UK Social Security and Child Support Commissioners' Decisions

You are here: BAILII >> Databases >> UK Social Security and Child Support Commissioners' Decisions >> Chief Adjudication Officer v. Bath [1999] UKSSCSC CG_11331_1995 (21 October 1999)
URL: http://www.bailii.org/uk/cases/UKSSCSC/1999/CG_11331_1995.html
Cite as: [1999] UKSSCSC CG_11331_1995

[New search] [Printable RTF version] [Help]

Chief Adjudication Officer v. Bath [1999] UKSSCSC CG_11331_1995 (21 October 1999)

R(G) 1/00

(Chief Adjudication Officer v. Bath)

Mr. M. J. Goodman CG/11331/1995
7.5.98

CA (Evans, Schiemann and Walker LJJ)
21.10.99

Validity of marriage - marriage ceremony not carried out in accordance with the statutory requirements - whether presumption of marriage from long cohabitation

The claimant and her husband went through a formal ceremony of marriage at a Sikh Temple in West London on 27 June 1956. There had been no other marriage ceremony and there was no marriage certificate. The claimant and her husband thereafter lived together in a monogamous relationship bringing up their children, until the husband's death in 1994. The Temple was registered for marriages in 1983. The claimant's claim for widow's benefit was refused on the ground that she had not been validly married as the Temple had not been registered at the time of her marriage. The tribunal upheld that decision. The claimant appealed to a Commissioner who held that, as there had been a bona fide ceremony of marriage followed by a long monogamous cohabitation, marriage should be presumed. The Chief Adjudication Officer appealed to the Court of Appeal.

Held, dismissing the appeal, that:

there was no statutory provision in the Marriage Act 1949 or elsewhere expressly rendering void a marriage in an unregistered building and it would be wrong not to extend to the parties to an irregular ceremony followed by long cohabitation the benefit of a presumption which would have applied if there had been no evidence of any ceremony at all;

R(G) 2/70 was distinguishable on its facts but (per Evans LJ) in so far as it decided that a marriage was void by reason of non-compliance with the relevant statutory requirements, even in the absence of express statutory provision to that effect, then it was not correct.

DECISION OF THE SOCIAL SECURITY COMMISSIONER

I allow the claimant's appeal from the decision of the social security appeal tribunal dated 3 May 1995 as that decision is erroneous in law and I set it aside. My decision is that, for the purpose of the claim made on 7 February 1994 for widow's benefit, it has been established that the claimant is "a woman who has been widowed" within the meaning of section 38(1) of the Social Security Contributions and Benefits Act 1992. Accordingly, provided the other requirements (e.g. contribution conditions) of that section are satisfied, the claimant is entitled to widow's pension in respect of the death of her husband on 23 January 1994 and the adjudication officer should forthwith make the appropriate award: Social Security Administration Act 1992, section 23.

This is an appeal to the Commissioner by the claimant, a woman born on 4 August 1939, from the unanimous decision of a social security appeal tribunal dated 3 May 1995 which dismissed her appeal from a decision of an adjudication officer issued on 12 August 1994, the relevant terms of which were, "It has not been established and cannot be presumed that there is a valid marriage between [the claimant] and [ZSB]. This is because there is no evidence of a valid ceremony in accordance with the Marriage Act 1949. Therefore it is not valid for social security purposes."

This case has, because of its complexity, been the subject of three oral hearings before me. At the conclusion of the first two hearings, on 28 August 1997 and 28 October 1997 respectively, I issued Directions requiring further written submissions and information. The last hearing, on 30 April 1998, was a joint hearing of this and another case (on Commissioners file CG/6957/1995) involving the same point, on which however a decision was deferred because in that case (unlike this one) there was a question whether or not there had been a marriage in India as well as any marriage in England. At the hearing on 30 April 1998, the claimant was present with her son but was not otherwise represented. On a previous occasion she had been represented by her daughter, who is a Solicitor. Detailed written representations have been made by her children on her behalf. I am grateful for them. At the hearing on 30 April 1998 Mr. Gurdial Singh, of Counsel, appeared for the claimant in the other case. He addressed me on the general issues which are common to both cases. I am much indebted to him for his researches and for his submissions. At all three hearings the adjudication officer was represented by Mr. Sriskandarajah, of the Office of the Solicitor to the Departments of Health and Social Security. I am also much indebted to him for his researches and submissions.

The claimant's husband died on 23 January 1994. A copy of his death certificate has been made part of the appeal papers in this case (No T23). Shortly thereafter, on 7 February 1994, the claimant made a claim for widow's benefit, stating that she had married her husband on 27 June 1956. The ceremony on which she relies as being a marriage ceremony in England (there being no marriage in any other country) took place on 27 June 1956 at a Sikh Temple in West London. A letter from the General Register Office for England and Wales dated 25 June 1997 indicates that that Temple was not registered for marriages until 28 September 1983.

Registration of buildings for marriages in England is provided for by the Marriage Act 1949 (which is still the principal Act), sections 41-52. It is clear from those sections that registration of a building for marriage purposes can operate only as from the date of registration and is not retrospective.

The position is therefore that when the claimant went through a formal ceremony of marriage according to Sikh religious rites (with witnesses, members of the family, etc, present) in the Temple in question on 27 June 1956, that Temple was not a building registered for marriages. A letter dated 19 September 1997 to Mr. Sriskandarajah from the Vice President and Authorised Person of the Temple in question states,

"It is a common knowledge that religious marriages are being performed in the Sikh Gurdwaras [i.e. Temples] all over the world. Although such marriages called "Anand Karj" are recognised in India but [sic] in U.K. these are treated as only blessings in the place of worship and I believe that the people undergoing such marriages are told about this."

In response to that, a written submission on the claimant's behalf states,

"I would also like to bring to the Commissioner's attention that the last paragraph of the letter from the Central Gurdwara, .. cannot be construed to reach [a conclusion adverse to the claimant]. A Sikh marriage, performed according to religious customs and rites, for any Sikh, is a marriage in itself and not merely blessings."

Whatever is the correct position, no doubt it would be possible after such a ceremony in an unregistered Sikh Temple for the parties to have gone thereafter to e.g. a Registrar's Office for a civil wedding ceremony. That however did not occur in the present case. It should also be noted that there is no certificate of marriage in this case. It has also been confirmed that there appears to be no Act of Parliament, Public or Private, relating to marriages in Sikh Temples.

After the ceremony in the Sikh Temple on 27 June 1956 the claimant and her "husband" lived together in what was clearly a "monogamous" relationship for some 37 years, bringing up the children of the "marriage". The husband's income tax and social security contributions were assessed on the basis that he was a married man.

