Sutton v The Queen

(1) Subject to the provisions of this Act, charges for more than one felony or for more than one misdemeanour, and charges for both felonies and misdemeanours, may be joined in the same information, if those charges are founded on the same facts, or form or are a part of a series of offences of the same or a similar character.

(2) Where before trial, or at any stage of a trial, the court is of opinion that a person accused may be prejudiced or embarrassed in his defence by reason of being charged with more than one offence in the same information, or that, for any other reason, it is desirable to direct that the person should be tried separately for any one or more offences charged in an information, the court may order a separate trial of any count or counts of such information.

(3) This section shall not affect any other provision of this Act or any other enactment permitting more than one charge to be joined in the same information.

Sub-section (1) introduced in South Australia an important rule of criminal pleading by the Crown, the operation of which largely depends upon the facts of the case (Ryan v. The Queen [25] ). The equivalent provision in the Indictments Act was considered in Ludlow v. Metropolitan Police Commissioner [26] where Lord Pearson pointed out that both the law and the facts "should be taken into account in deciding whether offences are similar or dissimilar in character", and I respectfully agree. If the offences are similar in character, they may constitute a series. "Series" does not import a clear criterion for determining what charges may be joined in the same information. Perhaps little more can be said about its meaning than Dixon J. said in Packett v. The King [27] , namely, that "it connotes some connection between the crimes". 1. (1982) 149 C.L.R. 1, at pp. 21-22. 2. [1971] A.C. 29, at p. 39. 3. (1937) 58 C.L.R. 190, at p. 207.

The facts to be taken into account in ascertaining whether the charges joined in the information satisfy the conditions stated in sub-s. (1) are the facts alleged by the Crown. They are the facts on which the indictment is framed and the only facts available to a trial judge upon which he can rule whether the joinder of charges is authorized by sub-s. (1). If the joinder of charges in an information is challenged, the judge will refer to the depositions taken on the committal proceedings and proofs of other evidence which the Crown proposes to tender at the trial in order to determine whether there is such a connection between the offences charged that they may rightly be regarded as a series. It is immaterial that those facts are disputed by the defence. The discretion conferred by sub-s. (2) to order separate trials "at any stage of a trial" provides protection for the accused in a case where the evidence adduced falls significantly short of the Crown allegations which warranted the joinder of the charges pursuant to sub-s. (1).

The eight counts in the present case were properly joined in the one information. The Crown case, as appears from the evidence at the trial, included evidence implicating the applicant in each set of alleged offences. Luba said that she had seen her assailant twice in the street after the evening when she was attacked and before the police told her that "they caught the guy". When she saw the applicant subsequently in court, she identified him. Veronica could not identify her assailant, but a woman with whom the applicant had been living at about the time when Veronica was raped said that the applicant had admitted to her that he had done "that rape at Woodville". Dianne had identified the applicant as her assailant shortly after the attack upon her when she was being driven in a police car past the Alberton Hotel. If those facts were accepted, the applicant had committed three sets of sexual offences against three young girls in a particular area of suburban Adelaide, in each case by the use of force and without the consent of his victim. There was a sufficient connexion among those alleged offences to treat them as a series and to charge them in the same information, though the trial judge had a discretion pursuant to sub-s. (2) to direct separate trials upon the counts or the sets of counts in that information.

When two or more counts constituting a series of offences of a similar character are joined in the same information, a real risk of prejudice to an accused person may arise from the adverse effect which evidence of his implication in one of the offences charged in the indictment is likely to have upon the jury's mind in deciding whether he is guilty of another of those offences. Where that evidence is not admissible towards proof of his guilt of the other offence, some step must be taken to protect the accused person against the risk of impermissible prejudice. Sometimes a direction to the jury is sufficient to guard against such a risk; sometimes it is not. Where a direction to the jury is not sufficient to guard against such a risk, an application for separate trials should generally be granted. There may be an exceptional case where countervailing factors appear but I do not presently foresee them. As Lord Cross of Chelsea observed in Director of Public Prosecutions v. Boardman [28] , so long as the general rule excluding similar fact evidence is maintained "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 by trying all the charges together". It would be misleading for a judge to require some further "special feature of the case" - the criterion adopted by Lord Pearson in Ludlow v. Metropolitan Police Commissioner [29] - before giving a direction for separate trials. The purpose of provisions such as s. 278 is to avoid the technicalities and rigid rules of criminal pleading and procedure, but not to impair the administration of criminal justice. Irvine C.J. pointed out in R. v. Brent [30] with respect to the corresponding Victorian provision, that it did not "intend to introduce any fundamental alteration in the general rule that the tribunal to determine the guilt or innocence of a person charged should not be affected by the evidence given in relation to another charge - certainly not the evidence given by another person in relation to another charge". The price of dispensing with the technicalities and rigid rules of criminal pleading and procedure is the imposition upon the trial judge of the onerous function of directing the course of proceedings to ensure that justice is properly administered. To that end he is given a discretion to order separate trials. Where the evidence admissible on one count is not admissible on another, and in consequence there is a real risk of impermissible prejudice to the accused, the sound exercise of the discretion generally (if not universally) requires a direction for separate trials (cf. per Lord Hailsham of St. Marylebone in Boardman [31] ; Novac [32] ).

1. [1975] A.C. 421, at p. 459. 2. [1971] A.C., at p. 41. 3. [1919] V.L.R. 46, at pp. 52-53. 4. [1975] A.C., at p. 447. 5. (1976) 65 Cr.App.R. 107, at p. 111.

Usually, there is no risk of impermissible prejudice arising from the conduct of a single trial upon an indictment containing more than one count where the evidence relating to the accused's implication in an offence charged in one count is admissible in proof of his guilt of the offence or offences charged in the other count or counts. The admissibility of the evidence relating to one count upon a trial for another count is thus a question of importance affecting both the exercise of the discretion under s. 278(2) and, in the event of a trial on multiple counts, on the direction given to the jury relating to the way in which they may use the evidence tendered in proof of each count at the trial.

In the present case, Mitchell J. declined to direct a separate trial for the offences charged in counts 2, 3 and 4, holding that the evidence implicating the accused in the commission of a set of offences against the one complainant was admissible in proof of his guilt of the offences against the other complainants. Her Honour said, inter alia:

He [i.e., counsel for the applicant] submitted however that the offences alleged in count 2 to 4 did not constitute part of a series and that, if they did, I should exercise my discretion to order a separate trial of those charges on the ground that there was a danger that the jury might use evidence which was not admissible in relation to the charges in count 1 or the charges in counts 5 to 8 in considering those charges and might wrongly convict the accused. Mr. Brebner [for the Crown] submitted that the evidence identifying the accused as the person who had committed the offences charged in counts 2 to 4 was admissible to prove his identity with the person in relation to whom the other charges had been brought and therefore there was no proper reason to order separate trials.

Evidence which tends to establish that an accused has committed the crime with which he is charged is admissible although it tends to show the commission of other crimes provided that it has "a strong degree of probative force" or "a really material bearing on the issues to be decided". See Makin v. Attorney-General (N.S.W.) [33] ; Markby v. The Queen [34] ; Reg. v. Perry [35] .

1. [1894] A.C. 57, at p. 65. 2. (1978) 140 C.L.R. 108, at p. 117. 3. (1981) 96 L.S.J.S. 205. Her Honour, considering the similarities between each of the sets of offences against the respective victims, thought that the evidence implicating the accused in one set of offences was admissible in evidence upon his trial for the offences charged in the other counts in the indictment. The principle which she applied is not open to doubt. It was recently reaffirmed by this Court in Perry v. The Queen [36] . In summing up to the jury, her Honour directed the jury's attention to the way in which the evidence admissible upon the counts charging one set of offences might be thought to be strongly probative of, or to have a material bearing on, the issues to be decided upon the other counts in the information. She said: Sometimes there may be such a striking similarity between two different acts that a jury may be satisfied beyond reasonable doubt that the person who committed one set of acts must have committed the other. That is to say that the accused has put a certain stamp upon the crime which makes it easily recognizable that he must have committed both sets of crimes. This could not be so if both sets are such that they may be explained by coincidence. There must be such a close similarity, such a clear underlying unity between both sets of acts as to make coincidence a very unlikely explanation for what happened.

