Introduction
In order for a trust to be valid, the ‘three certainties’ must be present: certainty of intention, certainty of subject matter, & certainty of object – (Lord Langdale in Knight v Knight).

CERTAINTY OF INTENTION:
The intent will probably only be found from words, either spoken or written. Unlike contract, the concept of a trust is too complex to be expressed otherwise than by words.
The words expressed by the alleged settlor must manifest an intention to create a trust. The mere fact that a right holder has an intention to create a trust but which he keeps to himself will not cause a trust to come into being – (Re Vandervell (No.2)).
The words used must objectively (as per Lord Diplock in Gissing v Gissing) show that the alleged settlor intended to subject the holder of the right, be it himself or third party; to a legally enforceable obligation rather than a moral one – (Re Adams and the Kensington Vestry: ‘in full confidence that she will do what is right as to the disposal thereof between my children, either in her lifetime or by will after her decease’ – there was no intention to create a trust).
Compare: Comiskey v Bowring-Hanbury (‘in full confidence that she will make such use of it as I should have made myself and that at her death she will devise it to such one or more of my nieces as she may think fit and in default of any disposition by her thereof by her will . . . . I hereby direct that all my estate and property acquired by her under this my will shall at her death be divided equally among the surviving said nieces’ – there was intention to create a trust, as it seemed clear, that the testator intended his nieces to take following her death.)
Mere intention that another should benefit is not enough – (Jones v Lock: ‘I give this to baby for himself’ – there was no intention to create a trust. Maitland points out: Men often mean to give things to their kinsfolk; they do not often mean to constitute themselves trustees.)
The least equivocal expression of such an intent will be found in the use of the word ‘trust’; although depending upon the context, other words may serve as well. e.g. Paul v Constance: ‘the money in the bank is as much yours as mine’.
Requirement of proof in some cases: Allegation of a declaration of trust regarding land, additionally needs to be manifested and proved by some writing, signed by some person able to declare such trust or by will – (Sec. 53 (1) (b) Law of Property Act 1925).
Allegation of declaration of trusts in will, is required to be proved by signed witnessed writing – (Sec. 9 Wills Act 1837).
Consequence of uncertainty of intention: If such an intent cannot be shown, no trust is created and the person who is in control of the rights becomes entitled to retain it beneficially.
Accordingly, if there has been a transfer, the transferee takes it beneficially – (Re Adams and the Kensington Vestry).
Where there is no transfer but only a failed allegation of self-declaration of trust, the settlor simply remains absolutely entitled – (Jones v Lock).
 
CERTAINTY OF SUBJECT MATTER:
Certainty of subject matter comprises of two separate but related concepts –
Certainty of trust rights &
Certainty of beneficial interest. 

CERTAINTY OF TRUST RIGHTS:
The rights which are to form the subject matter of the trust must be clearly identifiable – (Re London Wine: Wine bottles held on trust, were not separated from a larger lot, therefore trust bottles were not identifiable, as a result the trust failed , Re Goldcorp).
But, surprisingly in Hunter v Moss, a distinction was drawn between trust of tangible rights and intangible rights; and this case still stands for the proposition that, intangible trust rights need not be clearly identifiable. (50 shares were held on trust but were not separated from the others; it was held that, as each shares carried identical rights it did not matter which 50 were held on trust.)
Consequently it gave rise to practical difficulties, as it does not provide answer to the problem of dealing by someone in the position of Moss. i.e. What happens if he gives 50 of the shares to his mother as a birthday gift; does he give away trust shares or shares of his own? Even rule of tracing cannot provide answer to this. This issue is not resolved yet.
Conceptual uncertainty about the word used to describe trust rights/beneficial interest can mean there is no valid declaration of trust – (Palmer v Simmonds: ‘bulk’). However, where a court can determine an appropriate meaning of the word used; the declaration will be valid – (Re Golay: ‘reasonable income’ – the court was able to determine what was reasonable by reference to the beneficiary’s previous standard of living).
Residuary estate is always certain, because after satisfaction of all required obligation according to the will the trust right can be made certain.
‘Whatever is left’ trust is uncertain – (Sprange v Barnard); as it is not certain whether anything would left.
A trust cannot be created unless the rights that are to form the subject-matter of the trust exist at the time of declaration.
Consequence of uncertainty of trust right: If the trust right is uncertain, then no trust could have been intended by the settlor subject to exceptions described by Oakley. Thus, no trust is created.
Accordingly, if there has been a transfer, the transferee takes it beneficially – (Palmer v Simmonds).
Where there is no transfer but only a self-declaration of trust, the settlor simply remains absolutely entitled.

CERTAINTY OF BENFICIAL INTEREST:
Where the trust right is certain, but the interest to be acquired by the beneficiaries are uncertain, the express trust will fail, and a resulting trust will arise in favour of the transferee – (Boyce v Boyce).
However, such an uncertainty will not be a problem in case of discretionary trust, as these are left to be determined by the trustees themselves.

CERTAINTY OF OBJECT:
A non-charitable trust must have ascertained or ascertainable beneficiaries – (Lord Evershed MR in Re Endacott).
Where the beneficiaries are individually named, it does not matter that the whereabouts of the beneficiaries in not discoverable. In Re Gulbenkian, Lord Upjohn opined that, in the case of a beneficiary of a class gift whose whereabouts are uncertain, the trustees can apply to the court for a direction to pay his share into court. In short, the problem is correctly thought of as one confronting the trustees when faced with their duty to distribute the trust rights, not as a problem concerning the very existence of the trust.
Note that problems of certainty of objects only arise where the settlor uses a generic term to describe the class. No question of uncertainty of object arises where the beneficiaries are individually named.
 
TEST FOR DETERMINING CERTAINTY OF OBJECTS:
FOR FIXED TRUST:
 A fixed trust is one where the trust instrument specifies the share each beneficiary is to take. For a fixed trust to be valid the trustee must be able to compile a complete list of the beneficiaries – (IRC v Broadway Cottages); for the amount each beneficiary receives is dependant on the number of people in the class. This is known as ‘complete list’ test.
 
FOR DISCRETIONARY TRUST:
A discretionary trust is one where the trustees are given discretion as to who, within a class chosen by the settlor should receive trust rights, and/or how much each should receive. For a discretionary trust to be valid, the trustee must be able to determine whether any given individual is or is not a member of the class – (McPhail v Doulton); for they (and the court) need to know whether the proposed distribution is within their power. This is known as ‘is or is not’ test.
McPhail v Doulton abandoned ‘complete list’ test for discretionary trust as the amount each beneficiary receives is not contingent on the number of people in the class; and employed ‘is or is not’ test.
Sach LJ’s approach to ‘is or is not’ test:
The test is satisfied where the class of potential beneficiary is conceptually certain (a class is conceptually certain when the words used by the settlor or testator have a precise meaning in themselves), it is no objection that it may be difficult to establish whether or not any given person satisfies the description – (Sach LJ in Re Baden (No.2); this is the preferred analysis). If however, there is no possibility of adducing evidence to prove that anyone falls within the class, then the trust will fail; even if there is a perfectly conceptually certain class. (compare, difficult vs impossible)
It will be for the individual person to prove that he is within the class. If he can’t prove it, he is taken to be outside of the class. However, such evidential difficulty will defeat a fixed trust, as it will mean no complete list can be drawn.
Megaw LJ’s approach to ‘is or is not’ test:
The test is satisfied if it could be said with certainty that a substantial number of beneficiaries fall within the class – (Megaw LJ in Re Baden (No.2)).
Megaw LJ’s comments were made in the context of a conceptually certain class. However, it is not clear whether he would require this too.

Criticisms of Megaw LJ’s approach:
If this means that only those may take who are within the substantial number, then it appears to cut down the class contrary to the settlor’s intention; and reintroduces a version of the ‘complete list’ test.
Further, it gives no guidance to the trustees as to the extent of any survey he must take of the class before distributing, i.e. the extent of the consideration he must give to distributing to those not within the ‘substantial numbers’, yet who may fall within the class intended by the settlor.
Resolution of uncertainty of object by third party:
An opinion clause can not cure conceptual uncertainty, but may allow an individual to determine whether the concept applies in any particular case – (Re Coxen).
Case analysis: Re Wright (a trust for ‘such people and institution as my trustees think have helped me or my late husband’ – this was an invalid trust as there was conceptual uncertainty and the trustee’s opinion could not cure this)
Re Tuck (Wife of any heir must be of Jewish blood and worship according to Jewish faith, in case of any doubt the decision of the Chief Rabbi was to be conclusive – this was conceptually certain; accordingly the opinion clause cured evidential difficulties)
Re Leek (a trust for such persons as the company may consider to have a moral claim upon the settlor – this was an invalid trust as there was conceptual uncertainty and the company’s opinion could not cure this)

FOR POWER:
A power can be defined as an authority vested in a person to deal with or dispose of property not his own. For a power to be valid the done must be able to determine whether any given individual is or is not a member of the class – (Re Gulbenkian).

