The Secret Is Out There

The Secret Is Out There:
Searching for the Legal

Justification for the Doctrine
of Secret Trusts through
Analysis of the Case Law
G.W. Allan*
Abstract: The doctrine of secret trusts, which appears to operate in

defiance of the statutory formality requirements for testamentary dis-
positions, has proven notoriously difficult to justify, and an explanation
reconciling the main authorities has consistently eluded commentators.
Many of the academic opinions rely on widely accepted interpretations
of a small number of authorities that have been designated as leading
cases. The doctrine has, however, generated a substantial amount of
case law, much of which is rarely or never cited in the literature. This
paper seeks to provide a justification for the doctrine that is based on a
thorough analysis of all of the available case law. It is argued that, once
understood in their proper context, virtually all of the authorities can be
reconciled with one another, and that a coherent justification for the
enforcement of secret trusts can be gleaned.
Keywords: constructive trusts, equitable fraud, equity, express trusts,

formalities, fraud, secret trusts, testamentary dispositions, trusts, un-
conscionability, wills, Wills Act 1837
I. Introduction
In 1929, Lord Buckmaster observed that the doctrine of secret trusts
‘has for over 200 years been the subject of vexed controversy’.1
Eighty-two years later, the debate is far from settled, as is apparent
from the difficulties that academics have encountered in finding an
adequate justification for secret trusts.2 The doctrine continues to be
* Senior Lecturer, University of Wolverhampton; e-mail: gregory.allan@wlv.ac.uk.

This paper draws on a thesis supervised by Dr P. Walton. I am very grateful to
him for his assistance, as I am to Professor R. Gregory for his advice and insights.
1 Blackwell v Blackwell [1929] AC 235 at 318.
2 See below, nn. 9–18 and accompanying text, for a summary and examples of the
various academic views on the doctrine.
Common Law World Review 40 (2011) 311–344 311

DOI: 10.1350/clwr.2011.40.4.0222
COMMON LAW WORLD REVIEW
raised in the courts, most recently in De Bruyne v De Bruyne3 in the

Court of Appeal, in which the principles underpinning the doctrine
were explained in terms which are not consistent with some other
recent judgments.4 The reason why the doctrine has generated such
disagreement is simple to appreciate. Section 9 of the Wills Act 1837
(WA 1837) stipulates, essentially,5 that any attempt to determine the
destination of one’s property which is to take effect after one’s death
that does not appear in a valid will is void. Secret trusts appear to
operate in contravention of this rule, notwithstanding that it is a
fundamental rule of the British Constitution that judges do not have
the power to disapply Acts of Parliament.6 If the testator demon-
strates intention to subject the secret trustee to a trust obligation and
this intention is communicated to and accepted by the secret trustee
during the testator’s lifetime7 or, in the case of half-secret trusts, be-
fore or contemporaneously with the will’s execution,8 then the secret
trust is enforceable. Academic contributions to the debate, therefore,
have mainly focused on how and why secret trusts are enforced,
seemingly in defiance of clear statutory provisions.
Most discussions regarding secret trusts focus on two theories, the
‘fraud theory’ and the ‘dehors the will theory’. These are usually pre-
sented as competing theories,9 although they have also been pro-
posed as being complementary.10 The fraud theory proceeds along the
lines that because equity will not permit a secret trustee to perpetrate
a fraud by relying on s. 9 of the WA 1837 in order to avoid perform-
ance of the secret trust, it will be enforced, notwithstanding the stat-
ute. The generally accepted view is that fraud must involve personal
gain by the secret trustee.11 If fraud is understood in this way, secret
3 [2010] EWCA Civ 519. The doctrine also continues to be raised in the courts of
other common law jurisdictions. See, for example, Chinn v Hanreider 2009 BCSC
635 and Child Support Register & Kanavos & Ors [2009] FMCAfam 871 (20 August
2009).
4 Taylor v Revenue & Customs Commissioners [2008] STC (SCD) 1159 and Davies v
Revenue and Customs Commissioners [2009] UKFTT 138, TC.
5 The meaning of s. 9 of the WA 1837 is considered in more detail below in Part III.
6 See, for example, comments of Lord Reid in Madzimbamuto v Lardner-Burke
[1969] 1 AC 645 at 723.
7 These long-standing requirements were usefully summarized by Brightman J in
Ottaway v Norman [1972] Ch 698 at 702.
8 See Blackwell v Blackwell, above n. 1 at 334 and 339 per Viscount Sumner.
9 See, for example, D. Hodge, ‘Secret Trusts: The Fraud Theory Revisited’ [1980]
Conv 341; P. Critchley, ‘Instruments of Fraud, Testamentary Dispositions, and the
Doctrine of Secret Trusts’ (1999) 115 LQR 631; E. Challinor, ‘Debunking the Myth
of Secret Trusts’ [2005] Conv 492; R.A. Pearce and J. Stevens, The Law of Trusts
and Equitable Obligations, 3rd edn (Oxford University Press: Oxford, 2005) ch. 7;
J. Martin, Hanbury and Martin Modern Equity, 18th edn (Sweet & Maxwell:
London, 2009) ch. 5.
10 See especially J.G. Fleming, ‘Secret Trusts’ (1947) 12 Conv 28; S. Wilson, Todd and
Wilson’s Textbook on Trusts, 9th edn (Oxford University Press: Oxford, 2009)
ch. 10.
11 See, for example, L.A. Sheridan, ‘English and Irish Secret Trusts’ (1951) 11 LQR
314; J.A. Andrews, ‘Creating Secret Trusts’ (1963) 27 Conv 92; P. Matthews, ‘The
True Basis of the Half-Secret Trust?’ [1979] Conv 360; B. Perrins, ‘Secret Trusts:
312
THE SECRET IS OUT THERE
trusts are difficult to justify, especially half-secret trusts, as a half-