And that is what the Crown says here. The Crown says here it is so unlikely that you can disregard it that two or more people committed these crimes. If you decide the Crown is right (but you must bear in mind that it is not sufficient if the evidence simply raises or deepens the suspicion that this accused is guilty of all offences) it must make any other conclusion than guilty an affront to your common sense.

In this case the Crown says that, provided you are satisfied beyond reasonable doubt that the accused committed the crimes alleged in respect of one complainant, then the circumstances in which the other crimes were alleged to have been committed were so similar as to lead inevitably to the conclusion that he must have committed the other offences.

1. [1894] A.C. 57, at p. 65. 2. (1978) 140 C.L.R. 108, at p. 117. 3. (1981) 96 L.S.J.S. 205. 4. (1982) 150 C.L.R. 580.

If her Honour was right in holding that the evidence implicating the applicant in the commission of one of the sets of offences charged was admissible in proof of the issues to be decided upon the other counts in the information, no valid objection could be taken either to her Honour's refusal to direct separate trials or to the direction which she gave the jury as to the way in which they might use that evidence. But if the evidence were not so admissible, it was erroneous to refuse to grant separate trials and it was a misdirection to tell the jury that they were entitled to take account of the evidence implicating the accused in one set of offences in considering the issues arising upon the other counts in the information.

The applicant appealed to the Court of Criminal Appeal against his convictions. That Court agreed with Mitchell J. that the evidence implicating the applicant in one set of offences was admissible in proof of the issues arising upon the other counts. For reasons presently to be stated, I agree in that conclusion and it follows that, in my opinion, the appeal must fail. Ordinarily I would not favour granting special leave to appeal if the appeal is doomed to fail and the trial judge has correctly applied established principles of law. However, in the Court of Criminal Appeal Wells J., with whom Sangster and Mohr JJ. agreed, seemed to find it necessary to confine the application of the principles expressed in Perry to the facts of that case. The Court of Criminal Appeal, while not departing from Perry in the actual resolution of the present case, discerned in it some novel and undesirable departure from established principle. Properly understood, Perry does not exhibit a departure from established principle. It is necessary to grant special leave in order to restate the general applicability of what is expressed in Perry lest the relevant principles be given too narrow an application in South Australia.

The first principle of general application is that in a criminal trial evidence of the commission of offences other than the offence charged is prima facie inadmissible against an accused person. The chief reason for the prima facie inadmissibility of evidence of an offence other than the offence charged is this: it is thought that the antipathy which evidence of another offence is apt to engender may unjustly erode the presumption of innocence which protects an accused person at his trial; i.e., the evidence of the other offence may be regarded by the jury as being more probative of guilt of the offence charged than it can fairly be thought to be. This reason was assigned by Lord Simon of Glaisdale in Director of Public Prosecutions v. Kilbourne [37] and by Lord Hailsham in Boardman [38] . Lord Simon said:

But some relevant evidence is nevertheless inadmissible. To cite a famous passage from the opinion of Lord Herschell L.C. in Makin v. Attorney-General (N.S.W.) [39] : "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 in 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. On the other hand, the mere fact that the evidence adduced tends to show the commission of other crimes does not render it inadmissible if it is 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." That what was declared to be inadmissible in the first sentence of this passage is nevertheless relevant (i.e., logically probative) can be seen from numerous studies of offences in which recidivists are matched against first offenders, and by considering that it has never been doubted that evidence of motive (which can be viewed as propensity to commit the particular offence charged, in contradistinction to propensity to commit offences generally of the type charged) is relevant. All relevant evidence is prima facie admissible. The reason why the type of evidence referred to by Lord Herschell L.C. in the first sentence of the passage is inadmissible is, not because it is irrelevant, but because its logically probative significance is considered to be grossly outweighed by its prejudice to the accused, so that a fair trial is endangered if it is admitted; the law therefore exceptionally excludes this relevant evidence: whereas in the circumstances referred to in the second sentence the logically probative significance of the evidence is markedly greater. (See also Lord Moulton in R. v. Christie [40] .) Wells J. understood that the speeches in Boardman were directed to "fostering a fair and sensible exercise of the Christie discretion", i.e., the discretion to exclude "evidence which, though admissible, would probably have a prejudicial influence on the minds of the jury which would be out of proportion to its true evidential value" (per Lord Moulton in R. v. Christie [41] ). On this approach, the starting point is the relevance of the similar fact evidence and thus its admissibility unless excluded in exercise of the Christie discretion. This approach focused upon what his Honour thought to be a "unifying" and "ancestral" principle stated in these terms: Fundamentally, there is only one legal principle of evidence, and it is this: every item of evidence that is relevant to the issue - that is, that tends, of itself or in conjunction with other evidence, to prove or disprove a fact in issue - should be received and, quantum valeat, used for the purpose of that proof or disproof, unless some positive rule of law, based on judicially formulated policy, or on statute, prohibits that use. His Honour thought that a trial judge's approach to questions of admissibility should be to admit relevant evidence unless there is some applicable rule of exclusion: Once a disputed item of evidence is held to be relevant, the judicial mind must turn to the question whether any rule exists, by virtue of which the disputed item should be excluded, notwithstanding its relevance. This echoes what was said by Lord Goddard L.C.J. in R. v. Sims [42] : 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. That passage was criticized by the Privy Council inNoor Mohamed v. The King [43] , their Lordships approving the different approach taken by Kennedy J. in R. v. Bond [44] : [I]f, as is plain, we have to recognize the existence of certain circumstances in which justice cannot be attained at the trial without a disclosure of prior offences, the utmost vigilance at least should be maintained in restricting the number of such cases, and in seeing that the general rule of the criminal law of England, which (to the credit, in my opinion, of English justice) excludes evidence of prior offences, is not broken or frittered away by the creation of novel and anomalous exceptions. The criticism of Sims was accepted in Boardman [45] . 1. [1973] A.C. 729, at pp. 756-757. 2. [1975] A.C., at p. 451. 3. [1894] A.C., at p. 65. 4. [1914] A.C. 545, at pp. 559, 560. 5. [1914] A.C., at p. 559. 6. [1946] 1 K.B. 531, at p. 539. 7. [1949] A.C. 182, at pp. 194-196. 8. [1906] 2 K.B. 389, at p. 398. 9. [1975] A.C., at pp. 440, 449, 456.

The true approach focuses not so much upon the relevance of the evidence as upon the inherent tendency of evidence of other criminal offences unfairly to prejudice an accused person. The fundamental importance of ensuring the fairness of the trial for a particular offence requires that provisional paramountcy be given to the latter consideration and thus "the admission of similar fact evidence is the exception rather than the rule" (per Gibbs A.C.J. in Markby v. The Queen [46] ). The approach favoured by Wells J. is not consistent with settled authority.

