2 The prosecution resulted from a police raid on a house at Felixstow on 16 August 2001. The appellant lived on the premises and, although another person leased the house, it was the prosecution case that the appellant exercised control over the premises. Four other people, including Beuret, were staying at the house at the time of the raid.

3 Police officers searched the premises and located ingredients and equipment suitable for use in the manufacture of methylamphetamine. There were a number of utensils in the kitchen which bore traces of amphetamine. Furthermore, traces of powder and pseudoephedrine, an ingredient of methylamphetamine, were found on the hands of the appellant and the other occupants of the house. The prosecution alleged that these traces may have been deposited as a result of the occupants of the house crushing tablets as part of the process of manufacturing methylamphetamine. Some items found in the appellant's room, along with a list of names and amounts of money, were relied upon by the prosecution as supporting the inference that the appellant was involved in the sale of drugs.

4 The prosecution also led evidence to the effect that the appellant had purchased iodine, a chemical used in the manufacture of methylamphetamine, on two occasions approximately two months prior to the police raid.

5 When questioned by the police, the appellant denied any involvement in the manufacture of the drug. He said he had seen tablets around the house and was concerned about the activities of the other occupants. He had no knowledge of the various items found by the police. He said he had no connection with the drug trade.

6 In his opening address to the jury, the prosecutor explained that the prosecution case was based on inferences to be drawn from the finding of the various items in the house, the existence of the trace materials on the appellant's hands and the appellant's purchase of iodine. The prosecutor also told the jury that, if the appellant exercised control over the premises and knowingly permitted others to manufacture methylamphetamine, he could be found guilty on that basis alone.

7 Both the appellant and Beuret gave evidence. The appellant denied involvement in the manufacture of amphetamine. He also denied that he knowingly permitted others to manufacture the drug. He said he had been absent from the house for the three days prior to the raid. He claimed that items found in the home could have been left there by a man named McKenzie who had moved out of the house a few days before the police raid. The police officers found items in premises occupied by McKenzie which could have been used in the manufacture of amphetamines.

8 The appellant also stated in evidence that two former occupants of the house, Aaron Goodwin and Jennifer Warburton, used amphetamines and that he disapproved of them doing so. The appellant agreed that he had purchased iodine, but said he did so at the request of McKenzie after McKenzie told him he wanted it for a fish tank.

9 The appellant gave as a possible explanation for the traces of powder and pseudoephedrine on his hands, his handling of various utensils in the house.

10 The first ground of appeal complains that the trial judge instructed the jury in the course of his summing-up that the appellant could be found guilty of the offence charged on a factual basis which, according to the appellant, had not been relied upon by the prosecution at any stage throughout the trial.

11 The additional path to conviction identified by the trial judge arose in the following way. I have said that the co-accused Beuret gave evidence at the joint trial. He stated that on one occasion when he visited the appellant's house he saw the appellant add sugar to some oil on a plate. He said the oil was "a light orangey sort of colour" which was on an oval serving plate. The witness said that the appellant gave him some of the substance to try later on and it was methylamphetamine.

12 Beuret's counsel questioned the appellant about this incident in cross-examination, but the appellant denied it. The prosecutor did not ask the appellant any questions on the topic. In the course of his summing-up the trial judge said:

"Although the prosecutor did not expressly rely upon it in his closing address, you also have evidence from Beuret that he saw Franco add sugar to oil on a plate, which he later gave to Beuret to try, and Beuret identified it as meth. Franco denied that this had ever occurred. No other witness apparently saw it. If you accept Beuret on this point, it would be direct evidence of a step by Franco in the manufacture of meth."

13 Nothing further was said by the trial judge on this topic. Despite the brevity of the above passage, it is clear that the trial judge invited the jury to consider this incident as constituting a sufficient basis in itself for a conviction.

14 I have said that the prosecution did not rely on this evidence to support a conviction. It was conceded before this court that the prosecutor purposely refrained from doing so because he did not wish to rely on the co-accused Beuret as a witness of truth. There appears to have been no discussion between the trial judge and counsel before the summing-up as to whether the incident might found a conviction.

15 It must be said that the evidence of the incident deposed to by Beuret was left in a vague state, no doubt because the prosecution did not wish to rely on the evidence to support a conviction. In these circumstances, the appellant's counsel did not see the need to probe the matter in any depth.

16 Mr Beuret was not cross-examined as to when the incident took place. It is not clear how long after the addition of the sugar Beuret was given the substance to try. Nor is it clear how he was able to tell that it was methylamphetamine.