secret trustee, being identified on the face of the will as a trustee,
cannot take the secret trust property for himself, regardless of
whether or not the secret trust is performed. Therefore, those ad-
vocating the orthodox form of the fraud theory have proposed a
variety of reasons why half-secret trusts are enforced, most notably
that half-secret trusts are enforced because they are incorporated into
the testator’s will under the probate doctrine of incorporation by
reference. Unfortunately, this doctrine does not sit easily with the
requirements that secret trusts be communicated and accepted, or
with the absence of any requirement that a half-secret trust be re-
duced to writing.12 Generally, however, those who insist that fraud
must involve personal gain reach the conclusion that the fraud theory
cannot justify the enforcement of secret trusts.13 An alternative posi-
tion is that the concept of fraud in equity extends beyond fraudulent
enrichment, and that any failure by the secret trustee to perform the
secret trust constitutes a fraud on the testator and the secret bene-
ficiaries. The minority who subscribe to this view14 argue that the
fraud theory can explain the enforcement of both fully and half-secret
trusts because equitable fraud need not involve any element of per-
sonal gain. Advocates of the fraud theory in either form are not in
agreement regarding whether secret trusts are express or
constructive.15
The dehors the will theory, which is the most widely accepted justi-
fication for the doctrine, is based on the idea that the WA 1837 is
irrelevant to the enforcement of secret trusts. There is, however, a
dichotomy of opinion regarding precisely how and why a secret trust
should fall outside of the scope of the WA 1837. The favoured stand-
point is that a secret trust is an express inter vivos trust that remains
unconstituted until the death of the testator; this view renders the
fraud theory redundant, but raises awkward questions as to whether
secret trusts of land ought to comply with s. 53(1)(b) of the Law of
The Key to the Dehors?’ [1985] Conv 248; Critchley, above n. 9; Challinor, above
n. 9; B. McFarlane, ‘Constructive trusts arising on a receipt of property sub
conditione’ (2004) 120 LQR 667; Martin, above n. 9; Pearce and Stevens, above
n. 9.
12 See Sheridan, above n. 11; Matthews, above n. 11, on the incorporation by
reference theory. See below II.ii(b)(ii) for consideration of this theory, especially
n. 99 and accompanying text on the requirements for incorporation and n. 101
and accompanying text on orally created half-secret trusts. For an alternative
explanation of the enforcement of half-secret trusts, see Perrins, above n. 11.
13 See especially McFarlane, above n. 11; Challinor, above n. 9; Pearce and Stevens,
above n. 9; A. Hudson, Equity & Trusts, 6th edn (Cavendish: London, 2010) ch. 6.
Critchley, above n. 9, asserts that the fraud theory can provide a partial
justification, but cannot explain the enforcement of half-secret trusts or fully
secret trusts with honest secret trustees.
14 See Wilson, above n. 10; Fleming, above n. 10; Hodge, above n. 9; D. Hayton and
C. Mitchell, Hayton and Marshall Commentary and Cases on The Law of Trusts and
Equitable Remedies, 12th edn (Sweet & Maxwell: London, 2005) ch. 2, section 3.
15 For differing opinions, see, for example, Wilson, above n. 10; Hodge, above n. 9;
Hayton and Mitchell, above n. 14; Sheridan, above n. 11.
313
Property Act 1925 (LPA 1925).16 Others argue that a secret trust is
dehors the will because it is a constructive trust, although opinions
differ as to why this should be so.17 The main criticism of the dehors
theory is that secret trusts are by their very nature testamentary dis-
positions, and that therefore it is disingenuous to suggest that they fall
outside of the WA 1837.18
What emerges from the literature, then, is a bewildering array of
apparently irreconcilable theories and academic standpoints. They are
all, however, to a greater or lesser degree, united by a common short-
coming: lack of sufficient judicial authority. Given the number of cases
in existence, it seems peculiar that academics have not been inclined
to attempt to find a solution or unifying principle through careful
analysis of all of the case law. Yet it appears that this has not yet been
done. It is therefore proposed to concentrate on what has been said
by judges regarding secret trusts in order to determine how, if at all,
the doctrine is justifiable.
II. The Fraud Theory

