The real difficulty is in clause 4 of the will which reads as follows:

"4. I GIVE to my Trustee the residue of my real and personal estate

(b) To divide the balance then remaining equally (per capita)

The administration of clause 4(b) of the will has caused the executrix considerable difficulty. She approached the Master for directions and the Master authorised moneys to be expended on inquiries to see who came within clause 4(b) of the will. The deceased died aged 76 leaving no issue and was predeceased by both his parents. He had no brothers or sisters either of the whole or half blood, nor grandparents, uncles or aunts. He had no dependants. However, the executrix has made researches assisted by a genealogist and historical researcher and has to date found 1,225 persons who qualify as beneficiaries from the Castles' side of the family, and 450 Coghlans.

Evidence has been filed which shows clearly the problems the executrix has. The evidence, for instance, shows that a Mr Raper of South Australia swears that he is the father of Tayla Alish Quinn, his de facto wife Alisha Quinn being the mother of that child. There are other examples of ex-nuptial children whose surname bears no resemblance to anyone in the family tree, but who the executrix has reasonably concluded, after examination of their credentials, are legitimate claimants to a share.

In the month since the case was set down for trial and when I heard it, the number of beneficiaries who had been identified increased from 1,385 to 1,675. Although the law of diminishing returns will now apply, it is possible that more claimants will come to light even though the executrix and the genealogist have now been searching for some two years. The real question is, in all these circumstances is the gift in clause 4(b) valid?

On the hearing of the summons to determine that matter, Mr J B Whittle of counsel appeared for the plaintiff and Mr P Hallen, SC appeared for the defendant. The defendant was appointed by Master McLaughlin as a representative defendant of all the descendants of the four named individuals who were living on 25 July 1994.

Mr Whittle presented the material collected by the executrix and also provided the court with a most scholarly dissertation of the law. Mr Hallen, SC did not in any way dispute that the propositions put forward by Mr Whittle were the law of England, and agreed that there was no Australian authority that assisted the court. He argued that (a) on a proper reading of some of the English decisions the present gift still succeeded; and (b) the English decisions did not represent the law of Australia.

The basal problem is that the relevant cases tend to show that unless one can, with reasonable certainty, identify each and every beneficiary and the quantum of that beneficiary's share, then the trust fails. In the present case the consequence of the trust failing would appear to be that the testator's net estate, which is about $500,000, will go as bona vacantia.

The gift is to "issue". Mr Whittle submits that that word has its ordinary meaning, namely, descendants of all degrees; see Matthews v Williams [1941] HCA 32; (1941) 65 CLR 639, 650. There was some discussion as to whether issue included both legitimate children and ex-nuptial children. Under s 7(2) of the Children (Equality of Status) Act 1976, a reference to issue is to be construed as a reference to persons related in fact notwithstanding that that person or a person through whom the relationship is traced is or was an ex-nuptial child, unless the contrary intention appears. Mr Whittle drew my attention to this provision. Neither counsel argued that there was a contrary intention, Mr Whittle reminding me that if I had thoughts to the contrary, a representative of those claiming through ex-nuptial children would need to be added. I agree with counsels' assessment that persons who trace their relationship to the four named propositi through an ex-nuptial child fall within the class of issue.

Mr Whittle, in his submissions, put that it is necessary to separate trusts into two classes, viz: (a) fixed trusts; and (b) discretionary trusts. He then says, "Until 1971, the law required that in the case of both fixed and discretionary trusts the whole range of objects eligible for selection had to be ascertained or capable of being ascertained: see IRC v Broadway Cottages Trusts [1955] Ch 20, 35 (CA). Further, the trustee had to be able to ascertain the beneficiaries at the time the trust instrument came into force: see Re Hain's Settlement [1961] 1 All ER 848; [1961] 1 WLR 440. The compiling of a list of beneficiaries might be postponed where a gift was to be held on trust for unborn children, but at the very least, the trust instrument had to define with precision which unborn children were to take at some future time.

"So far as discretionary trusts are concerned this rule was swept away by the House of Lords in McPhail v Doulton [1970] UKHL 1; [1971] AC 424." Mr Whittle then points out that by a majority of three to two, the House held that a discretionary trust was valid "if it can be said with certainty that any given individual is or is not a member of the class" (see per Lord Wilberforce at

p 456). In so far as the Broadway Cottages case purported to lay down a contrary rule as to discretionary trusts, it was overruled.

