Sample Essay Questions

Question 1

In order for a settlor to create a valid inter vivos trust of property he owns absolutely, it is necessary both to constitute the trust perfectly and to meet certain statutory requirements. Discuss.

Answer Plan

  • Brief summary of the concept of a trust.
  • The need for title to the trust property to be effectively conveyed to the trustee, if not already present (Milroy v Lord (1862)).
  • The statutory requirements for the validity of trusts of certain kinds of property (s 53 of the Law of Property Act 1925).
  • The consequences of imperfect creation (brie?y).


The trust is a unique creation of the common law, and one of its most versatile concepts. Moreover, when the trustee holds property on trust for the bene?ciary, the trustee is submitting to one of the most powerful forms of legal obligation known to English law. A trustee may well ?nd himself bound by onerous duties, either imposed by the trust instrument, by statute, or by general principles of equity, and there are serious consequences awaiting any trustee who neglects his duties or breaches the terms of the trust. Likewise, the trustee is only the mere 'paper' owner of the property and, unless he is also a bene?ciary, will derive no bene?t from the trust property. The trustee holds the trust property 'on trust' for the bene?ciary for the bene?ciary's use and bene?t, and the trustee may even be prohibited from being paid for his labours (although this will be rare for professional trustees: see ss 28-30 of the Trustee Act 2000). Necessarily, therefore, there must be clarity and certainty when establishing a trust. It is vital to be clear that the person to whom title to the property is transferred (or who already possesses it) is indeed bound by a trust and so may not use the property for himself. Similarly, the rights of the bene?ciaries must be established with certainty so that they may be allowed to enforce the trust should the trustee fail to carry out its terms. In other words, the need to determine where the real ownership of property lies and the need to be certain about the nature of the ownership of the trustee and bene?ciary requires trusts to be created and evidenced with some measure of formality. Of course, that is not to say that all trusts must ful?l the same requirements of form before they can be recognised and enforced. After all,
it is common in English law for there to be different levels of formality when dealing with different types of property. For example, transactions concerning land - an immovable asset - have traditionally required a higher degree of formality than dealings with other kinds of property. Thus, it should come as no surprise to learn that the nature of the property which is to be the subject matter of the trust is relevant when considering how trusts may be validly created.

In considering the creation of trusts, it is important to realise that there are two distinct and separate issues which must be addressed. First, it is inherent in the concept of a trust that the trustee be invested with title to the trust property, either because he is already the owner of that property or because such title has been effectively conveyed to him by the settlor. Unless that is the case, the trust is said to be incompletely constituted and the bene?ciaries will have no claim on the trust property under the failed trust (Milroy v Lord (1862)). A recent innovation created by the Privy Council involves the occasion where the settlor appoints multiple owners as trustees, including himself. If he manifests an irrevocable intention to create a trust, his retention of the property will be construed as one of the trustees thereby constituting the trust, whether with or without a transfer to the other trustees. The maxim, 'equity regards as done that which ought to be done' will be applicable, see Choithram v Pagarani (2001). In cases where the trust property needs to be conveyed to a trustee, it is necessary to examine the particular type of trust property in order to determine what must be done in order effectively to transfer the title. The require­ments will be different for each type of property. Secondly, there are 'external' formality rules, imposed by statute, which regulate the way in which trusts per se can be created. These are to be found in the Law of Property Act (LPA) 1925 and they are designed to ensure that the creation of trusts of certain kinds of property is not open to doubt and to minimise the potential for fraud by the trustee. Failure to ful?l these requirements renders the trust unenforceable, even if title to the property is effectively vested in the trustee (s 53 of the LPA 1925).
The ?rst issue is the creation of a trust by establishing that the trustee has title to the property. Once again, a distinction needs to be drawn at the outset between situations where the existing owner of the trust property declares himself to be a trustee (self-declarations) and situations where the existing owner of property wishes to transfer the property to someone else as trustee (transfer and declaration). In the former case, the person who is to be the trustee already has title and, therefore, there is no need to transfer the property. Consequently, there is no need for any formality for this aspect of trust creation. All that is needed for an effective declaration of trust - or rather, a declaration by the current owner of himself as trustee - is some clear evidence of a present and irrevocable declaration of trust, as in Paul v Constance (1977), and this evidence may take any form. Indeed, there is no requirement that the declaration of trust ever be communicated to the intended bene?ciary (Middleton v Pollock (1876)). So, for example, if A, the owner of a book, declares that 'I now hold this book on trust for B', a trust will be created even if B is unaware of the fact, provided that there are no external formality requirements imposed by statute for the creation of trusts of books (which there are not). It is important to note, however, that for a declaration of trust by the present owner of the property to be effective, the circumstances of the case must not reveal a failed attempt by that owner to create a trust by the transfer of the trust property to another as trustee. In other words, declaration of oneself as trustee requires no formality, because the declarer already owns the property; but such a declaration will not be validly made if the owner had tried and failed to create a trust by the other method - transfer of property to a trustee - especially if that failure is due to a defect in the formality requirements for the transfer of that property (Milroy v Lord (1862)). The intention to declare oneself a trustee is very different from the intention to transfer property to another as trustee and they are mutually exclusive.

The second way in which trusts may be created is the effective transfer of property to the trustee. As noted at the outset, unless the trustee has title to the property, the trust is incompletely constituted and the bene?ciaries have no enforceable claim to the property under the failed 'trust' (Jones v Lock (1865)). Thus, if the owner of property decides to create a trust by transferring that property to someone else as trustee, it is essential that the intended trustee obtains title to the property in the manner appropriate to the type of intended trust property. So, if S (the owner and 'settlor') decides to create a trust of his book by transferring the book to A on trust for B, legal title to the book must be effectively transferred to A if the trust is to be constituted and B is to have enforceable rights to it. In reality, then, it is essential to know how title to different types of property may be transferred. Indeed, there is nothing special about the rules we are about to consider, as they are the normal rules applicable to the transfer of title to property whenever and whyever it is conveyed. It is just that, with trusts, the trustee is receiving the property on behalf of someone else and not for his own use. Typical examples of the formality requirements for the transfer of title in property to a trustee are: a deed or registered disposition for land (unregistered and registered respectively); execution of a share transfer form and registration as owner for stocks and shares; written assignment for choses in action; and delivery of possession or a deed of gift for personal property. It is clear, then, that the particular requirements for the effective transfer of title to a trustee depend upon the nature of the property being transferred and, in this sense, the formality requirements for the creation of a trust are really the formality requirements for the effective transfer of ownership of property generally.