i. Is Fraud Relevant at all to the Doctrine of Secret Trusts?
In Re Snowden,19 Megarry VC dismissed fraud as being merely ‘the
historical origin of the doctrine’,20 explaining that nowadays, ‘secret
trusts may be established in cases where there is no possibility of
fraud’.21 This is frequently accepted as an accurate statement.22 Ana-
lysis of the authorities, however, reveals Snowden23 to be anomalous.
Forty-two cases spanning 326 years24 have been identified in which
16 In particular, Pearce and Stevens, above n. 9 at 228, claim that there is ‘an
overwhelming consensus’ that secret trusts are express inter vivos trusts.
Critchley, above n. 9 at 640, describes this as the more ‘sophisticated’ version of
the dehors theory. See also D. Kincaid, ‘The Tangled Web: the Relationship
between a Secret Trust and a Will’ [2000] Conv 420. Note that s. 53(1)(b) LPA 1925
requires that ‘a declaration of a trust respecting any land or any interest therein
must be manifested and proved by some writing …’. The question of the relevance
of s. 53(1)(b) is addressed below at Part III.iii.
17 See especially McFarlane, above n. 11.
18 See especially Critchley, above n. 9.
19 [1979] Ch 528.
20 Ibid. at 535. Such a view is not unique to English law. See also C. Rickett,
‘Thoughts on Secret Trusts from New Zealand’ [1996] Conv 302.
21 Ibid. at 535.
22 For example, Megarry VC’s comments have recently been cited with approval in
two tribunal cases, Taylor v Revenue & Customs Commissioners, above n. 4 and
Davies v Revenue and Customs Commissioners, above n. 4. See also, for example,
comments of Kincaid, above n. 16 at 440; Pearce and Stevens, above n. 9 at 224;
Martin, above n. 9 at 161; Wilson, above n. 10 at 257. See also below n. 28. The
question of the standard of proof to be applied in secret trusts cases was also
raised by Megarry VC. This aspect is discussed in detail below at II.i(b)(ii).
23 Above n. 19.
24 From Thynn v Thynn (1684) 1 Vern 296 to De Bruyne v De Bruyne, above n. 3. See
Appendix for details.
314
the prevention of fraud is referred to as being the underlying justifi-

cation for the doctrine, including House of Lords’ judgments concern-
ing both fully25 and half-secret trusts.26 It is therefore curious that
Snowden,27 a first instance decision, has been afforded such signifi-
cance, not just by academics, but also in the courts of other common
law jurisdictions,28 and it is submitted that any assertions that fraud is
no longer relevant to the enforcement of secret trusts can be de-
bunked immediately by sheer weight of authority. Serious questions
remain, however, regarding the nature of the fraud. These questions
can best be answered by analysing the development of the doctrine in
relation to fraud.
ii. What is the Nature of the Fraud?
(a) The Eighteenth-century Judgments29

One of the clearest early judgments is Reech v Kennegal30 in which the
secret trustee conceded that he had broken the promise made to the
testator that he would perform the secret trust, but insisted that a
mere breach of promise was not a fraud. Lord Hardwicke rejected
this, stating that, unlike the common law, ‘this court considers it as a
fraud upon the testator . . . as well as the [secret beneficiary], by [the
secret trustee] representing as if there was no occasion to alter the
will’.31 Although the secret trustee stood to gain from breaking his
promise, Lord Hardwicke made no mention of personal gain or
fraudulent enrichment, emphasizing that any failure to perform the
secret trust constituted fraud in equity. Similarly, in Oldham v Litch-
ford,32 a secret trustee’s failure to adhere to the trust was described as
25 McCormick v Grogan (1869) LR 4 HL 82 and Cullen v Attorney General for Ireland

