Palmer, The Scope of the Similar Fact Rule, Adelaide Law Review, vol. 16 (1994) 161, at p. 169. 6. Perry (1982), 150 C.L.R., at p. 594, per Murphy J. 7. Perry (1982), 150 C.L.R., at p. 594, per Murphy J. 8. R. v Bond [1906] 2 KB. 389, at p. 398, per Kennedy J.; Piragoff, op. cit., at pp. 4-5; Cross on Evidence, 2nd Aust. ed. (1991), par. 21145; Ligertwood, op. cit., p. 81. 9. See Cross on Evidence, op. cit., par. 21150.

This Court has said that, if evidence that discloses the bad character of the accused is relevant to the charge "for some reason other than that he has committed crimes in the past or has a criminal disposition" [12] , it may be admissible. It will be admissible as a matter of law and not discretion [13] if the probative force of the evidence "outweighs" [14] or "clearly transcends" [15] its prejudicial effect. As the reasoning of the Court's decision in Harriman v The Queen [16] shows, the principles concerning the admissibility of evidence revealing other acts of misconduct are not confined to the so-called similar fact cases. They apply in any criminal trial where the Crown wishes to lead evidence tending to show that the accused has been guilty of wrongful acts other than those with which the accused is charged [17] .

1. Markby (1978), 140 C.L.R., at p. 116, per Gibbs A.C.J. with whose judgment Stephen, Jacobs and Aickin JJ. agreed. See also Perry (1982), 150 C.L.R., at p. 585, per Gibbs C.J.; p. 603, per Wilson J.; p. 609, per Brennan J. 2. Boardman [1975] AC., at p. 457, per Lord Cross of Chelsea; Markby (1978), 140 C.L.R., at p. 117, per Gibbs A.C.J.; Perry (1982), 150 C.L.R., at p. 585, per Gibbs C.J. 3. Perry (1982), 150 C.L.R., at p. 604, per Wilson J.; Sutton v The Queen (1984), 152 C.L.R. 528, at p. 564, per Dawson J.; Thompson v The Queen (1989), 169 C.L.R. 1, at p. 16, per Mason C.J. and Dawson J.; Harriman v The Queen (1989), 167 C.L.R. 590, at p. 598, per Dawson J.; B. v The Queen (1992), 175 C.L.R. 599, at p. 608, per Brennan J.; p. 618, per Dawson and Gaudron JJ. 4. Perry (1982), 150 C.L.R., at p. 609, per Brennan J.; Sutton (1984), 152 C.L.R., at p. 547, per Brennan J.; p. 560, per Deane J.; Hoch v The Queen (1988), 165 C.L.R. 292, at p. 300, per Brennan and Dawson JJ.; Thompson v The Queen (1989), 169 C.L.R., at p. 16, per Mason C.J. and Dawson J.; Harriman (1989), 167 C.L.R., at p. 633, per McHugh J. 5. (1989) 167 C.L.R. 590. 6. The contrary view expressed in Reg. v Rogerson & Paltos (1992), 65 A. Crim. R. 530, at p. 543 is incorrect.

In Markby [18] , Gibbs A.C.J. said that evidence that is relevant for a reason other than that it shows the criminal propensity of the accused is admissible as a matter of law if it has a "strong degree of probative force" [19] or "a really material bearing on the issues" [20] or is "so very relevant that to exclude it would be an affront to common sense" [21] . By criminal propensity, his Honour obviously meant an assumed tendency to commit criminal acts based on the fact that the accused has committed criminal acts in the past. On this view, evidence relevant for a reason other than that it shows the criminal or discreditable propensity of the accused is admissible as a matter of law, notwithstanding that it discloses that propensity [22] of the accused, if the evidence is strongly probative of that person's guilt. On this view, there is no requirement to weigh the probative value of the evidence against the risk of prejudice that its admission might generate. But Gibbs A.C.J. went on to say that the trial judge retains a discretion to exclude the evidence if the risk of prejudice outweighs its probative effect [18] - the so-called Christie [23] discretion. Gibbs C.J. repeated these views in Perry v The Queen [1] and Sutton v The Queen [2] .

1. (1978) 140 C.L.R., at p. 117. 2. Boardman [1975] AC., at p. 444, per Lord Wilberforce. 3. ibid., at p. 439, per Lord Morris of Borth-y-Gest citing Harris v Director of Public Prosecutions [1952] AC. 694, at p. 710. 4. ibid., at p. 456, per Lord Cross. 5. Character evidence indicates a person's general or usual traits or tendencies and, therefore, that person's propensities. In this branch of the law, however, the term "criminal propensity" is frequently used in a specific way to describe the manner in which a person commits criminal or wrongful acts. Thus, "[w]hile one would not normally say that a person has a "character" trait to commit a crime wearing a particular disguise, it seems appropriate to use the term "propensity" if he has committed a number of crimes in that disguise." (Australian Law Reform Commission, Report No. 26 (Interim), Evidence , (1985), vol. 2 at p. 223.) 6. (1978) 140 C.L.R., at p. 117. 7. R. v Christie [1914] AC. 545. 8. (1982) 150 C.L.R., at p. 586. 9. (1984) 152 C.L.R., at pp. 533-534.

Recent authority in this Court [3] , however, rejects the proposition that evidence revealing that the accused has committed acts of misconduct other than those the subject of the charge is admissible as a matter of law as long as it is strongly probative of that person's guilt. The recent cases emphasize that as a matter of law and not discretion the probative value of evidence revealing bad character or criminal propensity must be sufficiently strong to outweigh or clearly transcend the prejudicial effect of the evidence. But Markby has not been overruled on this point. In the practical administration of criminal justice, the difference between these two views is of real significance. If it is a condition of admissibility that the prosecution must show that the probative value of the evidence outweighs its prejudicial effect, the onus is on the prosecution to prove that condition. If the evidence is admissible merely because it has strong probative force, the onus is on the accused to show that evidence otherwise admissible should be rejected. Perhaps even more importantly, if this class of evidence is excluded as a matter of discretion, appellate review of the trial judge's discretion will be more limited than it will be if it is excluded as a matter of law [4] .

1. Perry (1982), 150 C.L.R., at p. 604, per Wilson J.; p. 609, per Brennan J.; Sutton (1984), 152 C.L.R., at p. 547, per Brennan J.; p. 560, per Deane J.; p. 564, per Dawson J.; Hoch (1988), 165 C.L.R., at p. 300, per Brennan and Dawson JJ.; Thompson v The Queen (1989), 169 C.L.R., at p. 16, per Mason C.J. and Dawson J.; Harriman (1989), 167 C.L.R., at p. 598, per Dawson J.; p. 633, per McHugh J.; B. (1992), 175 C.L.R., at p. 608, per Brennan J.; p. 618, per Dawson and Gaudron JJ. 2. In Reg. v Straffen [1952] 2 QB. 911, at p. 913, for example, counsel for the accused accepted that the prejudicial effect of the evidence went to the exercise of discretion and not admissibility. Consequently, he felt unable to challenge the trial judge's exercise of discretion.

Having regard to the recent decisions in this Court, the prejudicial effect of the evidence must be regarded as going to the issue of admissibility. In Sutton [5] , Dawson J. thought that the trial judge retained a discretion to reject the evidence even if its probative value outweighed or transcended its prejudicial effect. But his Honour went on to express the view that he found it difficult to envisage circumstances that could call for the exercise of the discretion. In my opinion, once it is accepted that the prejudicial effect of the evidence is a matter going to admissibility, no scope remains for the exercise of the discretion to reject probative evidence in criminal trials on the ground that it is unduly prejudicial to the accused.

The standard of proof for admissibility of evidence disclosing criminal propensity

In Perry [6] , Murphy J. thought that the risk of prejudice from admitting similar fact evidence is so great that the tribunal of fact should use such evidence against the accused only when there is "no rational explanation consistent with the accused's innocence". In Sutton [7] , Dawson J. went further. His Honour said that evidence was not sufficiently probative of guilt to overcome the risk of prejudice and be admissible against the accused unless there was no rational view of that evidence that was consistent with the innocence of the accused. This test was applied by Mason C.J., Wilson and Gaudron JJ. in Hoch v The Queen [8] and by Dawson, Toohey and Gaudron JJ. in Harriman [9] . If these statements represent the law, it is difficult to see why the recent cases have departed from the tests laid down by Gibbs C.J. in Markby and insisted that to be admissible the probative value of the evidence must outweigh or transcend its prejudicial effect.