Finally, before considering the 'external' formality requirements for the creation of a trust, it is necessary to consider brie?y one or two exceptions to the general principle just considered. First, a trust may be held to have been validly constituted, despite the fact that the trustee has not formally received title, if failure to be invested with that title is because of non-compliance with some condition outside the control of the settlor or trustee. Thus, in Re Rose (1952), a trust of shares in a private company was held to be perfectly constituted despite the fact that the legal title of the trustee had not been formally con?rmed at the relevant time by entry in the share register of the company. This was because transfer of legal title was by registration in the company's register but its directors had a discretion to refuse such registration and registration had not yet taken place. Given, then, that complete transfer of legal title was outside the settlor's (and trustee's) control, lack of compli­ance was not destructive of the trust. Similarly, in Mascall v Mascall (1985) title to registered land was taken to have been effectively transferred by completion of the relevant land transfer form even though registration of that title by the registrar had not yet occurred. Likewise, in Pennington v Waine (2002), the Court of Appeal decided that the execution and delivery of share transfer forms to an intermediary for the purpose of registering the new owner was suf?cient to transfer the equitable interest. However, it is unclear whether legal title will be taken to have been effectively transferred if subsequent registration (of the private company's shares or registered land) does not take place and so the exception really allows the trust to be constituted in advance of the time that it technically occurs. Secondly, according to Re Ralli (1964), it may be immaterial that the trustee acquires title to the trust property in a manner different from that which the settlor originally intended. So, even though the settlor has failed to transfer title to the trustee during his lifetime, if the intended trustee is also the settlor's executor under his will, when the settlor dies, the executor (and trustee) will obtain legal title by virtue of his position as executor and the trust will thereby be constituted, albeit in a different manner from that which was intended. Thirdly, there are several other methods by which title to property may be effectively transferred from one person to another without the normal formality rules for that type of property being satis?ed. However, these are usually relevant when the transferee obtains the property absolutely and not as trustee - that is, they are used to make gifts, not to perfect trusts. Examples include the principles of donatio mortis causa and the law of proprietary estoppel.

We shall look now at the other major requirement for the valid creation of a trust: the external formality rules imposed by statute. In essence, these rules are necessary to ensure certainty in dealings with certain kinds of property, especially land. In fact, the position in respect of trusts of property created inter vivos is relatively straight­forward. First, assuming that title to the property is with, or has been transferred to, the trustee, there are no further formal requirements for the creation of trusts of personalty (that is, not land or interests in land). So, trusts of property other than land may be created orally or in writing and all that is needed is a declaration of self as trustee or an effective transfer of title to another as trustee. In the case of land, however, the creation of a trust must be 'manifested and proved' in writing. Although the trust does not have to be in writing, there must be written evidence of it, even if that evidence is not contemporaneous with the date of the creation of the trust (s 53(1)(b) of the LPA 1925 and Rouchefoucauld v Boustead (1897)). Failure to comply with this evidential requirement renders the trust of land unenforceable, although there are exceptions for resulting or constructive trusts of land (s 53(2) of the LPA 1925 and Pettitt v Pettitt (1970)). Likewise, the court may, in exceptional circumstances, allow oral evidence to prove the existence of a trust of land if this is necessary to prevent fraud by the trustee, as where the trustee dishonestly claims that there is no trust and that he may keep the property because of an absence of the necessary writing (Rouchefoucauld v Boustead (1897)).

These, then, are the necessary formalities which must be met before a settlor can create an inter vivos trust of property which he owns absolutely. Of course, if the trust is by will, then different considerations apply (see s 9 of the Wills Act 1837, as amended) and if the subject matter of the trust is itself an equitable interest (so that the trustee holds an equitable title on trust for another), there may be further formality requirements springing from the requirements of writing found in s 53(1)(c) of the LPA 1925. Finally, it is relevant to note that, in some circumstances, it may be dif?cult to distinguish between, on the one hand, the creation of a trust of pure personalty (no writing) or of land (evidenced in writing) and, on the other, the transfer of an equitable interest in personalty or land under a trust which already exists. The difference is, however, crucial, for the transfer of any equitable interest under an existing trust - be it of personalty or realty - must actually be in writing under s 53(1)(c) of the LPA 1925. As the cases of Grey v IRC (1960) and Vandervell v IRC (1967) demonstrate, such a distinction is not always easy to draw.

Note: This question is quite speci?c as it does not require consideration of trusts by will (testamentary trusts) or where the settlor has only a purely equitable interest. Under­standing the distinction between creation of trusts by declaration and by transfer of property is important, especially as the methods are mutually exclusive.

Question 2

Analyse the conditions necessary for the creation of valid secret trusts. Are there any signi?cant differences between fully secret trusts and half-secret trusts?

Answer Plan

  • Testamentary trusts.
  • Secret trusts as an exception to these formality requirements.
  • Fully secret trusts: apparently absolute gifts and acceptance of the trust obligation.
  • Half-secret trusts: the trust but no bene?ciary; communication and accept­ance rules.
  • Differences between fully secret and half-secret trusts.