(1866) LR 1 HL 190.
26 Blackwell v Blackwell, above n. 1.
27 Above n. 19.
28 Re Snowden has been applied in several judgments in other common law
jurisdictions. See, for example, Glasspool v Glasspool Estate 1988 CanLII 1438, BC
SC; Bellinger v Nuytten Estate 2002 BCSC 571; Chinn v Hanrieder 2009 BCSC 635;
Child Support Register & Kanavos & Ors [2009] FMCAfam 871 (20 August 2009);
Howell v Hyde [2003] NSWSC 732; Brown v Pourau [1995] 1 NZLR 352; Quinn v
Dean, unreported High Court Wellington Registry A 123/84 July 1986, cited in
N. Richardson, ‘Secret Trusts in New Zealand’ [1995] Canterbury Law Review 6.
29 The Statute of Frauds came into force in 1677. The Statute introduced compulsory
formality requirements for testamentary dispositions of land (s. 5) and also greatly
restricted the circumstances in which nuncupative wills of personalty could be
validly created. Although cases concerning secret trusts began to appear almost
immediately, the reports from this period are extremely brief. In both Thynn v
Thynn, above n. 24, and Devenish v Baines (1689) Prec Ch 3, reference is made to
the prevention of ‘fraud’ as being the reason why the trusts were enforced, and in
Pring v Pring (1689) 2 Vern 99, the only relevant half-secret trust case from this
period, it was held to be ‘unjust’ for the secret trustee to rely on the Statute of
Frauds to avoid performance of the secret trust. Other early cases include Dutton
v Pool (1677) 1 Ventr 318, Chamberlaine v Chamberlaine (1678) 2 Freeman 34 and
Crook v Brooking (1688) 2 Vern 50.
30 (1748) 1 Ves Sen 123.
31 Ibid. at 125.
32 (1705) 2 Freem Chy 284.
315
‘a fraud upon the testator and the legatee’.33 It is therefore unsurpris-

ing that in Drakeford v Wilks,34 Lord Hardwicke, in order to prevent
fraud on the testator, enforced a secret trust even though the secret
trustee had died without gaining personally or even attempting to do
so. It was not thought, however, that equity could disregard the stat-
ute simply to give effect to the testator’s intentions. In Whitton v
Russell,35 because it was clear to the testator by the time of his death
that the secret trustees did not intend to fulfil the promise, Lord Hard-
wicke refused to enforce the purported secret trust, explaining that
the testator was not ‘drawn by this promise, not to add the legacy to
this codicil’.36 In this case, fraud on the testator was not present. As
Lord Hardwicke pertinently stated, ‘every breach of promise is not to
be called a fraud’.37
During the eighteenth century, then, it was established that secret
trusts were enforced to prevent the fraud which arises if the secret
trustee’s undertaking, upon which the testator relied when executing
his will, is not carried out.38 Any assertion that the historical basis of
the doctrine is the prevention of fraudulent enrichment39 is incorrect.
Strangely, of those who argue that fraud need not involve personal
gain,40 not one cites any of these eighteenth-century cases.
(b) The Nineteenth-century Judgments