FOR GIFTS SUBJECT TO A CONDITION PRECEDENT:
A condition precedent is a condition that must be fulfilled for a gift to take effect. On the other hand, a condition subsequent is a condition of defeasance – the gift will come to an end if the condition occurs.
In the case of a gift with a condition precedent that defines a class, an ‘is or is not’ test rather than ‘complete list’ test is appropriate, and the court will be liberal in determining criteria for vague terms – (Re Barlow: objects were ‘friends’ – this would definitely have been held uncertain had it been a discretionary trust, but it was held certain for gifts with condition precedent and therefore valid. It means a degree of conceptual uncertainty will not invalidate such a gift). However, Re Barlow was criticized by Emery.
Consequence of uncertainty of object:
In case of self declaration of trust which fails for uncertainty of objects, everything remains as before.
However, in case of a declaration accompanying a transfer which fails for such uncertainty, an automatic resulting trust will arise in favour of the transferor. Exactly why this ‘automatic resulting trust’ arises is a matter of great controversy.

ADMINISTRATIVE UNWORKABILITY:
 This requirement arises out of the abandonment of the ‘complete list’ test and adoption of the ‘is or is not’ test for discretionary trusts, and the consequent acceptance of validity of conceptually certain classes which is huge.
A discretionary trust can also be invalidated for being administratively unworkable – (R v District Auditor exp West Yorkshire Metropolitan CC). However, unworkabiltiy can be avoided if the settlor gives some instruction as to the principles upon which the trustee should exercise their discretion.
The concept of administrative unworkability is not clear as to what criteria will render a trust unworkable. However it is clear ‘size’ alone should not invalidate a discretionary trust. In Mcphail V Doulton, the extremely large class of ‘relatives’ did not invalidate a discretionary trust. Swadling suggests the problem is not size; rather absence of a ‘core class’ capable of being surveyed. If a ‘core class’ of objects can be identified within the larger class to whom the trustees may primarily devote their survey of objects before distribution, then the trust will not be administratively unworkable – although no case has explicitly stated this.
Administrative unworkability can invalidate a discretionary trust but not mere powers – (Re Hay); because duties of a discretionary trustee are more stringent than a fiduciary power holder (bare power holder has no duty at all; on the other hand the fiduciary power has a duty to consider), and that the beneficiaries of a discretionary trust have more rights of enforcement than objects of fiduciary powers. However; this reasons are not convincing.

CAPRICIOUSNESS:
A discretionary trust or fiduciary power (a power given to an individual virtute officio, such as a trustee) can also be invalidated for being capricious – (Re Manistry, Re Hay); as this renders performance of the discretionary trust or fiduciary power impossible. Unlike personal power, fiduciary power holder has a duty to consider the exercise of the power. Personal power holder has no duty to exercise it, or even to consider whether he should exercise it.
A discretionary trust or fiduciary power is capricious, where terms of the discretionary trust or fiduciary power, negatives any sensible intention on the settlor’s part, and precludes any sensible consideration by the trustess or donees.

FORMALITIES
SEC. 53(1)(b) LAW OF PROPERTY ACT 1925: Declaration of trust regarding land must be ‘manifested and proved’ by some writing.
It is no objection that the written evidence comes into being after the declaration of trust, as the statute says nothing about the time when it is to come.
This section (along with Sec. 53(1)(c) LPA 1925) does not apply to resulting, implied or constructive trusts – (Sec. 53(2) LPA 1925).
‘Rouchefoucauld v Boustead’ exception:
Later CA in Rouchefoucauld v Boustead developed that – this section does not apply where its application results in a fraud being committed. In such cases, oral evidence will be admissible to prove the existence of a trust. (However, it should be noted that the reasoning in Rouchefoucauld is circular – it will only be a fraud if the defendant is a trustee, but at the point where argument over admissibility is made, we do not yet have evidence showing that he is a trustee!).
However, this appears to disregard the statute as the trust being enforced is an express trust, on the basis that it is being proved by evidence of declaration, though oral.
So, later cases like Bannister v Bannister, Paragon v Thakarer classified such trusts as constructive trust, without giving any reason for such classification to utilise the statutory exception laid down in Sec. 53(2).
However, it should be noted that CA in Rouchefoucauld specifically and moreover logically stated that the trust they were enforcing was an express one.
‘Constructive trust’ approach does not fit well in a three party case (where A transfers land to B upon trust for C). Constructive trust can provide remedy only to an extent which is required to remove the unconscionability, it cannot go beyond that which means – B will hold the land on constructive trust for A not for C; unless he suffered any detriment.
 
Effect of non-compliance with sec. 53(1)(b):
Since Sec. 53(1)(b) is a rule concerning proof rather than enforceability, the logical effect of non-compliance will be – there is no trust at all in the eyes of the law (Swadling’s view – the preferable one).
However, Martin suggests non-compliance with Sec. 53(1)(b) renders a trust ‘valid but unenforceable’, based on an analogy with the old rule on contracts for the sale of interests in land.
The advantage of adopting Martin’s view is that, oral evidence will be then admissible to prove the failure of the express trust, which will then give rise to an automatic resulting trust. Such a result cannot be achieved by adopting Swadling’s view as it states non-compliance results the court to conclude there is no trust at all. There cannot be a failure of trust, without being a trust.

SEC. 53(1) (C) LAW OF PROPERTY ACT 1925:
Disposition of a subsisting (Note, this section does not apply to the original creation of a trust, but it is activated only when a beneficiary under a trust seeks to dispose off his interest) equitable interest must be in writing.
This section also does not apply to resulting, implied or constructive trust – (Sec. 53(2) LPA 1925).
The effect of non-compliance with this provision is that the purported disposition is void. There will be no disposition until the writing is executed.
It is assumed in all the leading cases (Grey, Vandervell and Oughtred) that, Sec. 53(1)(c) is not confined to trusts of land only. But, Sec. 205 (x) of LPA 1925 defines equitable interest as ‘equitable interest …………… in or over land or in the proceeds of sale thereof’. There is an argument on this basis that the leading cases in this area were decided ‘per incuriam’.
However, the better view is – since the section specifically mentions land in parts (a) and (b), but does not in part (c); the obvious interpretative conclusion to draw is that clause (c) is not to be confined to land. Further, it is argued if proceeds of sale of a land can be seen as an equitable interest for the purpose of Sec. 53(1)(c), why other personality would not?
GREY v IRC: Where the beneficial owner gives oral direction to the bare trustee, to hold the trust rights on different trust that he also administers, is void unless in writing.
Sec. 53(1)(c) was designed to protect trustees from false allegation by someone claiming to be an assignee of the beneficiary’s interest. The difficulty for trustee in such a case is that, if he pays out to the false assignee he thereby commits a breach of trust, and consequently incurs a liability to the true beneficiary to reinstate the fund. Since in Grey, the oral direction came from the beneficiary himself, the trustee knew it was genuine. So, there can be no question of providing protection to the trustee. Thus, it is submitted that the subsection should have held inapplicable in this case.

But, the HL in Grey took a literal approach to the section. Lord Radcliffe said that the word ‘disposition’ in Sec. 53(1)(c) covered all means and devices whereby an equitable owner effectively transfers his interest to someone else. Thus, it is clear from Grey that, a direction to a trustee by the equitable owner to hold on trust for a third party; or to transfer the interest to third party both is disposition for the purpose of this section and need to be in writing to be effective.
As the oral direction was not effective to pass the beneficial interest, it passed through the confirmatory deed which was executed to testify the giving of the direction. So. Mr. Hunter was held liable to pay tax upon this confirmatory deed as it passed beneficial interest.
Mr. Hunter could have avoided tax by following process: He could have orally declared himself a trustee of his shares for his grandchildren, instead of transferring legal title to the trustees. The effect of such a declaration would have been to create an original trust, rather than disposition of equitable interest under a subsisting trust. So, Sec. 53(1)(c) would not apply,  and Mr. Hunter could have still avoided the tax. Further if Mr. Hunter still wanted the original trustees to hold on trust, he could have transferred the legal title to them in a subsequent transaction, attracting only nominal stamp duty.