According to Pennycuick J in Re Tyler (1967), the 'particular principles of law applicable to secret trusts are really concerned only with trusts created by will'. By virtue of s 9 of the Wills Act 1837 (as amended), no will shall be valid unless, inter alia, it is in writing and signed by the testator and properly witnessed. Further, all wills are public documents and any bequests and devises contained therein may be on view for all the world to see. Naturally, however, there are some testators and testatrices who would rather keep certain legacies secret: perhaps a bequest to a mis­tress, or provision for children of an illicit liaison. The equitable doctrine of secret trusts allows shy testators to make 'private' bequests and so avoid embarrassment to many parties. In essence, secret trusts are those trusts which, although contained in a will, are valid without the need to satisfy the conditions laid down in s 9 of the Wills Act. They allow the testator or testatrix to make such dispositions as he or she pleases without the details being made public, save only to the person who is bound to put into effect those wishes.

Secret trusts fall into two categories: fully secret trusts and half-secret trusts. In a fully secret trust, property is given by will to a legatee absolutely without mention of any trust: the gift appears to be absolute and there is no trust on the face of the will (Thynn v Thynn (1684)). However, at any time during the testator's lifetime, the details of a trust and the intended bene?ciaries thereof may have been communicated to the legatee - orally or in writing - and, if the legatee agrees, he will hold the property on trust for the 'hidden' bene?ciaries when the testator dies. Both the very existence of the trust and the identity of the bene?ciaries are 'secret'. Of course, the onerous trust obligation cannot be imposed on a legatee of an absolute gift without that person's consent, and so it is essential that the fact of the trust and the details thereof are communicated to the legatee/trustee and accepted by him during the tes­tator's lifetime (McCormick v Grogan (1869)), although either before or after the date of the will is acceptable (Wallgrave v Tebbs (1855)). Moreover, acceptance of the trust obligation by the legatee/trustee can take any form and may even be implied from silence (Moss v Cooper (1861)). Once the trustee has accepted the trust obligation, he is usually barred from receiving any personal bene?t from the legacy, even though it appears to have been given to him absolutely. Thus, if the testator has communicated the fact of the trust to the legatee/trustee, but not the identity of the bene?ciaries, the trustee will hold the property on trust for those entitled to the residuary estate or the next of kin, as the case may be (Re Boyes (1884)). The only exception to this is where the fully secret trustee can prove that he or she also was intended to be a bene?ciary under the secret trust, a dif?cult, though not impossible, task (Irvine v Sullivan (1869)).

Turning now to half-secret trusts, it will be apparent that they are quite distinct from fully secret trusts, both as to the conditions for their validity and in the way they operate. In a half-secret trust, the bequest or legacy is given in the will 'on trust' to a trustee, but the identity of the bene?ciaries is not disclosed (Blackwell v Blackwell (1929)). In other words, it is clear that the recipient is not entitled to take the gift absolutely, although for whom the trustee holds the property is secret. Crucially, in order for a half-secret trust to be valid - that is, enforceable by the bene?ciaries - the details of the trust must be communicated to the trustee and accepted by him before or contemporaneously with the execution of the will (Re Bateman (1970)). It is not enough that the identity of the bene?ciaries be communicated at any point prior to the testator's death, although it is suf?cient if the trustee is given exclusive access to this information before the will is made, as where he is handed a sealed envelope or key to a safe at the relevant time, even if this is not to be opened until after the testator's death (Re Boyes (1884)). Similarly, because the actuality of the trust is declared on the face of the will, a half-secret trust will be invalid if the information communicated to the trustee contradicts the will (Re Keen (1937)).

At ?rst sight, then, both fully secret and half-secret trusts appear to offer the testator two different routes to achieve the same end: the distribution of testament­ary property to unnamed persons. Of course, the fully secret trust offers more secrecy, because there is no written evidence of a trust at all, but this is not the only difference between the two concepts. We have seen already that there is a major difference in the communication rules for each type of trust, with the half-secret trust suffering from a more stringent requirement. For the fully secret trust, however, the fact that communication may occur at any time before or after the date of the will has important consequences. For example, it is often the case that a testator executes his will many years before his death but the law of fully secret trusts allows him to redirect his property by gaining a legatee's consent to the imposition of a trust for another person at any time after that. It is obvious that this is not at all consistent with the policy of the Wills Act 1837 - after all, the testator could have executed an amended will or codicil - and there is a strong case for preventing the misuse of the principles of fully secret trusts in this way. Similarly, there is no Re Keen rule in fully secret trusts that requires the terms of the will and the details of the trust to be consistent although, given that the will says nothing about a trust in the ?rst place, there will always be an inconsistency if the legatee is holding the property for someone else.

In addition to these differences concerning the circumstances in which fully secret and half-secret trusts may come into existence, there are differences in the way they operate, not all of them logical or justi?able. Thus, if a fully secret trustee does not accept the trust obligation, the property remains his under the will because the absolute legacy takes effect. On the other hand, a half-secret trustee, being declared a trustee in the will, is forever barred from taking the property and must hold it for the residuary estate (Re Keen (1937)). Nor may a half-secret trustee qualify as a bene?ciary under the trust (Re Rees (1950)), whereas this appears possible under a fully secret trust (Irvine v Sullivan (1869)). Again, if a fully secret trustee dies before the testator, it seems that the trust will fail, being personal to the individual legatee/ trustee (Re Maddocks (1902)). This is not so with a half-secret trust, where equity will intervene on the ground that 'a trust will not fail for want of a trustee' (Sonley v Clock Makers Company (1780)), although it is not clear why equity cannot intervene to save a fully secret trust in like manner. The same difference in result appears to exist if the trustee dies or revokes his acceptance of the trust before the testator's death, with the fully secret trust failing (although possibly the fully secret trustee cannot keep the property for himself) and the half-secret trust being supported by the appointment of a new trustee. Similarly, whereas a fully secret trust can be revoked and replaced by a new fully secret trust by the testator at any time up to his death, a half-secret trust can only be revoked prior to death (thus leaving the property held on resulting trust) because it is impossible to communicate details of new bene?ciaries after the will has been executed.