The adjudication officer's written submission to the tribunal appears to have been on the basis that the ceremony at the Sikh Temple on 27 June 1956 was not a valid marriage (a) because the building was not registered for marriages and (b) because in any event the ceremony was not apt to create a marriage. Point (b) has not been pursued at the stage of appeal to the Commissioner in either of these two cases. I consider that to be a correct course of action. The real issue is whether or not what was undoubtedly according to Sikh religious rites a marriage ceremony was invalidated by its taking place in a building, the Temple, which was not registered for marriages at the time, though it became registered subsequently. I have dealt with the case on that basis. In any event, in so far as there is any question mark over the actual nature of the ceremony, that is in my view answered by the presumption from cohabitation to which I refer below.

The tribunal of 3 May 1995, while saying that they had "enormous sympathy with the appellant's predicament", gave as their reasons for dismissing the appeal, "Unfortunately, at the time of her marriage ceremony the Sikh Temple was not registered for performing marriages, nor had the marriage been registered in a Registry Office. As a result there has not been a valid ceremony in accordance with the Marriage Act 1949, and this meant that the appellant is not entitled to widow's pension." It is only fair to the tribunal to say that the issues that have subsequently been gone into at the stage of the appeal of the Commissioner were not all before the tribunal. I have held their decision to be erroneous in law, not because of any deficiency in findings of fact and reasons for decision, but because I have come to the conclusion that the fact that the Sikh Temple was not at the relevant date in 1956 a registered building is not fatal to the claim by the claimant to be the widow of Mr. ZSB because there operates in her favour the presumption of marriage from long cohabitation, to which I refer below.

Before I deal with that presumption, I ought perhaps to note that reference has been made by the parties to the provisions of section 49 of the Marriage Act 1949 ("Void marriages") paragraphs (a), (b) and (e) of which provide that if a person "knowingly and wilfully" intermarries either without notice being given to the Registrar or without the notified Registrar issuing a certificate or in a building other than that specified in the notice to the Registrar, the marriage shall be "void". However, I do not consider that (contrary to Mr. Singh's submissions) the undoubted fact that the claimant here had no "mens rea" , i.e. did not "knowingly and wilfully" violate marriage law, means that the marriage is automatically valid. Section 49 is not an enabling but a penalising section. The scheme of the 1949 Act is to classify marriages into certain categories, of which the present one is Part III of the Act, i.e. "Marriage under Superintendent Registrar's Certificate." There was no such marriage in 1956 in the ceremony in the Sikh Temple because the Registrar was never notified. Section 49 is directed to infringements of the detailed rules concerned with marriages under Part III of the Act not with marriages altogether outside the preview of Part III of the Act. Nevertheless, the fact that "mens rea" is needed for a marriage to be void in those circumstances is an indication of a theme which undoubtedly runs throughout the legislation and the case law. That there is that it is not only in the interests of the parties but is also in the public interest that wherever possible bona fide ceremonies of marriage should be upheld. That is particularly so in my view in a case of this kind where the claimant was one of the first Sikh immigrants to this country. She bona fide married according to the customs and rites of her own religion. That deserves the greatest possible recognition in this country.

During the course of these proceedings, I requested submissions from the parties as to whether there was, ".. any room for the operation in this case of the presumption of marriage from cohabitation after ceremony (see Halsbury's Laws of England, Fourth Edition, Vo1. 22, para. 993)?" The presumption of marriage can in fact arise, not only under the case law summarised in paragraph 993 of Halsbury's Laws, but also paragraph 992. I now set out both those paragraphs (though not the footnotes referring to the case-law on which they are based). They read,

"Presumption of marriage

Presumption from cohabitation without ceremony

Where a man and a woman have cohabited for such a length of time and in such circumstances as to have acquired the reputation of being man and wife, a lawful marriage between them will be presumed, even if there is no positive evidence of any marriage ceremony having taken place, and the presumption can be rebutted only by strong and weighty evidence to the contrary.

Presumption from cohabitation after ceremony.

Where there is evidence of a ceremony of marriage having been gone through, followed by the cohabitation of the parties, everything necessary for the validity of the marriage will be presumed in the absence of decisive evidence to the contrary, even though it may be necessary to presume the grant of a special licence, or the death of a former spouse."

Prior to the hearing before me on 30 April 1998, submissions had been solely on paragraph 993, "Presumption from cohabitation after ceremony". However, at the hearing on 30 April 1998 the matter was widened because rightly in my view Mr. Gurdial Singh submitted that paragraph 992 (presumption where no ceremony) was also relevant. I consider both of those paragraphs now.

Paragraphs 992 and 993 are both concerned with a single common law presumption from cohabitation of a relevant length of time and other circumstances. (There is no doubt here that the relevant factual circumstances are present - 37 years of "monogamy" with children of the union). After an examination of the case law which is cited by Halsbury to substantiate these two paragraphs, it appears to me that the presumption has some elements of a presumption of law rather than just a presumption of fact. In some ways it is akin to the position in Scotland where a valid marriage may be proved by "cohabitation with habit and repute" (see R(G) 1/71; R(G) 4/84; and R(G) 5/83) . The adjudication officer, however, submits that there is no room for the presumption in the circumstances of paragraph 993, i.e. "Presumption from cohabitation after ceremony", because the presumption is rebutted by the undoubted fact that the "marriage" in 1956 in the Sikh Temple took place in an unregistered building. However, that fact does not stop the presumption from otherwise arising. It could possibly for example be presumed that there had been some kind of civil ceremony afterwards although there is no evidence of this.

The extensive case-law cited by Halsbury to support paragraph 993 assists and in particular the High Court's decision in Re Shephard, George v. Thyer [1904] 1Ch. 456. In that case it was held that the presumption of marriage from long cohabitation applied even where it was known and admitted that the only marriage ceremony was a purported marriage in France which was invalid according to the laws of France. That decision seems to me to be completely in point and, although technically not binding on me, has of course high persuasive force. Nevertheless, I note that in paragraph 11 of R(G) 2/70, a decision of Sir Rawden Temple a former Chief Commissioner and an expert on matrimonial law, it is said,