Mr Whitte then submits, "However, after a considerable amount of controversy amongst academic commentators and text book writers, it now seems that McPhail v Doulton left the test relating to fixed trusts undisturbed: see Halsbury's Laws of England (4th edition) volume 48 para 554; Pettit Equity and the Law of Trusts (7th edition) page 48; Underhill & Hayton, Law of Trusts and Trustees (15th edition) pages 66-67; Parker & Mellows: The Modern Law of Trusts (6th edition) p 102; Hanbury and Martin: Modern Equity (14th edition) pages 100-101; Cohen: "Certainly Uncertain: The Discretionary Trust" (1971) 24 Current Legal Problems 133 at 145; Martin: "Certainty of Objects - What is Heresy?" (1984) The Conveyancer 304; Hayton (1984) The Conveyancer 307; Emery: "The Most Hallowed Principle - Certainty of Beneficiaries of Trusts and Powers of Appointment (1982) 98 LQR 551 at 559.

"It follows that in a fixed trust there must be sufficient certainty as to the identity of beneficiaries to allow the trustee to draw up a complete list of all beneficiaries who are to take under the trust. (See at first instance the Broadway Cottages case [1954] 1 WLR 659 at 664 per Winn-Parry J). The reason why the authorities speak of a complete list is because it is only when such a list has been compiled that the trust fund will be able to be divided into the shares for each beneficiary which are dictated by the instrument creating the trust."

Mr Whittle suggests that it is appropriate that one follow the analysis made by Mr C T Emery in his article referred to above, that is, that there are four separate and distinct elements involved in the notion of certainty, viz: (a) "conceptual certainty"; (b) "evidential certainty"; (c) "ascertainability"; and (d) "administrative workability". Mr Whittle then correctly submits that the present problem is one of evidential certainty, that is, even though one knows the concept of the class, is it possible to gather facts which will demonstrate all the members of the class? If it is not, then the gift fails because not only must one identify whether a particular person is a member of the class but also the quantum of that person's share of the fund. The latter cannot occur unless one can establish who all the members of the class are. Thus, one must prepare a complete list of all the beneficiaries or else the whole trust fails.

The first of the arguments put by Mr Hallen, SC, deals with just how certain one must be before disqualifying a trust under this principle of evidential certainty.

In Re Eden [1957] 1 WLR 788; [1957] 2 All ER 430, Wynn-Parry, J said that even if the whole of the fund is expended in undertaking inquiries to identify the whole range of objects, that should have to be done before the court could decide that the gift was bad for impossibility. However, in Re Saxone Shoe Co Ltd's Trust Deed [1962] 1 WLR 943; [1962] 2 All ER 904, Cross, J declined to follow Re Eden (WLR 954; All ER 913) and set out the test at WLR 955 (All ER 913) that "It is necessary for the court to be satisfied in these cases that a complete list of the beneficiaries could probably be compiled." (Note that the All ER report uses "can" in lieu of "could").

The facts of the Saxone case were that the trust deed benefited the former employees of a particular company or their dependants as the trustees in their discretion thought appropriate. It was held that it was necessary to be able to ascertain who all the beneficiaries were before one could work out who was entitled to be considered in the exercise of the discretion. In that circumstance, Cross, J said at WLR 953 (All ER 912), "Apart from authority, I should have thought that the answer depended on what was probable and not on what was theoretically possible. When the court directs an inquiry to ascertain next of kin, or distributes a fund on the master's certificate given in answer to the inquiry, it is always possible that the class will never be or has not been completely ascertained. ... All that is necessary is that the court should be satisfied that it is probable that the class can be or has been ascertained. Conversely, if the trustees had tried to ascertain the class of beneficiaries in the Broadway Cottages case, or the court had directed an inquiry to ascertain them, many answers would have been received to the inquiries made or advertisements issued, and it might have been that in fact the answers received would have covered the whole class. It would have been in the very highest degree improbable but not impossible. Nevertheless, I should have thought that the admission that the class was unascertainable was rightly made, because plainly no one would ever have been able to say in that case that any list compiled was probably complete." His Lordship then reviewed the cases and concluded that nothing in the cases affected that proposition.

Mr Hallen, SC says that in the instant case I should conclude as a matter of fact that after the activities of the executrix and the genealogist, the class and each member of it could probably be ascertained.

