High Court of Australia Fullagar, Kitto and Taylor JJ. Watts v Watts [1953] HCA 18 ORDER Appeal dismissed with costs. Cur. adv. vult. The following written judgments were delivered:— April 27 Fullagar J. I agree with the judgment of my brothers Kitto and Taylor, which I have had the advantage of reading. I wish to add only a few words. The question whether this or that expression should be used in an attempt to define the standard of proof required for a finding of adultery in divorce proceedings has never seemed to me to be a question of great practical importance, for the very simple reason that no responsible tribunal will deliberately declare that adultery has been committed in the absence of a real conviction, induced by evidence, that it has been committed. But, in my opinion, apart from such a special case as Preston-Jones v. Preston-Jones [4] , the standard of proof appropriate in divorce cases, where adultery is in issue, is not soundly or correctly described in the time-honoured formula which has been used for generations in charging juries in the criminal courts. The purpose of that formula is to bring home to juries the strength of the presumption of innocence in a criminal case, and to indicate to them that, if there is a reasonable hypothesis consistent with innocence, they should not return a verdict which will involve punishment and perhaps severe punishment. It is a traditional and sound formula in that jurisdiction, but to introduce it into the divorce jurisdiction would be contrary to long established practice in this country and to much English authority which is cited in Briginshaw v. Briginshaw [5] . For the rest, with all respect, I do not think that the general position with regard to standards of proof can be more accurately or more clearly stated than in the well known passage in the judgment of Dixon J. in Briginshaw v. Briginshaw [6] . With regard to the particular class of case with which we are now concerned, I agree with my brothers that the position can hardly be better put than in the words of Sir William Scott in Loveden v. Loveden [1] . 1. (1951) A.C. 391. 2. (1938) 60 C.L.R. 336, at pp. 363-366. 3. (1938) 60 C.L.R., at pp. 360 et seq. 4. (1810) 2 Hag. Con. 1, at p. 3 [161 E.R. 648, at pp. 648-649]. I conclude, then, that the central proposition of Mr. Joske's argument is sound, and that the learned trial judge in this case did misdirect himself with regard to the standard of proof to be applied by him to the evidence before him. I am nevertheless clearly of the opinion that we should not order a rehearing: still less should we ourselves pronounce a decree in favour of the petitioner. The doubts which his Honour expressed at the conclusion of his judgment are such that it appears plain to me that, even if he had approached the case without reference to a too stringent standard of proof, the result must have been the same. But there is more to be said than that. His Honour accepted certain evidence given by and for the respondent and co-respondent as to the events of the critical afternoon. He was, so far as we can tell, entirely justified in accepting that evidence; it was supported by the undoubted fact that the co-respondent was admitted to hospital on the following day. The evidence so accepted was not necessarily inconsistent with adultery having been committed during the afternoon. But it tended to support an innocent explanation of the presence of the respondent in the co-respondent's bedroom, and, if it was true, it became highly unsafe to infer from the general circumstances and from evidence by no means satisfactory in itself, that adultery had in fact taken place. Reading the evidence with the assistance of his Honour's view of much of it, one cannot escape the conclusion that the petition was rightly dismissed. It is possible that, as Mr. Joske urged, his Honour was mistaken as to the nature of the lock on the bedroom door, but I am quite unable to think that the absence of such a mistake could have made, or should have made, any difference. I think mention should be made of one other point made by Mr. Joske. Certain letters written by the co-respondent to the respondent were put in evidence. There was obvious prevarication by each of them when faced with these letters in the witness box. The learned judge found, as one would have expected, that the letters had been written as they appeared, and he went on to express the opinion that they afforded "the most cogent evidence of a guilty relationship between the respondent and the co-respondent". He added, however, that they did not prove that adultery had been committed during the critical afternoon at Hadley's Hotel. Founding himself on his Honour's opinion as to the evidentiary effect of the letters, Mr. Joske argued that they ought to have led his Honour to infer that adultery had occurred on that afternoon at Hadley's Hotel or had occurred at other times and places as charged in the supplemental petition. The argument is, of course, not without force, but I think that it breaks down on the threshold. For I am not able to agree with the learned trial judge that the letters afford cogent evidence of an adulterous association (assuming that that is what his Honour meant). They are such as to arouse a degree of suspicion, which the prevarication of the writer and the intended recipient does nothing to dispel. But it is impossible, in my opinion, to regard then as justifying a finding that an adulterous association existed or