Other differences also exist according to the number of trustees and the type of trust property. Thus, legacies to two or more persons in the will as trustees (that is, a half-secret trust) will usually be to them as joint tenants and, therefore, a valid acceptance of the trust obligation by any one of them will bind them all (Re Stead (1900)). However, in fully secret trusts, if there are two legatees/trustees, the gift in the will may be to them as tenants in common or as joint tenants because they are not declared trustees. If it is to them as tenants in common, only those who actually accept the trust are bound and the others may keep their own share (Tee v Ferris (1856)). If, on the other hand, the gift is made to the legatees/trustees as joint ten­ants, an acceptance by any one before the will is made will bind all, irrespective of the wishes of the others, although acceptance after the will binds only those so accepting (Re Stead (1900)). Finally, there may well be differences in respect of the additional
formality requirements necessary to establish secret trusts of land, although the issue is shrouded in doubt. So, while it has been suggested that fully secret trusts of land do not need to comply with s 53(1)(b) of the Law of Property Act (LPA) 1925, and, thus, do not need to be evidenced in writing (Ottaway v Norman (1972)), there is clear authority that half-secret trusts of land must meet this requirement (Re Baillie (1886)). It is uncertain whether this is a justi?able distinction between fully and half-secret trusts of land and there is much to be said for the Ottaway approach, as it seems illogical that a half-secret trust will be valid despite non-compliance with the formalities of the Wills Act 1837, but then fail because of lack of compliance with s 53(1)(b) of the LPA 1925. Indeed, it may well be that a half-secret trust of land can avoid being caught by s 53(1)(b) of the LPA 1925 on the ground that 'equity will not permit a statute (s 53) to be an instrument of fraud'. Be that as it may, the point was not fully argued in Ottaway v Norman and an apparent distinction remains.

In conclusion, then, although fully secret and half-secret trusts share the same general characteristic of being exempt from the formalities of the Wills Act, they are not so similar when it comes to the conditions necessary for their existence, nor the way in which they operate. Whether this be right or wrong, it is reason enough to require a distinction to be made between the two concepts.

Question 3

To what extent is it necessary to be able to distinguish between a trust and a power, particularly when considering the need for certainty of objects under a discretionary trust or power of appointment?

Answer Plan

  • Trusts and powers in outline: certainty of bene?ciaries.
  • Discretionary trusts.
  • Powers: types of powers.
  • Similarities and opportunities for confusion.
  • Differences in subsidiary certainty rules: administrative unworkability and capriciousness.


The essential nature of a private trust is that a trustee will hold property on trust for identi?able bene?ciaries. Save in the case of charities and a small number of exceptional situations, it is not possible to have a trust for a pure purpose. The essential duty imposed upon a trustee is to distribute the trust property according to the wishes of the settlor or testator and in conformity with the terms of the trust, and this is a duty which cannot be delegated to an agent or professional adviser: s 11(2) of the Trustee Act 2000. In this, they will be supervised by the court and be subject to suit by any disaffected bene?ciaries: the trust is mandatory and must be carried out. For these reasons, it is of paramount importance that the objects of the trust - the bene?ciaries - should be readily identi?able. In the language of Lord Langdale in Knight v Knight (1840), 'every trust must have certain objects'. In many cases, of course, the bene?ciaries of a trust will be named by the settlor and then there is no doubt that the certainty requirement is ful?lled, as where property is left 'on trust for Mr Smith and Mrs Jones'. However, in other cases, the settlor or testator may have decided to identify the bene?ciaries by a class description without naming them individually, as where property is given on trust 'for my relatives' or 'for my friends'. Clearly, gifts on trust for classes of persons who are not named must also have certain objects, for how else will the trustees know how to carry out the trust and the court be sure that they have done so properly?

In similar fashion, the creation of a 'power of appointment' will only be valid if the objects of the power can be ascertained with certainty (Re Gulbenkian (1970)). A power of appointment is literally a power given to a person (who may or may not be a trustee) to decide the destination of property, often the property of a testator or settlor. It is similar to a trust save only that the donee of the power (the person who exercises the power) is not under a duty to distribute the property but has complete freedom whether to do so or not. The power is said to be voluntary, and if it is not exercised the donee of the power may keep the property himself, or the settlor/ testator may have provided for a 'gift over in default of appointment' (that is, a direction as to where the property should go if the power is not exercised). In general terms, powers of appointment fall into three classes: ?rst, general powers of appointment, where the donee has power to appoint the property among the whole world, including himself, and so such powers are akin to absolute ownership of the property (for example, Re Beatty (decd) (1990)); secondly, hybrid powers of appoint­ment, where the donee of the power has power to appoint the property among the whole world except a de?ned class of individuals (for example, Re Hay (1981)); and thirdly, special powers of appointment, where the donee has power to appoint the property among a de?ned class of individuals (for example, Re Gulbenkian (1970)). Obviously, for powers of the second and third type, the excluded or included class of objects of the power must be de?ned with certainty in order that it is possible to decide whether the power has been validly exercised or not. An example is where the donee has power to appoint 'among his friends' or 'to anyone except his brothers'.

There is, then, a fundamental difference in principle between a trust and a power: the former is mandatory and must be carried out, while the latter is voluntary and may go unexecuted. Unfortunately, in practice, it is very easy to confuse trusts and powers, especially since most powers are given to people who are otherwise trustees and who hold the property itself on trust but have a power (not a trust obligation) over it (being so called '?duciary powers'). In practice, very real problems arise because special powers given to trustees can be virtually indistinguishable from discretionary trusts. A discretionary trust exists where trustees are under a duty to distribute trust property among a class of bene?ciaries (which a court will enforce), but have a choice as to which of those bene?ciaries shall actually bene?t and in what proportions. However, as we have seen, the donee of the power does not have to make any appointment at all and the court will not compel him to do so. In any given case, deciding whether there is a trust or a power - and consequently a duty to distribute or not - depends on the court's construction of the instrument which establishes the arrangement. On this, views can legitimately differ, as in McPhail v Doulton (1971), where the question of construction as a trust or power went all the way to the House of Lords.