VANDERVELL v IRC:
Where the absolute beneficial owner gives oral instruction to the bare trustee, to transfer the property to a third party intending his beneficial interest to pass to the same third party, the passing of the beneficial interest need not be in writing.
Sec. 53(1) (c) was designed to protect trustees. In a situation, where the legal & equitable title is vested in the same person (as equitable interest normally passes with the transfer of legal interest) there is no trust and therefore no trustees to protect. This is the reason why Sec. 53(1) (c) was held inapplicable to the Vandervell scenario.
Green points out – in Grey, though it seems that only the equitable interest was disposed off by Mr. Hunter, but it is not true. The legal title was in fact transferred from the trustees of Mr. Hunter to the trustees of his grandchildren; though in this case they happened to be same parties. Nonetheless, they were acting in different capacities, so equitable interest was not divorced from the legal estate and both moved together. Thus, the ‘purposive approach’ employed in Vandervell should have held equally applicable to Grey.
In Vandervell the HL in propounding the rule of law might have ignored two important things: First, it will equally apply to chattels, which can be transferred out of the trust by delivery – the result is that a trustee may give away trust rights on the basis of oral direction with no writing whatsoever. Secondly, though a legal transfer normally includes the beneficial interest, it is questionable whether it does in this case, where the beneficiary has not expressed in writing his intention to give up the beneficial interest. Without any writing why it should not be presumed that – when the trustee T transfers the legal title to X, that X takes the legal title because he has been appointed as a new trustee to replace T, in which case the beneficial interest would of course remain attached to the property?
However, Mr. Vandervell lost the case for the option to purchase, as he meant the VFTC to hold the option on trust, but he did not mention who the beneficiaries would be. As a result, there was a resulting trust for him. Thus, he was not able to divest himself fully of the beneficial interest. Accordingly, he was held liable to pay tax by virtue of Sec. 415 Income Tax Act 1952.
 
OUGHTRED v IRC:
An oral contract to transfer subsisting equitable interest which is unique (called specifically enforceable contract) gives rise to a constructive trust in favour of the trustee. Accordingly the transfer takes place under the constructive trust. This transfer need not be in writing by virtue of Sec. 53(2) – (Lord Radcliffe in his dissenting judgement in Oughtred v IRC, Neville v Wilson).
Note, this proposition was not accepted in Oughtred v IRC. In Oughtred v IRC, after entering such a specifically enforceable contract; the trustees formally transferred the legal title of the shares to Mrs. Oughtred. IRC claimed tax on the basis of this formal share transfer. The HL held that this document is taxable. Had it not been held, tax avoidance would have been made extremely simple. But, later CA in Neville v Wilson accepted it.
Note, now contracts for the sale of land are specifically enforceable only if they are in writing – (Sec. 2 of Law of Property (Miscellaneous Provisions) Act 1989).

GRAINGE v WILBERFORCE:
Where a beneficiary declares himself a trustee of his equitable interest for another, with no active duties to perform; the declaration of sub-trust must be in writing, as the original beneficiary drops out of the picture – (Re Lashmar, Grainge v Wilberforce). The head-trustee would then hold the equitable interest for the sub-beneficiary.
Nelson v Greenings and Skyes dismissed the bare trust/active duties distinction, and disapproved ‘automatic dropping out of the picture’ of the sub-trustee. It is impractical for the head-trustee to take over the sub-trust as he may incur liability as a ‘trustee de son tort’ for any misapplication of or loss to the trust rights although he has no obligation under the sub-trust. He could always choose to pay the original beneficiary/sub-trustee and let him deal with his own sub-trust.
In any case, it is arguable that any disposition of one’s equitable interest under trust should require writing on policy consideration.

CONSTITUTION
Constitution of trust means that the settlor must transfer the rights which are to form the subject matter of the trust to the intended trustee, according to the proper requirements for transferring that particular type of rights.
Once a trust is perfectly constituted, the beneficiaries of that trust can enforce it against the trustee, whether or not they have given value to anyone in exchange for its creation. Like an outright gift, a settlor cannot revoke a perfectly constituted trust, on the ground that the beneficiaries given nothing in return. It matters not in the case of a perfectly constituted trust, that the beneficiaries are volunteers.
Where the settlor himself is to be the sole trustee, then no transfer is required as the rights are already vested in him. Note, trust would be constituted, where there is more than one trustee and the trust right has been vested in only one trustee – (Choithram v Pagrani: the settlor himself was to be a trustee among others, but failed to transfer the rights properly to the trustees. However, since he himself was the settlor, no separate vesting was required in respect of him. Therefore, the trust was held to be perfectly constituted, despite the failure to transfer the rights to the other trustees).
The issue of constitution arises where the settlor wants a third party to act as trustee – in that case the settlor has to transfer the rights according to the proper requirements, for different type of rights.
If it is a title to land which is unregistered, then the execution of a deed is necessary.
If it is a title to land which is registered, then the trustee will need to be registered as proprietor of the title. This can only be affected by the Chief Land Registrar, and he will do so on the instruction of the settlor given by the completion of the correct form of transfer and lodging it with the Land Registry.
If it is a title to a chattel (including cash) then it will have to be conveyed by either deed or delivery.
Equitable interests under trusts can only be transferred by writing, signed by the alleged transferor – (Sec. 53(1) (c) LPA 1925).
Shares can only be transferred (in a way similar to registered titles to land), by the appropriate form of transfer, followed by the making of an entry in the company’s books by the company secretary.
In the case of public companies traded on stock exchanges in the UK, there is now a computer-automated system of transfer of shares. But, this does not detract from the principle that legal title passes only on registration.
A debt can only be released at law by written instrument.
These methods discussed above are also those which need to be used to make outright gifts of the various types of rights. Where these methods are not complied with the general rule is that, equity will not assist a volunteer to perfect an imperfect trust/gift – as enunciated in Milroy v Lord.
Milroy v Lord further holds that, an imperfect attempt to create a trust using a third party as trustee, will not be interpreted as a declaration by the settlor of himself as trustee. The reason being, an intention to constitute another as trustee is inconsistent, with any argument that the settlor intended to make himself as trustee. Same reasoning applies in respect of the argument that, failed attempt to make gifts can be seen as having declared himself as a trustee – as seen in Richards v Delbridge. As pointed out in Milroy v Lord, if such arguments were accepted there would be no such thing as an imperfect gift.