had ever existed. In my opinion the appeal should be dismissed. Kitto and Taylor JJ. This is an appeal from an order of the Supreme Court of Tasmania dismissing a petition and supplemental petition by the appellant for the dissolution of his marriage with his wife, the respondent, upon the ground of her adultery. The petitioner alleged that on 4th July 1950, at Hobart in Tasmania, the respondent committed adultery with the co-respondent, whilst the supplemental petition alleged other acts of adultery on a number of occasions between 6th October 1950 and 10th February 1951. The hearing of the suit occupied some twelve days and a considerable body of evidence relevant to the allegations referred to was put before the Court, but in the concluding stages of the trial counsel for the appellant conceded that his client could not succeed in the suit unless the Court found that adultery was committed as alleged in the petition on 4th July 1950. The learned trial judge took the view that the effect of this concession was to narrow the case down to the question whether adultery with the co-respondent took place on that occasion and it was to this particular issue that he thereafter devoted his attention. On this appeal counsel for the appellant contended that this concession was wrongly made and that the learned trial judge, having before him the whole of the evidence relevant to a consideration of all of the issues arising on the petition and the supplemental petition, should have addressed himself to a consideration and determination of all of these issues and that since he did not do so, this Court should itself determine them, or, at least, direct a new trial thereon. Whilst the Court is not precluded on this appeal from considering these issues it should not, in our opinion, accede to the appellant's submission. If it appeared that the concession referred to had been made in such circumstances that further consideration of the abandoned issues was necessary to do justice between the parties our view would be different, but it appears to have been made deliberately and for the very purpose of defining the issues for the final consideration of the learned trial judge. Moreover, upon a consideration of the whole of the evidence, we are of the opinion that the concession was not induced by any error on the part of counsel and, from the views expressed by the learned judge concerning the credibility of the witnesses concerned, it appears probable that if these issues had not, in effect, been withdrawn from his determination they would have been resolved adversely to the appellant. In reaching a conclusion on the issue whether adultery took place on 4th July 1950 his Honour said that, having considered the whole of the evidence, he was not satisfied beyond reasonable doubt that it took place. The italics are ours and serve to emphasise a substantial ground upon which his Honour's adverse finding on this issue was challenged by the appellant. In the early part of his reasons his Honour discussed the question of the degree or standard of proof required in suits for dissolution of marriage on the ground of adultery and concluded that in view of the decision in Preston-Jones v. Preston-Jones [1] , he should hold, contrary to Briginshaw v. Briginshaw [2] , that the standard of proof required is proof beyond reasonable doubt. That decision, was, he thought, such a decision as Dixon J. had in contemplation in Wright v. Wright [3] when he said, after referring to Ginesi v. Ginesi [4] ; "Some other decision, more particularly of the House of Lords, may make it necessary for us to reconsider Briginshaw v. Briginshaw [2] , but I do not think Ginesi v. Ginesi [4] does so" [5] . Since Preston-Jones v. Preston-Jones [1] was decided views similar to that of the trial judge in this suit have also been taken in the Supreme Courts of New South Wales and Queensland. Richardson J. considered there was a conflict between the decision in that case and Briginshaw v. Briginshaw [2] (see Stone v. Stone [6] ), whilst the same conclusion was reached by the Full Court of Queensland in Mackie v. Mackie [7] , and in the result the standard of proof beyond reasonable doubt was adopted, the latter Court holding that this is the standard to be applied "not only for adultery but for all matrimonial offences". In these circumstances it is necessary for us to examine the reasons in Preston-Jones v. Preston-Jones [1] for the purpose of seeing exactly what that case did decide. 1. (1951) A.C. 391. 2. (1938) 60 C.L.R. 336. 3. (1948) 77 C.L.R. 191. 4. (1948) P. 179. 5. (1938) 60 C.L.R. 336. 6. (1948) P. 179. 7. (1948) 77 C.L.R., at p. 211. 8. (1951) A.C. 391. 9. (1938) 60 C.L.R. 336. 10. (1952) 69 W.N. (N.S.W.) 275. 11. (1952) Q.S.R. 25. 