17 It does not appear that Mr Camilleri, the scientific expert called by the prosecution, was asked questions about the manufacturing process with this particular incident in mind. He gave general evidence about the process of manufacture of methylamphetamine. Mr Camilleri explained that one of the methods for manufacturing this drug was to begin by extracting pseudoephedrine from one of a number of relatively common pharmaceutical tablets. He said the tablets are usually crushed into a powdered form before being dissolved by a solvent and then filtered. Eventually the pseudoephedrine is reduced to powdered form. This is then treated with chemicals to produce methylamphetamine. Following this, any acid in the methylamphetamine is neutralised and, under steam distillation, a liquid is produced. The end result of this stage of the process is a methylamphetamine base which consists of an oil. The final stage involves acidifying the base and filtering it to form the methylamphetamine salt. According to the evidence, this is then often cut down with some substance such as glucose. However Mr Camilleri did not refer to glucose or sugar being added directly to the methylamphetamine oil.

18 At the conclusion of the summing-up, counsel who represented the appellant at the trial drew the trial judge's attention to the fact that Mr Beuret's description did not fit in with the process of manufacture described by Mr Camilleri. He also said that there was no evidence as to the identity of the oil. In the course of the discussion between the trial judge and defence counsel the following exchange took place:

"HIS HONOUR: I think the answer to your point, Mr McShane, is probably at p 955 of the transcript, where Beuret says:

'A I knew what he was doing because he gave me some afterwards to try.

That seems to be fairly explicit evidence that Beuret was identifying what the sugar was being added to and something that turned out to be methylamphetamine.

MR MCSHANE: That is what, on the evidence of Camilleri -

HIS HONOUR: They said there were a number of possibilities to manufacture. Nobody excluded it.

HIS HONOUR: You didn't cross-examine on the basis that or seek them to be recalled to put this possibility to them. If I redirect the jury, it seems I have to read to them what Beuret said at p 955. The whole thing is predicated upon them being able to accept Beuret's evidence, I have made that quite clear.

MR MCSHANE: You certainly have. I raise it though on two basis - and I appreciate where your Honour is coming from - that he took some and used it. I suppose I would argue it is still not clear that what he says he was supplied with was from that specific source or whether it was just some other source of methylamphetamine.

HIS HONOUR: 'He gave me some of it afterwards to try.' I would have thought it was sufficient for a jury to find the step that had gone before is something towards producing what Beuret identified as methylamphetamine.

MR MCSHANE: I appreciate that. I don't want to be pedantic about it, we went through it in great deal, I indicated to your Honour we have had the scientific evidence and there is no other evidence as far as from any of the defendants as to how to mix a dry substance with an oil. The inference there is that simply doesn't occur. The oil produced from the manufacturing process has got to be mixed to the dry substance, that was the evidence.

HIS HONOUR: I accept that may be a problem. If I redirect on it - and I have to do it in some detail on the evidence - I am then in fact highlighting something that Mr Kimber [the prosecutor] didn't express in his address against your client. I am really wondering whether that is tactically desirable.

MR MCSHANE: It is a bit of a 22 [sic] perhaps. Perhaps I had better leave that matter. I apologise for wasting your Honour's time.

HIS HONOUR: I would much prefer they are raised now rather than before the Court of Criminal Appeal. In that case there is nothing on which I intend to redirect."

19 It is a fundamental principle of fairness in all criminal proceedings that an accused person be given adequate particulars of the charge brought by the prosecution. In the absence of pleadings of the type required in civil proceedings, particulars beyond the short statement of particulars in the information may be supplied by the prosecution in response to a defence request or by direction of the court. However, the normal course in a jury trial is for the prosecution to provide the necessary particulars by means of its opening address together with the information contained in declarations of witnesses supplied to the defence.

20 Although it is desirable that the prosecution should indicate at the commencement of the trial the basis upon which it seeks a conviction, there will be cases in which the course of the evidence gives rise to the possibility of a further basis for conviction on the offence charged or the basis for a conviction on an alternative offence which is available at law. An important consideration for the trial judge in deciding whether to leave to the jury a path to conviction not previously mentioned is whether the accused has had an adequate opportunity to test evidence relevant to such reasoning, to call evidence relevant to it, and to address the jury on it.

21 A trial judge is not bound by the prosecutor's formulation of its case. It may be necessary for the judge to consider whether it is appropriate to leave alternative verdicts to the jury or direct on alternative paths to conviction irrespective of whether such alternatives are relied upon by the prosecution. However, it is essential when considering such a course to have regard to whether unfairness would result if, through no fault of the defence, the alternative basis had not been properly addressed during the trial.