Apart from the prejudicial effect of evidence of other offences, there are further reasons why the admission of any kind of similar fact evidence is the exception rather than the rule. Those reasons were discussed in Duff v. The Queen [47] , citing what Street C.J. had said in Kitto v. Gilbert [48] :

The rule of law in such cases is stated accurately in the Laws of England (vol. 13, p. 450), and is that "facts similar to, but not part of, the same transaction as the main fact are not, in general, admissible to prove the occurrence of the main fact." The rule, I think, is sometimes rested upon the ground of remoteness or want of reasonable connection between the principal and evidential facts in such cases, sometimes upon the inconvenience that would result if a different rule prevailed, and sometimes upon the tendency that there would be to confuse juries by raising collateral issues; but, on whatever ground it may be rested, it is a well established rule. 1. (1979) 39 F.L.R. 315, at p. 348 et seq.; 28 A.L.R. 663, at p. 689 et seq. 2. (1926) 26 S.R. (N.S.W.) 441, at p. 447.

Before the trial judge is at liberty to admit similar fact evidence he must be satisfied that the probative force of the evidence clearly transcends its merely prejudicial effect (Perry [49] ). It is the probative force (or cogency) of the evidence in comparison with the impermissible prejudice that it may produce which determines admissibility - "a question of law, not of discretion" as Lord Cross said in Boardman [50] , thought it is a question of degree. It is therefore wrong to approach the admissibility of similar fact evidence as though the primary principle is the discretion to which Lord Moulton referred in Christie . To treat the merely prejudicial effect of evidence of other criminal offences as no more than a factor to be taken into account in exercising a discretion would be to diminish the guarantee of a fair trial which the exclusionary rule is calculated to secure. The prima facie rule of exclusion is not open to doubt. In adopting this approach, Perry is clearly in the mainstream of settled authority [51] . Evidence of similar facts consisting in proof of the commission of offences other than the offence charged is therefore a particular category of circumstantial evidence which is subject to a special exclusionary rule. It is wrong to regard evidence of that kind as undifferentiated from circumstantial evidence generally and to hold, as Wells J. held, that there are no categories of circumstantial evidence.

1. (1982) 150 C.L.R., at pp. 604-605, 610. 2. [1975] A.C., at p. 457. 3. (1982) 150 C.L.R., at pp. 585, 592, 604-605, 609-610.

The second principle of general application is really reciprocal to the first. It is stated by Gibbs A.C.J. in Markby [52] :

To be admissible the evidence must have "a strong degree of probative force" (per Lord Wilberforce in Reg. v. Boardman [53] ), or "a really material bearing on the issues to be decided" (per Lord Morris of Borth-y-Gest [54] , citing Harris v. Director of Public Prosecutions [55] ); it may not be going too far to say that it will be admissible only if it is "so very relevant that to exclude it would be an affront to common sense" (see per Lord Cross in Reg. v. Boardman [56] ; and see per Lord Hailsham of St. Marylebone [57] ). The cited passages admit of different interpretations. For my part, I prefer the formulation by Lord Cross. That formulation does not suggest that admissibility depends on the significance in the eyes of the jury of the similar fact evidence in proving the guilt of the accused. Admissibility depends on cogency to prove another fact, not upon the likelihood that the jury will convict the accused person if they accept the similar fact evidence: cf. Vaitos [58] . The cogency of similar fact evidence is assessed by reference to its connexion with a fact in issue or to a fact in the chain of proof of a fact in issue. The requirement of cogency to prove such a fact is not satisfied by the capacity of the evidence to engender mere prejudice; it must go clearly beyond that to be admissible. Provided the similar fact evidence exhibits a sufficient connexion with the fact to be proved - a question of degree to be assessed by the trial judge - and the fact to be proved is or is relevant to a fact in issue in the trial, the similar fact evidence is admissible. The phrases collected by Gibbs A.C.J. in the passage cited from Markby are indicative of the degree of the connexion which is sufficient. The Full Court of the Supreme Court of Victoria in Reg. v. Chee [59] appeared to give the Markby criterion a limited application, but a differently constituted Full Court in Vaitos has accepted its general applicability. Wells J., though attracted by Chee , acknowledged the binding force of Markby. Perry accords, of course, with the Boardman and Markby criterion. 1. (1978) 140 C.L.R., at p. 117. 2. [1975] A.C., at p. 444. 3. [1975] A.C., at p. 439. 4. [1952] A.C. 694, at p. 710. 5. [1975] A.C., at p. 456. 6. [1975] A.C., at pp. 452-453. 7. (1981) 4 A.Crim.R. 238, at p. 274. 8. [1980] V.R. 303.

The third principle of general application is that the cogency of evidence of the commission of other offences is to be ascertained by reference to the whole body of proof in the case viewed in the light of experience. To determine the cogency of such evidence, it is necessary to identify and to keep steadily in mind the particular fact which the evidence is tendered to prove and how it tends to prove it. Evidence which would rebut a suggestion of accident once an accused is identified may not go far towards establishing his identity as the offender. Although questions of admissibility have usually arisen in cases where the Crown seeks to tender evidence of the commission of offences similar to the offence charged, the cogency of such evidence depends upon the connexion between the facts which that evidence proves directly and the fact which is said to be provable inferentially from them. The connexion between the offences which that evidence proves and the offence charged is not the precise question for consideration.

Striking similarity between the alleged similar fact and the fact to be proved inferentially is a frequent, but not the only, indication of that cogency or probative force which is required if the evidence is to be taken out of the primary exclusionary rule: see Perry [60] . In Perry the fact in issue to be proved inferentially was whether Mrs. Perry had administered to Mr. Perry the arsenic which he had undoubtedly ingested. On appeal to this Court it was held that parts of the allegedly similar fact evidence which the Crown had tendered were not probative or sufficiently probative of that fact to escape the prima facie rule of exclusion. The cogency of the evidence to which objection was taken was evaluated in the context of the other facts established by the whole of the evidence.

1. (1982) 150 C.L.R., at pp. 586-587, 593, 604-605, 609-610.

The third principle of general application occasioned no difficulty to the Court of Criminal Appeal in the present case but that Court found some difficulty in reconciling that principle with the result in Perry. Their Honours were at pains to show that the cogency of similar fact evidence can be derived from the context of the rest of the evidence in the case. Wells J. wrote:

Where a charge has been satisfactorily made out, it will be found that every fact relates, mediately or immediately, to every other fact so as to create one comprehensive and coherent whole, in which the constituent parts of the evidence are mutually interdependent, and is such that a hypothesis based on any one selected group of facts finds its verification in all the rest. From a purely formal point of view, the underlying reasoning that supports the conclusion is perpetually and convincingly circular; in this respect, it differs fundamentally from syllogistic reasoning which may be thought of, metaphorically, as a steady advance in one direction (emphasis added). The thrust of his Honour's observations can be accepted. It is often impossible to perceive the similarity between a fact proved directly in evidence and a fact to be proved inferentially or to determine the cogency of the former to prove the latter unless the fact proved directly is seen in the context of the other evidence in the case. Where the Crown seeks to prove or confirm a fact by inference drawn from similar facts, the fact to be proved or confirmed may be likened to a scene depicted in a mosaic and the pieces of evidence - direct and circumstantial - may be likened to the tiles of which the mosaic is composed. The pieces of similar fact evidence may, like the tiles, reveal their significance only in context with the rest of the evidence, appearing insignificant in isolation but clearly revealing the guilty inference when all the evidentiary tiles are locked together.