12. (1951) A.C. 391. In the first place, it should be observed that, in the circumstances of that case, a finding that the adultery alleged had taken place inevitably operated to bastardize the respondent's child. There is no doubt that in such cases a very high degree of proof is required and this requirement is undoubtedly reflected in their Lordships' reasons. A similar position arose in Gaskill v. Gaskill [1] where Lord Birkenhead said that he could find the respondent guilty only if he came to the conclusion that it was impossible, having regard to the then present state of medical knowledge and belief, that the petitioner could be the father of the child. In Preston-Jones v. Preston-Jones [2] Lord MacDermott, who dealt at some length with the standard of proof required in that case, expressed the view that Lord Birkenhead did not intend to specify a standard of proof beyond that calling for proof beyond reasonable doubt. He said: "Lord Birkenhead had previously indicated [3] his unqualified acceptance of the views expressed by Lord Lyndhurst, L.C., in Morris v. Davies [4] , and I do not find in that case or in the Banbury Peerage case [5] , also referred to by Lord Birkenhead, warrant for Mr. Middleton's contention. The evidence must, no doubt, be clear and satisfactory, beyond a mere balance of probabilities, and conclusive in the sense that it will satisfy what Lord Stowell, when Sir William Scott, described in Loveden v. Loveden [6] as "the guarded discretion of a reasonable and just man" " [7] . Dixon J., in Briginshaw v. Briginshaw [8] , referred to Lord Birkenhead's observations in Gaskill v. Gaskill [1] and said that the latter had applied to matrimonial causes the rule relating to legitimacy, namely, "that to bastardize a child conceived and born during wedlock it is not enough to establish a mere preponderence of probability in favour of the inference that the husband did not beget the child; the presumption of legitimacy is not rebutted unless the proof excludes all reasonable doubt" [9] . Dixon J. proceeded: "The use of the phraseology of the criminal jurisdiction is due to Lord Lyndhurst in Morris v. Davies [10] , a case the course of which is fully examined by Cussen J. in In the Estate of L. [11] . Cussen J. concludes his consideration of the legitimacy rule by saying:—"The expression "beyond reasonable doubt" recalls the ordinary direction in criminal cases that it is necessary that the jury should be satisfied of guilt beyond reasonable doubt before they disregard the primary presumption of innocence. It may be that the origin of the rules in cases like the present is that adultery was, and to a certain extent still is, regarded as an offence, and is not to be imputed on a mere balance of probabilities as in an ordinary civil case [1] . This does not appear to me necessarily to imply that his Honour considered that always and for all purposes adultery must be established beyond reasonable doubt" " [2] . 1. (1921) P. 425. 2. (1951) A.C. 391. 3. (1921) P., at p. 434. 4. (1837) 5 Cl. & F. 163 [7 E.R. 365]. 5. (1823) 1 Sim. & St. 153 [57 E.R. 62]. 6. (1810) 2 Hag. Con. 1, at p. 3 [161 E.R. 648, at p. 649]. 7. (1951) A.C., at pp. 416-417. 8. (1938) 60 C.L.R. 336. 9. (1921) P. 425. 10. (1938) 60 C.L.R., at p. 367. 11. (1837) 5 Cl. & F., at p. 215 [7 E.R., at p. 385]. 12. (1919) V.L.R. 17, at p. 30. 13. (1919) V.L.R., at p. 36. 14. (1938) 60 C.L.R., at p. 367. A further reference to the origin of the "legitimacy rule" is to be found in Piggott v. Piggott [3] which was decided only six months after Briginshaw v. Briginshaw [4] . There Dixon J. said that the husband was bound to "adduce evidence which produces a clear and satisfactory inference contrary to the primary presumption. It will not be sufficient unless it produces a moral or judicial conviction, so that the tribunal is satisfied beyond reasonable doubt" [5] . "This standard of persuasion, exceptional in civil matters, was" his Honour continued, "established in Morris v. Davies [6] " [5] . But the exceptional requirement that in such cases proof beyond reasonable doubt is required affords no ground for asserting that the standard of proof in other suits for dissolution of marriage should be denoted by the expression "proof beyond reasonable doubt". There is no doubt that in the particular circumstances of Preston-Jones v. Preston-Jones [7] their Lordships did apply this standard of proof, but in so doing, it seems to us, they assumed that this was the appropriate standard of proof in the circumstances of the case before them and did not, as a final appeal tribunal, consider or decide that this standard of proof was applicable in all suits for dissolution on the ground of adultery or for any other matrimonial offence. Lord Morton made it quite clear that their Lordships were not primarily concerned with the problem which arose in Briginshaw v. Briginshaw [4] when he said: "In Ginesi v. Ginesi [8] the Court of Appeal, after a survey of the authorities, held that a petitioner must prove adultery "beyond reasonable doubt". In my view the burden of proof is certainly no heavier than this, and counsel for the appellant did not contend that it was any lighter. Assuming that this burden lay on the appellant, did he discharge it?" [9] . And the statement by Lord Simonds that the question, as he saw it, was "whether the court ought to accept this evidence as adequate to justify a finding that beyond all reasonable doubt the child was not the child of the husband" [1] rather indicates that his Lordship did not intend, at least originally, to express any final opinion as to the standard of proof required generally in suits for dissolution of marriage. His Lordship had already emphasised that the result of a finding of adultery in cases such as that then before their Lordships would be, in effect, to bastardize the child. It is true, however, that Lord Simonds added that since preparing his opinion he had had the advantage of reading that of Lord MacDermott, and we understand his Lordship to say that he concurred in the views expressed by Lord MacDermott on the standard of proof required in such suits. Lord Normand's statement that the commissioner: "if he had proceeded on the evidence and