(i) Fully Secret Trust Cases
Because the dichotomy between fully and half-secret trusts emerged
in the nineteenth century, they will now be considered separately. In
Muckleston v Brown,41 a case concerning the Mortmain Act 1736,42
Lord Eldon stated that secret trusts are enforced ‘on the ground that
33 Ibid. at 285 per Wright LK.
34 (1747) 3 Atk 539.
35 (1739) 1 Atk 448.
36 Ibid. at 449.
37 Ibid. at 449. This refutes arguments such as that of Challinor, above n. 9 at 297,
who considers the ‘fraud on the testator’ argument to be ‘a bald assertion that a
testator’s wishes should be put into effect in a manner that is not acceptable’.
38 The comments of Lord Arden in Barrow v Greenough (1796) 3 Ves Jun 152 at 154
and Lord Cowper in Sellack v Harris (1708) 20 Eq C Abr. 46 at 48 also support the
‘fraud on the testator’ approach. Cf. Jones v Nabbs (1718) Gilb Rep 146, a
judgment that is incongruous with all other judgments from the era and has not
been followed.
39 See, for example, Andrews, above n. 11; Challinor, above n. 9; Pearce and
Stevens, above n. 9; P. Pettit, Equity and the Law of Trusts, 10th edn (Oxford
University Press: Oxford, 2005) ch. 7; Martin, above n. 9; Hudson, above n. 13. See
also Rickett, above n. 20.
40 See above n. 14.
41 (1801) 6 Ves Jun 53.
42 The category of secret trust cases, referred to throughout this work as ‘Mortmain
cases’, must be explained here. The Mortmain Act 1736 essentially prohibited
devises of land to charities or in trust for charitable purposes (see section 1).
There are a number of cases in which a testator devised land to a legatee, subject
to an agreement outside of the will that the legatee would use the land for
charitable purposes. Typically, the action would be brought by the residuary
legatees who would claim that the legatee in question took the land subject to a
316
that testator would not have devised the estate to [the secret trustee],
unless he had undertaken to pay that sum. The principle is that the
statute shall not be used to cover a fraud’,43 and in Chamberlain v
Agar,44 Plumer VC referred to ‘. . . that Species of Fraud, which con-
sists of not complying with a Promise, on which the Testator relied
[when determining his testamentary dispositions]’.45 This approach
was continued later in the century. In Jones v Badley,46 for example,
Lord Cairns, paraphrasing Sir Page Wood VC in Wallgrave v Tebbs,47
who was in turn quoting Turner LJ in Russell v Jackson,48 stated that
the doctrine exists ‘for the prevention of fraud’,49 explaining that
when a person undertakes:
. . . that he will carry the testator’s intention into effect, and the property
is left to him upon the faith of that promise or undertaking, it is in effect
a case of trust . . . [because] no one can doubt that, if the devisee had
stated that he would not carry into effect the intentions of the testator,
the disposition in his favour would not have been found in the will.50
This view was approved in the House of Lords in Cullen v Attorney-
General for Ireland51 by Lord Westbury, who described a secret bene-
ficiary’s interest as being created by the secret beneficiary’s promise,
‘the breach of which confidence would amount to a fraud’.52
The House of Lords case of McCormick v Grogan,53 particularly
Lord Westbury’s judgment, is generally regarded as authority that
fraud must involve personal gain by the secret trustee,54 and that an
unusually high standard of proof must be applied in secret trusts
cases. If phrases from the judgment are taken in isolation, this view
appears reasonable. Lord Hatherley stated that secret trusts should
only be enforced in ‘clear cases of fraud’55 when there has been a
binding secret trust for charitable purposes. If this claim was upheld, the land in
question would be held by the would-be secret trustee on resulting trust for the
testator’s estate on the grounds that it was an illegal devise.
43 Above n. 41 at 69. Lord Eldon echoes this reasoning in Stickland v Aldridge (1804)
9 Ves Jun 517 at 519.
44 (1813) 2 Ves & Bea 259.
45 Ibid. at 262.
46 (1868) LR 3 Ch App 362.
47 (1855) 2 K & J 313 at 321.
48 (1852) 10 Hare 204 at 211–12.
49 Above n. 46 at 364.
50 Ibid. The original passage from Sir Page Wood was also quoted with approval by
Lord Romily in Proby v Landor (1860) LR 3 Ch App 362 and in Rowbotham v
Dunnett (1878) 8 Ch D 430. This shows the great consistency during this period.
51 Above n. 25.
52 Ibid. at 198.
53 Above n. 25.
54 See Re Snowden, above n. 19, and, by extension, those judgments relying on Re
Snowden. See above n. 22 and n. 28 for example. This idea is extremely prevalent
among commentators. See, especially, Challinor, above n. 9; Pearce and Stevens,
above n. 9; Martin, above n. 9; Hudson, above n. 13. Note also that Wilson, above
n. 10 at 251 and 253, proposes that Lord Westbury may have merely meant that it
must be possible to demonstrate that the secret trustee intended to deceive the
testator.
55 Above n. 25 at 89.
317
‘fraudulent inducement’56 by the secret trustee. Lord Westbury used