EXCEPTIONS TO THE GENERAL RULE:
There are six situations in which courts have departed from the rule in Milroy v Lord –
Detrimental reliance: Where there is an imperfect gift/trust, there may be detrimental reliance on the part of the intended done/beneficiary. Detrimental reliance may lead the court to order the perfection of the imperfect gift/trust – (Dillwyn v Llewelyn, Pascoe v Turner). In the meantime, the purported transferor will hold the promised right on constructive trust for the intended transferee.
Although, it appears both conscious and justified, it might be asked why the law does not simply respond by forcing the defective donor to make the donee’s loss (compensation) or to give up his own gain (restitution) rather than making good his expectation.
The rule in Re Rose: An imperfect gift/trust will be perfected, where the alleged donor/settlor has done everything in his power to transfer the rights, but the transfer failed, because it required the assistance of a third party which was not provided (such assistance is required to transfer shares or a title to registered land, see above) – (Re Rose). The alleged donor/settlor will then hold the rights on constructive trust for the intended transferee.
The reason was said to be found in notions of common sense, which is of course, no reason at all.
The rule in Strong v Bird: Where the alleged donor intends to make a gift during his lifetime, but fails to vest the legal estate in the intended donee, the gift may still be perfected, if legal title subsequently vests in the donee by being appointed as an executor of the alleged donor (now deceased); provided the donor had a continuing intention to make the gift up until death – (Strong v Bird); as it is not possible for the executor to sue himself to recover the rights.
Following Re Ralli, in which Buckley J made reference to the rule in Strong v Bird in finding that a trust was constituted when the property fortuitously came into the hands of the trustee, it appears that the rule applies to perfect, not only imperfect gifts but imperfect trusts as well.
The rule in Re Ralli: Where the alleged settlor intends to create a trust during his lifetime, but fails to vest the legal estate in the intended trustee, the trust may still be perfected, if legal title subsequently vests in the trustee by being appointed as an executor of the alleged settlor – (Re Ralli).
Although, this looks similar to Strong v Bird, it would seem to form a separate rule, for it applies even though there is no continuing evidence of an intention to give. Indeed, given that, Re Ralli involved an unperformed promise to give, rather than a failed donation, any talk of a continuing intention is nonsense.
However, this was only an obiter pronouncement of Buckley J. Further, there is much talk in his lordship’s judgment of vague notions of conscience. Moreover, the decision is difficult to square with the earlier decisions of Re Brook, which was not cited to the judge.
Donationes mortis causa: Donationes mortis causa (singular donatio mortis causa) also called ‘deathbed gifts’, are gifts that are made inter vivos, but which are condition, only taking effect on death. If the donor revives and demands the property back, he is entitled to it. The typical case is – where I hand you my Rolex watch and tell you that, ‘if I do not survive the dangerous operation I am about to undergo, the watch is yours to keep’.
The problem with such a gift is that, it would seem to be a testamentary disposition, and therefore subject to the provisions of the Wills Act 1837, which prohibit oral wills. Yet, the courts have held such gifts valid.
The condition for the operation of the rule are laid down in Cain v Moon – (a) the gift must have been made ‘in contemplation though not necessarily in expectation of death’; (b) the donee must in some respect receive the property in question before the death of the donor; & (c) the gift must be made conditional on the donor’s impending death.
What does the requirement (b) means depends on the nature of the property. In respect of chattels, the donor must hand over either the chattel itself, or the means of getting control over it, e.g. key of a car (Woodard v Woodard). Where the subject matter concerns with chose in action, some ‘indicia (evidence) of title’ must be transferred, e.g. in case of shares the delivery of share certificate (Dufficy v Mollica); in respect of land the transfer of the title deeds (Sen v Headly: the CA recognised a donatio mortis causa of land for the first time).
Unconscionability: Court will order perfection of an imperfect gift, where it would be unconscionable for the alleged donor to resile from his gift – (Pennington v Waine).
However, no reason was given as to why it was unconscionable on the facts of the case. It is not clear why the gift has to be perfected, rather than compensating the relying party for his loss or stripping the donor of any extra advantage he would receive if the gift were now treated as invalid. Moreover, no member of the court in this case seems to have noticed that the facts of this case were very similar to that of Milroy v Lord. Furthermore, the court relied on Choithram v Pagarani as an authority, but this was a case involving express trust, while Pennington was a case of constructive trust.

COVENANTS TO SETTLE/PROMISES TO CREATE A TRUST
PROMISES IN DEEDS: A promise to create a trust (or to make a gift) contained in a deed, can be enforced at common law by those who are party to it, whether or not they provided any consideration for that – (Cannon v Hartley). On the other hand, equity will not enforce a promise merely because it is made by deed. In such a case, equity leaves the promise to his remedy at common law.
The typical case we are concerned with is that – where one person voluntarily (without consideration) promises another (i.e. intended trustee) by deed, that he will transfer rights to him in order to hold it on trust for a third party (i.e. intended beneficiary). Now, the question is, can the promise be enforced where the right holder fails to transfer the rights properly?
Enforcement by the intended beneficiary: The immediate problem for the intended beneficiary is that, he is not a party to the deed. So, he cannot enforce the promise. However, there are three exceptional situations, in which the intended beneficiary will nevertheless be able to enforce the promise.
Where the intended beneficiary is made party by statute: If the Contracts (Rights of Third Parties) Act 1999 were to apply to voluntary covenants (which is doubtful) then, the intended beneficiary can enforce the promise by virtue of Sec. 1 CRTA 1999 which provides that – a third party may in his own right enforce a term of the contract if the contract expressly provides that he may, or if the term purports to confer a benefit on him; unless on a proper construction of the contract it appears that the parties did not intend the term to be enforceable by the third party. However, it is doubtful whether CRTA 1999 applies to voluntary covenants.
Where the intended beneficiary can sue under this Act, he can obtain any remedy that would have been available had he been a party to the contract – (Sec. 1(5) CRTA 1999). This might be interpreted as meaning that the third party may obtain specific performance, if the subject matter is such that the remedy of damages would be inadequate. However, it is established that a volunteer may not obtain specific performance. It is difficult to see why a third party, who has not given consideration, should be in a better position under the 1999 Act.
Where the intended beneficiary is within the ‘marriage consideration’: At one time, it was traditional upon marriage for the woman’s father to set up a trust. The beneficiaries of that trust would be the husband and wife for their joint lives, the survivor for their life, remainder to any children of the marriage and in the event of there being no children, to the wife’s next of kin. The wife would also enter into a covenant, whereby she promised to convey to the trustees any rights she might later receive above a certain value to be held on the same trusts, problem ensues when she fails to do it.
Unlike common law, equity regards marriage as ‘the most valuable consideration imaginable’ – (AG v Jacobs-Smith).
The parties to the marriage and the issue of it are within the ‘marriage consideration’, that is to say – equity treats them as if they had given consideration; as the very reason for entering into the covenant is to provide for the family. But note, next of kin are not considered as within the marriage consideration – (Re Plumptre).
They are, therefore, able to obtain specific performance of the covenant where the purported settlor fails to transfer the rights – (Pullan v Koe). Pullan v Koe also envisaged, equity would not only insist on specific performance of the covenant, but also treat the trust as constituted by way of constructive trust the minute the covenant could be performed.
The doctrine of ‘marriage consideration’ applies only to marriage settlements; that is settlements made in contemplation of marriage. A husband and wife who are already married, who set up a trust for themselves and their children, do not create a marriage settlement; accordingly the children are not within any marriage consideration.
Trust of the right to sue on the covenant: An argument has sometimes been made that, the intended trustee being a party to the covenants, holds his common law right to sue on the covenants, on trust for the intended beneficiary. If this argument is made out, then intended beneficiary can compel the intended trustee to exercise his right and sue alleged settlor. The assumption is that, the intended trustee would recover substantial damages at common law, which he would then hold on trust for the intended beneficiary.
However, the argument is unlikely to succeed for two reasons as illustrated by Re Cook. First, requisite intention to create a trust of the right to sue will generally be missing. Secondly, a trust of the right to sue is only possible in the case where the covenant relates to present, rather than future rights.
Enforcement by the intended trustee: Being a party to the covenant, the intended trustee undoubtedly has a claim at law for breach of covenant. But, cases like Re Pryce, Re Kay, Re Cook suggest that, he will be prevented by a court of equity from suing the alleged settlor at law, because to do so would give the intended beneficiary by indirect means what he could not obtain directly.
It might be argued that, in deciding the point Eve J in Re Pryce went beyond the rule ‘equity will not assist a volunteer to perfect an imperfect gift’, as the trustees were not asking for equity’s assistance, merely a ruling on whether they had to sue. These cases have never been overruled, though it should be noted that they are only decisions at first instance.
Even if a court were to accept that these authorities were wrongly decided, and accordingly were the intended trustee allowed to sue, a further question arises as to what damages he would recover; substantial or nominal?
The normal measure of damages will be loss of expectation, i.e. substantial damages. There is an argument which says that the intended trustee suffers no loss of expectation, because had the covenant been performed, he would now be a trustee and a trustee makes no personal gain from his trusteeship. Thus, all that he is entitled to is an order for nominal damages. For this reason, Eve J was correct in Re Pryce to deny the intended trustee to sue the alleged settlor at law, but not for those reasons given in that case.
The problem with this argument is that, it forgets that the intended trustee’s claim is being brought at law, and at law, he would hold any rights transferred to him for himself and not for the intended beneficiary; as the common law does not recognize trusts. Thus, in the common law’s eyes he has suffered a substantial loss and he should recover substantial damages.
Even if intended trustee were to recover substantial damages, there is an argument that he would hold the damages on resulting trust for the alleged settlor. The reason is said to be the failure (for the reasons identified in Re Cook) of trust of the right to sue on the covenant in favour of the intended beneficiary, accordingly a resulting trust of the right to sue in favour of the alleged settlor arises. Since, the right to sue is held on trust for alleged settlor, so will be any damages acquired through the exercise of that right.
This reasoning justifies the decision of Re Pryce as this would appear that the intended trustee is immediately liable to repay the damages to the alleged settlor, the person he had just sued to recover it. This circularity of action is of no body’s interest.
However, the difficulty with this argument is that, any trust in favour of the intended beneficiary cannot be said to have failed at all. It was simply the case that, no trust arose in favour of him, because there was no declaration of trust in his favour. In other words, the trust did not fail, only the argument that there was a trust. The damages will instead be held by the intended trustee on trust for the intended beneficiary, for the damages is simply the law’s substitute for performance of the covenant.