1. (1982) 150 C.L.R., at p. 596. 2. (1984) 152 C.L.R., at p. 564. 3. (1988) 165 C.L.R., at p. 296. 4. (1989) 167 C.L.R., at pp. 602, 607 and 613 respectively.

If evidence revealing criminal propensity is not admissible unless the evidence is consistent only with the guilt of the accused, the requirement that the probative value "outweigh" or "transcend" the prejudicial effect is superfluous. The evidence either meets the no rational explanation test or it does not. There is nothing to be weighed - at all events by the trial judge. The law has already done the weighing. This means that, even in cases where the risk of prejudice is very small, the prosecution cannot use the evidence unless it satisfies the stringent no rational explanation test. It cannot use the evidence even though in a practical sense its probative value outweighs its prejudicial effect.

Moreover, in Harriman [10] , Dawson J., correctly in my view, assumed that the exclusionary rules concerning evidence of criminal propensity are not confined to similar fact evidence. Accordingly, if the no rational explanation test constitutes a condition of admissibility, evidence that the proceeds of a robbery carried out by the accused were found at the scene of a murder would not be admissible to prove the murder charge unless it satisfied that test [11] . Furthermore, if the no rational explanation test is the law, practitioners and trial judges have long been engaged in a futile exercise. They have wasted much time and effort in attempting to identify the prejudice to the accused that might arise if the evidence is admitted, the weight to be given to the evidence, and whether that weight outweighs the likely prejudice. A further reason for concluding that the no rational explanation test is not applicable to every tender of such evidence is that it requires the trial judge and, on appeal, the Court of Criminal Appeal to perform the task of the jury and evaluate the strength of the evidence. I do not recollect any other area of the Australian law of evidence where the test that the judge applies to the admissibility of a class of evidence is the same test that the jury must apply to the question of guilt if the evidence is admitted.

1. ibid., at p. 601. 2. cf. the facts of Reg. v O'Meally [No. 2] [1953] VL.R. 30.

In my opinion, the standard of proof required to admit evidence disclosing a person's criminal or discreditable propensities varies according to the reasoning process to be employed, the nature of the evidence, and the degree of potential risk to a fair trial if the evidence is admitted. It will be necessary later in this judgment to return to this question.

In Perry [12] , Murphy J. said that, despite protests to the contrary, evidence has been tendered in similar fact cases to prove the guilt of an accused by showing that the accused was a person likely to have committed the offence for which he or she was being tried. In Harriman , Dawson J. challenged the conventional view that propensity evidence was not admissible to prove the guilt of the accused. His Honour said [13] that neither Markby nor Hoch intended "to say that propensity evidence must be relevant otherwise than as evidence of disposition before it can be admitted". Dawson J. went on to say: "A close examination of the cases decided in an effort (ultimately unsuccessful) to avoid the forbidden chain of reasoning will show that when propensity evidence was admitted it was in general because of its relevance as propensity evidence, whatever other label was put upon it." [13] His Honour thought [13] that Reg. v Straffen [14] and Makin v Attorney-General (N.S.W.) [15] were examples of cases where the probative value of the evidence lay in proof of the accused's disposition. In B. v The Queen [16] , Dawson and Gaudron JJ. said that evidence of propensity may be admissible if it "goes beyond showing a mere propensity to commit crime or a particular type of crime and points in some other way to the commission of the offences in question, then it will be admissible if its probative value for that purpose outweighs its prejudicial effect". However, their Honours went on to say [16] :

That is not to say that it may not still be evidence of propensity or disposition, but it will then be evidence of propensity or disposition of a particular kind which in the circumstances has a degree of relevance justifying its admission. 1. (1982) 150 C.L.R., at p. 593. 2. (1989) 167 C.L.R., at p. 600. 3. (1989) 167 C.L.R., at p. 600. 4. (1989) 167 C.L.R., at p. 600. 5. [1952] 2 Q.B. 911. 6. [1894] A.C. 57. 7. (1992) 175 C.L.R. 599, at p. 618. 8. (1992) 175 C.L.R. 599, at p. 618.

Since Markby , however, this Court has rejected the use of an accused person's propensity to commit crime as a reasoning factor in determining guilt or innocence. If propensity reasoning is a legitimate form of reasoning in criminal courts in this country, the principle on which Markby was decided needs modification. In Markby , the Court unanimously held that on a charge of murder, arising out of a killing occurring during a sale of drugs to the deceased, evidence was not admissible that the accused had been involved in earlier incidents where a person buying drugs had been cheated or robbed. Gibbs A.C.J., with whose judgment Stephen, Jacobs and Aickin JJ. agreed, said [17] that the principles applicable had been "authoritatively stated in Makin " [18] . His Honour said that "evidence of similar facts is not admissible if it shows only that the accused had a propensity or disposition to commit crime, or crime of a particular kind, or that he was the sort of person likely to commit the crime charged" [17] . But Gibbs A.C.J., echoing the words of Lord Herschell in Makin [19] , said that evidence of similar facts is admissible "if it is relevant in some other way" [17] . His Honour then went on to make it clear that the evidence must be relevant in some way other than through disposition by saying "that is, if it tends to show that he is guilty of the crime charged for some reason other than that he has committed crimes in the past or has a criminal disposition". As Dr. J.R.S. Forbes has pointed out [20] , " Markby , unlike Makin , explicitly forbids reasoning from disposition in any case ". Nevertheless, the statement of principle by Gibbs A.C.J. accorded with what Lord du Parcq, speaking for the Judicial Committee, had said in Noor Mohamed v The King [21] .

1. Markby (1978), 140 C.L.R., at p. 116. 2. [1894] A.C., at p. 65. 3. Markby (1978), 140 C.L.R., at p. 116. 4. In Makin [1894] AC., at p. 65, Lord Herschell giving the Advice of the Privy Council said: "It is undoubtedly not competent for the prosecution to adduce evidence tending to shew that the accused had been guilty of criminal acts other than those covered by the indictment, for the purpose of leading to the conclusion that the accused is a person likely from his criminal conduct or character to have committed the offence for which he is being tried." But Lord Herschell immediately qualified that general proposition. Taken literally, the qualification negates the general proposition. His Lordship said: "On the other hand, the mere fact that the evidence adduced tends to shew the commission of other crimes does not render it inadmissible if it be relevant to an issue before the jury, and it may be so relevant if it bears upon the question whether the acts alleged to constitute the crime charged in the indictment were designed or accidental, or to rebut a defence which would otherwise be open to the accused." Lord Herschell made no reference to the need to weigh the cogency of the evidence of other criminal conduct against the prejudice that its admission was likely to generate. 5. Markby (1978), 140 C.L.R., at p. 116. 6. Similar Facts (1987), p. 28.

The facts proved as to the death of Gooriah would certainly tend to deepen that suspicion, and might well tilt the balance against the accused in the estimation of the jury. It by no means follows that this evidence ought to be admitted. If an examination of it shows that it is impressive just because it appears to demonstrate, in the words of Lord Herschell in Makin's case "that the accused is a person likely from his criminal conduct or character to have committed the offence for which he is being tried," and if it is otherwise of no real substance, then it was certainly wrongly admitted.

In Perry [22] , the accused was charged with attempting to murder her third husband in 1978 by administering arsenic to him. The Crown sought to prove the charge, inter alia, by evidence that a de facto husband of the accused who had died of barbiturate poisoning in 1970 had symptoms consistent with arsenic poisoning. The Crown also tendered evidence that her second husband had died of arsenic poisoning in 1961 and that her brother had died of arsenic poisoning in 1962. All members of the Court held that the evidence relating to the de facto husband was inadmissible and the Court was equally divided as to whether the evidence concerning the brother was admissible. By majority, the Court held that the evidence concerning the second husband was admissible. Gibbs C.J. stated the law in terms similar to those which he had used in Markby . Wilson J. and Brennan J. in separate judgments [23] accepted that the law was correctly stated by Gibbs A.C.J. in Markby .

1. (1982) 150 C.L.R. 580. 2. ibid., at pp. 603 and 609 respectively.