It is now clear that there are crucial differences between a discretionary trust and a power and, although some case law has tended to blur the distinction somewhat by suggesting that a court might exercise a ?duciary power (one given to a trustee) where the donee is unable or unwilling to do so (Mettoy Pension Trustees Ltd v Evans (1990); Re William Makin & Sons Ltd (1993)), it is still important to be able to distinguish between the two concepts. For example, the scope of the duties of a trustee under a discretionary trust are wider than those of a donee of a power, even a ?duciary one. Obviously, discretionary trustees must carry out the trust and distri­bute the property, but their duty to consider when and to whom to distribute the property is much wider than that of a donee of a power. A donee of a ?duciary power has merely to consider from time to time whether to exercise the power, and his survey of the range of objects of the class need not be as extensive as that undertaken by a discretionary trustee (McPhail v Doulton (1971)). Indeed, the donee of a pure power (that is, a donee not in a ?duciary position) seems to be under minimal duties, perhaps only to protect those entitled in default of appointment by not exercising the power in favour of anyone outside the class (Mettoy Pension Trustees Ltd).

In one respect it has been recognised that discretionary trusts and powers of appointment should be subject to the same principles. The House of Lords, in McPhail v Doulton (1971), decided that the test for deciding whether the objects of a discretionary trust were suf?ciently certain and the test for deciding whether the objects of a power were suf?ciently certain, should be the same, not least because the classi?cation of a disposition as a discretionary trust or a ?duciary power usually turns on ?ne points of construction. It would be inequitable to make their validity (if the tests were different) depend on such subjective issues. Consequently, according to McPhail, the test for certainty of objects of a discretionary trust and a power is the Re Gulbenkian test, formally applicable only to powers: namely, that it must be possible to say with certainty whether any given individual is, or is not, a member of the class of bene?ciaries or objects (the 'given postulant' test). No higher test was
needed for discretionary trusts (as was previously the case: IRC v Broadway Cottages Trust (1955)) because, even if the court was called on to execute the discretionary trust, it did not have to divide the trust property equally and did not need to know all the members of the class: that, after all, was not what the settlor intended.

The assimilation of the tests for certainty for these two concepts has brought considerable relief to those charged with making the dif?cult distinction in practice between a discretionary trust and a power. Of course, as we have seen, fundamental differences between the concepts remain and there will always be cases where it is important to make the distinction per se. For example, it seems that 'administrative unworkability' cannot invalidate a power in the same way that it can invalidate a discretionary trust, almost certainly because of the less pressing duties of a donee of a power (R v District Auditor ex p West Yorkshire MC (1986)). Likewise, if 'capricious­ness' has any application to powers (for example, Re Manisty (1974)), it is unlikely that it adds anything to the rules concerning discretionary trusts. However, be that as it may, the decision in McPhail should be welcomed because, although there are differences between discretionary trusts and powers, those differences are now related primarily to their operation, not to their validity.

Question 4

Assess the validity of the following dispositions in the will of Thomas, who died in December 2006:

(a) £10,000 to my Aunt Agatha, knowing that she will use the money in order to secure the future of my daughters;

(b) my houses in Penzance and St Ives on trust for my daughters, Amanda and Barbara, for their lifetimes and thence in equal shares between such of my other kinsfolk now living as may be resident in the county of Cornwall, save only that no person of the Protestant religion shall be entitled to any portion; and

(c) the residue of my estate to my Executors for such of my colleagues at work as they shall in their discretion think ?t.

Thomas worked for the National Health Service at various hospitals throughout his life, but spent the last 10 years of his working life in Penzance General Hospital, which was destroyed by ?re just after his retirement in 2004.

Answer Plan

  • Certainty of intention: construction of the gift.
  • Certainty of objects: ?xed trusts; conditions precedent.
  • Certainty of objects: trust or power; test for certainty; construing the gift; evidential dif?culties; administrative unworkability.


It is a cardinal principle of the law of equity that a trust may only be valid - that is, enforceable by the bene?ciaries against the trustee - if it has been created with certainty of intention, certainty of subject matter and certainty of objects (Knight v Knight (1840)). Along with the bene?ciary principle, these three certainties repre­sent the inherent attributes of a trust. The requirement for certainty of intention (or certainty of words) ensures that the potential trustee of property is clearly placed under a duty to use the property in the way intended by the settlor or testator. Such certainty exists where the words used by the settlor or testator reveal the imposition of a trust obligation on the intended trustee as distinct from an intention to transfer the property to that person absolutely. It is what distinguishes a transfer by way of trust from a transfer by way of absolute gift (Re Conolly (1909)). The second of the three certainties, certainty of subject matter, requires that both the property which is to be the subject matter of the trust (Palmer v Simmonds (1854)) and the extent of the intended equitable owners' bene?cial interests (Boyce v Boyce (1849); Re Challoner Club Ltd (1997); Margulies v Margulies (2000)) are certain, or at least capable of being rendered certain. This will ensure that the trustee is aware of the precise scope of his or her duties under the trust. Thirdly, there must be certainty of objects in order that the trustee should know in whose favour he or she must perform the trust, so identifying who may apply to the court should the trustee fail in this duty. The precise requirements of certainty of objects vary for different kinds of trusts and powers but, in all cases, a settlor's or testator's failure to de?ne their objects with suf?cient certainty will mean that the property is held on resulting trust for the settlor or residuary legatees (McPhail v Doulton (1971)). This particular problem concerns various aspects of the 'three certainties', and each limb of Thomas's testamentary disposition will be considered in turn.

(a) £10,000 to Aunt Agatha

The issue raised by this disposition is whether the gift of £10,000 to Aunt Agatha is by way of trust or as an absolute gift - in other words, whether there is suf?cient certainty of intention to create a trust. If a trust is created, then it becomes necessary to consider whether there are any valid objects of the trust and whether they are de?ned with suf?cient certainty.