A problem always arises these days with tracing descendants of a male person because one can never be one hundred percent certain that the male person has not sired an ex-nuptial child. However, when considering this type of problem, the Public Trustee and other persons who are involved in the administration of the estate of intestate males investigate what evidence there is and come to the view that it is either probable or not that the intestate sired a child and distribute the estate accordingly. Mr Hallen, SC says that in the instant case much the same sort of principle applies. That is, if after extensive inquiries, and he says that extensive inquiries have been made over the last two years, the evidence does not show the probability of more ex-nuptial children being identified, then one just discards that possibility because one is dealing in probabilities, not possibilities.

I consider that this is a good argument as far as it goes. However, in the instant case the researches of the genealogist show that more likely than not there are other people out there who have not yet been ascertained. She may or may not be able to identify those people.

A common sense answer to this sort of question would be to say that if after, say, another six months, the genealogist has not uncovered any further beneficiaries, then the distribution should be made to those who at that stage had been discovered.

Unfortunately, the inquiry must be whether as at the date of the trust, in this case the date of death of the testator, which is assumed to be 25 July 1994, a complete list of the beneficiaries could probably be compiled. The importance of making the inquiry as at the date of the trust is clear from the cases; see for instance Re Leverhulme [1943] 2 All ER 274. One can use what has in fact happened with the inquiries made by the genealogist to say the probabilities as at 1994 and one could perhaps say what is a reasonable time for the inquiries to be made, but it does not seem to me that one now can set a deadline for the ascertainment of the beneficiaries.

Accordingly, I do not consider that I can accede to Mr Hallen, SC's submission, at least in its starkest form.

Before considering whether I should follow the English line of authority, I need to consider the thought of Cross, J that the exercise is to a degree like a next-of-kin enquiry before a Master: Saxone supra p 912.

The 7th edition of Daniell's Chancery Practice (Stevens and Sons Ltd, London, 1901) sets out what happens when there is an inquiry for next-of-kin at pp 823-4. Essentially, if the evidence adduced by the parties is insufficient to enable the inquiries to be satisfactorily answered, then an advertisement is placed, prepared by the parties and approved by the Master, fixing a time within which each claimant is to come in and prove his or her claim, appointing a day for adjudicating thereon. Unless otherwise ordered, all persons who do not come in and prove their claims within that time will be excluded from the benefit of the order.

If some similar procedure were put in place in the instant case, one could cut off all claims not asserted by a particular date. However, the analogy is really not complete because in the case of the Master, the people who had not put in their claims were not barred, they merely were excluded from the judgment so that they would need to mount a Re Diplock type action against the beneficiaries. Of course, a similar procedure could be allowed here, but the problem is that if the English cases are followed the rule is that unless there is evidential certainty, the whole trust is invalid.

A good reason for not following the English cases would be that in the discretionary trust situation, the certainty of beneficiary rules have been considerably modified; see McPhail v Doulton supra. However, as Mr Whittle says, the leading English texts on trust law have considered whether the certainty of beneficiary rule still applies with fixed trusts and have almost unanimously come to the view that it does. There is need for certainty in the law of trusts throughout the common law world, and in my view, if the majority of the greatest academic minds on the subject have reached this view it would need very considerable arguments to the contrary to persuade a single Judge in Equity in New South Wales not to follow that line. Accordingy, the McPhail case does not seem to me a sufficient reason to depart from the rule as stated by Cross, J in Saxone's case.

It is then useful to see how the English line of cases has fared in other countries. According to Scott on Trusts 3rd ed (Little Brown & Co, Boston, 1967) para 122, "Where the trustee is directed to divide the trust property equally among the members of a class, and the class is so indefinite that it would be impossible or impracticable to divide the trust property in equal shares among them, the trust fails." The authority given (in the supplement) is Re Gulbenkian's Settlement [1970] AC 508.

In New Zealand, Tipping, J said though possibly obiter in Re Beckbessinger [1993] 2 NZLR 362, 369-70, after citing Mr Emery's article, "Where ... there is what is called a fixed trust it must be possible to compile a complete list of all beneficiaries. A fixed trust is where the trust property must be distributed amongst all members of the class either equally or in specified proportions. Obviously in that circumstance the trustee cannot perform the task unless all members of the class can be identified with certainty."