The difficulty which the Court of Criminal Appeal had in reconciling the third principle with the result in Perry stems, I suspect, from a failure to distinguish between the context of the facts proved directly by the evidence and a wider context which includes the guilt of the accused of the offence charged. Though it is right to determine the admissibility of a piece of similar fact evidence by reference to the whole of the body of proof, it is wrong to regard the body of proof as encompassing the guilt of the accused - the very issue which is to be proved by inference from the facts directly proved or provable in evidence. Gibbs C.J. said in Perry [61] :

In other words, it was necessary to assume the guilt of the applicant of the offences of which she was charged in order to render admissible the evidence regarding the death of Duncan. Such a line of reasoning is obviously objectionable. And I said [62] : To seek to prove a fact issue by a chain of reasoning which assumes the truth of that fact is, of course, a fallacy, repugnant alike to logic and to the practical processes of criminal courts. Although Wells J. accepted the latter proposition in terms, his Honour said: it is possible to pose many cases in which the primary facts proved include facts that may, of themselves, be colourless or amount to matter of bare suspicion only, and yet it would be manifestly opposed to principle to exclude evidence of them on the ground that (per Gibbs C.J.) it is "necessary to assume the guilt of the [accused] to render [the impugned evidence] admissible". Later, in a passage following the passage cited above [63] , his Honour said: Given, then, that a criminal offence C is to be proved, and that there have been received in aid of proof similar fact transactions P, Q, R, and S, there is no reason why, for example, P should not be used to confirm Q; P and Q (so confirmed) to confirm R; R (so confirmed) and P and Q to confirm S, and P, Q, R, and S, to assist in the proof of C, which, in turn (viewed in the totality of the case), confirms P, Q, R, and S. In such a case, striking similarities between P and Q may include one factor that is only weakly proved in Q but strongly proved in P. P may thus reinforce the factor in Q, and the two (including Q so reinforced) act in aid of proof in the balance of the case. To assert the validity of such a proposition is to do no more than apply to the facts the elements of inductive reasoning. 1. (1982) 150 C.L.R., at pp. 589-590. 2. (1982) 150 C.L.R., at p. 612. 3. Ante, p. 550.

If his Honour is merely describing the consistency which must exist between the body of proof and the jury's verdict "where a charge has been satisfactorily made out" (the phrase introducing this passage), this passage does not relate to admissibility; accordingly, it would not be to the point to repeat that the cogency of evidence (P, Q, R and S) tendered to assist in the proof of a further fact (C) cannot be affected by treating the further fact as though it were part of the body of proof. But if his Honour is saying that it is a legitimate means of inductive reasoning to find that the cogency of evidence to prove a further fact is enhanced by having regard "in the totality of the case" to the further fact, I cannot agree: the further fact is simply not part of the evidence in the case. When the inference of guilt is properly drawn it is, of course, supported by the primary facts established in the admitted evidence, but one cannot attribute cogency to evidence tendered to establish a primary fact by assuming that the fact has been proved. If proof of the fact is assumed in order to give cogency to similar fact evidence and thereby to make it admissible to prove the same fact, there is an impermissible circularity in reasoning of the kind which led to the rejection of Mr. Duncan's symptoms in Perry [64] . The Court of Criminal Appeal's difficulty in perceiving the circularity inherent in admitting that evidence in Perry need not detain us, for our present concern is not with the facts of that case but with the general application of the principles which the judgments express.

It is a canon of logic, rather than of law, that one cannot prove a fact by a chain of reasoning which assumes the truth of that fact. That canon has a particular application in determining the cogency and hence the admissibility of similar fact evidence. When the Crown seeks to tender similar fact evidence as the foundation for inferring a fact to be proved in a trial, it is erroneous to assume the truth of the fact to be proved in determining the cogency of the evidence. That is a proposition of general application; it is not limited to the facts of Perry or to cases bearing a close factual similarity to it. That proposition does not preclude reference to direct evidence of the fact to be proved in determining the cogency of similar fact evidence. Similar fact evidence tending to confirm the existence or occurrence of such a fact may be confirmed by direct evidence of the same fact.

It remains to apply these principles to the present case. The starting point is to identify the fact to be proved and how the allegedly similar facts may be thought to prove it. The fact which the allegedly similar facts were tendered to prove is that one offender committed the three sets of offences. If, but only if, that fact were proved, it would be permissible for the jury to regard proof of the applicant's guilt of one set of charges as proof of his guilt of the others.

Luba, Veronica and Dianne had each been grabbed from behind and threatened with death; they had each been abducted from the street and taken to the grounds of a school; they had each been subjected to oral intercourse or an attempt at oral intercourse against her will; Veronica and Dianne had each been subject to anal and vaginal intercourse against her will. The offences had occurred in the same suburban area (within two miles) and at intervals of five to six weeks between the sets of offences. Similar descriptions of an assailant had been given by each of the victims. If these facts support the inference that one offender committed the three sets of offences, it would be right for the jury to infer that the applicant, if guilty of one set of offences, was guilty of all. Mitchell J. thought that there were sufficient similarities among the attacks upon the three girls to support an inference that the same offender had committed them. Her Honour therefore left it to the jury to say whether, as a matter of fact, that inference should be drawn. She said:

Some of the common features are of little significance. For example, the wearing of jeans is common to a large section of members of the community, as is the smoking of cigarettes. There was however striking similarity in the method of abduction and the method of conducting the alleged victim to the places where the offences were said to have been committed; a similarity in the fact that in each case the offences were alleged to have been committed in the grounds of a school at night and finally the commission in the first [semble, second] and third sets of charges of acts of vaginal, oral and anal intercourse. I was of the opinion that the similarities between the circumstances of the incidents deposed to by the witnesses for the prosecution were so striking that it should be left to the jury to decide whether, upon identification to its satisfaction of the accused as the person who committed one set of rapes, it was satisfied that he must have committed the other set of rapes and must have been guilty of the attempted rape.

In reviewing her Honour's conclusion, I am guided by what Lord Wilberforce said in Boardman [65] :

If this test is to be applied fairly, much depends in the first place upon the experience and common sense of the judge. As was said by Lord Simon of Glaisdale in Reg. v. Kilbourne [66] , in judging whether one fact is probative of another, experience plays as large a place as logic. And in matters of experience it is for the judge to keep close to current mores. What is striking in one age is normal in another: the perversions of yesterday may be the routine or the fashion of tomorrow. The ultimate test has to be applied by the jury using similar qualities of experience and common sense after a fair presentation of the dangers either way of admission or of rejection. Finally, whether the judge has properly used and stated the ingredients of experience and common sense may be reviewed by the Court of Appeal. The great experience of Mitchell J. and of the members of the Court of Criminal Appeal has led each of their Honours to conclude that there is such a similarity between the sets of the offences that the jury might rightly have concluded that the same offender had committed all three sets. Where experience in the day to day work of the criminal courts contributes so much to the resolution of questions of this kind, this Court should be loath to interfere with the conclusions of the Supreme Courts. Though I should have thought that the case is close to the boarder line, I think their Honours placed it on the correct side. At all events, having regard to the reasons of Mitchell J. and of the Court of Criminal Appeal for admitting the evidence on all counts, I would not depart from the view they have expressed. 1. [1975] A.C., at pp. 444-445. 2. [1973] A.C., at p. 756.

Special leave to appeal should be granted. I would dismiss the appeal.

This is an application for special leave to appeal from a decision of the Full Court of the Supreme Court of South Australia (Wells, Sangster and Mohr JJ.) dismissing an appeal from the applicant's conviction on each of seven charges of rape and of one charge of attempted rape. The eight charges, which were heard together, arose from three separate occurrences. Each occurrence involved a sexual attack upon a different teen-aged girl. The applicant seeks to raise three issues relating to the conduct by the learned trial judge (Mitchell J.) of his trial. Those issues are:

1. Whether Mitchell J. was in error in ruling that the charges against the applicant formed or were "a part of a series of offences of the same or similar character" within s. 278(1) of the Criminal Law Consolidation Act 1935 S.A..

2. If the answer to 1. is in the negative whether her Honour should, in any event, have ordered that the applicant be tried separately in respect of the charges relating to each occurrence.

3. Whether her Honour was in error in failing to direct the jury that, in considering the charges relating to a particular occurrence, they should ignore the evidence in relation to the other two occurrences.