The imposition of a trust over the money requires Thomas to have imposed a mandatory obligation upon Aunt Agatha to carry out his wishes. This is purely a matter of construction of the words used (Lambe v Eames (1871)), although it is clear that precatory words (words of expression, hope or desire) will not of themselves impose a trust on the recipient of property in the absence of corroborative evidence (Re Conolly (1909)). In our case, the words used by Thomas are stronger than preca­tory words: Thomas does not 'hope' or 'wish' that Agatha will carry out his wishes, he
'knows' that she will. However, it is still not certain that this is enough to constitute a trust obligation. For example, in Re Adams and the Kensington Vestry (1884), a gift 'in full con?dence that . . .' was held to be absolute and not mandatory in the way of a trust, and in Margulies v Margulies (2000) some communications from the testator were mandatory and some were precatory and the court decided that this demon­strated an overall lack of intention to create a trust. On balance, it is likely that the bequest to Aunt Agatha will be construed as an absolute gift, essentially because of the absence of any other indication that this is intended to be a trust. Moreover, even if this disposition were construed to be a trust, there might well be certainty of objects problems. While it is obvious that a trust in favour of 'my daughters' is perfectly certain as to objects, the disposition in this case is 'in order to secure the future of my daughters'. A literal reading of this suggests that this is a trust for purposes connected with Thomas's daughters, and trusts for purposes are void (Re Endacott (1960)). Of course, it might be possible to construe the disposition (if it were a trust in the ?rst place) in favour of the daughters per se (compare Re Osoba (1979)) or as a trust saved by the Re Denley principle. However, these dif?culties in de?ning the objects of the trust precisely only add to the doubts surrounding the lack of certainty of intention (e.g., Re Kayford (1975)). Consequently, in all probability, this is an absolute gift to Aunt Agatha, who may use the property for the bene?t of Thomas's daughters if she chooses, but cannot be compelled to do so (Lassence v Tierney (1849)).

(b) The houses in Penzance and St Ives

It is clear from the wording of the second disposition in Thomas's will that a trust is intended. Although use of the word 'trust' does not always impose a trust in law (Tito v Waddell (No 2) (1977)), there is nothing here to suggest otherwise. Likewise, the subject matter of the trust is certain, always assuming Thomas did own a house in Penzance and St Ives. The problem is, then, one of certainty of objects. Clearly, there is no dif?culty with the life interests given to Thomas's daughters, who are both named. The disposition will operate under the Trusts of Land and Appointment of Trustees Act 1996, with Amanda and Barbara being given the rights to posses­sion, etc, established by that statute. The problem arises with the class of persons entitled to the reversionary interests, being 'my kinsfolk now living as may be resident in the county of Cornwall' and then subject to an exclusion against any person of the Protestant religion.
First, it is necessary to determine the nature of the trust affecting the two houses for this will help determine whether there is certainty of objects of the reversionary class. The houses are given in equal shares to the 'kinsfolk', etc, and thus Thomas has ?xed in advanced the share of each person within the class. The trustees have no discretion to apportion the trust property among the class but must divide it up equally. This is a ?xed trust and because the court must be able to execute the trust in default of the trustees and divide the property equally, the test of certainty of the objects is the 'complete list' test laid down in IRC v Broadway Cottages Trust (1955).

It must be possible to draw up a complete list of all Thomas's 'kinsfolk' who currently reside in the county of Cornwall. This may prove dif?cult, not because of the residence restriction, for that should be easy enough to determine, but because 'kinsfolk' is an imprecise concept. It will only be possible to draw up a list of kinsfolk if we know what 'kinsfolk' actually means. Such inherent uncertainty in the concept used by Thomas to de?ne his class may prove fatal unless the court is prepared to rede?ne the concept for the trustees in the same way that the Court of Appeal rede?ned 'relatives' in Re Baden (No 2) (1973). This should not be ruled out since a court of equity will generally prefer validity to invalidity, especially if the trustees' duties are not otherwise dif?cult to perform evidentially. Finally, even if the court adopts a benevolent attitude to the ?xed trust, there is still the requirement that no person may have a share if he or she is 'of the Protestant religion'. Clearly, the point is that 'not being a Protestant' is to be a condition precedent for entry to the class of bene?ciaries. Consequently, the scope of the condition precedent must also be certain because, otherwise, it will be impossible to determine who has been excluded. The test of certainty for conditions precedent is that it must be possible to say with certainty whether one person would or would not ful?l the condition (Re Allen (1959) and Re Barlow (1979)). As is clear from Re Tuck (1978), a condition precedent related to religion can be regarded as certain under this test although, in that case, a third person was given the task of deciding who fell within the religious condition. Subject then to it being possible to de?ne what quali?es a person as 'a Protestant', the condition precedent will be valid. If the ?xed trust for the class fails for other reasons, the validity or not of the condition precedent is irrelevant.

(c) The residue of Thomas's estate

The gift of residue for 'such of my colleagues at work' raises, once again, the problem of certainty of objects, those objects being de?ned by reference to a class description. As regards this disposition, it is clear that the executors are given a discretion as to whom from among the class they shall select to receive a portion of the residue, and in what proportions. So, this part of Thomas's will discloses either a discretionary trust for the class or a special power of appointment given to the trustees to appoint among the class. Of course, the difference is crucial so far as the executors are concerned because, if this is a discretionary trust, they are under a mandatory obligation to make a selection from among the class and distribute the property whereas, if this is a power, they may decide not to distribute and cannot be compelled to do so. Whether this disposition discloses a trust or a power is a matter of construction and, as McPhail v Doulton (1972) shows, the distinction is not always easy to draw. In our case, it is important that there is no gift over in default of appointment, perhaps suggesting that the executors must choose from among the class, although the absence of a gift over does not always foretell a trust, since the testator may be content to allow the property to revert to next of kin if the power is not exercised (Re Weekes (1897)). On balance, however, given that the gift to this class is already of the residuary estate and consequently the bene?ciaries in default would be dif?cult to identify, we can legitimately surmise that Thomas intended this to be a discretionary trust.