"11. It would not, however, be right to part with this aspect of the case without referring to Re Shephard, George v. Thyer [1904] 1Ch.456. There the parties gave evidence that the only ceremony of marriage through which they went took place in France. The case was argued and decided on the basis, accepted by the learned judge, that expert evidence showed that the ceremony could not have been valid, and it was submitted that if the parties were not married by that ceremony they were not married at all. Nevertheless in the parties' long cohabitation after the ceremony, a marriage was presumed. I regret that I do not find any great assistance in the case. It is clearly unsupported by Sastry Velaider Aronegary v. Sembecutty Vaigalie (1881) 6 App. Cas. 364 which it purports to follow, and where the decision depended upon a ceremony of marriage, prima facie valid. If the only marriage ceremony which the parties claimed to have gone through is invalid, the whole force of any presumption in favour of their marriage is destroyed and it is impossible to treat an invalid marriage as presumptive evidence of a valid marriage. An explanation of Re Shephard to be found in the books, is that the learned judge must have rejected the evidence of the parties as to what occurred in France. But the claimant's evidence in this case is quite clear, and was accepted, and I feel unable (with respect to the learned judge) to follow Re Shephard, which was undoubtedly decided as a hard case. And hard cases are, I think, to be cautiously regarded. Where there is evidence that the marriage asserted can only derive from the particular ceremony having validly taken place, it seems to me that the formal validity of that ceremony is unaffected by the fact of cohabitation thereafter. If such cohabitation is relevant, it may have reference to the possibility that the parties thought they were married, and validly married. But it has no bearing on the validity."

I naturally pay the greatest of respect to those statements by the learned Commissioner and I bear in mind the system of precedent of Commissioners' decisions, enjoined by a Tribunal of Commissioners in R(I) 12/75, paragraphs 21 et seq. Nevertheless, in the present case, I consider R(G) 2/70 distinguishable. In that case the claimant was relying on an alleged ceremony of marriage in an English registry office and not, as here, on a marriage according to different religious rites in a Temple of that religion. Moreover Re Shephard is, so far as I am concerned, a High Court decision, in pari materia, which I should follow unless there is any good reason why I should not. I have carefully read the entirety of the Law Report of Re Shephard. It seems to me that it carries the proposition for which it is cited and that I should follow it. There does not have been drawn to the attention of the learned Commissioner who decided R(G) 2/70 the possibility that, where a ceremony that the parties have relied on is shown to have been invalid, one ought then also to consider the authorities summarised in paragraph 992 of Halsbury's Laws i.e. "Presumption from cohabitation without ceremony" (my underlining).

I now therefore turn to the point (strongly made by Mr. Singh) as to the application of paragraph 992, "Presumption from cohabitation without ceremony". It should be remarked that, if a person is able to have a marriage presumed where there has been no ceremony at all (paragraph 992), it would hardly be equitable that a person who has gone through a bona fide ceremony thought to be a marriage (and entered into long cohabitation on the strength of it) should be in a worse position. The legal position must in my view be this. First one asks whether or not a particular ceremony can be presumed to be have been a valid marriage. If it is decided that it cannot be so presumed, then one treats the case thereafter as if there has been no ceremony at all and looks at the remainder of the facts to see whether there can be presumption from cohabitation without ceremony. If there has been a long "monogamous" cohabitation with nothing to cast any doubt on the facts of the situation (i.e. the strong and weighty evidence to the contrary to which Halsbury refers), then the fact that there may have been an invalid ceremony in the first place does not in my view prevent there being a presumption of marriage from the facts themselves. I note for example that Halsbury cites in the notes to paragraph 992 the case of Re Green, Noyes v. Pitkin (1909) 25 TLR 222 for the proposition that the presumption from cohabitation without ceremony was applied in a case where there had apparently been a foreign "marriage". It would be erroneous in my view to treat the two situations referred to in paragraphs 992 and 993 of Halsbury as if they were mutually exclusive. They are but classifications of one overall common law presumption, which in my view should undoubtedly on the facts of this case be applied. For that reason I have held the claimant's claim to be the widow of the deceased to be justified.

Lastly I should comment that all of the arguments in this case have been based on the undoubted position that, when asking whether or not a person is a "widow" under social security legislation, it has to be shown that there had been a valid marriage according to English law if the marriage took place in England i.e. the lex loci celebrationis must be observed. That "lex" includes of course the common law presumption of marriage from long cohabitation. That there must be a valid English marriage, according to the lex loci, is a proposition which was recently affirmed by the Court of Appeal in Fuljan Bibi v. Chief Adjudication Officer, 25 June 1997 in the context of polygamous marriages. There was quoted at paragraphs D and E of page 6 of the transcript of that decision the well known definition of marriage given by Lord Penzance in Hyde v. Hyde and Woodmansee [1886] LR1 P and D 130 at 133, defining marriage as follows,

"I conceive that marriage, as understood in Christendom, may for this purpose be defined as the voluntary union for life of one man and one woman, to the exclusion of all others."

In the present case there is nothing to suggest that the Sikh marriage ceremony and the consequences thereof in the eyes of the Sikh religious authorities was other than such a voluntary union for life of one man and one woman to the exclusion of all others. In my view that "marriage" is validated by the common law presumption from long cohabitation, in pursuance of the policy of the law that, in the absence of the clearest possible reason why there should not be such a presumption, a ceremony of "marriage" bona fide entered by parties who thereafter who live monogamously and bring up children of the union should be respected and accorded the proper legal status of marriage.

Date: 7 May 1998 (signed) Mr. M. J. Goodman

Commissioner

The Chief Adjudication Officer appealed to the Court of Appeal. The decision of the Court of Appeal follows.

DECISION OF THE COURT OF APPEAL

Mr. J. R. McManus QC (Instructed by the Department of Social Security) for the Appellant.

The Respondent did not appear and was not represented.

LORD JUSTICE EVANS:

The Respondent to this appeal, Kirpal Kaur Bath, did not appear before us. She had succeeded in her appeal to the Social Security Commissioner, where she was assisted by counsel, and we were not told why she was not also represented in this Court.

The reason may well be lack of funds. She is a lady now aged 59 who in 1956 at the age of 16 went through a Sikh marriage ceremony with Zora Singh Bath, who was then aged 19 and was also a Sikh. They lived together as man and wife for 37 years until he died on 23 January 1994. They had two sons, born in 1963 and 1965. He built up a successful cash and carry business and in later years, as her sons grew up, she helped him in the business. He paid income tax and social security contributions on the basis, which was never queried, that he was a married man. When he died, she applied for the pension payable to a widow under section 38 of the Social Security Contributions and Benefits Act 1992. That was on 3 February 1994.

Her application was refused and the refusal was belatedly justified on the ground that she was not a widow, because there was "no evidence of a valid [marriage] ceremony in accordance with the Marriage Act 1949" (DSS legal advice 5 August 1994). She appealed to the SSAT where the adjudication officer said that he had been advised that the ceremony could not be accepted as a valid marriage. He continued "It has been established that the ceremony did not take place in a registered building, and it was also established that the ceremony had not been registered in a Registry Office". The Appeal Tribunal found that :-

"The appellant went through a Sikh marriage ceremony at a Sikh temple which was not registered at the material time. The marriage was not registered in a Registry Office".