In my opinion, Markby and Perry were decided on the principle that evidence of an accused person's criminal conduct is only admissible when it is relevant to the proof of guilt in some way or for some reason other than that the accused is a person who, because of his or her criminal disposition, is likely to have committed the crime in question. Arguably, proof of guilt by proof of the accused's disposition was not the issue in either case. But the role of this Court in declaring the law for the nation would be unduly circumscribed if the principles that govern its decisions were not authoritative or binding beyond the particular facts of the case in which the principle was formulated and applied. Unless subsequent decisions of this Court have undermined their authority, Markby and Perry should be treated as establishing, as Ms. Shaw contended, that evidence of criminal conduct is only admissible if it is relevant in some way other than establishing that the accused has a disposition to commit crime or a particular kind of crime.

In my opinion, none of the subsequent decisions of this Court overrule Markby or Perry . Nor is the ratio decidendi of any of the subsequent decisions inconsistent with Markby or Perry . In Sutton , the Court held that evidence in respect of a number of charges of rape was admissible in proof of each charge of rape because of the striking similarity of each offence. Gibbs C.J. and Brennan J. applied the principles that they had applied in the earlier cases. Deane J. said [24] that the effect of Makin is that evidence that the accused is a person of bad character is deemed irrelevant in English law. But his Honour also said that evidence of bad character "may be admitted if it is relevant for other reasons" [25] . At least in the context of similar facts cases, Deane J. made it plain [26] that it is probability reasoning not propensity reasoning that makes the damaging evidence admissible. Dawson J. accepted [27] that the law was as stated in Makin , Reg. v Boardman [28] , Markby and Perry .

1. Sutton (1984), 152 C.L.R., at p. 556, citing Boardman [1975] AC., at p. 461, per Lord Salmon. 2. ibid., citing Noor Mohamed , [1949] A.C., at p. 195. 3. ibid., at p. 557. 4. ibid., at p. 563. 5. [1975] A.C. 421.

In Hoch , the Court held that a number of charges of indecent assault should not have been heard together unless the trial judge had found that there was no reasonable basis for concluding that the complainants had concocted their similar stories. In a joint judgment, Mason C.J., Wilson and Gaudron JJ. stated the principles for the admissibility of similar fact evidence in terms of probability and not propensity reasoning. Their Honours said [29] :

The basis for the admission of similar fact evidence lies in its possessing a particular probative value or cogency by reason that it reveals a pattern of activity such that, if accepted, it bears no reasonable explanation other than the inculpation of the accused person in the offence charged: see Dixon J.'s discussion in Martin v Osborne [30] . In that same case Evatt J. pointed out that it bears that probative value or cogency not as a matter of deductive logic but by reason that it allows for "admeasuring the probability or improbability of the fact or event in issue, if we are given the fact or facts sought to be adduced in evidence" [31] . 1. (1988) 165 C.L.R., at p. 294. 2. (1936) 55 C.L.R. 367, at p. 375. 3. ibid., at p. 385.

In Thompson v The Queen [32] where the accused was charged with the murders of two sisters, the Court upheld the admissibility of evidence of subsequent murders by the accused of another sister and her family on the ground that there was a striking similarity in the manner that the various deaths had occurred. Probability reasoning and not propensity reasoning was used. Mason C.J. and Dawson J. said [33] that the "similarity between the circumstances of those murders and the circumstances of the earlier deaths of the two sisters was sufficiently striking to eliminate coincidence as a reasonable hypothesis and to lead to the conclusion that the two sisters did not die accidentally but at the hand of the applicant". Brennan J. [34] agreed that the evidence was admissible because of the striking similarity of the circumstances of the murders. Deane J. [35] and Gaudron J. [36] thought that the evidence was admissible not because of the striking similarity between the circumstances of the various murders but by reason of "the underlying unity" of the two incidents. Gaudron J. said [36] that the "inextricable connexion or underlying unity, rather than points of similarity, raises the improbability of the deaths having occurred in a manner other than as alleged in the prosecution case and gives the evidence that probative force which renders it admissible".

1. (1989) 169 C.L.R. 1. 2. (1989) 169 C.L.R., at p. 18. 3. ibid., at pp. 31-32. 4. ibid., at p. 32. 5. ibid., at p. 40. 6. ibid., at p. 40.

In Harriman , where the appellant was charged with conspiring to bring heroin into Australia, the Court upheld the admissibility of evidence that disclosed that the appellant had been involved in other heroin dealings with the co-conspirator. The Court held that the evidence was admissible to show the relationship between those two persons and to negative the suggestion that the appellant had innocently accompanied the co-conspirator to a town in Thailand where the co-conspirator had purchased heroin. Apart from Dawson J., no member of the Court thought that the evidence was admissible as propensity evidence. Moreover, four members of the Court said that evidence is not admissible if it does no more than merely show that the accused has a propensity to commit crime [37] .

1. Harriman (1989), 167 C.L.R., at p. 594, per Brennan J.; p. 607, per Toohey J.; p. 613, per Gaudron J.; p. 627, per McHugh J.

Apart from the statements of Dawson J., nothing in the cases subsequent to Markby and Perry undermines or raises any doubt about the principles formulated by Gibbs A.C.J. in Markby . Indeed, in Perry and other cases, the majority of the judges have accepted the validity of those principles. But that is not the end of the matter. The question arises whether those principles accurately reflect the way that this branch of the law of evidence has operated in Australia and in England in this century and whether they need to be modified. To those questions I now turn.

Prior to the decision of the House of Lords in Boardman , the admissibility of what is often misleadingly called similar fact evidence depended upon whether the evidence fell into one of the exceptions to the rule that it was "not competent for the prosecution to adduce evidence tending to shew that the accused has been guilty of criminal acts other than those covered by the indictment" [38] . Recognized exceptions were those in which evidence was tendered to prove system, identity, knowledge or intent or which rebutted "defences" such as accident or innocent association. The list of exceptions was not closed [39] . But the tendency of the courts was to determine whether the evidence fell into a recognized category. If it did that was usually the end of the matter. Nevertheless, at least after the decision of the Privy Council in Noor Mohamed [40] , it was recognized that the trial judge retained the ordinary discretion applicable in criminal trials to reject the tender of evidence that was legally admissible if its prejudicial effect outweighed its probative value.

1. Makin [1894] AC., at p. 65. 2. Harris v Director of Public Prosecutions [1952] AC., at p. 705. 3. [1949] A.C., at p. 192.

In 1974 in Boardman [41] , the House of Lords, although not openly departing from the law as laid down in Makin , substantially reformulated the conditions for admitting evidence of criminal conduct in proof of a criminal charge when that conduct is not the subject of the charge. In Boardman , the House unanimously held that evidence of other criminal conduct is not admissible if it does no more than prove that, because of the accused's criminal character or disposition, he or she is a person who was likely to have committed the crime in question. Like Lord Herschell in Makin , however, all members of the House accepted that in some circumstances evidence of criminal conduct, other than the conduct that is the subject of the charge, is admissible to prove an accused's guilt. Only Lord Hailsham of St Marylebone specifically excluded reasoning from propensity. His Lordship said [42] :

It is perhaps helpful to remind oneself that what is not to be admitted is a chain of reasoning and not necessarily a state of facts. If the inadmissible chain of reasoning is the only purpose for which the evidence is adduced as a matter of law, the evidence itself is not admissible. If there is some other relevant, probative purpose than for the forbidden type of reasoning, the evidence is admitted, but should be made subject to a warning from the judge that the jury must eschew the forbidden reasoning. 1. [1975] A.C. 421. 2. ibid., at p. 453.

Lord Morris of Borth-y-Gest said [43] that evidence of other crimes may be admitted if it "is sufficiently substantial having regard to the purpose to which it is professedly directed to make it desirable in the interest of justice that it should be admitted". Lord Wilberforce said [44] that the "basic principle must be that the admission of similar fact evidence (of the kind now in question) is exceptional and requires a strong degree of probative force". Lord Cross of Chelsea said [45] :

The question must always be whether the similar fact evidence taken together with the other evidence would do no more than raise or strengthen a suspicion that the accused committed the offence with which he is charged or would point so strongly to his guilt that only an ultra-cautious jury, if they accepted it as true, would acquit in face of it. 1. [1975] A.C., at pp. 438-439. 2. ibid., at p. 444. 3. ibid., at p. 457.