even stronger language, emphasizing ‘the criminal character of
fraud’,57 explaining that the doctrine is based on ‘personal fraud’58
which is only present if ‘a malus animus is proved by the clearest and
most indisputable evidence’59 that the secret trustee ‘knew that the
testator or the intestate was beguiled and deceived by his conduct’.60
A different picture emerges when the judgments are read fully.
Lord Hatherley, once again,61 followed the established view of fraud,
explaining that a secret trust arises due to the ‘fraud thus committed
by the heir in inducing the testator to die intestate, upon the faith of
the heir’s representations that he would carry all such wishes as were
confided to him into effect’,62 thus clarifying his comments regarding
the need for a fraudulent inducement. He made no statement regard-
ing the standard of proof to be applied.
Lord Westbury similarly explained that a secret trust will be im-
posed when the testator ‘communicates the disposition . . . and the
disponee assents to it, either expressly, or by any mode of action
which the disponee knows must give to the testator the impression
and belief that he fully assents to the request’.63 What he undoubtedly
meant was that once it is established that a secret trustee has led the
testator to believe that he will perform the secret trust by expressly or
implicitly acceding to the testator’s requests, the court will insist that
he fulfils his promise, otherwise the testator will have been ‘beguiled
and deceived’64 and the fraud will have been ‘proved by the clearest
and indisputable evidence’.65 Lord Westbury did not mention the need
for an especially high standard of proof. Rather, he was explaining the
nature of the fraud and why communication and acceptance are
essential requirements for valid secret trusts. If a secret trustee does
not perform a properly communicated and accepted secret trust, he is
committing a personal fraud or a ‘malus animus’.66 His use of the word
‘criminal’67 to describe fraud, when read in the context of the rest of
his speech and the other authorities from the period, is likely to be a
56 Ibid.
57 Ibid. at 97.
58 Ibid.
59 Ibid.
60 Ibid. at 98.
61 Note that Lord Hatherley was known as Sir Page Wood prior to becoming Lord
Chancellor. See his judgments in Wallgrave v Tebbs, above n. 47 at 321; Tee v
Ferris at (1856) 2 K & J 357 at 367–8.
62 Above n. 25 at 88.
63 Ibid. at 97.
64 Ibid. at 98.
65 Ibid. at 97.
66 Ibid. This undermines the assertion by Martin, above n. 9 at 171, that ‘it is not
sufficient to say as did Lord Sumner in Blackwell v Blackwell that secret trusts are
based on the essential elements of intention, communication and acquiescence’.
See below, Part V, for further explanation as to the requirements for secret trusts.
67 Ibid.
318
mere reinforcement of the idea that it would be dishonest for the

secret trustee to renege on his promise.
Lord Westbury’s explanation of the legatee being converted into a
trustee merely by his acceptance of the secret trust obligation is irre-
concilable with the view that he regarded equitable fraud as synonym-
ous with fraud in common law or even criminal law. In fact, as
Solicitor-General,68 it appears that Lord Westbury stated that, in
secret trusts cases, there must be an arrangement, ‘the abandonment
of which by [the secret trustee] would amount to a fraud on the
testator’69 and that this ‘fraud or malus animus lay in the [secret
trustee] inducing, by his promise, the testator to confer on her bounty,
which, otherwise, he would not have conferred’.70 The interpretation
of Lord Westbury’s speech submitted here is also corroborated by his
comments on the nature of secret trusts in Cullen,71 and by the fact
that Lord Cairns, in McCormick,72 professed concurrence with Lord
Westbury.73 It is peculiar that, given how frequently Lord Westbury is
cited as the leading authority on the fraud theory,74 this interpretation
of his speech has apparently never before been proposed. The ortho-
dox interpretation of Lord Westbury’s judgment has long provided
ammunition for critics of the fraud theory, enabling them to bolster
their position with House of Lords’ authority. That interpretation, it is
submitted, is incorrect, and arguments and authorities relying on it75
should be reconsidered.
Support for this view can be gleaned from Norris v Frazer,76 in
which the secret trustee was a married woman. Upon the death of the
testator, her husband, who had never promised to perform the secret
trust, became entitled to the trust property in right of her. It is appar-
ent from the facts of the case that neither the husband nor the wife
had sought to enrich themselves fraudulently. The wife, not having
control over the secret trust property, was not in a position to gain
personally, and the husband’s conduct was praised by Bacon VC as
having been ‘frank and honourable and fair in every respect’.77 Never-
theless, the secret trust was enforced because ‘a more direct, a more
personal fraud could not be committed than for [the wife] to refuse to
perform that promise which she made to the testator on his death
68 Lord Westbury, then Sir Richard Bethell, served as Solicitor General between
1852 and 1856.
69 Lomax v Ripley (1855) 3 Sm & Gif 48 at 63. It appears from the report that these
are Sir Richard Bethell’s words. If not, the only other likely explanation is that this
is part of what is apparently a joint submission from counsels for the various
defendants.
70 Ibid. at 64.
71 Above n. 25 at 198.
72 Above n. 25 at 99. Lord Cairns did not give a full speech.
73 Lord Cairns’s view of the nature of the fraud is clear. See above, text to n. 50.
74 See above n. 54.
75 Ibid. Also, as regards the standard of proof, see above n. 22 and n. 28.
76 (1873) LR 15 Eq 318.
77 Ibid. at 330.
319
bed’.78 Tellingly, Bacon VC cited Lord Westbury’s speech in