Its decision was:

"Appeal disallowed. It has not been established, and cannot be presumed that there is a valid marriage between Kirpal Kaur Bath and Zora Singh Bath. This is because there is no evidence of a valid ceremony in accordance with the Marriage Act 1949, and therefore it is not valid for Social Security purposes. As a result, Widow's Pension is not payable on the contributions of Zorah Singh Bath because it has not been proved that she is his widow".

The reasons given were:

"The Tribunal had enormous sympathy with the appellant's predicament. Unfortunately, at the time of her marriage ceremony the Sikh temple was not registered for performing marriages, nor had the marriage been registered in a Registry Office. As a result there had not been a valid ceremony in accordance

with the Marriage act 1949, and this meant that the appellant is not entitled to widow's pension"

That decision was dated 24 May 1995. On 1 November 1995 she applied for leave to appeal out of time to a Social Security Commissioner. Her application was considered sympathetically and she was given leave on 15 December 1995. The Commissioner's decision is dated 7 May 1998. Part of the delay was caused by the department's failure to comply with its obligation to submit observations to the Commissioner. They were due by 14 March 1996 but not received until August or September (though bearing the date 19 March) because of "the need to seek legal advice from the department's solicitors" (letter seeking an extension of time dated 14 March 1996). The reasons given for asking the Commissioner to dismiss the appeal were that the marriage ceremony had taken place in England but in a form not known to English law. Even between persons domiciled in India, and even if the Sikh temple had been registered as a place where marriage ceremonies could take place, the marriage could not be valid.

The Commissioner himself asked whether there was any room for the operation in this case of the presumption of marriage from cohabitation after the ceremony, referring both parties to Halsbury's Laws of England (4^th ed.) Vol. 22 para. 993. That was on 23 May 1997. The adjudication officer responded on 18 July 1997 referring, for the first time, to Commissioner's Decision R(G) 2/70 a case decided by Sir Rawden Temple on 6 February 1970. This decision, it was suggested, "stands against the operation of the principles" stated in Halsbury.

The Commissioner allowed the appeal. He held:

"In my view that 'marriage' is validated by the common law presumption from long cohabitation, in pursuance of the policy of the law that, in the absence of the clearest possible reason why there should not be such a presumption, a ceremony of 'marriage' bona fide entered by parties who thereafter who live monogamously and bring up children of the union should be respected and accorded the proper legal status of marriage".

Notwithstanding this decision in her favour, dated 7 May 1998, Mrs. Bath has not been paid the pension to which the Commissioner held she is entitled, because, we were told, the department, now the Benefits Agency, exercised its power to withhold payment pending the present appeal. Nearly six years after her claim was first made, she has not received a penny.

The appellant was represented by Mr. J.R. McManus QC for whose assistance we are grateful. Rarely can it have been necessary for counsel, even one as skilled and experienced as Mr. McManus is, to seek to justify such an unattractive case. For 37 years two departments of government, the Inland Revenue and the Department of Social Security, treated Mr. Bath as a married man and claimed taxes and contributions accordingly. When Mrs. Bath after his death claimed the pension to which his widow is entitled, for the first time, the Agency said that he was never married and that their children are illegitimate. Mr. McManus submits that it is bound to refuse payment, even if the clearest possible estoppel would arise, were the defendant a private citizen, because as a public authority it has no power to make any payment which is not permitted by the statute.

We were also told that the Agency is placed in a situation of real difficulty by the Commissioner's decision, conflicting as it does or seems to do with the decision in R(G)2/70 nearly thirty years ago. Adjudication officers therefore need guidance from the Court, we were told, and other appeals are pending. Nevertheless, no attempt was made to have those other appeals heard at the same time as this one, even when it became known that the respondent would not be represented. This appeal could have been adjourned, perhaps so that an amicus could be instructed, but that would have overlooked the urgent need for the respondent to know where she stands and not to be any longer deprived of the pension to which she is entitled, if the Commissioner's decision is correct.

Even if adjudication officers are now aware of a conflict between the two Commissioners' decisions and they have need of the Court's guidance accordingly, it does not appear from the documents before us that these doubts played any part in the decision to refuse Mrs. Bath's application. Initially, the department confused Mr. Bath with his cousin, who had emigrated to Canada, and the internal memorandum begins "Although Kirpal's marriage to Zora Singh Bath on 27 June 1956 appears acceptable ...." (p.41). The adjudication officer's submission to the SSAT includes -

"3. Reported decisions of the Commissioner considered by the Adjudication officer to be relevant.

None"

The deponent's submission to the Commissioner, as noted above, was based on a different ground, and it was not until the Commissioner raised the issue of a presumption from cohabitation that any reference was made to the earlier (1970) decision.

Before us, Mr. McManus confined himself to a single submission. This was that the presumption of marriage arising from cohabitation was rebutted in the present case by the SSAT's finding that the Sikh temple in which the marriage ceremony took place in 1956 was not registered for such a purpose at that date under the Marriage Act 1949. The ceremony relied upon, therefore, was shown not to be a valid marriage ceremony, and no other ceremony could be presumed to have taken place at any other time.

The Commissioner's decision

The Commissioner identified the issue as follows:

"The real issue is whether or not what was undoubtedly according to Sikh religious rites a marriage ceremony was invalidated by its taking place in a building, the Temple, which was not registered for marriages at the time, though it became registered subsequently. I have dealt with the case on that basis" (paragraph 10)."

After setting out paragraphs 992 and 993 of Halsbury's Laws (4th ed. Vol 22), he referred to Re Shephard [1904] 1 Ch 456, a decision of Kekewich J and to R(G) 2/70 "a decision of Sir Rawden Temple, a former Chief Commissioner and an expert in matrimonial law". Then he said :-

"16. I naturally pay the greatest of respect to those statements by the learned Commissioner and I bear in mind the system of precedents of Commissioners' decisions, enjoined by a Tribunal of Commissioners in R(I) 12/75, paragraphs 21 et seq. Nevertheless, in the present case, I consider R(G) 2/70 distinguishable. In that case the claimant was relying on an alleged ceremony of marriage in an English registry office and not, as here, on a marriage according to different religious rites in a Temple of that religion. Moreover Re Shephard is, so far as I am concerned, a High Court decision, in pari materia, which I should follow unless there is any good reason why I should not. I have carefully read the entirety of the Law Report of Re Shephard. It seems too me that it carries the proposition for which it is cited and that I should follow it. There does not have been drawn to the attention of the Learned Commissioner who decided R(G) 2/70 the possibility that, where a ceremony that the parties have relied on is shown to have been invalid, one ought then also to consider the authorities summarised in paragraph 992 of Halsbury's Laws i.e. "Presumption from cohabitation without ceremony" (my underlining).