PROMISES FOR CONSIDERATION:
If consideration in normal contractual sense has been given in exchange for the promise to set up a trust (or to make a gift), and that promise has not been performed, the promise will be enforceable at law through a claim for damages or; if damages are thought to be an inadequate response, by an award of specific performance in equity.
The promise cannot be enforced by the intended beneficiary/done as he is not a party to it – (Re Cook). Re Cook was of course decided before the enactment of the CRTA 1999, and it should now be asked whether Sec. 1 of that Act might give the beneficiary/donee a right to sue in his or her own name.

PROMISES AND DETRIMENTAL RELIANCE:
Where a promise to create a trust/make a gift, has been made and the promisee has relied on the promise to his detriment, i.e. changed his position to his detriment in the reasonable belief that the promise would be performed; equity (though not the common law) will compel the promisor to perform his promise. Meanwhile, the promisor will arguably hold the right in question on constructive trust for the promisee.

SECRET TRUST
A full secret trust (FST) arises where a will states that some right is left to a person as an absolute gift, but the testator has agreed with the person that he is to hold the rights as trustee for another. There is therefore no mention of any trust in the will.
Example: A clause in John’s will states – ‘I leave £1000 to Smith absolutely’. However, John had told Smith that he is to hold the £1000 on trust for John’s friend Sarah. This is a FST, as the will makes no mention of any trust whatsoever, and it appears to be an absolute gift.
A half secret trust (HST) arises where some right is left by will to a person and the will states that he is to hold the rights on trust, but the terms of the trust are not declared in the will. Instead, the terms are agreed between the testator and the person. Therefore, here the will mentions the existence of a trust, but not the terms of the trust.
Example: A clause in Barbara’s will states – ‘I leave £1000 to Steve on trust for the purpose I have communicated to him’. Barbara had told Steve that he is to hold the £1000 on trust for David. This is a HST, as the fact that it is a trust is mentioned in the will, but not the details.
Now, look at the following provisions of the Wills Act 1837 –
All testamentary dispositions (that is, dispositions made in a will) must be in writing, and signed by the testator, in the presence of two or more attesting witnesses present at the same time – (Sec. 9 WA 1837).
Gifts to attesting witnesses and their spouses are void – (Sec. 15 WA 1837).
Any alteration of the will (a codicil) must be in the same form as the will, i.e. signed, witnessed and writing – (Sec. 20 WA 1837).
Thus, enforcement of secret trust seems to be inconsistent with the Sec. 9 of the WA 1837, as the evidence required to prove the making of the declaration of trust is not in the form sanctioned by Sec. 9.

JUSTIFICATION FOR ENFORCING SECRET TRUSTS:
The earlier justification for enforcing secret trust was based on ‘fraud theory’ developed in Rochefoucauld v Boustead, that is – a statutory provision designed to prevent fraud, cannot be used to commit a fraud. A legatee who takes on the basis that, he is a trustee, but who later seeks to rely on the statute to take absolutely – would be using the statute to perpetrate a fraud. Consequently, evidence not in the form sanctioned by Sec. 9 would be allowed.
However, there are a number of problems with this fraud theory –
It fails to address the point that such evidence is inherently unreliable.
The reasoning of this theory is circular, because it already bases on an assumption that a trust exists, in order to allow evidence to prove fraud.
Most importantly, it struggles to explain the admission of evidence in cases of HST. If the evidence is refused admission there, there will be an automatic resulting trust for the testator’s estate, and therefore no possibility of any personal gain by the trustee.
For that reason, Lord Buckmaster and Hailsham in Blackwell v Blackwell, had to redefine the fraud which the courts were trying to prevent is not just a personal gain to the trustee, but the defeating of the expectations of the secret beneficiaries or the disappointment of the wishes of the settlor. But, this expanded fraud theory is no less circular, than the previous ‘narrow’ fraud theory.
An alternative theory is that put forward by the HL in Cullen v AG for Northern Ireland, and adopted by Viscount Sumner and Lord Warrington in Blackwell v Blackwell; which is now conventionally called the ‘dehors (outside) the will’ theory.
Lord Warrington in Blackwell v Blackwell said – “What is enforced is not a trust imposed by the will, but one arising from the acceptance by the legatee of a trust communicated to him by the testator, on the faith of which acceptance, the will was made or left unrevoked”. Viscount Sumner expressed himself in similar terms – “It is communication of the purpose to the legatee, coupled with acquiescence or promise on his part that removes the matter from the provision of the Wills Act and brings it within the law of trusts”. Under this – secret trusts are regarded as inter vivos declaration of trust by the testator; the only atypical feature is that, the trusts are not constituted, i.e. the rights are not transferred into the hands of the trustee, until the testator’s death. Their essence is the acceptance of a personal obligation by the legatee.
However, there are a number of problems with this theory as well –
It still fails to address the fundamental objection that the evidence the court admits is inherently unreliable.
It assumes a false dichotomy between the law of trusts and the law of wills.
It does not explain, why the acceptance of the trust by the trustee should be important, there being no requirement in English law that a trustee accept the trust imposed upon him.
Most importantly, it is founded on an unduly narrow interpretation of what is a ‘will’. It assumes that the ‘will’ is the formal document executed by the testator. But, this is not what the statute means by a will, for, as we have seen, prior to the Statute of Frauds 1677, wills could be made orally. The ‘will’ mentioned in the statute is the totality of the testator’s wishes as to the distribution of his rights on his death, and the intention that certain rights be held on trust for others is undoubtedly part of that ‘will’.
Note that, both the fraud theory and outside the will theory are still in play, though some judges talk of the dehors (outside) the will theory as representing the ‘modern view’, which explains the admission of the otherwise inadmissible evidence.

REQUIREMENTS OF A SECRET TRUST:
In order to prove a FST or an HST, the evidence must show: (1) the intention of the testator to create a trust, satisfying the traditional requirements of three certainties; (2) timely communication of that intention to the intended trustee; & (3) timely acceptance by the intended trustee of the trust obligation – (Ottaway v Norman).
INTENTION:
The principle here is the same, as with any form of trust: there must be certainty of intention. Interestingly, if an attempt to create a HST fails for uncertainty of intention, the conclusion will be – there is an absolute gift on the face of the will; thus FST will be a possibility.

COMMUNICATION:
For FST, the communication must be made to the intended trustee, before the testator’s death – (Re Boyes, Wallgrave v Tebbs). The reason is, if the legatee did not know he was intended to be a trustee, he could hardly be said to have accepted or acquiesced in his appointment – (Re Boyes).
In case of FST, it is not sufficient to communicate merely the fact of the trust to the intended trustee, the terms of the trust must also be communicated to him – (Re Boyes). However, handing over the terms in a sealed envelope which is not to be opened until the testator’s death would suffice, provided the legatee knows that it contains the term of the trust – (Re Keen).
For HST, communication must be made before or at the time of the execution of the will – (Re Keen: this is known as the ‘broad ratio’ of Re Keen). In Re Keen, the testator gave a sum of money to the intended trustees ‘to be held upon trust and disposed of by them among such person, persons or charities as may be notified by me to them or either of them during my lifetime’. This would be void independently of the question of inconsistency discussed below. The reason being; otherwise the testator would be able to change his will without the execution of a codicil.
However, the difficulty with this reasoning is that, if the trust really does arise outside the will, then a change of mind over the terms of the trust is not a change in the will at all. Moreover, it would seem strange that a communication post-execution be acceptable in the case of a FST but not in the case of an HST. It should also be noted that the finding in Re Keen is arguably obiter as the secret trust was struck down on a different ground; which is known as the ‘narrow ratio’ of Re Keen. Nevertheless, in Re Bateman, the rule was applied as if no doubt could be entertained about it.
The ‘narrow ratio’ in Re Keen holds that, inconsistency between the terms of the will, and the evidence sought to be admitted to prove the existence of secret trust; is fatal. In Re Keen, the communication was meant to be done in future (‘may be notified by me to them during my lifetime’), but it was already done by a sealed envelope prior to the execution of the will. This inconsistency lead the court to hold that there was no secret trust.
Where there are intended to be two or more trustees of a FST, but communication is not made to all, Re Stead provides that, only the trustee to whom communication is made is bound, unless the trustees take as joint tenants and communication is made before the making of the will. There is no case law on this point for HST. In case of HST, the trustees will certainly take as joint tenants, since legal title can no longer be held under tenancy in common because of Sec. 1(6) TOLATA 1996. Now, if Re Keen is correctly decided, the communication will anyway have to precede the execution of the will; accordingly all will be bound.
A testator must communicate not only the trusts and the terms, but also the identity of the right to be held on trust – (Re Colin Cooper). In Re Colin Cooper, the testator attempted to add further property to an existing secret trust, but did not communicate about this additional property to the secret trustee. It was decided that, the trustee held the additional property (not all, only the additional property) on resulting trust, since it was a HST. Had it been a FST, the trustee would have taken absolutely. Alternatively, one might argue that, once obliged, even the fully secret trustee will hold the additional rights on resulting trust.