22 The function of the trial judge and the approach to be adopted by an appellate court in these circumstances were summarised by Street CJ in R v Solomon [1980] 1 NSWLR 321 at 327:

"Ordinarily, it is the province of the Crown to formulate and present the case for the prosecution which will, in due course, be summed up by the judge to the jury. There may, however, be other matters of fact or law which the trial judge, in the discharge of his duty to ensure a fair trial according to law, considers it necessary to put to the jury, even though not propounded or developed by the Crown. The fairness or unfairness of travelling beyond the ground covered by the Crown will, of course, be evaluated by the trial judge and will be to the forefront in his deciding how far, if at all, to put new considerations to the jury. It seems to me that, where a judge does cover fresh ground in terms that are correct in law and properly based on evidence in the case, at the highest his decision so to do will only be challengeable if it can be seen that the accused person was thereby placed at a tactical disadvantage. In such a situation, appellate intervention would be appropriate, not by reason of the judge having canvassed fresh ground simpliciter, but by reason of unfairness attending his so doing. The relevant unfairness will ordinarily be looked for in procedural considerations. The judge, drawing upon his own forensic experience, will be readily appreciative of the tactical considerations which will have governed counsel in the conduct of the case for the accused. Objections to evidence, lines of cross-examination and decisions upon the material to be advanced on behalf of the accused, not to mention the general trend of the final address to the jury made on behalf of the accused, will all, of course, have been governed by the nature of the Crown case as opened by the prosecutor and developed through evidence tendered on behalf of the Crown at the trial. It is readily understandable that, within these procedural and tactical fields, there could arise an element of real prejudice, if the judge, in his summing-up, raises new approaches available to, but not expressly relied upon, by the Crown. But in every case in which a question arises regarding the development of new approaches, the question concerning the judge at first instance, and on appeal the question concerning this Court, will be to determine whether in so doing there will be worked an unfairness to the accused."

23 In Solomon's case a conviction for murder was set aside because of the unfairness which resulted from the trial judge leaving reckless indifference as a basis for conviction when that basis had not been relied upon by the prosecution at the trial. The issue was not raised by the trial judge with counsel prior to the summing-up.

24 In King v The Queen [1986] HCA 59; (1986) 161 CLR 423 two men were charged with the murder of the wife of one of them. The prosecution was conducted on the basis that the husband was an accessory before the fact because he had procured the co-accused to commit the murder. In his summing-up the trial judge directed the jury that the husband could be convicted if he procured a person other than the co-accused to commit the offence.

25 The conviction of the husband was set aside. It was held that the introduction of this basis for conviction for the first time in the summing-up had resulted in possible prejudice to the husband's case. Dawson J commented (432) that had the husband's counsel known that the matter would be left to the jury on this basis he may well have cross-examined or addressed in a different way.

26 Other cases in which convictions have been quashed by reason of alternative paths to conviction being raised with the jury in the summing-up without prior warning to the defence include R v GAS [1998] 3 VR 862 and R v Carr [2000] TASSC 183; [2000] 117 A Crim R 272.

27 In my view the appellant's case was clearly prejudiced by reason of the fact that the incident deposed to by Beuret was left to the jury as a possible basis for conviction in circumstances where the appellant had no warning that it would be so left. The incident deposed to by Beuret was vague and lacking in detail. There was an apparent discrepancy between the description given by the witness and the stage at which, according to the expert evidence, glucose would have been used in the process. In any event, Beuret had a purpose to be served in shifting blame to the appellant. There was no discussion of any of these matters in the short reference to the incident which appears in the summing-up. It is not unreasonable to suppose that these matters would have been the subject of cross-examination and comment by the appellant's counsel if it had been known that the incident would be left to the jury as a possible basis for a conviction. The combined effect of the matter being led to the jury without warning and the lack of adequate instruction as to the incident itself gave rise to a serious procedural irregularity (cf. R v Heaton NSW CCA BC 9002369 1 June 1999 per Gleeson CJ at 8). I will return to the consequences of this irregularity after discussing the other grounds of appeal.

28 The next ground of appeal arises out of a direction given to the jury in relation to alleged lies told by various witnesses. The trial judge said:

"I suggest there is no doubt that some of the defence witnesses, being the occupants of the house, have told you substantial lies about what occurred, or did not occur, in that Felixstow house. The differences in their stories were not merely matters of detail, so that the variations could all be accounted for by faulty but honest recollections of what occurred. You must also consider the possibility that they have all lied to you and no-one has told you the whole truth. It may be that some of the occupants of the house told you the truth about what others did in the house, but not about their own activities. It may be that some of them invented or exaggerated their evidence of what others did to deflect blame from themselves or their friends.