The reasoning which led to his decision, quoted above, was as follows:

"It should be remarked that if a person is able to have a marriage presumed where there has been no ceremony at all (paragraph 992), it would hardly be equitable that a person who has gone through a bona fide ceremony thought to be a marriage (and entered into long cohabitation on the strength of it) should be in a worse position. The legal position must in my view be this. First one asks whether or not a particular ceremony can be presumed to be have been a valid marriage. If it is decided that it cannot be so presumed, then one treats the case thereafter as if there has been no ceremony at all and looks at the remainder of the facts to see whether there can be presumption from cohabitation without ceremony. If there has been a long 'monogamous' cohabitation with nothing to cast any doubt on the facts of the situation (i.e. the strong and weighty evidence to the contrary to which Halsbury refers), then the fact that there may have been an invalid ceremony in the first place does not in my view prevent there being a presumption of marriage from the facts themselves." (paragraph 17).

The decision also includes references to a submission made under section 49 of the Marriage Act 1949 which provides that certain marriages are void when the parties to them "knowingly and wilfully" inter-marry without specified provisions of the Act being complied with, (paragraph 12), and to the common law concept of marriage (paragraph 18), to both of which I shall refer below.

Further facts

These were not in dispute. The Sikh marriage ceremony took place at the Sikh Temple, the Central Gurdwara, at 79 Sinclair Road, London W.14. There was a marriage ceremony duly administered by a Sikh priest in accordance with the Sikh custom and religion. There were up to fifty persons present and photographs were taken. In a letter dated 1 November 1995, Mrs. Bath wrote :-

"Like most other first wave immigrants, we had no knowledge of the law, were completely illiterate and led to believe, by our elders and peers, that we had followed the correct procedures. Both my husband and I have always been religious and we would have been very concerned if our marriage had not been carried out in accordance with the Sikh custom and religion or thought that our marriage was not valid".

Enquiries revealed that the temple moved from Sinclair Road to 62 Queensdale Road, London W1 sometime between 1956 and 1983. It was registered for marriages at the new address with effect from 26 September 1983. There was no evidence that it was not registered at either address before 1983, apart from the following correspondence. On 29 September 1994 the District Manager of the Benefits Agency wrote to the Temple asking "Could you please advise whether, in 1956, this Sikh temple was a registered building for performing marriages?" and President replied, by letter dated 4 October :-

"..... please note that I came in this country in 1960 and at that time the Sikh temple as far as I know was not registered in performing the marriages. Some other old members of the temple had confirmed me this point".

Under section 41 of the Marriage Act 1949, the Registrar General was required to keep a book in which buildings which were places of religious worship could be registered for the purposes of the Act. There was no evidence as to the state of the book at any time. On the other hand, the respondent accepted in her letter dated 1 November 1995 that the building was not registered, and the Commissioner proceeded on the basis that it was not (paragraph 6). He referred to a later letter from the Office for National Statistics dated 25 June 1997 which confirmed that Queensdale Road was registered on 28 September 1983, but which made no express reference to the position of Sinclair Road before that date.

Marriage Act 1949

This was a consolidation Act which permits the solemnisation of a marriage in a building, being a place of worship registered for that purpose (section 41) "according to such form and ceremony as those persons [the couple being married] may see fit to adopt" (section 44(1)) provided that a Registrar or an authorised person is present (section 43) and subject to certain requirements including the presence of two or more witnesses (section 44(2)) and an exchange of appropriate undertakings (section 44(3)).

Section 49 reads:

"49.Void marriages

If any persons knowingly and wilfully inter-marry under the provisions of this Part of this Act:

......

(e) In any place other than the church, chapel, registered building office or other place specified in the notice of marriage and certificate of the superintendent registrar:

(f) In the case of a marriage in a registered building (not being a marriage in the presence of an authorised person), in the absence of the registrar of the registration district in which the registered building is situated ; ....

The marriage shall be void".

This section, therefore, renders the marriage void, notwithstanding the exchange of vows, if the parties to it have "knowingly and wilfully" failed to comply with the relevant statutory provisions. There is no other provision in the Act which renders void or invalid a marriage which is not carried out in accordance with the statutory requirements as to place, persons present, etc., and so, in particular, there is no statutory provision to that effect when the parties concerned are unaware that there has been a failure to comply with the Act in some respect.

There can be no suggestion in the present case that either of the couple was aware of any defect. On the contrary, they were aged 16 and 19, they had recently come to this country and the ceremony was conducted by a senior churchman at the Temple in the presence of their family and friends. A marriage contract was made between them. There is no statutory provision which renders their marriage void.

Authorities

The common law presumption of marriage is set out in two paragraphs in Halsbury's Laws (above), each accompanied by extensive citations of decided cases-

"992. Presumption from co-habitation without ceremony.

Where a man and a woman have cohabited for such a length of time and in such circumstances as to have acquired the reputation of being man and wife, a lawful marriage between them will be presumed, even if there is no positive evidence of any marriage ceremony having taken place, and the presumption can be rebutted only by strong and weighty evidence to the contrary.

Presumption from co-habitation after ceremony.

Where there is evidence of a ceremony of marriage having been gone through, followed by the cohabitation of the parties, everything necessary for the validity of the marriage will be presumed in the absence of decisive evidence to the contrary, even though it may be necessary to presume the grant of a Special Licence or the death of a former spouse. In most cases a certificate of marriage will be available, and this will usually suffice to prove the marriage".

The authorities referred to include Piers v. Piers (1849) 2 HL Cas. 331, where the grant of a Special Licence by the bishop was presumed, even though the bishop testified his belief that he had not granted the licence, 30 years after the event (see also Hill v. Hill [1959] 1. WLR 127) and R. v. Mainwaring (1856) 7 Cox CC192, where the due registration of a dissenting chapel was presumed. In my judgment, these authorities show that the common law presumed from the fact of extended cohabitation as man and wife that the parties had each agreed to cohabit on that basis, and the presumption was extended to include an inference that the statutory requirements first introduced by Lord Hardwicke's Marriage Act 1753 had been duly complied with; but in each case the presumption was capable of being rebutted by clear and convincing evidence. It is understandable why clear evidence was required to rebut the presumption after a long period of unchallenged cohabitation as man and wife, because the evidence in rebuttal would by definition refer to events many years in the past and might be concerned with matters that were not easily susceptible of proof at that distance of time.