The three occurrences were in successive months (August, September and October) of 1981. Each occurred in either the suburb of Alberton or the suburb of Woodville which are in the same general area of Adelaide. In each, the victim was attacked as she walked alone along a public street at night by being grabbed by the assailant, either from behind or from the side, with one hand being placed over her mouth and another around her neck or body. In each, she was told not to scream or she would be killed and was restrained by one arm while she was forced to walk a short distance to the place where she was sexually molested. In each, that place was in the grounds of a State school: the first and third being in the grounds of the Alberton Primary School; the second being in the grounds of the Woodville Primary School. In each, after rape or other sexual molestation, the assailant left her at the scene of the crime and set off in the direction of Port Road which runs through both Alberton and Woodville. In each case, the assailant was alone, on foot and apparently unarmed. In each case, the assailant was described by the victim as being of medium build and as being either of "about six feet" in height or a "bit smaller than" an indicated person who was about 6′2″ tall. The victims' accounts of the assailant's hair, were in general accord: they varied from "blond" to "dark blond" and from "shoulder length" to "on" or "just above" the shoulder. In the case of two of the three occurrences, the victim had identified the applicant as the person who had assaulted her: in one case, shortly after the offences occurred; in the other, approximately a month after the occurrence, when the victim saw the applicant on two occasions in the street. In the case of the third occurrence (the Woodville rape) there was evidence from a young woman who lived with the applicant from August until October 1981 that the applicant had admitted to her that he was the assailant.

At the trial, Mitchell J. came to the conclusion that the similarities between the three occurrences were so striking as to warrant an inference by the jury that the assailant in each case was the same person or, to use her Honour's words, that "upon identification to its satisfaction of the accused as the person who committed one set of rapes, it was satisfied that he must have committed the other set of rapes and must have been guilty of the attempted rape". I have, in the preceding paragraph, summarized what I see as the main similarities between the three occurrences. In the context that the three incidents occurred in successive months in the same general area of Adelaide, those similarities were, in my view, such as to justify her Honour in reaching the conclusion that it was open to the jury to draw such an inference. Her Honour adopted the approach that it followed, from that conclusion, that any evidence which was admissible in relation to the charge or charges arising from any one of the three occurrences was properly admissible on the charge or charges relating to the other two. The underlying logic of that approach is clear enough: if it were inferred by the jury that the same person was the assailant in each case, evidence that the applicant was the assailant in any one case was likewise probative of his having been the assailant in the other two. A question arises whether that approach is consistent with the law.

In a case where there is evidence connecting an accused with the particular offence with which he is charged, evidence that he has committed other offences my be relevant to the question of his guilt of the charged offence even though it does no more than establish that he is the sort of person who is likely to have committed that offence. Such evidence of propensity to crime or to a particular kind of crime is not, however, made admissible merely by reason of its character as such. The classic statement of the general proposition to that effect is that of Lord Herschell L.C., speaking for the Judicial Committee in Makin v. Attorney-General (N.S.W.) [67] :

It is undoubtedly 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, 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 (emphasis added). That statement of the proposition was as one of two related but "independent" propositions which "together cover the entire field" (per Lord Hailsham of St. Marylebone, Director of Public Prosecutions v. Boardman [68] ). Lord Herschell's second proposition was: 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 It is plain, both from the reference to "purpose" in the first proposition and from the wording of the second proposition, that Lord Herschell L.C. did not propound the first proposition as a general positive rule of exclusion of evidence of commission of other crimes and the second proposition as but an exception grafted upon the first. To the contrary, he cited with approval the statement of Maule J. in Reg. v. Dossett [70] in which the combined effect of the two propositions is expressed in the form of a denial of any positive rule of exclusion: "[a]lthough the evidence offered may be proof of another felony, that circumstance does not render it inadmissible, if the evidence be otherwise receivable." Put another way, the effect of what was said in Makin is that evidence that an accused "is a man of bad character with a disposition to commit crimes" is, in itself, "deemed to be irrelevant in English law" (per Lord Salmon, Director of Public Prosecutions v. Boardman [71] ; see also per Latham C.J., Sodeman v. The King [72] ), but "may be admitted if it is relevant for other reasons" (Noor Mohamed v. The King [73] ). 1. [1894] A.C. 57, at p. 65. 2. [1975] A.C. 421, at p. 452. 3. (1846) 2 C. & K. 306, at p. 307 [175 E.R. 126, at p. 127]. 4. [1975] A.C., at p. 461. 5. (1936) 55 C.L.R. 192, at p. 206. 6. [1949] A.C. 182, at p. 195.

In Makin [74] , Lord Herschell L.C. instanced two cases where evidence of the commission of other offences may be "relevant to an issue before the jury", namely, where "it bears upon the question whether the acts alleged to constitute the crime were designed or accidental, or to rebut a defence which would otherwise be open to the accused". Those instances were not intended to be and are not exhaustive (see Harris v. Director of Public Prosecutions [75] ). The authorities establish that evidence of the commission of other offences may be admissible if it be probative not merely of propensity to commit crime or a particular crime but of the actual commission by the accused of the act or acts which are said to have constituted the basis of the offence with which he is charged or of some other fact to be proved (see, e.g. R. v. Sims [76] ; Director of Public Prosecutions v. Boardman).

1. [1894] A.C., at p. 65. 2. [1952] A.C. 694, at p. 705. 3. [1946] 1 K.B. 531.

The distinction between evidence of "mere propensity" and evidence which is relevant for the reason that it is probative of a fact which is in issue is probably best illustrated by example. Evidence that an accused has, in the past, murdered a number of people is, in itself, not admissible on his trial on a charge of having committed another murder or on a charge of some different kind of offence. Such evidence may establish a propensity to crime or to the particular kind of crime. It is not however, in itself, probative of his guilt of the offence with which he is charged or of any fact which is in actual issue on his trial for that offence: at best, it is productive of suspicion or prejudice on the basis that the past offences indicate that he is the sort of person who is likely to be guilty of that offence. 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 coincidence, the same person committed all offences or if the similarity between the evidence of different witnesses as to different offences which the accused is said by them to have committed is so striking "that they must, when judged by experience and common sense, either all be true, or have arisen from a cause common to the witnesses or from pure coincidence" (Boardman [77] ), 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.

One finds, in judgments in "similar fact" cases since the decision in Makin v. Attorney-General, an increasing tendency to emphasize the fundamental character of the first of Lord Herschell's propositions, often at the cost of the second of them. In 1934, Viscount Sankey L.C., speaking for the Judicial Committee in words which were subsequently adopted by Latham C.J. in this Court, described that first proposition as "one of the most deeply rooted and jealously guarded principles of our criminal law" (Maxwell v. Director of Public Prosecutions [78] ; Burrows v. The King [79] ). Recognition of the importance of the first proposition and concern lest "the protection afforded by" [it] "be gravely impaired" (Noor Mohamed v. The King [80] ) have combined to cause it to be accepted as not merely one of two complementary propositions relating to prima facie relevance but as a substantive rule of exclusion. Thus it has been held that the fact that evidence may be relevant to some theoretical issue is insufficient to except it from the general rule of exclusion unless it can be seen to be in truth material to an actual issue which is raised in substance as distinct from being merely a formal issue raised by, for example, a plea of not guilty (Thompson v. The King [81] ; Noor Mohamed v. The King [82] ). In Markby v. The Queen [83] , the first proposition was stated by Gibbs A.C.J., with whose judgment Stephen, Jacobs and Aickin JJ. agreed, to be a general "rule of exclusion" which is a "rule of law". His Honour also stated that the question whether "similar fact evidence" was admissible should be approached on the basis that "the admission of [such] evidence is the exception rather than the rule". That approach of the present Chief Justice accords with that of Lord Wilberforce in Director of Public Prosecutions v. Boardman [84] . I agree with it. The present-day rationale of that general "rule of exclusion" should be seen as being that evidence of mere propensity to commit criminal offences, either generally or of the type charged, is liable to erode the presumption of innocence by substituting trial by prejudice and suspicion for trial by reference to the question whether the proven facts establish beyond reasonable doubt that the accused is guilty of the offence of which he stands charged.