The prosecutor suggested in his closing address that if you find that any of the occupants of the house on 16 August had lied to you, you should be very careful in accepting anything else which that witness said to you whether it was in favour of or against the prosecution case. While it is entirely up to you what you accept or do not accept on the facts, on the evidence it may be well prudent for you not to accept anything else an established liar has said to you unless that evidence is supported by some other evidence which you do accept."

29 In the passage from the prosecutor's address referred to by the trial judge the prosecutor said:

"If you find that a witness has lied to you deliberately, then that will obviously affect the credibility of that witness generally and you will need to be very careful before you place too much reliance upon that witness's evidence.

For example, you might think that Mr Franco has clearly lied to you about aspects of what was taking place at that house."

After giving examples of alleged lies told by the appellant, the prosecutor said:

"It is a matter for you, but you might think that Mr Franco has clearly told you some lies in the course of his evidence.

That will be an important assessment to make, ladies and gentlemen, because if he had lied to you you will need to be very, very careful before you use anything that he said against Mr Beuret in assessing the case against Mr Beuret."

30 Counsel for the Director of Public Prosecutions conceded on appeal that the trial judge's remarks, particularly when taken in conjunction with the reference to the prosecutor's address, would be understood by the jury as encompassing a situation which might apply if the appellant told lies. It was conceded that it was open to the jury on this direction to conclude that, if they found that the appellant had lied about any of these matters, it would be appropriate for them not to accept anything else that the appellant said to them unless the evidence was supported by reliable evidence from another source.

31 Whereas a direction along these lines might be appropriate in certain cases when instructing the jury on the assessment of the evidence of defence witnesses, it is contrary to principle to place a qualification of this nature on the circumstances in which the appellant's evidence and that of the defence witnesses is to be considered. The jury were required to give effect to any aspect of the evidence which gave rise to a reasonable doubt as to the guilt of the appellant. Although there was no onus on the appellant, evidence giving rise to reasonable doubt could come from the appellant or witnesses called by him. It was inappropriate to restrict the evidence of the appellant and his witnesses available to be taken into account for this purpose to evidence which was supported by other evidence.

32 In my view this was a misdirection on a matter of fundamental importance to the assessment of the evidence and was capable of causing a miscarriage of justice.

33 A further complaint made on appeal arose out of the trial judge's directions on the topic of lies. It was not suggested by the prosecutor at the trial that any lies told by the appellant disclosed a consciousness of guilt. In all the circumstances, it was unnecessary and inappropriate for the jury to be given the type of direction referred to in Edwards v The Queen [1993] HCA 63; (1993) 178 CLR 193. However, it was argued that the trial judge should have directed the jury along the lines suggested in Zoneff v The Queen (2000) 200 CLR 234 at 245:

"You have heard a lot of questions, which attribute lies to the accused. You will make up your own mind about whether he was telling lies and if he was, whether he was doing so deliberately. It is for you to decide what significance those suggested lies have in relation to the issues in the case but I give you this warning: do not follow a process of reasoning to the effect that just because a person is shown to have told a lie about something, that is evidence of guilt."

34 Such a direction is appropriate in cases where there is a risk of misunderstanding the significance of possible lies. In the circumstances of this case, I do not regard the giving of such a direction as essential, although I think it would have been preferable if the trial judge had told the jury that the significance of possible lies by the appellant was restricted to the issue of credibility.

35 A further ground of appeal complains that the trial judge erred in not directing the jury on the dangers of failing to act on the evidence of Mr Beuret.

36 There is no requirement in South Australia to give a full corroboration warning in the case of an accomplice who is charged as an accused in the same prosecution and who gives evidence at the joint trial: R v Webb and Hay (1992) 59 SASR 563 at 581; [1994] HCA 30; (1994) 181 CLR 41 at 93.

37 The approach which has been adopted by this court and approved by the High Court in R v Webb and Hay is based on the following extract from R v Henning NSW CCA (unreported) 11 May 1990:

"But different principles apply when the supposed accomplice who gives evidence against a co-accused is himself an accused giving evidence in his own case. It would be difficult indeed to seek to apply inflexible rules to such situations. For the interests of justice will almost certainly require different responses in different circumstances. Considerable latitude must be allowed in order to enable trial judges to address the situation in a manner which will adapt to the competing interests in the particular case."

38 In my view there was no requirement in the present case to give a corroboration warning in relation to the evidence of Mr Beuret.

39 The next ground of appeal asserts that the prosecution case depended substantially upon circumstantial evidence and, as a consequence, the trial judge was required to direct the jury in accordance with the expanded direction on the onus and standard of proof referred to in authorities such as Peacock v R [1911] HCA 66; (1911) 13 CLR 619 and Plomp v R [1963] HCA 44; (1963) 110 CLR 234.