The presumption was supplied by the House of Lords in a Scottish Case, Captain de Thoern v. A. - G (1876) 1 App. Cas. 686, where the marriage in 1862 was invalid, not because of any defect in form or as to any statutory requirement, but because the husband was not finally divorced from his previous wife. This was because the time for appealing from the divorce decree had not expired and both parties honestly believed that there was no obstacle to their union. The husband died five years later, and there were three sons of the marriage (one was born posthumously) and in 1872 the sons asked for a declaration of their own legitimacy. The Lord Chancellor, Lord Cairns, held that there "was ample ground for presuming, according to the law of Scotland, that marriage by consent of which cohabitation with habit and repute is evidence" (p.689). He then rejected the submission that such consent could not be inferred in the circumstances of that case because the parties were under the impression that the ceremony of marriage was a valid ceremony, therefore the need for a fresh contract did not arise. He referred to two authorities, Piers v. Piers (above) where "It was held in a most striking way as a general rule that the presumption of marriage is not the same as the presumption raised with regard to other facts, which may be presumed either the one-way or the other; that the presumption of marriage is something much stronger, and that from cohabitation with reputation a marriage is presumed unless there is the strong and cogent evidence to the contrary" (p.689-690).

The second case was Breadalbane [1866] LR2 HLS 269 where "the presumption was held to be one that not only might be drawn but ought to be drawn from the cohabitation with habit and repute, although in that case the cohabitation commenced with the ceremony of marriage which not only was invalid by reason of the real husband of the woman being alive at the time, but was known to both parties to be invalid". In that case, Lord Westbury had said :-

"There is nothing to warrant the proposition that the subsequent conduct of the parties shall be rendered ineffectual to prove marriage by reason of the existence at a previous period of some bar to the interchange of consent. It would be very unfortunate if it were so. Marriage may be contracted between parties in a foreign land where certain observations are required which from ignorance or mistake may not have been fulfilled ..... The parties having cohabited on the strength of an imperfect celebration, may afterwards come to Scotland and reside there for years, continuing the same course of life ..... I think a sounder rule and principle of law will be that you must infer the consent to have been given first moment when you find the parties able to enter into the contract" (p.691).

In Sastry etc. v. Sembecutty Vaigalie (1881) LLR. 6 App. Cas. 364 the Privy Council was concerned with the law of Ceylon, where the parties had gone through a form of marriage and shown an intention to be married. They not only lived together as man and wife, but there was strong evidence to show that there was a legal marriage (p.366). On the other hand, the wife herself gave evidence which raised doubts as to whether the ceremonies had been fully performed (p.367). The judgment, given by Sir Barnes Peacock, referred to the submission of Doctor Phillimore regarding Roman-Dutch law, and continued:

"It does not, therefore, appear to their Lordships that the law of Ceylon is different from that which prevails in this country; namely that where a man and woman are proved to have lived together as man and wife, the law will presume, unless the contrary be clearly proved, that they were living together in consequence of a valid marriage, and not in a state of concubinage". (p.371).

That authority was cited to Kekewich J in In re. Shephard (above). An English man and woman travelled to France with the intention of getting married and they purported to go through a form of marriage there. They then lived together in England as man and wife for 30 years and had several children. The question of legitimacy arose. There was a conflict of evidence as to whether the French ceremony was valid. The judge was prepared to assume that the marriage, as alleged, was "impossible according to French law and according to the habits of law abiding people in France" (p.462). After quoting from the Sastry judgment (above) he continued:

"So that I have here a case of the highest authority getting rid of the fact of marriage, and recognition of children ; it does not show that either of these is essential ; but the parties were living as man and wife for the time mentioned in the report, and it was held that the presumption of marriage must prevail. Now here I have the intention to marry; but that there is not a shadow of doubt. Some evidence about which there is a great deal of doubt. There is a somewhat romantic story, doubtful in its details, of a marriage de facto of something gone through to perfect the intention of marriage, and I have some evidence of recognition of children. Now, after 30 years, the Court has been asked to say that because the marriage has not been proved, and cannot be proved, these children are not to be admitted to share. I think I should be going against the authorities if I came to any such conclusion ...." (p.463-4)

Finally, in In re Green (1909) 25 TCR 222 Swinfen Eady J held that a marriage by repute which was valid by the law of the foreign domicil of the parties should also be recognised as valid by the English Courts. The husband's first wife died in 1842. He went through a form of marriage with his deceased's wife's sister in 1850. The marriage was invalid under English law for that reason. In addition, he found that "his social position and comfort were affected by the view taken of the connexion" in England, and so in 1855 he emigrated to Brooklyn and was joined there by his second wife and their children. They became domiciled there and lived as man and wife until she died in 1907. Their child whose legitimacy was in question was born in 1861.

The judge held that the relevant period was from 1855 until 1861 and that the marriage was recognised by the local law. The report concludes -

"Then it was suggested that without some formal ceremony the law of England would not recognise a foreign marriage; but his Lordship was of opinion that what was recognised by the law of the domicile would be held good by the law of the domicile" (p.223)

Decision R(G) 2/70 (Sir Rawden Temple)

There was a ceremony of marriage of an unusual kind on 12 July 1947. There was a formal exchange of voluntary consent in the presence of witnesses, before a person whom the woman believed was a registrar. She bona fide believed that it was a valid ceremony. She was uncertain where it took place, but she said positively that it was not at the building which was the local register office. Nor was it shown that the requirements for a common law marriage in England were satisfied, because no episcopally ordained priest was present (paragraph 10, citing R. v. Millis (1844) 10 Cl. & Fn. 534). The Chief Commissioner said this:

"It seems to me, on the evidence as to what took place, that the "marriage" was not celebrated in an authorised mode in a stipulated place by a proper official, and that it was void ipso jure",

and he concluded:

"As I see it, if the claimant was not married by the ceremony of 12 July 1947 she was not married at all, and, with regret, I have felt obliged to come to the conclusion that she was not so married" (ibid.).

He then considered the judgment in Re Shephard. He did not find any great assistance in it and felt unable to follow it. "Where there is evidence that the marriage asserted can only derive from a particular ceremony having validly taken place, it seems to me that the formal validity of that ceremony is unaffected by the fact of cohabitation thereafter .... such cohabitation .... has no bearing on the validity of the ceremony itself" (page 11).