McCormick79 as his main authority. Similarly, in Re Boyes,80 Kay J,
who cited McCormick,81 stated that the secret trustee’s promise is
‘binding the conscience of the donee, on the ground that otherwise a
fraud would be committed, because it is to be presumed that if it had
not been for such promise the testator would not have made or would
have revoked the gift’.82 Evidently, in the late nineteenth century, Lord
Westbury was not regarded as having introduced a requirement that
fraud must entail personal gain.
In conclusion, the uniformity regarding fraud in relation to fully
secret trusts during this period is extraordinary. The underlying
justification for the enforcement of fully secret trusts was repeatedly
held to be the instrument of fraud principle, and fraud was repeatedly
held to arise upon any breach by the secret trustee of an undertaking
on which the testator relied when deciding how to devise his prop-
erty. There was no inconsistency during this period, not even in
McCormick.83
(ii) Half-secret Trust Cases

Although Muckleston v Brown84 has been mentioned in relation to
fully secret trusts,85 in fact it was unclear in this case whether or not
the secret trustees took as trustees on the face of the will.86 Lord Eldon
apparently did not consider this issue to be material; he explained that
secret trusts are enforced to prevent fraud on the testator87 without
giving any indication that the question of whether or not the trust was
fully or half-secret was a crucial one. The obvious inference to draw is
that Lord Eldon considered that the instrument of fraud principle
applies to all secret trusts. This line of reasoning was followed in
Podmore v Gunning.88 Although the alleged half-secret trust was not
upheld, Shadwell VC stated that, had sufficient evidence been ad-
duced, he would have given effect to the half-secret trust on the in-
strument of fraud principle.89
78 Ibid. at 331.
79 Above n. 25.
80 (1884) LR 26 Ch D 531.
81 Above n. 25.
82 Above n. 80 at 535.
83 Above n. 25.
84 Above n. 41.
85 See above, text to nn. 41 and 43.
86 The testator’s residuary estate of realty, including a manor and an estate in a place
called Overseal, was devised, apparently absolutely, to the defendants. A
subsequent codicil, however, devised a farm to the defendants ‘upon trust for the
like uses and purposes as my manor and estate at Overseal now stand limited’.
The bill alleged that the estate at Overseal was subject to a secret trust in favour
of a charity. See above n. 41 at 64–6 for Lord Eldon’s consideration of this point.
87 Above n. 41 at 69.
88 (1836) 8 Sim 644.
89 Ibid. at 656–60.
320
In Smith v Attersoll,90 a half-secret trust was enforced, although

Lord Gifford did not explain why. He did, however, describe the secret
beneficiaries ‘not as legatees, but as cestuis que trust’91 and he de-
scribed the letter containing the half-secret trust’s details as ‘not to be
considered as testamentary’,92 citing a secret trust case as authority.93
This suggests that the half-secret trust was not incorporated into the
will by reference, as has been claimed,94 rather that it was enforced on
the instrument of fraud principle.
Another case which has been cited by proponents of the incorpora-
tion by reference theory95 is Johnson v Ball,96 in which a half-secret
trust that was not committed to writing until after the will’s execution
was held void on the ground that to enforce it ‘would be to receive, as
part of or as codicils to the will, papers subsequent in date to the will,
which are unattested’.97 Parker VC proceeded to explain that fully
secret trust cases ‘have no application to the present; nor . . . have
those cases cited in the argument, in which the will refers to a trust
created by the testator by communication with the legatee ante-
cedently to or contemporaneously with the will’.98 Parker VC’s de-
scription of validly communicated half-secret trusts as being created
‘by communication’99 strongly suggests that he did not consider them
to be incorporated into the will by reference. It should be noted that
although the details of the half-secret trusts had actually been orally
communicated to the trustees prior to the will’s execution, the will
referred to the half-secret trusts ‘being appointed by letter’. Thus, the
non-enforcement of the orally communicated trusts should not be
seen as evidence in favour of the incorporation argument because
enforcing these trusts would have been inconsistent with the will.100 It
should also be noted that, in Irvine v Sullivan,101 a half-secret trust that
was orally communicated prior to the will’s execution but not reduced
to writing until afterwards was enforced without question. In fact,
90 (1826) 1 Russ 266.