ACCEPTANCE:
The rights will only be subject to a secret trust, if the intended trustee accepted that he would hold it on trust – (Ottaway v Norman). Silence will count as an implied acceptance – Moss v Cooper.
 
FAILURE OF A SECRET TRUST:
 In the case of FST, where the testator fails to communicate it to the intended trustee before his death, then the trustee will take the rights absolutely – (Wallgrave v Tebbs, Proby v Landor).
However, if the legatee is told that he is to hold on trust, but the terms of the trust are never communicated to him within the testator’s lifetime, he will hold the right on resulting trust for the testator’s residuary legatee or intestate successors – (Re Boyes); because, the legatee had been told of the existence of the trust and hence, it would be fraudulent for him to take beneficially.
In the case of HST, the result of any failure of the testator to specify the trusts will be that, the trustees will hold the rights on resulting trust for the testator’s residuary legatee or intestate successors – (Re Colin Cooper).

PREDECEASE OF THE SECRET BENEFICIARY:
 If the intended secret beneficiary predeceases the testator, his interest under the trust would lapse, and fall into the residue. However in Re Gardner (No.2) Romer J advocated that – secret trusts are inter vivos trusts, so the rights should go to the beneficiary’s estate. But, this decision is doubtful on the basis that, the intended trust would not be effected until the testator’s death.

PREDECEASE OF THE SECRET TRUSTEE:
If an half secret trustee dies before the testator, the trust will not fail, provided the terms can be ascertained; as equity will not allow a trust to fail for want of trustees.
However, the secret trust doctrine seems to be based on an acceptance or acquiescence by a particular trustee and it is because of that agreement that the situation is taken out of the province of the statutory rules on formality, if that is right, then the doctrine of lapse should apply and the gift should fall into the residue.
Under the law of succession, a bequest lapses and fails if the legatee predeceases the testator. As a matter of logic, the predecease of a full secret trustee should therefore cause the trust to fail. This was accepted in an obiter pronouncement by Cozens-Hardy MR in Re Maddock where he suggested, if a full secret trustee dies before the testator, the trust will fail and the rights will fall into residue.

WITNESSING BY SECRET BENEFICIARY:
Sec. 15 applies only in respect of gifts received under a will, but the dehors the will theory suggests that the beneficiary receives the gift under a trust which operates outside the will. Accordingly, Sec. 15 would not invalidate a secret trust, where the secret beneficiary witnesses the will – (Re Young: although this was a case of HST, the principle is equally applicable to a FST).

WITNESSING BY SECRET TRUSTEE:
Sec. 15 will invalidate a FST where the trustee witnesses the will, as the trustee takes beneficially on the face of the will.
Sec. 15 should not invalidate a HST where the trustee witnesses the will, because the will itself shows that he is intended to be a trustee. But, the dehors story says that, he is not a trustee under the will. So, Sec. 15 should also apply to him as much as to a trustee of FST.
Of course, one might argue that Sec. 15 should not apply to any trustee, secret or not, since none takes a beneficial interest, but such a view cannot stand alongside Re Young, because that would be to have it both ways.

DISCLAIMER BY THE SECRET TRUSTEE:
AFTER DEATH OF THE TESTATOR: In Re Maddock, Cozen-Hardy MR was of the view that a disclaimer by the full secret trustee after the testator’s death has the effect of invalidating the trust, seemingly on the basis that it depends upon the existence of a personal obligation binding the trustee.
Despite obiter dicta to the contrary by Cozen-Hardy MR in Re Maddock, Lord Buckmaster in Blackwell v Blackwell stated obiter that a FST will not fail if the trustee disclaims the trust subsequent to the testator’s death. This is logical, because the trust becomes constituted by the will on the death of the testator, so it should not be allowed to fail for the want of trustee. Oakley argues that there should be a constructive trust on the authority of Bannister v Bannister. But, Bannister was concerned only with a question of admissibility of parol evidence to prove a declaration of trust, and does not lay down a general rule. In any case, it would seem to be nothing more than the imposition of a remedial constructive trust.
In the event of a disclaimer by the trustee in HST, it is assumed that equity will not allow a trust to fail for want of trustee.

BEFORE DEATH OF THE TESTATOR:
A secret trustee is entitled to revoke his previous acceptance of the trust at any time prior to the death of the testator, provided it is communicated to the testator.
In case of FST, if the testator has had time to amend his will after the trustee’s revocation of acceptance before his own death, but leaves it unaltered; the absolute gift on the face of the will takes effect. On the other hand, if he had had no time to make alternative arrangements before his own death, after the trustee’s action; equity may act to prevent fraud or inequitable conduct on the part of the legatee by providing a new trustee or compelling the legatee to hold the rights on resulting trust for the testator’s estate.
In case of HST, the rights will be held on resulting trust for the testator’s estate.
But note, since both the fraud theory and dehors theory emphasise the fact that, the trigger for admissibility is the acceptance of the trust by the secret trustee, it can be argued that where acceptance can be revoked, the justification for admitting the evidence contrary to the WA 1837 is gone.

CAN A SECRET TRUSTEE BENEFIT FROM THE TRUST:
Trustee of a HST cannot bring evidence to show his entitlement to the trust rights, because to do so would be inconsistent with the terms of the will, which identifies him as a trustee – (Re Rees). In Re Tayler, Pennycuick J stated obiter that he did not find this reasoning ‘easy’ and that in principle, evidence is admissible to prove all the terms of the trust, including a trust in favour of the trustee.
Swadling attempted to reconcile these two cases on the basis that, in Re Rees the secret trustee was a professional one, being a solicitor; if the settlor really wanted him to take beneficially then the trustee should have made it in the will. On the other hand, in Re Tayler, the secret trustee was just a friend of the testator, not his solicitor.
In cases of FST, there is no objection to the intended trustees leading evidence to prove that they were intended to benefit in accordance with the intention of the testator, for the will does not name them as trustees but as apparent beneficiaries; accordingly, there will be no contradiction of the will. This was seen in Irvine v Sullivan.

WHAT TYPE OF TRUST IS ULTIMATELY ENFORCED?
Whether secret trusts are express trust or constructive trust, becomes an important question in the case of a secret trust of a title to land; as Sec. 53(1)(b) LPA 1925 provides that only written and signed evidence is admissible to prove a declaration of trust respecting a title of land. Thus, many scholars and judges think that secret trusts are constructive trust, to utilize the exemption set out in Sec. 53(2).
However, secret trusts are undoubtedly express trusts, since both the fraud theory and dehors the will theory are justifications (although not very strong ones) for the admission of evidence of a declaration of trust not in a form sanctioned by the statute.
That does not mean that the evidence of the declaration will be inadmissible, for the fraud theory in Rochefoucauld could be harnessed, at least in the expanded form used in Blackwell v Blackwell, to directly overcome the Sec. 53(1)(b) rule.
Moreover, Paul Todd argues, FST and HST depend on a principle of equity which, if not defeated by Sec. 9 Wills Act 1837, will not be defeated by Sec. 53(1)(b) LPA 1925 or any other provision intended to prevent fraud, irrespective of whether the trusts are express or constructive.