The parties had cohabited as man and wife for 21 years, until his death, and they had two children. The Commissioner dealt with this aspect as follows :-

"13. The local tribunal applied the principle that where there is doubtful or inconclusive evidence of a ceremony of marriage having been performed, possibly imperfectly, and cohabitation thereafter, the validity of the marriage will be presumed in the absence of evidence to the contrary. Russell v. Attorney-General [1949] page 391; Piers v. Piers (1849) 2 HL Cas. 331. On the evidence they accepted, this could only have application to the ceremony, as described by the claimant, occurring on 12 July 1947; and if I am correct in my view of the ceremony, there was, I think, no room for the application of any such principle and presumption to save the invalid ceremony. I do not see how the presumption could have been involved rejecting the claimant's evidence".

Conclusions

These authorities show that when the man and women have cohabited as man and wife for a significant period there is a strong presumption that they have agreed to do so, in proper form. Thus in Scotland (Captain de Thoren) and Brooklyn, N.Y. (In re Green) it was sufficient that their agreement could be implied. When there is, as there is in England, a legal requirement that the marriage ceremony shall take a certain form, then the presumption operates to show that the proper form was observed, and it can only be displaced by what I would call positive, not merely "clear", evidence (see the authorities cited in support of Halsbury's Laws paragraph 993). How positive, and how clear, must depend among other things upon the strength of the evidence which gives rise to the presumption - primarily, the length of cohabitation and evidence that the parties regarded themselves and were treated by others as man and wife.

I would agree with the Commissioner's Decision R(G) 2/70 that when there is positive evidence that the statutory requirements were not complied with, then the presumption cannot be relied upon to establish that they were. However, that is not the end of the matter. Mr. McManus' submission assumes that if the place of worship where the ceremony takes place is not registered in accordance with the Act then the marriage is "invalid", or more precisely, that it does not count as a marriage for the purposes of the social security legislation regarding widows' pensions. This leads to the remarkable conclusion, apart from the injustice of which the Respondent complains, that a valid marriage ceremony may be presumed where there is no evidence from the surviving widow, as where she refuses to co-operate by providing information, but it may not be presumed when, as here, she gives a truthful account which leads to the discovery that there was a failure to comply with the requirements of the Act, of which she was and remained unaware. Likewise, when as here she accepts the truth of what the department alleges against her, she is in a worse position than if she had insisted on their producing positive evidence of non-registration (which they have not done). Even more remarkable is the consequence referred to by the Commissioner (paragraph 17). A marriage could be presumed from long cohabitation when there was no ceremony, but not when a bona fide ceremony failed to comply with all the requirements of the Act, perhaps for some trivial reason.

In my judgment, the submission is fallacious, for this reason. There is no statutory provision that a marriage, otherwise carried out in proper form, by an authorised celebrant and at a place of worship eligible to be registered under the Act, is invalid merely on the ground that the building was not registered, for whatever reason. The marriage was not void or voidable under the law in force before 1971 (see Halsbury's Laws (4^th ed.) Vol. 13 paras. 538-9) nor was it rendered void by section 49 of the 1949 Act.

The Commissioner in Decision R(G) 2/70 cited the judgment of Ormrod J. in Collett v. Collett [1968] P 482 for the proposition "The general tendency of the law as it has been developed has been to preserve marriages where the ceremonial aspects were in order". Ormrod J held:

"In my judgment, the principle which emerges from the corpus of legislation regulating the formation of marriages in England and from the reported cases arising therefrom, is that if a ceremony of marriage has actually taken place which, as a ceremony, would be sufficient to constitute a valid marriage, the courts will hold the marriage valid unless constrained by express statutory enactment to hold otherwise"

The requisites of a valid marriage are set out in Halsbury's Laws (4^th. ed.) Vol. 22 para. 907. They include "(b) that certain forms and ceremonies should be observed". The principle stated by Ormrod J is adopted in Halsbury, and there is no reference to any statutory provisions which might constrain the conclusion that the marriage is invalid if the place of worship where the ceremony is held is not registered.

For this reason, I would hold that Mr. and Mrs. Bath were validly married in 1956 by reason of the ceremony at the Sikh temple at Sinclair Road, notwithstanding that the Temple was not registered at that date pursuant to the Act. This assumes, of course, that they were unaware of that fact. The contrary is not and cannot be suggested. It is noteworthy that guilty knowledge by both parties is necessary for the marriage to be void under section 49 (Halsbury's Laws Vol. 22 paragraph 914
no. 23). Unless both participate, it is not invalid (ibid.).

Where does this leave Commissioner's Decision R(G) 2/70? Clearly, in my view, the basis of the decision was the positive evidence that the ceremony had taken place elsewhere than at a Registered Office. No priest officiated, as was necessary for a common law marriage, and it was not shown that a Registrar was present (though I wonder why the common law presumption did not assist the applicant to establish this). He found, therefore, that the ceremonial aspects were not in order (his words in paragraph 13) and that the presumption could not displace positive evidence that the ceremony took place where it was not permitted by the Act. He did not consider whether, in the absence of statutory provision, the fact that the unidentified building was not a registered building was sufficient to render the marriage "invalid", but this may have been because the claimant did not assert that she thought that the building was a Register Office (see paragraph 3) nor that "she went to any place other than A", which was not (paragraph 6).

That case, therefore, was clearly distinguishable on its facts from the present case. If it decides that a marriage is "invalid," meaning void or of no effect, by reason of non-compliance with any of the statutory requirements, even in the absence of express statutory provision to that effect, then in my respectful view that is not a correct statement of the law. I appreciate that we have not heard full argument on what may be an important general question of law, and I can only express this as my present view. The valuable Bromley's Family Law (9^th ed.) at page 34 et seq and 82 et seq does not indicate that it is wrong.

The present case

The Commissioner based his decision on the common law presumption arising from long cohabitation and held that the presumption overcomes the admitted fact that the requirements of a valid ceremony were not complied with in one respect. In my view, his conclusion was correct, for the reason which I have summarised above, namely, that the failure to comply was not sufficient to render the marriage "invalid" under the statute. I agree with him, however, that the presumption has an important part to play. There is no evidence whether the celebrant was a duly authorised person under the Act, and it must certainly be presumed that he was. I would also be prepared to hold, notwithstanding the finding that the Temple was not registered under the Act, that the evidence relied upon by the department was not sufficiently positive to displace a presumption that the building was registered. The letters from the current senior officers of the Temple in its new location are not even the best evidence, which could and should be produced, in the form of the book, which there was a statutory obligation to keep. The applicant simply does not know, and her "admission" therefore is of no weight as evidence in support of the department's case.