not upon his own speculations of what may be possible, and if he had not misdirected himself that he should not be satisfied without the evidence of a scientist, could not reasonably have arrived at any other conclusion than that the child was not the child of the appellant" [2] , does not suggest any general rule as to the standard of proof required in suits for dissolution of marriage and so far as we can see there is no other passage in his Lordship's opinion which throws any light on this problem. Lord Oaksey expressly confined his observations to the type of case before their Lordships. He said: "The respondent in this appeal has twice been successful in the courts of first instance and now for the first time she is by your Lordships' House to be found guilty of adultery and her child is in effect to be bastardized." 1. (1938) 61 C.L.R. 378. 2. (1938) 60 C.L.R. 336. 3. (1938) 61 C.L.R., at p. 415. 4. (1837) 5 Cl. & F. 163 [7 E.R. 365]. 5. (1938) 61 C.L.R., at p. 415. 6. (1951) A.C. 391. 7. (1938) 60 C.L.R. 336. 8. (1948) P. 179. 9. (1951) A.C., at p. 412. 10. (1951) A.C., at p. 403. 11. (1951) A.C., at p. 409. "She is a wife against whose marital conduct nothing has been said or suggested except that her child was born 360 days after her husband had access to her. In such circumstances the law, as I understand it, has always been that the onus upon the husband in a divorce petition for adultery is as heavy as the onus which rests upon the prosecution in criminal cases. That onus is generally described as being a duty to prove guilt beyond reasonable doubt but what is reasonable doubt is always difficult to decide and varies in practice according to the nature of the case and the punishment which may be awarded. The principle on which this rule of proof depends is that it is better that many criminals should be acquitted than that one innocent person should be convicted. But the onus in such a case as the present is not founded solely on such considerations but upon the interest of the child and the interest of the State in matters of legitimacy, since the decision involves not only the wife's chastity and status but in effect the legitimacy of her child" [1] . His Lordship thought that to hold that the evidence in the case discharged "the onus which rests upon the prosecution in criminal cases or upon any party who seeks to prove illegitimacy is, in my respectful opinion, to depart altogether from the practice in such cases" [2] . It seems to us that Lord MacDermott alone proceeded expressly to deal with the standard of proof required to establish, in suits for dissolution of marriage, the commission of some matrimonial offence and there is no doubt that his Lordship took the view that the standard required is proof beyond reasonable doubt. But it is equally true that his Lordship considered that, upon its true construction, there is no real distinction between the expressions "satisfied" and "satisfied beyond reasonable doubt". These brief references to the views expressed by their Lordships do not indicate to us any necessity for reconsideration of the considered decision of the High Court in Briginshaw v. Briginshaw [3] . 1. (1951) A.C., at p. 409. 2. (1951) A.C., at p. 411. 3. (1938) 60 C.L.R. 336. We agree with respect with the observation of Lord MacDermott that the evidence in such cases, "must, no doubt, be clear and satisfactory, beyond a mere balance of probabilities, and conclusive in the sense that it will satisfy "the guarded discretion of a reasonable and just man" " [4] . But in saying this, we do not intend to agree that such a degree of proof is required in such cases as is indicated in the criminal jurisdiction by the expression "proof beyond reasonable doubt". The statute requires that the Court, before making a decree, shall be satisfied on the evidence that the case of the petitioner has been proved, and we are content to conclude that "satisfied" means satisfied having regard to the gravity of the issues involved. This aspect of the matter is fully discussed in Briginshaw v. Briginshaw [3] and we do not wish to add anything to what was then said. If "satisfied beyond reasonable doubt" means no more than this, then to substitute that expression, imported from the criminal law, for the words "satisfied on the evidence" would, at the most, as Denning L.J. said in Davis v. Davis [5] , afford a "temptation to give effect to shadowy or fanciful doubts", whilst if it means something different from "satisfied on the evidence" there is no warrant for any such substitution and consequential amendment of the statute. We should, perhaps, add that we respectfully agree with Lord MacDermott's observation that the matter cannot be better put than it was in Loveden v. Loveden [6] . 1. (1951) A.C., at p. 417. 2. (1938) 60 C.L.R. 336. 3. (1950) P. 125, at p. 129. 4. (1810) 2 Hag. Con. 1 [161 E.R. 648]. But although the learned trial judge was in error in adopting the standard of proof which he did, the appeal should, in our opinion, be dismissed. It is quite clear from the learned trial judge's reasons that he was left in such a state of doubt that he could not have been satisfied on any test that adultery took place on 4th July 1950, and this would be sufficient to dispose of the appeal. Such a conclusion would, in our opinion, have been most reasonable; indeed, the evidence tends to establish that probably it did not take place. For the reasons given we are of the opinion that the appeal should be dismissed.