1. [1935] A.C. 309, at p. 317. 2. (1937) 58 C.L.R. 349, at p. 253. 3. [1949] A.C., at p. 191. 4. [1918] A.C. 221, at p. 232. 5. [1949] A.C., at pp. 191-192. 6. (1978) 140 C.L.R. 108, at p. 117. 7. [1975] A.C., at p. 444.

The fact that evidence that an accused has committed other offences may be technically admissible, as probative of a fact which is actually in issue and as thereby coming within the exception to the general rule of exclusion, does not necessarily mean that a trial judge should allow it to be led. It is the duty of the presiding judge at a criminal trial to set the essentials of justice above the strict rules of evidence if the strict application of the latter would operate unfairly against the accused (Harris v. Director of Public Prosecutions [85] ). If, in a particular case, the probable prejudicial effect of evidence that the accused has committed other crimes "would be out of proportion to its true evidential value" (per Lord Moulton, R. v. Christie [86] ), the trial judge may and should exclude it. The power to exclude such evidence is a discretionary one. In a case where the admission of evidence of the commission by the accused of other criminal offences is likely to be gravely prejudicial, it is a matter for the trial judge to consider whether the evidentiary value of the evidence is sufficiently substantial, having regard to the purpose for which it is technically admissible, to make it desirable in the interests of justice that it should be admitted.

1. [1952] A.C., at p. 707. 2. [1914] A.C. 545, at p. 559.

In recent cases in this Court (Markby [1] and Perry v. The Queen [2] ) and in the House of Lords (Boardman [3] ), one finds statements to the effect that "similar fact evidence" is only admissible if it possesses "a strong degree of probative force". Such statements may well serve as timely admonitions of the need to safeguard against erosion of the general rule of exclusion by ensuring that the admission of "similar fact evidence" is restricted to the case where it is properly and positively seen, when viewed in the context of the evidence as a whole, as having "a really material bearing on the actual issues to be decided" (per Lord Morris of Borth-y-Gest, Boardman [4] ) as distinct from merely establishing bad character or propensity to commit crime or the particular type of crime. The general rule is that "similar fact evidence" is inadmissible. It is for the prosecution to demonstrate to the trial judge that any such evidence comes within the exception to that prima facie and fundamental rule of exclusion. That having been said however, I find difficulty with the proposition that a body of similar fact evidence which, in the above sense, is of "distinct" probative force is not admissible unless that probative force is "strong". It is true, as Brennan J. pointed out in Perry [5] , that the probative force of such evidence "depends upon the connexion, judged according to experience and common sense, between the facts directly established by the evidence and the fact in issue, which may be the conduct charged, the identity of the culprit or his state of mind" and that the "connexion is a question of degree". It is, for example, possible that there will be a need for strong or striking similarity between different offences or between the evidence of different witnesses about acts of the accused on different occasions or a need that there be strong evidence that the accused committed a "similar fact" offence before the relevant body of "similar fact" evidence will be admissible as having distinct probative force on a fact which is actually in issue. Once the necessary standard of similarity or proof which is necessary to give the evidence such probative force on a fact in issue is met however, it is difficult to see any proper basis in principle or authority for imposing a further requirement that the probative force of the similar fact evidence should be strong. I would prefer to say, adopting the words of Brennan J. in Perry [5a] , that what is necessary for the "similar fact evidence" to be admissible is that the "probative force of the evidence clearly transcends" - by which I understand his Honour to mean goes beyond - "the merely prejudicial effect" of proof of prior criminality or propensity. If the prosecution is unable to demonstrate that, in the circumstances of the particular case and in the context of the evidence as a whole, the "similar fact" evidence clearly has probative force, on a fact in issue, which is independent of or goes beyond its prejudicial effect as evidence of mere propensity, the "similar fact" evidence will not have been shown to come within the exception of the rule of exclusion and should be rejected.

1. (1978) 140 C.L.R., at p. 117. 2. (1982) 150 C.L.R., at pp. 586-587, 589, 604. 3. [1975] A.C., at p. 444. 4. [1975] A.C., at p. 439. 5. (1982) 150 C.L.R., at p. 609. 6. (1980) 150 C.L.R., at p. 609.

It follows from what has been said above that, in the present case, evidence in relation to one occurrence was, under the general rule of exclusion of evidence that an accused has committed other crimes, prima facie inadmissible on the charge or charges relating to another occurrence. Such evidence was only admissible if it was relevant to an issue which was actually raised in the trial of the accused on the charge in relation to which it was tendered and if it was not inadmissible on other grounds such as, for example, that it was hearsay.

The principal issue - indeed, the only real issue - between the Crown and the applicant in relation to each of the eight charges against him was the identity of the assailant. Once the conclusion is reached that the similarity between the three occurrences was so striking as to warrant a conclusion by the jury that the assailant on each occasion was the same person, it is apparent that evidence that the applicant was the assailant in the case of any one of the assaults was relevant and possessed probative force in relation to the question whether the applicant was the assailant in relation to the other occurrences. That is not to say that the evidence was admissible on the basis of an assumption that the applicant was guilty of the offence which was the subject of a particular charge or of the offence or offences which related to other occurrences. It is simply to recognize that, once the similarity between the three occurrences is accepted as being such as to warrant a conclusion by the jury that the assailant on each occasion was the same, evidence that the applicant was the assailant on one of the other occasions is material and has probative force on the question whether he was the assailant on the occasion to which the charge relates and evidence that he was the assailant on the occasion to which the charge relates is material and has probative force on the question whether he was the assailant on the other occasions.

It follows that Mitchell J. was correct in holding that, in the circumstances of the present case, evidence in relation to the other occurrences was admissible on the charge or charges relating to each occurrence for the reason that it came within the exception to the general rule excluding evidence of other criminal offences. It was common ground on the hearing of the present application that the consequence of a decision to that effect was that all three issues which the applicant wishes to raise must be resolved adversely to him. I would, accordingly, refuse the application for special leave to appeal.

This is an application for special leave to appeal against convictions for sexual offences against three different girls on three different occasions. In relation to the first girl, the applicant was convicted upon one count of attempted rape by oral sexual intercourse. In relation to the second girl, he was convicted of three counts of rape consisting of oral, anal and vaginal intercourse and in relation to the third girl, he was convicted of four counts of rape consisting of oral and anal intercourse and two separate acts of vaginal intercourse.

The questions which arise upon this application are whether it was proper to join all the counts in the same information and, if so, whether the trial judge nevertheless erred in the exercise of her discretion in refusing to sever the information and direct separate trials in relation to some of the counts. Even if the charges were properly joined and the trial judge did not err in the exercise of her discretion, a third question arises whether the trial judge should have directed the jury to consider the evidence separately in relation to the counts involving each of the girls.

At the trial, counsel for the accused submitted that the counts in relation to the second girl should be tried separately from the counts in relation to the other two girls which he conceded were properly joined. The basis of this concession was that the offences against the first and the third girls were committed in the same place, namely, the grounds of the Alberton Primary School, although on different occasions. The offences against the second girl took place at another location, namely, the grounds of the Woodville School. This submission was rejected by the learned trial judge who ruled that evidence in relation to each of the counts involving one girl was admissible in relation to the counts involving the other girls and that for that reason all of the counts were properly joined in the same information and that she should not exercise her discretion to sever the information.