Finally, the Commissioner referred to the position at common law (paragraph 18). There is a complication in the present case, because the celebrant was not an episcopally appointed priest (R. v. Millis, above) although he may be presumed to have been duly authorised and the proper person to perform the ceremony according to Sikh rites and custom. This could give rise to questions of domicil, which it is unnecessary for us to enter into. Likewise, Re Shepard was concerned with a marriage ceremony conducted overseas, and I do not regard it as relevant here.

For these reasons, in my judgment, the appeal fails. I trust that Mrs. Bath will be paid her pension forthwith.

LORD JUSTICE SCHIEMANN:

I have had the advantage of reading in draft the judgments of Evans and Robert Walker LJJ. I agree with them that this appeal should be dismissed. As a relative stranger in the area of matrimonial law I prefer to rest my reasoning on the narrower grounds set out by Robert Walker LJ.

LORD JUSTICE ROBERT WALKER:

I agree with Evans LJ that this appeal should be dismissed. I reach that conclusion by reference to the presumption of marriage arising from long cohabitation, and the absence of compelling evidence to rebut that presumption.

Apart from the presumption, the law as to the validity of marriages solemnised (or said to have been solemnised) in England is now wholly statutory. The statutory rules are to be found (and were to be found in 1956, when Kirpal Kaur Bath went through a Sikh marriage ceremony with Zora Singh Bath at the Central Gurdwara, 79 Sinclair Road London W14) in the Marriage Act 1949 ("the 1949 Act"). The 1949 Act (which was a consolidating statute) has been amended, principally by statutes enacted in 1970, 1983 and 1986, but the amendments are relatively minor and not material for present purposes. Section 11(a)(iii) of the Matrimonial Causes Act 1973 does not take the matter any further.

However the 1949 Act and the statutes which it consolidated (and indeed all the statute law dealing with marriage in England, going back to Lord Hardwicke's Marriage Act of 1753) were built on a foundation of ecclesiastical law of great antiquity. Lord Hardwicke's Marriage Act was intended to curb the scandal of clandestine marriages (at a time when a married woman's fortune automatically became, at common law, the property of her husband). Other statutes were aimed at permitting persons of different religious affiliations (notably Quakers and Jews) to be lawfully married otherwise than in accordance with the rites of the Church of England. Many of the statutory provisions were permissive or directory in character. For instance s.-1 of the Marriage Act 1823 (now reproduced in s. 22 of the 1949 Act) required two or more witnesses, as well as the officiating clergyman, to be present at any marriage solemnised according to the rites of the Church of England. But it was held in Wing v Taylor (1861) 2 Sw. & Tr. 278 that non-compliance did not make a marriage void.

The scheme of the 1949 Act, so far as now material, is that Part I deals with restrictions on marriage (that is, capacity); Part II deals with marriage according to the rites of the Church of England; and Part III deals with marriage under a superintendent registrar's certificate. In Part I, sections 1 (prohibited degrees) and 2 (persons under sixteen) state expressly that non-compliance makes a marriage void. Section 3 requires parental consent (unless dispensed with) to the marriage under Part III of a person under full age, but s.-48(1)(b) makes plain that non-compliance does not invalidate a marriage. Section 4 (marriage may be solemnised between 8 am and 6 p.m.) appears to be directory only, although a deliberate breach exposes the offender to a severe criminal sanction (s. 75(1)(a)).

This absence of a clear and complete code as to which irregularities do, and which do not, make a marriage void is carried forward into Parts II and III of the 1949 Act. Parts II and III each contain one section (ss. 24 and 48) stating expressly that proof of compliance with certain procedural requirements is unnecessary, and that no "evidence shall be given to prove the contrary in any proceedings touching the validity of the marriage". Each also contains one section (ss. 25 and 49) stating expressly that a marriage is void if the parties to it "knowingly and wilfully intermarry" in contravention of the requirements specified in those respective sections. The cases of non-compliance listed in s. 49 include –

(a) without having given due notice of marriage to the superintendent registrar;

(b) without a certificate for marriage having been duly issued by the superintendent registrar to whom notice of marriage was given;

...

(e) in any place other than the church, chapel, registered building, office or other place specified in the notice of marriage and certificate of the superintendent registrar;

(f) in the case of a marriage in a registered building (not being a marriage in the presence of an authorised person), in the absence of a registrar of the registration district in which the registered building is situated;"

As Evans LJ has observed, the parties to the Sikh marriage ceremony were then aged 16 and 19. They had recently come to this country and they were unfamiliar with the English language and with English laws and customs. They intended to get married and they did not intend to break the law in any way. They did not therefore come within the scope of s. 49, and there is no statutory provision expressly rendering their marriage void.

Nevertheless there was (on the Chief Adjudication Officer's view of the facts, and apart from any presumption) a manifold non-compliance with the provisions of Part III of the 1949 Act: no notice of marriage under s. 27, no declaration under s. 28, no entry in the marriage notice book under s. 31, no certificate under s. 32, no registered building under s. 41, and no registrar or authorised person present under
s. 44. If in this case the husband and wife had been compelled by adverse circumstances to separate soon after the ceremony, so that no presumption arose from cohabitation, I feel real doubt whether they could have been regarded as lawfully married under English law, despite the logic of the argument based on the mental state required for a marriage to be void under s. 49.

In Collett v Collett 1968 P 482 ( a case on the Foreign Marriages Act 1892 concerned with a marriage conducted in dramatic circumstances at the British Consulate in Prague in 1948) Ormrod J distinguished between directory and essential provisions of that Act despite the presence of a limited provision (s. 13) comparable to ss. 24 and 48 of the 1949 Act, and the absence of any provision comparable to ss. 25 and 49. Ormrod J emphasised (in a passage which Evans LJ has already cited) the importance of upholding, wherever possible, the validity of marriages entered into in good faith.

This court did not hear any submissions on the decision of Judge Aglionby in Gereis v Yagoub [1997] 1 FLR 854 (discussed in Bromley's Family Law, 9^th ed pp 82-3) and I prefer to express no view on it except to note that some reliance seems to have been placed on the Christian character of the ceremony (at a Coptic Orthodox church not registered for marriages). But in this case it is unnecessary to go further into the issue of validity of an irregular marriage ceremony which is not followed by long cohabitation. Where there is an irregular ceremony which is followed by long cohabitation, it would be contrary to the general policy of the law to refuse to extend to the parties the benefit of a presumption which would apply to them if there were no evidence of any ceremony at all. I agree with Evans LJ, for the reasons which he gives, that there is insufficient evidence to rebut the presumption in this case.

Order: Appeal dismissed.