In my opinion, the principles of admissibility that were formulated in Boardman cannot be reconciled with the formulation in Markby . The Markby formulation is in accordance with the reasons of Lord Hailsham in Boardman but is inconsistent with the approach of the majority in that case. I cannot agree with the statement of Gibbs C.J. in Sutton [46] that Boardman did not formulate any new principles for the admissibility of evidence disclosing other acts of misconduct. Boardman plainly authorizes the use of propensity reasoning in appropriate cases.

To permit a tribunal of fact to use propensity reasoning undermines the fundamental principle that in a criminal trial guilt is not to be "inferred from the character and tendencies of the accused" [47] . But there are undoubtedly cases where the interests of justice require that evidence should be admitted even though the prosecution intends to rely on the criminal propensity of the accused. Day after day in the criminal courts, evidence of other criminal acts on the part of the accused is admitted to show "the relationship" between the accused and the complainant [48] . In these cases, the relevance of the evidence is that proof of the accused's propensity makes it more likely than not that the accused is guilty of the offence with which he or she is charged.

1. Dawson v The Queen (1961), 106 C.L.R., at p. 16. 2. See, e.g., Reg. v Garner (1963), 81 W.N. (Pt 1) (N.S.W.) 120 (evidence of a long course of cruelty and ill treatment); Reg. v Hissey (1973), 6 S.A.S.R. 280 (previous acts of violence in a case where the charge was murder by means of a blow to the abdomen); Reg. v Etherington (1982), 32 S.A.S.R. 230 (evidence of previous acts of sexual intercourse and sexual assault).

The leading case on relationship evidence in this Court is O'Leary v The King [49] where, on a charge of murdering a fellow employee at a bush camp, the Court upheld the admissibility of evidence proving that on the day and night of the killing the accused had violently assaulted other employees. Latham C.J. rejected [50] the argument that the evidence was admissible because it identified "the accused with the person who had committed the crime, the crime being of a special character presenting specific features which showed that it was committed by a person who had certain abnormal characteristics". His Honour said [51] that the evidence "that, on the day and the night of the killing of Ballard, he actually attacked particular fellow employees without cause is also evidence which goes to show the probability that he would attack some other fellow employee".

1. (1946) 73 C.L.R. 566. Although Dixon and Williams JJ. discussed the events as being part of "one transaction" and Latham C.J. and Rich J. as part of the relevant circumstances, the case was concerned with events occurring in the course of the relationship between the accused and his fellow employees over a period of many hours. 2. (1946) 73 C.L.R., at p. 574. 3. ibid., at p. 575.

The part which the prisoner took in the drunken orgy which, as the facts suggest, culminated in the fatal attack upon the deceased man would appear to me to be relevant to the question whether the prisoner was the assailant and, if so, whether he was at the time capable of forming, and did form, the intention which would make his crime murder. 1. ibid., at p. 577.

The fact that the appellant alone of all these inhabitants was in the course of the orgy committing acts of violence and threatening violence must have in these circumstances probative value as making it logically probable that he was the man who assaulted Ballard. 1. ibid., at p. 582.

Evidence is also frequently admitted in criminal trials to prove that the accused had possession of equipment needed to commit the crime in question even though that evidence discloses the criminal propensities of the accused. In practice, such evidence is led to prove that the accused's propensity makes it likely that he or she committed the crime in question. Thus evidence is admissible, virtually as of course, that the accused was in possession of instruments that might have been used to commit the crime in question. It is unreal to think that a jury uses such evidence simply to prove that the accused is a person who had the means to commit the crime in question [54] . Such evidence is not admissible, however, if it does no more than show that the accused has a propensity to commit crimes of that sort. Thus in Thompson and Wran v The Queen [55] , this Court said that, on a charge of breaking and entering and stealing arising out of safe blowing by means of explosives, evidence would be admissible that the accused were in possession of "a supply of gelignite, detonators, wires and batteries, suitable for the blowing of the safes". However, the Court held that evidence that the accused were in possession of tools that could be used to pick the locks of safes was not admissible. That evidence had no connexion with the crime charged and merely showed a criminal propensity to rob safes.

1. The rhetorical question - "For what purpose did the accused have this [equipment] if it was not to commit this sort of crime?" is not unknown in a prosecutor's address. I suspect that it would occur to few practitioners to question whether such a statement was consistent with Markby principles. 2. (1968) 117 C.L.R. 313, at p. 316.

However, the clearest case of the use of propensity evidence to prove guilt is one involving a sexual offence. It would be impossible to estimate the number of such cases where evidence has been admitted to prove that the accused had committed offences against the complainant on occasions other than the occasion that is the subject of the charge. R. v Ball [56] established that Makin did not prevent the giving of propensity evidence in such cases. In Ball , the accused were charged with incest. They were brother and sister. They lived in a house with one bedroom with a double bed that bore signs of two persons having occupied it. The House of Lords unanimously held that evidence was rightly admitted that previously they had lived in another house as man and wife and that the female defendant had given birth to a child and registered the male defendant as the father. Lord Loreburn L.C. said [57] :

[T]his evidence was clearly admissible on the issue that this crime was committed - not to prove the mens rea, as Darling J. considered, but to establish the guilty relations between the parties and the existence of a sexual passion between them as elements in proving that they had illicit connexion in fact on or between the dates charged. Since Ball , it has never been doubted that in a sexual offence case, evidence of the conduct of the parties is relevant to prove that the offence charged occurred, notwithstanding that the only reasoning process is propensity [58] . 1. [1911] A.C. 47. 2. ibid., at p. 71. 3. See, e.g., R. v Gellin (1913), 13 S.R. (N.S.W.) 271 and Reg. v Etherington (1982), 32 S.A.S.R. 230.

Ball and the cases that follow it establish beyond dispute that the Anglo-Australian law of evidence does not prohibit reasoning from criminal propensity. Indeed, in both Thompson v The King [59] and Straffen [60] the judges openly acknowledged that the accused's propensity to commit the crime in question was a legitimate matter to take into account. In Thompson v The King , the accused was convicted of committing acts of gross indecency with two boys. He was arrested when, according to the prosecution case, he kept a subsequent appointment with the boys. The accused set up an alibi and asserted that it was a case of mistaken identity. The House of Lords held that evidence was admissible against him that the accused was carrying powder puffs when he was arrested and had photographs of naked boys [61] in a drawer at his home. Lord Finlay L.C. said [62] that possession of the articles tended to show that the accused "had abnormal propensities of the same kind" as the person who committed the acts of gross indecency. Lord Sumner said [63] :

The actual criminal made an appointment to meet the same boys at the same time and place three days later and presumably for the same purpose. This tends to show that his act was not an isolated act, but was an incident in the habitual gratification of a particular propensity. The appellant, as his possession of the photographs tends to show, is a person with the same propensity. 1. [1918] A.C. 221, at pp. 225-226, 230, 231, 234, 235, 238. 2. [1952] 2 Q.B., at pp. 916-917. 3. According to Lord Atkinson all the photographs were "indecent in their attitude, and some apparently depraved in suggestion": Thompson v The King [1918] AC., at p. 230. 4. ibid., at p. 225. 5. ibid., at pp. 234-235. See also Lord Atkinson, at p. 230, Lord Parker, at p. 231, Lord Parmoor, at p. 238.

It may be that the evidence that was held to be rightly admitted in Thompson v The King would now be rejected. In Boardman [64] , Lord Cross said that what Lord Sumner had said in Thompson v The King [65] about homosexuality being "a particular perverted lust" "sounds nowadays like a voice from another world". But whether the evidence was rightly or wrongly admitted in Thompson , the Law Lords in that case saw no conflict between Makin and the use of propensity reasoning

1. [1975] A.C., at p. 458. 2. [1918] A.C., at p. 235.