Under s. 278(1) of the Criminal Law Consolidation Act 1935 S.A., which stems from the Indictments Act 1915 U.K., more than one charge may be joined in the same information if the charges are founded on the same facts, or form or are a part of a series of offences of the same or a similar character. Sub-section (2) of the same section confers a discretion upon the court to direct, before trial or at any stage of a trial, the separate trial of any count or counts in an information where it is desirable to do so whether because an accused person may otherwise be prejudiced or embarrassed in his defence or for any other reason.

The trial judge took the view that if the evidence admissible to prove the charges involving one girl was admissible to prove the charges involving the others, then the offences in question formed a series of offences of the same or a similar character and that there could be no basis for the exercise of her discretion to direct separate trials. There can be no doubt that the trial judge was correct in taking this view. Whatever might otherwise be considered to be a sufficient connexion between separate offences or sets of offences to constitute them a series, it will be enough if the evidence admissible to prove one offence is admissible in the proof of the others. Such evidence of similar facts will only be admissible, for reasons to which I shall immediately turn, if there is a clear connexion between the different offences. It is, therefore, appropriate to deal first with the question of admissibility.

The law does not admit evidence tending to show only that an accused person has committed other offences as proof of the commission of a particular offence with which he is charged. This is because mere proof of the other offences can do no more than demonstrate a criminal propensity and the prejudicial nature of evidence of this kind is far greater than any relevance it might have. To admit such evidence would be to invite the injury to proceed upon suspicion rather than proof. If, however, the evidence which tends to prove the commission of other offences has a sufficient additional probative value beyond showing a disposition to commit crime or a particular type of crime, then the evidence is admissible for that reason and not because it may show a criminal propensity. The cases in which similar fact evidence may have sufficient additional relevance to make it admissible are not confined, but recognized instances occur where the evidence is relevant to prove intent or to disprove accident or mistake, to prove identity or to disprove innocent association.

The authorities which now enable the law to be stated in this way start with Makin v. Attorney-General (N.S.W.) [6] and culminate in Director of Public Prosecutions v. Boardman [7] . They have been recently examined by this Court in Markby v. The Queen [8] and Perry v. The Queen [9] and there is no point in repeating the examination in this case. There is, however, one matter that has not been settled with any precision and that is the test to be applied in determining whether similar fact evidence has sufficient probative value or relevance, other than to establish bad character or a criminal disposition, to be admissible. The suggested standards range from the requirement that the evidence must have a really material bearing upon the issues to be decided to the requirement that it must be so very relevant that to exclude it would be an affront to common sense. See Harris v. Director of Public Prosecutions [10] ; Director of Public Prosecutions v. Boardman [11] .

1. [1894] A.C. 57. 2. [1975] A.C. 421. 3. (1978) 140 C.L.R. 108. 4. (1982) 150 C.L.R. 580. 5. [1952] A.C. 694, at p. 710. 6. [1975] A.C., at p. 456.

The test favoured by Gibbs C.J. in Perry v. The Queen was that the evidence should not be only technically relevant (otherwise than as showing a propensity); it must be really material; it must have strong probative force [12] . In the same case, Wilson J. thought that to be admissible the evidence must have a sufficiently high degree of probative force to outweigh the inevitable prejudice flowing from the nature of the evidence [13] . Similarly, Brennan J. would admit similar fact evidence only where its probative force clearly transcends its merely prejudicial effect [1] .

1. (1982) 150 C.L.R., at pp. 586-587. 2. (1982) 150 C.L.R., at p. 605. 3. (1982) 150 C.L.R., at p. 610.

Similar fact evidence is, of course, only circumstantial evidence but it is circumstantial evidence of a particular kind because of the inevitable prejudice, to use the words of Wilson J., which it carries with it. Although there is no necessary requirement of law or of practice, generally where the evidence in a case is circumstantial some direction ought to be given to that jury that before they convict the accused they must be satisfied that the facts are such as to be inconsistent with any other rational conclusion than the guilt of the accused. See Reg. v. Hodge [2] ; Plomp v. The Queen [3] ; Gant v. The Queen [4] . Such a direction is a particular expression of the mandatory direction that guilt must be proved beyond reasonable doubt, but it is appropriate in most cases of circumstantial evidence because it serves to emphasize the special care which must be taken in those cases.

1. (1838) 2 Lewin 227, at p. 228 [168 E.R. 1136, at p. 1137]. 2. (1963) 110 C.L.R. 234. 3. (1981) 147 C.L.R. 503.

The question which arises with similar fact evidence, which because of its prejudicial nature is treated as a special kind of circumstantial evidence, is not only what direction should be given to the jury but whether the evidence is admissible at all. Having regard to the various expressions which are used to lay down the test of admissibility, it seems to me that a trial judge may find assistance in arriving at the correct test in any particular case by applying the same standard as the jury must ultimately apply in dealing with circumstantial evidence. If in considering the admissibility of similar fact evidence the trial judge concludes that there is a rational view of that evidence which is inconsistent with the guilt of the accused, then he ought not admit it because in those circumstances the evidence cannot be said to have a sufficiently strong probative force. Prejudice may operate where neither logic nor experience necessarily require the answer that the evidence points to the guilt of the accused and that being so the probative force of the evidence will not outweigh or transcend its prejudicial effect.

Such an approach is, I think, consistent with what this Court said in Martin v. Osborne [5] in considering the admissibility of similar fact evidence and is certainly consistent with the frequently expressed view that similar fact evidence is dangerous and is to be treated with greater caution than other circumstantial evidence. Of course, the question is ultimately one for the jury but if in order to rule that the evidence is admissible the trial judge must conclude (as it is clear he must) that it has a stronger degree of probative force than would lead merely to the conclusion that it is capable of being regarded, apart from propensity, as pointing to guilt, then it seems to me to follow that before admitting the evidence the trial judge himself must conclude that a reasonable jury would, if they accept the evidence, regard it as being inconsistent with innocence.

This is to say no more than that to be admissible the similar fact evidence must give rise to a clear inference tending to establish an element of the offence charged and must not be reasonably explicable upon some other basis. Notwithstanding that the question is one of admissibility, it is nevertheless a matter of degree rather than deductive logic and it cannot be said that a clear inference arises unless "general human experience would be contradicted, if the proved facts were unaccompanied by the fact sought to be proved". See Morgan v. Babcock & Wilcox Ltd. [6] .

It should be added that even when the admissibility of similar fact evidence is decided as a question of law in favour of the Crown, there still remains the possibility, perhaps ordinarily no more than a theoretical possibility, that the evidence can be excluded by the trial judge in the ordinary exercise of his discretion to exclude evidence when its probative value is outweighed by its prejudicial effect. For my part I find it difficult to envisage circumstances in which this could happen because it seems to me that the whole purpose of restricting the admissibility of similar fact evidence is to ensure that it cannot be used unless its probative force is sufficiently strong to outweigh or transcend its prejudicial effect, but that is not to deny the possibility.