CAN A FST BE REVOKED BY THE TESTATOR?
 Whether a testator can revoke previously declared FST and revert to the position as it is expressed on the face of the will (i.e. an absolute gift to the legatee), depends on when the FST comes into existence. If for example, the FST comes into existence the moment the legatee/trustee accepts the trust obligation, then such an attempt will be ineffective; as once a trust is created, the beneficiaries rights becomes perfect (Milroy v Lord). On the other hand, if the FST comes into existence at the testaor’s death, then clearly the trust can be revoked at any time before the testator’s death.
Under the dehors theory, secret trusts arise on the testaor’s death. The judgement in Blackwell v Blackwell and Re Maddock supports this view. The logic of this approach is that the communication to the trustee during his lifetime, and acceptance thereof, is simply an inter vivos declaration of trust, and that the trust becomes constituted when the rights passes into the hands of the trustee on the testator’s death. Thus, the FST can be revoked at any time before death.
However, if the rationale for enforcing secret trusts is to prevent the trustee/legatee perpetrating a fraud by agreeing to accept a trust obligation and then claiming the trust rights himself; then this policy can be upheld only by deciding that the trust becomes constituted at the date of the acceptance of the obligation. Such a view of secret trust is implicit in Re Gardner (No. 2) and Sonley v Clock Makers Company. Thus, one might try to argue that FST is not revocable once declared. However, it should be noted that, Sonley was a case of HST and Re Gardner has attracted considerable academic criticisms.

PRIVATE PURPOSE TRUST
Look at the following examples:
X transfers property to Y, for him to hold as trustee for Z. This is a straightforward private trust, where property is to be held on trust for a particular beneficiary, and it is of course valid.
X transfers property to Y, for him to hold as trustee for charitable purposes. This is not a trust for particular persons, and is therefore not a private trust. Instead, it is a trust for purposes, and because those purposes are charitable, the trust is valid, even though there are no identifiable beneficiaries.
X transfers property to Y to hold ‘for the maintenance of good relations between nations ……….. The preservation of the independence of the newspapers and similar purposes’. This is also a trust for purposes, rather than persons; but here the purposes are not charitable. This is a non-charitable purpose or private purpose trust.
A private purpose trust is generally void as – it does not clearly indicate a class of beneficiaries, there is no one who can enforce the trust; and purposes generally do not expire within a limited time, thus tend to violate rule against perpetuity.
However, a power to carry out a purpose is perfectly valid. There is no similar problem of enforcement with respect to powers, for there is no duty to carry out powers, and so no problem of non-feaseance. As regards misfeaseance, those who would take in default of appointment have the power to bring the trustee to book, if he purports to exercise the power in a way which is outside the intended purpose.

TRUST OR POWER OF APPOINMENT?:
It may be unclear whether a settlor intends to impose a trust on the recipient or intends merely to give him a power of appointment. Consider: ‘ 100,000 to my trustees for distribution to such of my relations as my trustees shall in their absolute discretion thinks fit’ – does ‘for distribution’ mean ‘to be distributed’, an imperative direction imposing an obligation upon them to distribute; or is it rather to be interpreted as ‘available for distribution’, thus creating no obligation but giving the trustees a power that they may exercise if they so choose?
Several rules of construction may determine whether a trust or a power of appointment is intended:
If there is an explicit ‘gift over’ in default of appointment, e.g. ‘my shares in X company to my trustees, with power to appoint to my children in such portions as my trustees shall in their absolute discretion decide, and in default of appointment to King’s College London’ – there is a power of appointment, because if the settlor provides for the case where the trustees do not appoint, clearly they are under no duty to do so.
Note, the ‘gift over’ must specifically arise on default of appointment. A residuary gift is not a gift over for this purpose. Residuary clauses deal with failures of all kinds; nothing therefore can be inferred about any specific gift just because the will contains a residuary clause.
If there is no ‘gift over’ in default of appointment, one must determine the true intentions of the testator by construing the will, or settlement as a whole. For example, if the testator uses words that clearly indicate his intention to create a trust with respect to some of his gifts, but does not in the gift under consideration, the court is apt to conclude that there is no trust – the testator knew what words to use to create a trust and in respect of this gift did not (see Re Week).
The settlor’s use of word ‘power’ is not determinative, but words such as ‘shall’ or ‘to be’ as in ‘shall distribute’ or ‘to be divided amongst’, seem quite clearly to be imperative, strongly indicating the imposition of a duty, and thus a trust.
The court will not, however, validate a purpose trust by construing it as valid power for purposes – (Re Shaw).

EXAMPLES OF SOME ANOMALOUS VALID PURPOSE TRUSTS:
Trust for monument and graves: A trust for the erection or maintenance of monuments or graves with reasonable provision is valid – (Mussett v Bingle, Re Hooper); but not with very general provisions, such as ‘some useful memorial to myself’ – (Re Endacott).
These trusts can be charitable, where they are for a tomb which can be regarded as part of the fabric of the church – (Hoare v Osborne); or for the maintenance of a churchyard in general – (Re Vaughan).
Trust for individual animals: A trust for the benefit of specific animal is valid – (Re Dean: This case is somewhat dubious on perpetuity ground. This was not clearly discussed). Trust for animals in general, or particular species is valid as charitable trusts.
Trust for the saying of masses: A trust for the saying of masses in private is valid – (Bourne v Keane). Trust for the saying of masses in public is valid as charitable trusts – (Re Hetherington).
Apart from these situations, there are few more instances where private purpose trust was held valid; but these are not relevant for exam purpose.
It is clear that the courts will not allow any new exceptions, and will not extend the existing categories – (Re Endacott, Re Astor).
One important point to note that, yet these trusts have to comply with the rule against perpetuity to be valid.
Where an anomalous trust of one of these kinds is upheld, the court will make a Pettingall order (Pettingall v Pettingall) under which the trustee or executor of the will undertakes to carry out the purpose, and the court grants leave to those persons who would receive the funds if the gift had been declared invalid, to approach the court, if the trustee or executor fails to carry out the purpose or misuses the fund.
 
 
TRUST FOR PERSONS LIMITED BY A PURPOSE:
Where the beneficiary’s interest under the trust is determined by the expense of a certain benefit for him (e.g. trust for advancement, education, maintenance of settlor’s son), these are not purpose trusts, rather trust for persons limited by a purpose – often termed as ‘Re Sanderson’ type trust.
Accordingly, the trust may exhaust any funds set aside for it, or may fail to exhaust those funds. In case of Re Sanderson type trust, the beneficiary only has a right under the trust commensurate with the named expense; and any remaining funds must be disposed of by way of a gift over, or will otherwise go under an automatic resulting trust.
The principle of construction was laid down in Re Sanderson as follows: ‘if a gross sum be given, or if the whole income of the property be given, and a special purpose be assigned for that gift, the court always regards the gift as absolute, and the purpose merely as the motive for the gift, and therefore holds that the gift takes effect as to the whole sum or the whole income, as the case may be.’
Thus, in Re Andrew and Re Osoba, the beneficiary was allowed to take the surpluses. But in Re Abbott a true Re Sanderson type trust was found, on the basis that the beneficiaries in that case had died. Accordingly, there was a resulting trust in favour of the testator’s estate.
RE DENLEY APPROACH: A trust for non-charitable purpose which would otherwise be void may nevertheless be upheld, if the purpose is not of an abstract kind, rather it directly or indirectly confers benefit to ascertained/ascertainable individuals – (Re Denley: the case concerned a land which was to be held on trust ‘for the purpose of a recreation or sports ground primarily for the benefit of the employees of the company and secondarily for the benefit of such other person or persons (if any) as the trustees may allow to use the same’). Note, these trusts yet have to comply with the rule against perpetuity.
Subsequent commentary has tended to treat the case as merely one of a particular kind of discretionary trust (see Re Grant); or as a trust for persons, with the purpose being treated merely as a ‘superadded’ direction or motive for the gift (see Re Lipinski). In other words, the case appears to have been read so as to deny that it represents a departure from the beneficiary principle. The case of Re Denley therefore remains a dubious one.
 