In Straffen , the accused had been charged with the murder of two girls. He was found unfit to plead by reason of his insanity but was committed to an institution. He escaped and was at large for an afternoon. During this period another girl, Linda Bowyer, was murdered. When questioned by police, the accused denied that he had killed her but admitted that he had spoken to her. He also admitted that he had killed the other two girls. Each of the three killings had five similarities: (1) each victim was a young girl; (2) each was manually strangled; (3) none had been sexually assaulted; (4) in no case was there any evidence of a struggle; and (5) no attempt had been made to conceal any of the bodies although that could easily have been done. The Court of Criminal Appeal held that on a charge of having murdered Linda Bowyer evidence establishing the circumstances of the deaths of the other two girls was admissible against the accused. Slade J. who gave the judgment of the Court said [66] :

I can see no distinction in principle between the present case and Thompson v The King Abnormal propensity is a means of identification. In Thompson's case evidence was admitted to prove his identity which showed that he was a person who suffered from the abnormal propensity of homosexuality. It is an abnormal propensity to strangle young girls and to do so without any apparent motive, without any attempt at sexual interference, and to leave their dead bodies where they can be seen and where, presumably, their deaths would be detected. In the judgment of the court, that evidence was admissible because it tended to identify the person who murdered Linda Bowyer with the person who confessed in his statements to having murdered the other two girls a year before, in exactly similar circumstances. 1. Straffen [1952] 2 QB., at pp. 916-917.

In the light of the decisions in O'Leary, Ball, Thompson v The King and Straffen , the statements in Boardman , and the practice of the courts in admitting "relationship" evidence in the sexual offence and other cases, I think that it is impossible to maintain that the Anglo-Australian law of evidence prohibits the use of propensity reasoning in all circumstances. The rational development of this area of the law is not assisted by applying principles that do not accord with the decisions of the courts. In these circumstances [67] , the Court should modify the statements of principle formulated in Markby so that they accord with the reported decisions, the practices of the courts, and the need to do justice to the prosecutor as well as the accused.

1. cf. Hatzimanolis v A.N.I. Corporation Ltd. (1992), 173 C.L.R. 473, at p. 482.

But upon what basis should propensity reasoning be admitted? Plainly, it cannot be admitted merely because it has probative or even strong probative value. The risk of an unfair trial through the use of propensity reasoning is too great to allow such a low threshold of admissibility. Consequently, this Court has insisted that as a matter of law and not discretion the probative value of the evidence must outweigh or transcend its prejudicial effect [68] .

1. See, e.g., Perry (1982), 150 C.L.R., at p. 604, per Wilson J.; p. 609, per Brennan J.; Sutton (1984), 152 C.L.R., at p. 547, per Brennan J.; p. 560, per Deane J.; p. 564, per Dawson J.; Hoch (1988), 165 C.L.R., at p. 300, per Brennan and Dawson JJ.; Thompson v The Queen (1989), 169 C.L.R., at p. 16, per Mason C.J. and Dawson J.; Harriman (1989), 167 C.L.R., at p. 598, per Dawson J.; p. 633, per McHugh J.; B. (1992), 175 C.L.R., at p. 608, per Brennan J.; p. 618, per Dawson and Gaudron JJ.

Nevertheless, the proposition that the probative value of the evidence must outweigh its prejudicial effect is one that can be easily misunderstood. The use of the term "outweigh" suggests an almost arithmetical computation. But prejudicial effect and probative value are incommensurables. They have no standard of comparison. The probative value of the evidence goes to proof of an issue, the prejudicial effect to the fairness of the trial. In criminal trials, the prejudicial effect of evidence is not concerned with the cogency of its proof but with the risk that the jury will use the evidence or be affected by it in a way that the law does not permit. In no sense does the probative value of evidence disclosing propensity, when admitted, outweigh its prejudicial effect. On the contrary, in many cases the probative value either creates or reinforces the prejudicial effect of the evidence. In my view, evidence that discloses the criminal or discreditable propensity of the accused is admitted not because its probative value outweighs its prejudicial effect but because the interests of justice [69] require its admission despite the risk, or in some cases the inevitability, that the fair trial of the charge will be prejudiced.

1. cf. Lord Morris in Boardman [1975] AC., at p. 439.

If there is a real risk that the admission of such evidence may prejudice the fair trial of the criminal charge before the court, the interests of justice require the trial judge to make a value judgment, not a mathematical calculation. The judge must compare the probative strength of the evidence with the degree of risk of an unfair trial if the evidence is admitted. Admitting the evidence will serve the interests of justice only if the judge concludes that the probative force of the evidence compared to the degree of risk of an unfair trial is such that fair minded people would think that the public interest in adducing all relevant evidence of guilt must have priority over the risk of an unfair trial.

If the evidence does no more than prove a mere propensity to commit crimes of the kind in question, it will never have sufficient probative force to make it admissible. If it does have the required degree of probative force, it will be because it is relevant for a reason other than proof of propensity or because it colours one's perception of the other evidence to such an extent that it can be confidently inferred that the accused gave effect to the propensity on the occasion in question. Evidence of a striking similarity between the commission of the crime and the method used by the accused will frequently be sufficient to prove that inference. But striking similarity is not the exclusive test. The circumstantial force of the other evidence together with the propensity evidence may prove the inference. Ball is a good example. Without the propensity evidence, the prosecution had proved no more than an opportunity to commit incest. Moreover, the propensity evidence had no intrinsic mark that connected it with the other evidence. However, the combination of the propensity evidence and the other evidence pointed irresistibly to incest having occurred. Proof of the offence depended on probability reasoning not similar facts. For propensity evidence to be admissible, however, it will need to have "specific probative value in relation to the crime charged " [70] . Even then, propensity evidence will not be admitted if the prejudicial value of the evidence makes it contrary to the interests of justice to admit it.

1. T.R.S. Allan, Similar Fact Evidence and Disposition; Law, Discretion and Admissibility, Modern Law Review, vol. 48 (1985) 253, at p. 255.

If the risk of an unfair trial is very high, the probative value of evidence disclosing criminal propensity may need to be so cogent that it makes the guilt of the accused a virtual certainty. In cases where the risk of an unfair trial is very small, however, the evidence may be admitted although it is merely probative of the accused's guilt. Each case turns on its own facts. But the judge must bear in mind that the admission of evidence revealing criminal propensity is exceptional. Further, as Lord Cross pointed out in Boardman [71] , while there remains a general rule against the admission of other acts of misconduct, "the courts ought to strive to give effect to it loyally and not, while paying lip service to it, in effect let in the inadmissible evidence".

Thus, where the prosecution case depends entirely on propensity reasoning [72] , the evidence will need to be very cogent to be admitted. When propensity reasoning is relied upon, the danger is high that the tribunal will convict simply because of the accused's propensity instead of using it as an evidentiary factor. Consequently, in such a case the evidence will need to be so cogent that, when related to the other evidence, there is no rational explanation of the prosecution case that is consistent with the innocence of the accused. However, I do not think that evidence disclosing or tending to prove other criminal or wrongful conduct, and consequently the criminal or discreditable propensity of the accused, must always meet this high standard. In the relationship cases, for example, propensity reasoning may simply reinforce or explain other evidence that directly implicates the accused. In such cases, it would be contrary to both the practice of the criminal courts and the interests of justice to use the no rational explanation test as the condition of admissibility of such evidence. In other cases, particularly those where the evidence is admissible for a reason other than the accused's propensity, the risk of prejudice may be so small that justice both to the accused and to the prosecution can be done by admitting evidence that is probative of guilt and warning the jury that they must not use the evidence in the way that is likely to create prejudice.

It follows that in each case where evidence is tendered that discloses, directly or indirectly, the criminal propensities of the accused, it is necessary to identify the nature of the risk, if any, to which the admission of the evidence gives rise. In similar fact cases, for example, evidence is often admitted for the reason that the association of the accused with so many similar deaths, injuries or losses, as the case may be, makes it highly improbable that there is any innocent explanation for the accused's involvement in the matter [73] . In these cases, the propensity of the accused will usually only be established by the verdict. The risk of prejudice is much less than in a case like Thompson or Straffen .

1. Makin [1894] AC. 57 (the baby-farming case) and R. v Smith (1915), 11 Cr. App. R. 229 (the brides in the bath case) are examples.