In considering the admissibility of similar fact evidence the inference to be drawn from the fact when proved is one thing; the strength of the evidence required to prove the fact is another. There must be clear evidence which if accepted by the jury establishes that fact. No difficulty will ordinarily arise where the evidence, if accepted, directly establishes that fact but where the fact itself is a matter of inference then the inference must be capable of being clearly drawn from the evidence relied upon before that evidence is admissible. If it were not so, the requirement that evidence of similar facts should have strong probative force before being admissible would be considerably undermined. The proposition may be illustrated by reference to Perry v. The Queen [7] . In that case the accused was charged with the attempted murder of her husband by administering arsenic to him. Each of the similar facts sought to be proved was an instance of poisoning, by the ingestion of arsenic, of a person with whom the accused had a close association at the relevant time. The whole of the Court was of the view that evidence of one particular instance was inadmissible, not because that instance if proved would not constitute a similar fact carrying the necessary clear inference that the accused was guilty of the offence with which she was charged, but because the available evidence, even if accepted, did not prove with sufficient strength the ingestion of arsenic in order to provide a factual basis - a similar fact - upon which to found the ultimate inference. The instance to which I refer is that of Duncan. There was no direct evidence that the illness which he suffered was due to arsenical poisoning. If it was, it was a matter of inference and the evidence relied upon to support that inference was, even in the context of the other evidence (including that the ingestion of arsenic by the accused's husband), insufficient to establish with sufficient probative force the fact from which a further inference could be drawn that the accused had administered arsenic to her husband. And that, of course, had to be taken into account in determining whether the evidence in relation to Duncan was admissible. The question was not reached whether the similar fact which the Crown sought to prove in relation to him was sufficiently probative of the guilt of the accused of the offence with which she was charged. The evidence, even if accepted, failed to establish the similar fact with sufficient certainty to warrant its admission. Clearly enough, if there had been sufficiently strong evidence to establish that Duncan's illness was due to arsenical poisoning a majority would have concluded that the evidence was admissible because the coincidence would, in all the circumstances, have been too great according to ordinary human experience to avoid the conclusion that the accused was implicated. In other words, it would then not have been possible reasonably to conclude that the ingestion of arsenic by Duncan was in the circumstances a matter of mere coincidence.

The same problem does not arise in the present case. The question which arises here is that of the identity of the accused with the person or persons who committed the offences against the complainants. There was at the trial no contest that those offences took place nor was there any contest that each of the complainants was truthful in her description of her assailant and of the circumstances in which each assault took place. The evidence of one set of offences was tendered in proof of the others upon the basis that the circumstances of each offence, including the description of the assailant, were so similar as to be beyond the realm of coincidence. The evidence, if admissible upon this basis, entitled the jury to conclude that the same person committed all three sets of offences and if they were satisfied beyond reasonable doubt that the accused was correctly identified in any one instance, they were entitled to convict him, as they did, of all the offences with which he was charged.

The evidence was admissible on the basis upon which it was tendered if the proof of the identity of the accused in relation to one set of offences was sufficiently probative of his identity in relation to the others. That is to say, proof of the accused's identity in one instance must have given rise to a clear inference that he was the offender in the other instances. In cases such as this it is sometimes said that the circumstances of the various offences must be strikingly similar. See, e.g.,Director of Public Prosecutions v. Boardman [8] . That, however, is merely to posit another test of no particular precision. See Scarrott v. The Queen [9] , per Scarman L.J. In my view it was for the trial judge to determine admissibility by deciding whether the evidence was, if accepted, such that a reasonable jury ought to conclude that the similarity between the descriptions of the various offences was beyond mere coincidence and that there was no reasonable explanation of that similarity consistent with the innocence of the accused.

1. [1975] A.C. 421. 2. (1977) 65 Cr.App.R. 125, at p. 129.

The trial judge in fact used the test of striking similarity but it is clear, in my view, that in applying that test she came to a conclusion which justified the admission of the evidence. After listing a number of aspects of the offences she said:

Some of the common features are of little significance. For example, the wearing of jeans is common to a large section of members of the community, as is the smoking of cigarettes. There was however striking similarity in the method of abduction and the method of conducting the alleged victim to the places where the offences were said to have been committed; a similarity in the fact that in each case the offences were alleged to have been committed in the grounds of a school at night and finally the commission in the first [semble, second] and third sets of charges of acts of vaginal, oral and anal intercourse. I was of the opinion that the similarities between the circumstances of the incidents deposed to by the witnesses for the prosecution were so striking that it should be left to the jury to decide whether, upon identification to its satisfaction of the accused as the person who committed one set of rapes, it was satisfied that he must have committed the other set of rapes and must have been guilty of the attempted rape.

That the trial judge did not err in her approach to the admission of the similar fact evidence in this case is, I think, enough to require the refusal of special leave to appeal. However, I would add that in giving the matter consideration I have reached the conclusion, not without hesitation, that her decision to admit the evidence was correct. It is not, I think, a case in which any one circumstance common to the various offences was sufficiently striking to eliminate any reasonable possibility of coincidence. Rather it was the accumulation of common circumstances which had that effect. It is, of course, possible to consider each similar fact in the context of other similar facts and conclude that, looked at together, mere coincidence is not a reasonable hypothesis but be unable to reach that conclusion viewing each set of similar facts separately. Despite some suggestion to the contrary in the Court below, Perry v. The Queen did not require each set of similar facts in that case to be regarded separately in order to arrive at their probative value; it was merely that the impugned evidence, even when viewed in the context of the other evidence, did not have the required probative force. It was not, for example, held that the evidence of illness on the part of Duncan could not be viewed in the context of the illness of the accused's husband which was clearly due to arsenical poisoning. It was merely that even when so viewed the cause of Duncan's illness could not be sufficiently established as being the ingestion of arsenic to enable any further inference to be drawn with the requisite clarity. It was not, of course, possible to assume for the purpose of assisting in the drawing of that further inference that the accused administered arsenic to her husband for that was the very inference sought to be drawn from the evidence. But as Brennan J. [10] pointed out it is always the case that "evidence of a series of occurrences exhibiting a more attenuated similarity may be admissible because the frequency of the occurrence of the similar facts enhances the probative force of the evidence, though the necessary probative force would be lacking if the similar fact had occurred but once or on a few occasions only (cf. Reg. v. Scarrott [11] )".

1. (1982) 150 C.L.R., at p. 610. 2. [1978] Q.B. 1016, at pp. 1022-1023.

In the present case, each of the complainants was between the ages of thirteen and eighteen years at the time the offences were committed. Each of the offences was committed in a school yard, two sets of offences in the same school yard. The first complainant identified the accused as her assailant. The second complainant said that the assailant was in his early twenties and the third said he was eighteen or twenty. Each of the complainants estimated his height to be about six feet and said that he was of medium build. The first and third complainants said that their assailant had blond hair; the second complainant said that her assailant had light brown or dark blond hair. All three said his hair was, in effect, shoulder length. The first complainant said that her assailant's hair was straight but the other two complainants said their assailant's hair was wavy. Each of the first two complainants said that her assailant smelt of cigarette smoke. The third said she could not smell anything on the breath of her assailant but said that he asked her if she had seen his matches. In each case the assailant was a lone offender on foot who grabbed his victim from behind or after he had been behind her and placed his hand (in two instances his right hand was specified) over her mouth. In the case of two of the complainants the offender placed his left arm around each girl, in the one instance around her arms or body or both and in the other instance pinning her arms. In the case of the third girl the assailant placed his hand around her neck. Each girl was told not to scream or to shut up or she would be killed. In each case the complainant was restrained by one of the offender's arms whilst she was taken to the place of the offence or offences. In each case the place of the offence or offences was a school yard within a short distance from the place of abduction. Each offence took place within a confined geographical area. The offence or offences in relation to each of the complainants took place at intervals of some five or six weeks. In each case the assailant left the complainant at the place of the offence or offences and made off in the same direction. The first complainant's assailant attempted oral sexual intercourse. In the case of each of the second and third girls the assailant had oral, anal and vaginal sexual intercourse. All of these circumstances, when considered together as they ought to have been, were, in my view, sufficient to eliminate coincidence and to require as the only reasonable finding the conclusion that the same man committed the offence or offences against each of the three complainants. The evidence in each case was for this reason admissible in the others and for this reason the offences formed a series of offences of the same or similar character. Since the evidence was admissible in this way no prejudice or embarrassment could be shown to arise from the trial of all the charges against the accused together.