40 Furthermore, it was argued that the jury should have been directed in accordance with the detailed directions appropriate for circumstantial evidence cases discussed in Chamberlain v The Queen (No 2) [1984] HCA 7; (1984) 153 CLR 521 and Shepherd v The Queen [1990] HCA 56; (1990) 170 CLR 573.

41 The trial judge expressed the view when discussing the matter with counsel at the conclusion of his summing-up that this was not a case in which the major thrust of the prosecution case was based upon circumstantial evidence. I am unable to agree with that observation. It was rightly conceded on the hearing of the appeal that the prosecution case was based substantially on circumstantial evidence. The first limb of the prosecution case was dependent upon inferences to be drawn from the location of various implements and ingredients in the house, the finding of the list of names and financial references and the fact that the appellant had purchased iodine on previous occasions. The alternative prosecution case was based on circumstantial evidence from which the jury were asked to infer that the appellant, as occupier of the premises, must have been aware that drugs were being manufactured on the premises and that he condoned or permitted this activity.

42 In each of these instances it would have been appropriate to direct the jury that the primary facts were to be proved by the prosecution beyond reasonable doubt and, further, that to bring in a verdict of guilty it had to be established not only that guilt should be a rational inference, but that it was the only rational inference which the circumstances would allow.

43 It would also have been appropriate for the trial judge to direct the jury on the nature of circumstantial evidence generally as well as identifying in some detail the various items of circumstantial evidence relied upon by the prosecution together with the appellant's answers to or explanations for these items of evidence.

44 These matters were not canvassed in the summing-up to the extent to which the circumstances of the case required.

45 There is a further difficulty in relation to the directions on the alternative basis for conviction relied upon by the prosecution, namely, the claim that the appellant, as the occupier of the premises, knew that others were manufacturing methylamphetamine and that he permitted them to do so. The trial judge instructed the jury on the appellant's answer to this allegation in the following terms:

"Franco's evidence was if there were any steps being taken for the manufacture of meth in the house by the other occupants he was unaware of that. If you find that to be proved this third element would not be proved against Franco on this alternative basis. What, if any, weight that you give to his evidence that he did not know what the others were doing is for you to decide."

46 The direction conveys the notion that the appellant was required to prove his lack of knowledge before being entitled to an acquittal. However, this appears to have been a slip of the tongue by the trial judge and, in my view, the jury would not have overlooked the directions on onus and standard of proof which appear elsewhere in the summing-up.

47 Finally, it was argued that the trial judge erred in failing to direct the jury against propensity reasoning. Evidence of past sales of methylamphetamine by the appellant was led as part of the circumstantial evidence against him. It was not suggested at the hearing of the appeal that the evidence was inadmissible for this purpose. When directing the jury on the evidence the trial judge said:

"In this trial each accused is only charged with the offence of taking part in the manufacture of methylamphetamine. Neither is charged before you with any other offence. If you think either accused has committed any other drug offence, you must not convict him of the offence charged simply because you think he is guilty of some offence and you want to convict him of something, but only if the offence charged is actually proved to you. In particular, you have heard evidence which, if you accept it, would show that each accused had sold illegal drugs which, in itself, is a serious but different criminal offence. Each denied it.

However, the charge here is not selling drugs but taking part in the manufacture of a drug. Mere proof of selling a drug does not prove that the seller took part in the manufacture of it. You must not convict either accused of the offence charged if you were to find no more proved against him than that he had sold an illegal drug."

48 The trial judge went on to direct the jury on the use which could be made of this evidence.

49 It would have been preferable if the trial judge had made it clear that the jury could not reason from the evidence of alleged drug dealing that the appellant's character was such that he might commit the offence charged. However, the legitimate use to which the evidence could be put was made clear to the jury and, in all the circumstances, I do not think that the failure to give further directions of the type suggested would have contributed to a miscarriage of justice.

50 I have expressed the view that the raising of an alternative path to conviction for the first time in the summing-up resulted in unfairness to the appellant and that it gave rise to a serious procedural irregularity. I cannot exclude it as a possible cause of miscarriage of justice. I am also of the view that the directions on possible lies told by the appellant and his witnesses and the consequent effect that such evidence might have on the assessment of the defence case was also capable of causing miscarriage of justice.

51 For these reasons I would allow the appeal, set aside the conviction and order a new trial.

52 DEBELLE J. I have had the advantage of reading the reasons of Duggan J. I agree with them and with the orders he proposes.