Makin is the classic example. It was a case involving objective improbability reasoning, not propensity reasoning [74] . The propensity of the accused to kill the babies was only established by the conclusion that it was probable to the point of certainty that so many babies including the baby the subject of the indictment could not have died by accident. Accordingly, they must have been murdered by the Makins. It was the verdict that established the accused's propensity. I am unable to agree with the statement of Dawson J. in Harriman that Makin was "a case in which the probative value of the other instances of baby-farming activities on the part of the two accused lay in establishing a disposition on their part to engage in such activities which, together with the other evidence in the case, pointed with great force to their having murdered the one child with whose murder they were charged" [75] . On this approach, the jury would find the disposition from the uncharged deaths and then use that disposition to conclude that the accused had murdered the child the subject of the charge. That is certainly not the way that the courts have hitherto dealt with cases giving rise to objective improbability reasoning. Indeed where there is only one similar instance beside the charged instance relied upon, it would be a very strange reasoning process. The risk of prejudice in true similar fact cases is not from propensity reasoning but from the fact, as Murphy J. pointed out in Perry [76] , that "[c]ommon assumptions about improbability of sequences are often wrong." A jury may wrongly give the similar fact evidence far more weight than it deserves.

1. See Piragoff, op. cit., p. 109. 2. (1989) 167 C.L.R., at p. 600. 3. (1982) 150 C.L.R., at p. 594.

In other cases involving similar facts, however, the accused may have admitted the facts of a similar incident or the facts of a similar incident may clearly point to wrongdoing on the part of the accused in relation to that incident. In such cases, there is a risk that, instead of relying on probability reasoning, the jury will simply rely on the propensity of the accused as revealed by the incident that is admitted or proved. It follows that the nature of the prejudice and the degree of risk of an unfair trial will always depend on the facts of each case.

It also follows that I am unable to agree with those statements in this Court [1] that suggest that evidence that discloses the criminal propensity of the accused cannot be admitted unless that evidence together with the other evidence denies any rational explanation of the accused's conduct that is consistent with his or her innocence. That rule will be generally applicable when the Crown is relying on the accused's criminal propensity because the risk of prejudice from propensity reasoning is so high. But in the relationship cases, for example, where evidence of propensity is relied on as confirmatory or explanatory of evidence implicating the accused, I do not think that such a high standard is either required or appropriate. Similarly, in cases where the accused's propensity is disclosed, but is not the basis of any reasoning process, a standard of proof lower than the no rational explanation standard may suffice for admission.

1. Sutton (1984), 152 C.L.R., at p. 564; Hoch (1988), 165 C.L.R., at p. 296; Harriman (1989), 167 C.L.R., at pp. 602, 607, 613.

Cox J., the learned trial judge, made a number of specific findings after hearing over 100 witnesses for the prosecution. In making those findings, his Honour said that he thought all witnesses had given their evidence honestly. Conflicts in the evidence were the result of the fallibility of human recollection. His Honour found that:

Michael Black was at Sturt Reserve by 1.30 p.m. He did not ride his bicycle to Thiele Reserve after that time.

In all likelihood, the man that Michael Black was seen talking to at Sturt Reserve at about 2.25 p.m. was the appellant.

A witness, Mr. Bernardi, saw Michael Black's bicycle leaning against the railing near the toilet block at Thiele Reserve before 3.26 p.m. The bicycle was never lying on the bitumen nor was it anywhere else in Thiele Reserve, as other witnesses had deposed.

Drowning, as an explanation of Michael Black's disappearance, was not a reasonable possibility.

Abduction was conceded to be the only rational alternative to drowning and any abduction was very likely to have been for a sexual purpose.

It was very likely that Michael Black left Sturt Reserve in a vehicle driven by his abductor.

Cox J. said that the evidence showed that the appellant had a disposition for young boys and was willing to pursue it by the extreme means of kidnapping and rape. His Honour said that it was erroneous to suppose that it was "always necessary to point to some additional characteristic or application of discreditable propensity evidence to demonstrate its relevance to the crime charged". Hoch made it plain, his Honour said, that more than "appreciable probative value" was needed to admit the evidence. The learned judge said that in this case there was "more". He then referred to various matters.

It is not clear to me whether his Honour was making findings about the matters to which he then referred because he began his discussion of them by saying that "[t]here is evidence from which the jury could find". But in the context of his judgment, I think that it should be inferred that his Honour in referring to those matters was making findings of fact. Matters to which his Honour referred included the following:

On the day preceding Michael Black's disappearance, the appellant had been informed that Thiele Reserve was a nude bathing spot and had been given directions as to how to get there.

On the same day at Sturt Reserve he had spoken to children who were strangers to him. He had asked them to tell him if they could think of any place where he could go for entertainment. He also invited them to go with him in his van and show him those places.

The appellant spoke to Michael Black on the day that he disappeared, patted his dog and lent him a knife to clean a fish. He later told his daughter that he had seen Michael Black twice and a hand gesture indicated that he had seen him on one of these occasions at Thiele Reserve.

There was no clear evidence of Michael Black having been seen after 2.25 p.m. A van similar to the accused's van was driven away from the direction of Sturt Reserve "in a hurry about 2.45".

Between 3 p.m. and 3.30 p.m., Mrs. Gould heard the sound of a vehicle arriving at the northern, i.e. toilet block, end of Thiele Reserve and leaving soon afterwards. At the same time she heard the agitated barking of a dog. Her description of the vehicle's sound fitted the accused's van "though plainly it could have fitted a large number of other vehicles as well".

Early in his ruling, Cox J. had said that the test of admissibility for propensity evidence was "exceptionally high" and that this Court had said the test was "whether there is a rational view of the evidence that is inconsistent with the guilt of the accused" [2] . His Honour said that that test had to be applied not only to the H. evidence but also "to the primary issue whether Michael Black was, indeed, abducted by anyone". The learned judge held that the H. evidence was admissible, saying:

In my opinion, the evidence of the accused's presence and behaviour in the area on January 17 and 18 including his dealings with Michael Black on the afternoon of January 18, the opportunity his van gave him to carry out an abduction, the vehicle and dog noises heard coming from Thiele Reserve, and the very unusual nature of the particular crime in question combine to render the evidence of the [H.] abduction admissible in proof of the identity of Michael Black's abductor. The nature of the crime in this case is of great importance. The more unusual the type of crime, the more difficult it may be to accept mere coincidence as a reasonable explanation. Less additional evidence may be needed in such a case to prove a connexion between the propensity evidence and the alleged instant crime and so establish overall the requisite high degree of proof We are not concerned here with a commonplace crime, such as housebreaking or rape, but with the abduction and detention, at least overnight, for sexual purposes of a young boy of pre-puberal or barely puberal age - quite an unusual crime in this community. 1. Citing Hoch (1988), 165 C.L.R., at p. 296.

His Honour went on to say that it would be an affront to common-sense to postulate two persons in Michael Black's vicinity at Murray Bridge at about the same time that afternoon, almost certainly at Sturt Reserve, and each of them having a propensity to kidnap and sexually assault young boys and each having the physical means of doing so. His Honour concluded his ruling by saying:

Obviously the prejudicial effect of admitting the [H.] evidence would be very considerable. It is only the cogency of the evidence on the issue of the identity of Michael Black's abductor that could justify the Court's receiving it. I consider that its high probative value transcends its prejudicial effect or, to put it another way, that its probative force is so strong that in all the circumstances it would not be unfair to admit it. I would not exclude it on discretionary grounds.

This last passage indicates that his Honour thought that he had a discretion to exclude the evidence. But it is not clear whether the last sentence in that passage was a new and independent conclusion or whether it was a conclusion drawn from the preceding sentences. If it was a conclusion drawn from the preceding sentences, it was erroneous. If it was such a conclusion, the learned trial judge has regarded the prejudicial effect of the evidence as a matter that goes to the exercise of the discretion to reject evidence that is admissible and not to the condition of its admissibility. I have already referred to the recent cases in this Court that hold that prejudice goes to admissibility. In the context of the present ruling, the importance of the distinction lies in the onus of proof.

On reading his Honour's ruling as a whole, I incline to the view that he did see prejudice as going to the issue of discretion and that would mean that his Honour has put the onus on the appellant to justify the exclusion of relevant evidence [3] . Nevertheless, I do not think that that error, if there be one, affects the admissibility of the evidence. That is because his Honour applied the no rational explanation test as the standard for admitting the evidence. And, as I have indicated, that makes irrelevant the question of weighing probative value against prejudice.