UNINCORPORATED ASSOCIATION (UA)
An UA exists where two or more persons are bound together for one or more common purpose by mutual understandings, each having mutual duties and obligations, in an organization which has rules identifying in whom control of the organization and its funds is vested, and which can be joined or left at will – (Conservative and Unionist v Burrell).
UA is one of interest here because, if an individual wishes to devote his money to the carrying out of a purpose, in particular after his death, then one way of doing so, is to give money to an UA the purpose of which is the same of the transferor.
But the problem is, UA itself has no legal personality, and so has no capacity to hold rights. Such a transfer to UA, thus, appears to be void, as a transfer of rights to a tree or pet cat would be.
However, the court found a way of validating such transfers as people insist on trying to transfer rights to UA. Their method is to say that, though the association itself has no legal personality, its members or more often its officers do. The transfer is, therefore, construed as one to the members or officers of the association.
 
CONSTRUCTION OF A GIFT TO UA:
 Initially, the courts viewed such transfers as made to all the members personally (either directly as co-owner of the right; or through the medium of a trust, the treasurer or other officers of the club acting as trustees). This interpretation would allow each member of the association to claim their share of the right, and use it as they wished. Obviously, this would have been against the intention of the donor.
This problem led the court to regard the transfer as made to the members on trust not for themselves, but for the purpose of the association. This interpretation turned out to be a disastrous for the transferor. Because, the transfer would have been invalid as a private purpose trust, unless the purposes were exclusively charitable, or a type falling within the Re Denley principle.
 
The more recent and preferable approach is to construe such transfers as made to the members themselves, but subject to their contractual obligation between themselves to use the rights so given, to promote the purpose of the association – (Re Recher, Re Liplinski). This is known as the ‘contract-holding theory’.
This construction works, as it uses a contract to control the expenditure of the fund, rather than the terms of a trust, so that the carrying out of the purpose is enforced by way of compliance with a contract, not by way of a trustee’s carrying out a private purpose trust.
However, there are two versions of this theory – purely contractual approach & bare trust/contractual mandate solution. 
Purely Contractual Approach: This has been interpreted to mean – the members all individually hold the rights as co-owners, so there is no trust, but they are bound by their contract between themselves, to deal with the rights as decided by the association rules.
The problem with such a purely contractual approach is that, a person cannot give rights subject to another’s contract (i.e. you cannot give a friend rights to be held on the terms of a contract with another party, for a contract is a personal obligation between individuals, and you are not a privy to their contract). Although, Matthews provided a solution to this (by incorporating a provision to the association’s rules that any gift or contractual payment received are taken members individually subject to their contract); this seems fanciful.
Bare Trust/Contractual Mandate Solution: The best interpretation is – the rights are held as an ‘accretion to the funds of the association’; which means such transfers are construed as a gift to the treasurer-trustee of the UA, to hold the rights on the same trust as he does the other rights of the UA.
One thing to note in particular, the contract-holding theory does not ensure that a donor who donates money to an association for the use of the association in its activities will get his wish, for an association may always change its rules or goals, and thus devote its funds to other activities, or even to decide disband and split up the remaining funds at any given time. In light of this, Swadling has proposed that, a disgruntled donor may claim that his gift was made on condition that it be used for a specified purpose. This is not a transfer on purpose trust, but a conditional gift recognized at common law. If this construction can be put on the gift, and the gift is not devoted to the purpose, the donor can bring a personal action under the law of unjust enrichment, for restitution of an equivalent sum.
Another possibility which is related to the mandate idea is to utilize the type of trust upheld by the HL in Barclays Bank v Quistclose, as a means of achieving an abstract purpose trust. According to Lord Millett’s analysis, a borrower holds rights on resulting trust for the lender but with a power (or in some case, a duty) to carry out the lender’s revocable mandate. It is not a purpose trust, but simply a trust subject to a power to apply the rights for a purpose.
To achieve the bare trust/contractual mandate solution, the courts have been prepared to ignore words which in other contexts would suggest that a true purpose trust is intended (for example, ‘solely’ in Re Lipinski). In Re Lipinski, Oliver J reasoned that, since a valid gift may be made to an UA as a simple accretion to the funds, why should a gift specifying a purpose be invalid? He also considered the expressed purpose could be treated merely as a motive for the gift, not as a trust obligation. Alternatively, he said that the gift could be held valid on the authority of Re Denley.
But, the court will not construe a gift as one to the members, where the members do not have complete control of the rules of the association – (Re Grant: the testator left his entire real and personal estate to the Chertsey and Walton Constituency Labour Party. Its rules were capable of being altered by an outside body, the National Labour Party. Accordingly, the gift failed).

THE CONSTRUCTION OF TRANSFERS TO POLITICAL PARTIES:
In Re Grant, Vinelott J held that the local Labour Party was not an UA, as its rules were capable of being altered by the National Labour Party. But, this reasoning is doubtful. In Conservative and Unionist v Burrell, the court held that, political parties are ‘political movements’, not UA. Since, movements do not have the membership of an association, the bare trust/contractual mandate solution could not work. But, again it is not clear, why political parties cannot be associations with a definite membership, even if they are political.
Instead, in Burrell the court explains a gift to the party as given under an agency arrangement, the treasurer of the party to use the funds for the purposes of the party, as an agent of the donor. This construction in unsatisfactory in more ways than one, failing in particular to account for testamentary gifts, for when a will comes into operation the testator is dead, and a dead person cannot be a principal for any agent.

DISTRIBUTION OF RIGHTS UPON DISSOLUTION OF UA:
An association may be wounded up in a formal manner, but in the absence of formal dissolution the question arises as to the circumstances which will justify a finding that the UA has ceased to exist. In Re G.K.N. Bolts and Nuts Ltd, it was held that ‘mere inactivity is not sufficient, unless it is so prolonged or so circumstanced that the only reasonable interference is spontaneous dissolution in which case the court must select a date’.
Where the association’s purposes were charitable and the rights were held by its officers/members on trust for charitable purposes, any remaining funds will be applied cy-près (an old French legal term meaning, approximately ‘as near as possible’) the same purpose – (Re Vernan, Re Finger). This means, the surplus funds will be applied to a different but similar purpose.
However, where the rights were held on trust for valid non-charitable purposes, different rules apply. If the rules of the association make provision regarding the distribution of rights on dissolution, those rules will be decisive, and the court will merely give effect to such rules.
But, in the event where the rules are silent as to the destination of the surplus funds on dissolution, traditionally surplus funds were held on resulting trust for members in proportion to their contribution – (Re Printers, Re Hobourn); on the basis that, the funds were held on purpose trust, regarding the purpose as the purpose of the association, which then fails because of the dissolution. But this is wrong, as the winding up of the association will not in itself be a winding up of the purpose.
An alternative approach was seen in Cunnack v Edwards, where a surplus remained in a friendly society’s funds after the death of the last widow annuitant. The CA held that the members had contributed on the basis of contract, i.e. each payment was made in consideration for the payment of an annuity to the subscriber’s widow. Each member had therefore enjoyed their full contractual entitlement from the fund. Thus, the surplus went bona vacantia to the Crown.
However, donors donate to the purpose trust thereby contracting with the trustees to obtain a particular personal benefit out of the purpose – makes a nonsense of the idea that the money is held genuinely on purpose trust. Cunnack v Edwards is an anomalous case, and it is difficult to put any reasonable construction on it.
This approach was also used in Re West Sussex, to deal with parts of the surplus remaining on the dissolution of a police benevolent fund. Goff J held that, members’ contributions, together with funds raised by way of entertainments and raffles, had all been given on a contractual basis, and should therefore go bona vacantia to the Crown. The part which represented the contributions of identifiable donors, however, was held for such donors on a resulting trust.
It is difficult to see why third party contributors, even if identifiable, should have any claim in such circumstances. The validity of the initial gift is usually explained on the basis that, it is an absolute gift to the members of the association. If that is so, such contributions should be dealt with on the same basis as the rest of the funds. Thus, it is submitted that there is no room for a resulting trust for third parties.
The more acceptable modern solution to this issue is the one advocated in Re Bucks, which applied the modern contract-holding theory and held that – on dissolution, the funds were held by the members of the association outright.
Under either version of the contract-holding theory, the surplus funds will simply be distributed among the membership existing at the time of the dissolution. Because, under the contract-holding theory the members of the association hold the rights outright. When the association dissolves, their contractual obligation to use the rights in a particular way simply disappears, so they can distribute the rights to themselves in equal shares, or do anything else with them they want, so long as they do it consensually according to the rules of the association.