Since McPhail v Doulton, the test for certainty of objects of a discretionary trust and a power have been assimilated and, even if this is a power, the test we must apply is whether it is possible to say with certainty whether any given person is, or is not, a member of the class (McPhail). Unfortunately, although this test is easy to state, it is dif?cult to apply because the leading case on its application (Re Baden (No 2) (1973)) gives us three alternative approaches. According to Stamp LJ, the test is satis?ed only if it is possible to say in fact whether any given person is, or is not, a member of the class and this requires both the class to be de?ned with precision and there to be enough evidence available to make positive or negative choices in respect of all potential applicants. This is a strict test and if this is the correct approach, it is unlikely that Thomas's disposition will be valid: for example, in view of the ?re at his last place of work, will there be suf?cient employment records to indicate with whom he worked, even if it were possible to say who then quali?ed as a 'colleague'? Secondly, even if we take Sachs LJ's approach, it may not be possible to hold this discretionary trust valid. In his view, the test is satis?ed if it is possible to say, in theory, whether any given person is, or is not, a member of the class, irrespective of whether there is enough evidence to make such a decision. The class must be con­ceptually certain in the sense that the concept used by the settlor must be precise. The corollary is that evidential dif?culties will not invalidate a trust because any person who cannot prove that he or she is a member of the class will be deemed to be outside it. However, it is likely that Thomas's disposition cannot satisfy even this less strict version of the 'is or is not' test: that is, the concept of a 'colleague' may be too imprecise to allow us to say who is or is not within its ambit, even in theory. The only hope would be if the court were prepared to reinterpret 'colleague' in some way that made it more certain, perhaps as 'fellow employee' (Re Baden (No 2) (1973)). Finally, we come to Megaw LJ who proposed the least strict version of the 'is or is not' test, believing it to be satis?ed if it could be said of a substantial number of persons that they were inside the class, even if it could not be said of every potential person whether he or she was or not. It may be that Thomas's dispositions would satisfy this version of the test for it seems possible to say of many people that they were his 'colleagues', even if we could not say de?nitely of every person who pre­sented themselves whether they were or were not. Unfortunately, however, Megaw LJ's approach does seem at variance with that of the House of Lords in McPhail, and it is the view of the 'given postulant' test that is commonly criticised (being vague and imprecise). Most commentators prefer the approach of Sachs LJ.

The conclusion is then that Thomas's third disposition fails for uncertainty of objects. It may be that a court would be prepared to rede?ne the concept of 'colleague' so as to make it more certain and then apply the 'is or is not' test to the class description so rede?ned (as in Re Baden), but that cannot be guaranteed. Indeed, even then, the executors face one more hurdle, for it may be that the class of the discretion­ary trust is 'administratively unworkable' and so void for 'secondary uncertainty'. This concept was expounded by Lord Wilberforce in McPhail and it is clear from R v District Auditor ex p West Yorkshire MC (1986) that if the settlor or testator stipulates a class so large that the trustees cannot effectively ful?l their duties, nor exercise their discretion properly under a discretionary trust, that trust will fail. In our case, Thomas's large number of former places of work may bring the class description within this principle. This will be a matter of judgment, and many people would argue that a trust should not be void for administrative unworkability simply because the class is large, but only when this makes it impossible for the trustees to carry out their duties properly.

Note: This is a typical problem on the three certainties, with concentration on certainty of objects. When considering certainty of objects, it is important to appreciate that the three versions of the Baden test may cause different results for the validity of the trust or power. All must be considered and Baden should not be regarded as having a clear ratio decidendi.

Question 5

X, who has recently died, made a will in 2005 in which he made the following gifts:

(i) £5,000 to C and D upon trust for such charitable or benevolent objects as Z, my trustee shall select.

(ii) £250,000 to the University of London upon trust to establish and main­tain in perpetuity a School of Law Reform.

(iii) £3,000 to the trustees of the South Blankshire Methodist Conference for the promotion of physical recreation for Methodists in South Blankshire.
Consider whether these gifts are charitable.

Answer Plan

  • Gifts devoted exclusively for charitable purposes.
  • Trusts for the advancement of education.
  • Exemption of charities from the perpetuity rule - excessive duration principle.
  • Trusts for the promotion of sport.
  • Recreational Charities Act 1958.
  • Public bene?t test.


(i) £5,000 to C and D upon trust

Section 1 of the Charities Act 2006 de?nes a charity as any institution which is established for charitable purposes only and is subject to the jurisdiction of the High Court. Section 2 identi?es 13 purposes that are treated as charitable and declares that the institution in question is required to satisfy the public bene?t test. The ?rst question in issue is whether this gift of £5,000 is charitable. In the problem, the will transfers '£5,000 to C and D upon trust for such charitable or benevolent objects.' This raises the question as to whether the test for charitable objects is satis?ed.

Charitable trusts, like private trusts, are subject to a test of certainty of objects. A charitable trust is subject to a unique test for certainty of objects, namely, whether the objects are exclusively charitable. This is af?rmed in s 1(1)(a) of the Charities Act 2006. In other words, if the trust funds may be used solely for charitable purposes the test will be satis?ed. Indeed, it is unnecessary for the settlor or testator to specify the charitable objects which are intended to acquire the trust property, provided that the trust instrument manifests a clear intention to devote the funds for 'charitable purposes' the test will be satis?ed.