1. Although his Honour did not mention onus, it is impossible to believe that so experienced a judge would not have applied the settled rule that the onus is on the accused to show that admissible evidence should be excluded on discretionary grounds.

There is, however, another reason why in my respectful opinion his Honour's approach must be regarded as erroneous. His Honour took the view that the H. evidence could only be used against the appellant if there were anterior findings that Michael Black had been abducted for sexual purposes and then murdered. His Honour directed the jury that they had to be satisfied beyond reasonable doubt about these matters before they could use the H. evidence to identify the appellant as the abductor. Consequently, a finding that Michael Black was abducted for sexual purposes was necessary if the appellant was to be identified with the crime and that finding had to be established beyond reasonable doubt to his Honour's satisfaction before the H. evidence could be admitted. But with great respect to his Honour, I am unable to accept that, independently of the H. evidence, the other evidence established beyond reasonable doubt that Michael Black was abducted for sexual purposes.

Although one may speculate that Michael Black must have been abducted for sexual purposes, I do not think that that conclusion rises above surmise or speculation. Without the H. evidence, there is no evidence upon which it could be found beyond a reasonable doubt that Michael Black was abducted for sexual purposes. If the appellant had been indicted for sexually assaulting as well as murdering Michael Black and the evidence showed that H. had been abducted but not sexually assaulted, I do not see how a jury could be satisfied beyond reasonable doubt that Michael Black had been sexually assaulted. Admittedly, abducting for sexual purposes and sexual assault are two different matters. But in the context of this case it is a distinction without a difference.

In my view, however, the learned trial judge took too narrow a view of the use that could be made of the H. evidence. I do not think that its use was limited to identifying the accused as the abductor, if there was one. The propensity evidence was also relevant to the question of whether there had been an abduction. The learned trial judge's approach was reminiscent of the approach of the House of Lords in Thompson v The King and that of Slade J. in Straffen . That approach was understandable when the admissibility of evidence disclosing criminal propensity was thought to depend upon categories of exceptions. But the great advance that Boardman made was to reject the category approach. In my opinion, the H. evidence, if it was admissible, and I think that it was, was admissible to prove the abduction, the purpose of the abduction and the murder as well as the fact that the appellant was the killer.

The evidence of H. and the statement that the appellant made to his then wife were admissible because combined with the other evidence they point irresistibly to the conclusion that the appellant abducted Michael Black and murdered him. There is no rational explanation of the evidence that is consistent with his innocence. Apart from the hypothesis that the appellant abducted and murdered Michael Black, the only arguable hypotheses that are open are that Michael Black drowned or that some other person abducted him and murdered him. When the evidence of H. and the appellant's then wife is taken into account, neither hypothesis is a rational possibility.

Admittedly, all the circumstances are not known. One of the dangers of circumstantial evidence is that the logical coherence of a theory that fits the known circumstances pulls the mind to conclusions that would never be reached if more circumstances were known. The natural tendency of a tribunal of fact in a circumstantial evidence case is to dismiss as fanciful or speculative alternative theories that require the existence of circumstances in addition to those that are known. In res ipsa loquitur cases, for example, defendants rarely persuade judges or juries that they are not negligent when the cause of the accident is not established. In criminal trials, the tendency to dismiss exculpatory theories as fanciful or speculative is often likely to be present because there would be no case to go to the jury unless there was a theory of the known circumstances that pointed to the accused. For that reason, many counsel, experienced in the conduct of criminal trials, would much prefer to defend a case based on eye-witness evidence than one based on circumstantial evidence. Nevertheless, while I am conscious of these dangers, I think that, once the H. evidence is admitted, there is no rational hypothesis on the evidence that is consistent with the appellant's innocence.

In my opinion, the known circumstances point to and are consistent with the hypothesis that the appellant invited Michael Black into his van at Sturt Reserve and either at that reserve or after leaving that place abducted him for sexual purposes, planted his bike and other belongings at Thiele Reserve to distract attention from the appellant's association with him at Sturt Reserve, and murdered him. That hypothesis is established beyond reasonable doubt because the only alternative hypotheses cannot be rationally maintained.

Michael Black is undoubtedly dead. No sighting of him has occurred since he was seen at Sturt Reserve on the afternoon of 18 January 1989. He was either murdered or he drowned. It is irrational to suppose that he could have met his death in some other way.

About 14 January 1989, the appellant left home by himself in his white Kombi van, a vehicle equipped with sleeping and cooking facilities and a television set. He returned home on 20 January. The ostensible purpose of his trip was the study of physics. He was at Sturt Reserve on 17 and 18 January and camped overnight there. He was present at the reserve on the afternoon that Michael Black disappeared. On his own admission, he spoke to Michael Black at 1.30 p.m. That is within an hour of the last reliable sighting of Michael Black. Whether or not the appellant abducted Michael Black, he must in fact have spoken to him later than 1.30 p.m. The evidence suggests that Michael did not arrive at the reserve much before 1.30 p.m., and that he did some fishing. According to the appellant [4] , he lent Michael a knife to kill a fish that Michael had caught. The appellant was also seen to pat Michael's dog. He admitted that Michael Black came to his van and watched television for a short period when he returned the knife. He admitted that he had in his van tape of the same kind that he had used to tie up H. at Port Noarlunga.

1. The appellant's daughter testified that the appellant had told her that he gave Michael Black a knife to clean a fish.

Cox J. said that there was no reliable evidence that Michael Black was seen at Sturt Reserve after about 2.25 p.m. that day [5] . His Honour thought that it was likely that the appellant was the person talking to Michael Black at that time. But the evidence does not fit the appellant's description. I would not accept that it was the appellant. Apart from this finding, however, I can see no basis for rejecting his Honour's conclusion concerning the primary facts of the case. He had the advantage of seeing and hearing the witnesses.

1. There was evidence from one witness that Michael Black was at Sturt Reserve "about 3 o'clock". But Cox J. obviously thought that this evidence was unreliable.

A van, similar to the appellant's van, was seen being driven away from the direction of Sturt Reserve in a hurry at about 2.45 p.m. The appellant admits that he had left Sturt Reserve by that time. He claimed, however, that he left the reserve between 2.15 p.m. and 2.30 p.m. Nevertheless, given the scope for error in accurately recalling the times of incidents, the close relationship between Michael Black's last sighting at Sturt Reserve, the sighting of a van similar to the appellant's being driven in a hurry at about 2.45 p.m. and the appellant's leaving that reserve at about 2.30 p.m. has probative value.

The evidence of H. and the appellant's wife established that the appellant had a propensity to abduct young boys from public places for the purpose of imprisoning them and sexually assaulting them. Apart from the present case, that propensity has manifested itself on one occasion only. But the statement to his wife established that the H. incident was not a one-off, spur of the moment aberration. It is impossible to believe the appellant's evidence that he had no intention of abducting H. when he induced him to enter the van and that the abduction was a spontaneous act. His credibility is such that his claim cannot stand with the evidence of H. and his wife. Her evidence established that the propensity of the appellant existed as at January 1989. The modus operandi that the appellant used when he abducted H. was to lure him into his Kombi van by falsely telling him that he wanted help to retrieve a set of keys, threaten him and then bind, gag and blindfold him. In addition, when the appellant abducted H., he threatened to hurt him if he tried anything. Later at his home, he threatened to kill or hurt H. unless H. took off his clothes. There is no reason to suppose that on other occasions he would not use a similar modus operandi if he gave effect to his propensity.

On the day prior to Michael Black's disappearance, the appellant spoke to two children whom he did not know and invited them to go with him in his van and show him places of interest. By itself, the invitation excites the suspicion that the motive for the invitation was sexual. The evidence of H. and the appellant's wife also makes it probable that at back of the invitation was a desire by a "lonely" [6] man to have some form of intimate, almost certainly sexual, contact with the children. It provides a solid ground for thinking that his association with Michael Black on the following day was also driven by a desire for the same form of contact.

1. The appellant told his wife that he committed the offence against H. because he was lonely.

The probabilities tell heavily against Michael Black going to Thiele Reserve for his own purposes. He did not like the place. He had been forbidden to cross the river, which he would have had to do to get to Thiele Reserve. He did not swim at Sturt Reserve, which was one of his favourite swimming places. He had no towel, nor, so far as is known, did he have a swimming costume. Why then would he leave Sturt Reserve and go to Thiele Reserve to swim? No one saw him there although two boys who were there knew him. If he drowned, he must have done so within a short period of arriving; otherwise they must have seen him.