I have also considered the Irish situation. Delany, Equity and the Law of Trusts in Ireland (Roundhall Sweet & Maxwell, Dublin, 1996) relies on the English law at p 77 of the work as representing the law of Ireland.

The only relevant Australian authority is Kinsela v Caldwell Mahoney, J - 28 June, 1973, unreported, and in the High Court where the appeal was dismissed [1975] HCA 10; (1975) 132 CLR 458. The High Court did not deal with the present problem. Mahoney, J considered the Gulbenkian case and also McPhail v Doulton. He considered that neither dealt with the principles to apply in a situation where a trust provided that the next-of-kin were to take. He did not consider any of the other fixed trust cases other than Muir v Inland Revenue Commissioners [1966] 1 WLR 251, but gave general approval to cases such as Leverhulme and the passage which I have quoted from Scott on Trusts. I will deal with the parts of his Honour's judgment that deal with the rationale for the certainty rule later. Suffice it to say he seems to have considered that generally speaking the English line of authority should be followed in Australia.

Accordingly, the English cases have been followed generally in the common law world and, unless there is some very good reason to the contrary, they should be followed here.

There are some reasons why the English cases should not be followed. First, a Court of Equity must always be on the qui vive to realize that community circumstances have changed. The most recent example of this is the decision of Carmody v Delehunt [1984] 1 NSWLR 667, affirmed by the High Court as Delehunt v Carmody [1986] HCA 67; (1986) 161 CLR 464, where the rule that equity follows the law meant that equity followed the statute law for the time being so that an amendment of the statute law as to joint tenancies at common law was followed in equity.

In England up until recently and also in New South Wales, there was a very strict principle which controlled the way in which land descended to heirs and to next-of-kin. All marriages were to be conducted in an open and notorious way; all land transactions were to be dealt with publicly so that there could not be any dispute as to who was entitled. This system broke down to a limited extent because there were some marriages which took place by special licence, or in the Fleet prison and there was the employment of uses and trusts which disguised beneficial landholdings though the holdings at law were still open and notorious. Where there was such a system, one could relatively easily work out from the documents who begat who. The registration system of births, deaths and marriages, first in the established church and later in secular registers appears to have been pretty good. Even in respect of what the English would call "colonial registers of birth" the records were fairly good as mostly the colonies were involved in fairly small communities.

However, these days, being married is not a necessary prerequisite to having children. Even married persons may give the child a surname other than the father's surname. Unmarried persons who produce a child often call the child by the mother's surname. In remote areas and some Aboriginal communities births are not always registered. All these factors produce greater uncertainties than would have been the case when the rules as to certainty were laid down.

Furthermore, the rules as to future interests have changed with the Perpetuities Act 1984. The law once insisted that an interest, if it vest at all, must certainly vest within a life or lives in being plus 21 years and judged the validity of an interest on whether, as at the date of inception of the trust, there was any possibility of the interest failing. The Act directed attention away from possibilities as at the date of the instrument to a wait-and-see rule. So

s 8 of the Perpetuities Act provides that an interest is to be treated as valid until such time (if any) as it becomes certain that it must vest after the end of the perpetuity period.

Although the authorities would show that I should follow the English rule, it seems to me that I would be justified in slightly modifying its operation for Australian conditions of the present day. The rule will be satisfied if, within a reasonable time after the gift comes into effect, the court can be satisfied on the balance of probabilities that the substantial majority of the beneficiaries have been ascertained and that no reasonable inquiries could be made which would improve the situation.

In making this modification, I go to the reason why the rule was introduced. This can be found in the unreported judgment of Mahoney, J in Kinsela v Caldwell at p 6 where his Honour says, "The kind of uncertainty, in relation to trusts, which is here in question, (viz uncertainty as to who is or who will be the individuals to be benefited) has been said to have its origin in the principle that, a court of equity does not recognise as valid a trust which it cannot both enforce and control: In re Astor's Settlement Trusts; Astor v Scholfield [1952] Ch 534 at 549 ..." Indeed, immediately after following those words, Roxburgh, J said in the Astor Settlement case "This seems to me to be good equity and good sense." Mahoney, J then went on to quote Lord Eldon in Morice v Bishop of Durham (1805) 10 Ves Jun 521, 539; [1805] EngR 97; 32 ER 947, 954, "As it is a maxim, that the execution of a trust shall be under the controul of the Court, it must be of such a nature, that it can be under that controul; so that the administration of it can be reviewed by the Court; or, if the trustee dies, the Court itself can execute the trust; and a due administration directed; and then, unless the subject and the objects can be ascertained, upon principles, familiar in other cases, it must be decided that the Court can neither reform mal-administration, nor direct a due administration. That is the principle of that case." See also Muir's case at