Referring back to the problem the question in turn concerns the construction of the expression, 'or'. If this conjunction is used disjunctively, as is the norm, it would follow that the test for certainty of charitable objects will not be satis?ed for benevo­lent objects that are not charitable are entitled to bene?t. In Chichester Diocesan Fund v Simpson (1944), a testator directed his executor to apply the residue of his estate 'for such charitable or benevolent objects' as they may select. The executors assumed that the clause created a valid charitable gift and distributed most of the funds to charitable bodies. The court decided that the clause did not create charitable gifts and therefore the gifts were void. A similar result was reached in Attorney General of the Bahamas v Royal Trust (1986). The effect will be that a resulting trust for the testator's residuary estate will arise. There is a possibility that the court may, on construction, decide that the non-charitable purposes are merely incidental to the main charitable purposes, see Verge v Somerville (1924). However, there is very little evidence on the facts of the problem that may support this contention. Alternatively, the legacy will be valid if the word, 'or' is construed conjunctively in the sense that only benevolent objects that are charitable are entitled to bene?t. Again, one would be hard pressed to convince a court of such construction. The court will look at all the circumstances of the case, including the entire will and evidence that exists outside the will, to ascertain the intention of the testator.

(ii) £250,000 to the University of London upon trust

In this problem, the issue is whether the donation of £250,000 to the University for the stated purpose is charitable. Does the gift advance education? The donation is to the University of London. This clearly is a charitable body that exists to advance education. If, as is possible, the University is the trustee to promote the stated purpose, the question arises as to whether the stated purpose is charitable? A School of Law Reform within the University exists as a department to undertake the task of examining the extent to which the current law is satisfactory and in appropriate cases to suggest proposals for reform. In the law of charities, education has been inter­preted generously and is not restricted to the classroom mode of disseminating knowledge, but requires some element of instruction or supervision. Thus, research is capable of being construed as the provision of education; see Re Hopkin's Will Trust (1964). In McGovern v Attorney General (1981), a de?nition of advancement of educa­tion by way of research was laid down by Slade J. He posited that the requirements are that: (a) the subject matter of the proposed research is a useful object of study; and (b) it must be contemplated that the knowledge acquired as a result of the research will be disseminated to others; and (c) the trust is for the bene?t of the
public, or a suf?ciently important section of the public. It would appear that a school to study the extent to which the law is in a satisfactory state would clearly satisfy the test laid down by Slade J.
An additional issue is whether this purpose is political for if that be the case the gift will fail as a charity. Political activities include attempts to change the law and gifts to further the objects of political parties. A trust for political purposes is incapable of subsisting as a charity for the court may not stultify itself by deciding that it is in the public good for the law to be changed, see National Anti-Vivisection Society v IRC (1948), concerning the activities of the stated society and McGovern v Attorney General (1981), where Amnesty International was declared to be a political organisation. The School exists only to study the extent to which the law is satisfac­tory; it is not designed to lobby Parliament for changes to be introduced. On this basis it is a charitable activity.

A further point involves the reference to the school being maintained perpetually. Would this infringe the perpetuity rule? Although charitable trusts, like private trusts, are subject to the rule against remote vesting. Charitable trusts, as distinct from private trusts, are not subject to the rule against excessive duration. Indeed, many charities (schools and universities) continue inde?nitely and rely heavily on perpetual donations. Accordingly, if the purposes are charitable the gift will not fail for infringing the perpetuity rule.

(iii) £3,000 to the trustees of the South Blankshire Methodist Conference

The purpose as stated in this problem raises a number of issues. Is the purpose as declared in the will, namely physical recreation, a charitable activity? In any event, would the public bene?t test be satis?ed? The donation contemplated appears to promote recreation under the guise of advancing religion. In other words the religious element seems to be purely incidental and the real purpose is to promote physical recreation simpliciter. A similar purpose was considered by the court in IRC v Baddeley (1955). Here the House of Lords decided that the gift was void, as recreational activities or sport are not within the preamble to the Statute of Elizabeth or the 'spirit and intendment of the preamble', see Re Nottage (1895) and IRC v City of Glasgow Police Athletic Association (1953). Section 2(2)(g) of the Charities Act 2006 includes the advancement of amateur sport within the list of charitable purposes. However, there is little evidence in the will to indicate that the funds are to be used for the promotion of amateur sport.

The Recreational Charities Act 1958 was passed in order to clarify the law regarding the provision of recreational activities. Section 1(1) of the Act stipulates that the provision of recreational facilities shall be charitable if two criteria are ful?lled, namely: (1) the public bene?t test is satis?ed; and (2) the facilities are pro­vided in the interests of social welfare. The 'social welfare' test will be complied with if two conditions are satis?ed as enacted in s 1(2). The ?rst requirement is continuous as stipulated in s 1(2)(a). The second requirement may be satis?ed in alternative ways either by proving that the facilities are available to a limited class of objects who have a need for such facilities by virtue of one or more of the factors enumerated within s 1(2)(b)(i) (such as a youth club or an organised outing for orphaned children) or 'the facilities are available to the entire public' (such as a public swimming pool or a public park) or 'female or male members of the public at large' (women's institutes, etc). These requirements are referred to as the 'basic conditions' (see s 2A of the 1958 Act as amended by the Charities Act 2006). In Guild v IRC (1992), the court decided that the test today is whether the facilities are provided with the purpose of improving the conditions of life of the bene?ciaries, irrespective of whether the participating members of society are disadvantaged or not. In short, the material issue concerns the nature of the facilities rather than the status of the participants. On this basis a strong case could be made out that the activities contemplated are within the Recreational Charities Act 1958.

The de?nitive issue in this problem is whether the public bene?t test will be satis?ed. The public bene?t test is used to distinguish a public trust from a private trust. A public trust is required to exist for the bene?t of the public (the community) or an appreciable section of society. A unique application of this test is reserved for trusts for the relief of poverty. The test of public bene?t is adopted in ss 2 and 3 of the Charities Act 2006. Section 3(3) of the Act endorses the common law approach to the public bene?t test that existed prior to the introduction of the 2006 Act. The satisfaction of the test is a question of law for the judge to decide on the evidence submitted to the court. In IRC v Baddeley (1955) the court decided that the test will not be satis?ed where the intended bene?ciaries comprise a class within a class. The facts of the problem refer to Methodists (a class) in South Blankshire (another class). In addition, the bene?ciaries must not be numerically negligible and comprise a suf?cient section of the community to be treated as charitable. These are ?exible questions for the court to decide.