The search evidence makes it highly improbable that Michael Black drowned. If he had drowned, the chance that his body would not have been recovered was very small. In addition, nothing in the evidence suggests that there was any particular danger associated with swimming in the area of the river near where his clothes were found, and he was a competent swimmer. If he went to Thiele Reserve for some purpose of his own, it is most unlikely that he drowned.

If Michael Black did not voluntarily go to Thiele Reserve, then the only rational conclusion is that his clothes, bike, rod and dog were taken there by an abductor. The conclusion that he was abducted that day by the appellant and did not drown is strongly confirmed by the evidence of Mrs. Gould and the evidence of Mr. Bernardi who saw the bike there some time before 3.26 p.m.

At about 3.15 p.m. Mrs. Gould heard a vehicle arrive at the end of the reserve where the bike and fishing rod were later found. The vehicle stayed for only a short period and its stay was marked by the distressed barking of a dog which continued after the vehicle had departed. That dog was beyond doubt Michael Black's dog. Mrs. Gould described the engine of the vehicle as having a "very rowdy sort of chugging engine". The vehicle took off quickly. Fourteen months later she took part in a test at Thiele Reserve to see whether she could identify the accused's vehicle by its noisy engine. Eight vehicles - five sedans, two Kombi vans and a panel van - took part in the test. The appellant's van was used twice in the test. It was the fourth and ninth vehicle tested. On both occasions she said that the noise from the appellant's van was "very similar indeed" to what she heard. She said that another vehicle, probably the panel van, bore a resemblance to what she heard. But she thought that the fourth vehicle was the most likely vehicle.

On the day before Michael Black's disappearance, the appellant had been informed that Thiele Reserve was a place where he could swim in the nude. He was given directions as to how to get there. A week later he took his wife and family through Thiele Reserve after a picnic at Sturt Reserve. A conversation with his wife indicated that he had been at Thiele Reserve before and liked it. This evidence makes it probable that he went there to swim on the day before Michael Black disappeared.

The probability that the appellant had already been to Thiele Reserve, the fact that his vehicle had admittedly left Sturt Reserve and was being driven in the Murray Bridge area at 3.15 p.m. and the evidence of Mrs. Gould make it likely that it was the appellant's vehicle that she heard at Thiele Reserve. Moreover, it appears from the summing up of Cox J. that another witness, Mr. Neindorf, saw a white Kombi van in the car park area of Thiele Reserve, and no one recollects any other vehicle entering or leaving the car park before 4.30 p.m. that might have been used to abduct Michael Black. Independently of the propensity evidence, there is a significant probability that the appellant drove to Thiele Reserve about 3.15 p.m., stayed shortly, and left hurriedly in circumstances that caused the distressed barking of Michael Black's dog.

Moreover, if the appellant killed Michael Black, he had a compelling motive to plant a false trail at Thiele Reserve to suggest that Michael Black had drowned there. Killers do not usually plant false trails unless they feel it is necessary to do so to distract attention from their association with or opportunity to kill their victims. If Michael Black did not drown, it is likely that his killer had some motive for planting a trail suggesting that he did. Of course, the motive may simply have been to suggest drowning so as to make it unlikely that there would be a hunt for the abductor. But the disappearance from Sturt Reserve would itself have raised the drowning theory. In addition, going to Thiele Reserve to plant a false trail with the bike, dog, clothes and fishing rod posed risks for the abductor. The evidence strongly suggests that Michael Black's dog was tied up at Thiele Reserve before it broke free. Both the place where it was tied up and the place where Michael's clothes were found meant that the abductor had to walk some distance from his or her vehicle. The motive that the appellant had for planting a false trail at Thiele Reserve was that too many people had or might have seen his association with Michael Black at Sturt Reserve. If he was the abductor, he had a strong motive to fabricate a persuasive case of drowning at a place away from Sturt Reserve, and Thiele Reserve was a swimming place to which he had been directed and where he very probably had been.

When all the circumstances including the appellant's propensity are considered, I think that the only rational conclusion is that the appellant abducted Michael Black for sexual purposes and later murdered him. Sexually assaulting young boys is regrettably not unknown in South Australia or elsewhere in this country. But luring boys into vehicles, tying them up, sexually assaulting and keeping them imprisoned is, as Cox J. said, an unusual crime in South Australia or elsewhere in this country. Unless Michael Black drowned, he must have been lured into a vehicle at Sturt Reserve together with his dog, bike and fishing gear, abducted, driven to Thiele Reserve and later murdered. The appellant had the propensity to abduct Michael Black. He had asked two children to travel with him in the van and the terms of his invitation to them suggest that the reasons he gave for making it were false. Given his propensity, it is likely that he was seeking sexual gratification from those children and it is a short step to conclude that that propensity was excited by his contact and association with a young boy like Michael Black. Furthermore, the appellant was with Michael Black shortly before his disappearance. He was not at Sturt Reserve after Michael Black disappeared.

The propensity evidence, the appellant's behaviour at Sturt Reserve, the proximity between the time that he left Sturt Reserve and the time that Michael Black disappeared together with Mrs. Gould's evidence establish a convincing case beyond a reasonable doubt that the appellant abducted Michael Black and planted a false trail at Thiele Reserve. If he did that, the conclusion that he also murdered Michael Black is irresistible. I do not think that it is of any importance that he did not murder H. Assuming that he had no intention to murder H., once the conclusion is reached that he abducted Michael Black, no rational conclusion is open other than that he also murdered him.

This is not a case where proof of the appellant's propensity did no more than show a mere propensity to commit crimes of a sexual nature. If the evidence had established that Michael Black had gone to Thiele Reserve to swim, the case would have been different. There would have been no false trail, and Mrs. Gould's evidence would not have had the same force. But the evidence points to an abduction at Sturt Reserve at a time when the appellant was present, the planting of a false trail at Thiele Reserve and the appellant's presence at Thiele Reserve where Michael Black's belongings were found. In addition, the appellant admits that at Sturt Reserve he spoke to Michael Black at a time shortly before he disappeared. There was, therefore, such a nexus between the other facts in the present case and the manner in which the accused's propensity manifested itself in the H. incident that evidence of his propensity had significant probative value. It was so significant that together with the other evidence there was no rational explanation consistent with the innocence of the appellant.

Once the propensity evidence is used alternative hypotheses and doubts that one might have about particular segments of the evidence are eliminated. Circumstantial evidence "works by cumulatively, in geometrical progression, eliminating other possibilities" [7] . Ultimately, the propensity and other evidence in this case combine to cut out all hypotheses other than the hypothesis that the appellant abducted and murdered Michael Black.

1. Reg. v Kilbourne [1973] AC. 729, at p. 758, per Lord Simon of Glaisdale.

The combination of that evidence eliminates drowning as a rational possibility. It also eliminates abduction by someone other than the appellant. The circumstances that point to the appellant leave no room for a rational hypothesis that a person other than the appellant was the abductor and murderer. It would be a remarkable coincidence, having regard to the facts that are known and can be reasonably inferred if, in addition to the appellant, there was present at Murray Bridge that day another person who abducted Michael Black and who knew, and had a reason to plant a false trail at, Thiele Reserve. It would be remarkable because that person, like the appellant, would have had a propensity to abduct a young boy, would have used a vehicle with a very rowdy chugging engine with an uneven beat like the appellant's, would have abducted his victim about the time that the appellant left Sturt Reserve and would have selected as his victim a boy that the appellant had associated with not long before. The hypothesis that another person abducted Michael Black is fanciful compared to the hypothesis that the appellant was the abductor.

As I have indicated, the learned trial judge directed the jury to examine the evidence in a framework different from that which I think should have been used. In most cases, that would mean that a new trial would have to be ordered. But in my opinion that order is not required in this case. The directions of his Honour erred in favour of the appellant. Upon the evidence which the jury must have accepted to convict the appellant, no reasonable jury could have failed to convict him if they had been directed in accordance with the principles and the legal framework that I think were the appropriate ones. In those circumstances, the directions of his Honour did not constitute any miscarriage of justice.