All this is good equity and good sense. Equity does not make a rule just for the sheer fun of making a rule, but it must have a purpose. The purpose of the rule as to certainty of beneficiaries is, in modern conditions, just as well catered for by the modification I have made as is the rule in its original form. Indeed it is better accommodated, because for instance in the present case, instead of the testator's intention being completely frustrated and the money passing as bona vacantia, at least 1,675 people who can be ascertained and do fit within the description will benefit as the testator intended notwithstanding that it might not be possible to make a complete list with certainty.

Accordingly, I must turn to the genealogist's affidavit. It is not quite clear from her evidence as to whether she approached the matter by taking the children begotten by each of the four propositi and then going through the stocks, but she must have done this exercise. A pedigree or family tree would have assisted considerably in following what she has done. Doubtless this can still be provided.

The main problems that the genealogist has found relate to (a) unregistered births; (b) production of ex-nuptial children, especially by males; (c) disappearing or disappeared beneficiaries; and (d) adoptions.

I believe the correct way of proceeding is to briefly discuss these four matters, publish these reasons and then stand the matter over for compliance with my further directions. I can then make an order in the form that the executrix (who has become a trustee) will be justified in distributing the estate in a particular way pursuant to s 63 of the Trustee Act 1925.

It must be remembered that the testator wanted a per capita distribution. This means that if any beneficiary disappears, there is a possibility that that beneficiary may have in the period since he or she disappeared, produced a number of issue, each of whom would take. Accordingly, it is not possible, as is often the case, to simply provide that the share of a person who is known to be within the class but who has disappeared be retained and then paid into unclaimed moneys.

As to category (a), the real problem is to be satisfied that the person is whom he or she claims to be. The executrix must act on the evidence and make her decision one way or the other. Doubtful cases (of which there should be very few) should have a share assigned to them and put in reserve pending further evidence. In many cases it will be more efficient to pay for the cost of the birth certificate etc out of the share that would otherwise go to that person.

As to (b), most of the known problems in this area have now been solved. The problem that is left is whether anybody produced an ex-nuptial child who is not known. After the lengthy inquiries that have been made, the problem can be addressed by advertisement and closing order.

As to (c), this is the real problem that remains. As I said earlier, the problem is not only that there might be a missing beneficiary, but that beneficiary may have produced further issue of unknown number. However, there would only appear to be one person still left in this category, Kathleen May O'Brien. This lady has not been seen since 1995, nor has her child Renae. They would appear to be the only two persons as at 1995 who fit into category (c) and the chances of other issue are virtually non-existent as at the relevant date. Accordingly, two shares should be put aside for them.

As to (d), unfortunate though it may be for the people concerned, if they were adopted (without any fault of their mother or themselves to quote one lady's evidence), they do not qualify as issue.

I congratulate the executrix, and those advising her, of the way they have gone about searching for beneficiaries. I believe that little further can be done. The genealogist may well bring her research up to date at, say, 2 November 1998 and in addition an advertisement should be placed in The Australian newspaper or some other nationally circulating newspaper requiring all persons who claim to be descendants of the four propositi to lodge their claim with the solicitors for the executrix by, say, 2 November 1998.

I will then have the matter listed before me on, say, Tuesday 24 November 1998 at 9.30 am to give to the executrix the necessary directions under s 63 of the Trustee Act. Those directions will be that she would be justified in (a) putting aside X shares for persons who she considers may be legitimate claimants but who have not yet fully satisfied her by evidence; and (b) distributing the balance to those persons who as at 2 November 1998 have satisfied her of their eligibility. A declaration as to clause 3 of the will can also be made then. The costs of all parties to the proceedings on the indemnity basis should be paid out of the estate.

I would add that some years ago many of the problems in this class of case were dealt with by the executrix taking out insurance against missing beneficiaries. If such insurance is still available, it may well be economical to take it out especially as the insurer may consider that the vast amount of work that has been done may mean that it has a relatively small risk. This matter might be explored, if it has not already been explored, before the next day the case is mentioned.