91 Ibid. at 271.
92 Ibid. at 270–1.
93 Jones v Nabbs, see above at n. 38. Note that other cases were also cited, but none
on incorporation by reference.
94 See Matthews, above n. 11 at 363.
95 Ibid.
96 (1851) 5 De G & Sm 84.
97 Ibid. at 90–1.
98 Ibid. at 91.
99 There is no requirement that, for a document to be incorporated into a will by
reference, it must be communicated to the legatee. See In bonis Smart [1902] P 238
PDAD at 240, cited in Martin, above n. 9 at 157, for a summary of the
requirements for a document to be incorporated into the will by reference.
100 See Re Keen [1937] Ch 236, for a further example of an apparently properly
communicated half-secret trust not being enforced due to inconsistency with the
will. In Keen, the enforcement of half-secret trusts was unequivocally attributed to
the prevention of fraud.
101 (1869) LR 8 Eq 673. See Re Young [1952] Ch 344 for another example of an orally
communicated half-secret trust being enforced. These seriously undermine the
incorporation by reference argument.
321
there is only one case, Re Baillie,102 in which the court declined to

enforce an orally communicated half-secret trust. The doctrine of in-
corporation by reference was not mentioned, however, and North J’s
reasoning was unclear.
Another instructive case is Briggs v Penny,103 in which although the
half-secret trust’s terms had been put in writing prior to the will’s
execution, no communication appeared to have taken place. Knight-
Bruce VC explained that unattested papers can be admitted either
under ‘the power of a testator to incorporate in his will another exist-
ing paper’104 or for ‘the prevention of fraud, by compelling a legatee to
perform, after the testator’s death, a promise made by him to the
testator, upon the faith of which the testator, to the knowledge of the
legatee, gave the legacy’.105 That he ordered the determination of
whether communication had been made106 indicates that he would
have been prepared to enforce the half-secret trusts for the preven-
tion of fraud should sufficient evidence of communication have been
established.
In Re Fleetwood,107 a properly communicated half-secret trust was
enforced. Hall VC, quoting directly from an Irish case, Riordan v
Banon,108 made it clear that half-secret trusts are enforced to prevent
fraud on the testator, stating that:
. . . the instruments of fraud [principle] appears to me to apply to cases
where the will shews some trust was intended, as well as to those where
this does not appear upon it. The testator, at least when his purpose is
communicated to and accepted by the proposed legatee, makes the
disposition to him on the faith of his carrying out his promise, and it
would be a fraud in him to refuse to perform that promise.109
He also described half-secret trusts as being created ‘by communica-

tion’,110 which is consistent with the fraud theory, as well as pointing
out that were the half-secret trust not to be enforced, ‘the residuary
legatees would stand to profit from the secret trustee’s fraud’.111
Finally, Hall VC held that one of the secret beneficiaries could not take
102 (1886) 2 TLR 660.

103 (1849) 3 De G & Sm 525.
104 Ibid. at 547.
105 Ibid.
106 Ibid. at 548.
107 (1880) LR 15 Ch D 594.
108 10 Ir Eq Rep 649.
109 Above n. 107 at 607.
110 Ibid. at 604.
111 That equity will not allow one man to profit from another’s fraud (the rule in
Huguenin v Baseley (1807) 14 Ves Jun 273) is a long-standing principle that has
been invoked several times in secret trusts cases (e.g. in Tee v Ferris, above n. 61
at 367 per Page Wood VC; Blackwell v Blackwell, above n. 1 at 241 per Lord
Warrington).
322
her interest because she had witnessed the will.112 Although this sug-
gests that the half-secret trust was considered to be part of the testa-
mentary disposition and was thus incorporated into it,113 Hall VC’s
unequivocal comments regarding fraud demonstrate otherwise.
In summary, the weight of authority from the nineteenth century
favours the view that the fraud theory applies to half as well as to fully
secret trusts, notwithstanding academic opinions to the contrary.114
This argument is considerably reinforced when these cases are read in
the light of the fully secret trust cases from the period,115 in which
fraud was consistently explained as not requiring any personal gain.
On the other hand, by the end of the nineteenth century, no half-
secret trust had actually been enforced under the incorporation by
reference principle or indeed for any reason other than to prevent the
statute being used as an instrument of fraud.
(c) The Twentieth-century Judgments

(i) Fully Secret Trust Cases
During this period, the fraud on the testator justification was en-
dorsed four times at first instance116 and thrice by the majority in the
Court of Appeal.117 In Re Stead,118 for example, a secret trust was
enforced to prevent fraud on the testator. Farwell J saw no contra-
diction in citing McCormick119 as authority that the doctrine can apply
only ‘in clear cases of fraud’120 while at the same time giving ‘the
fullest credit to the [secret trustee] for desiring to speak the truth’.121 It
is also notable that in Tharp v Tharp,122 Neville J applied the doctrine
112 Presumably due to s. 15 of the WA 1837. See below, text to n. 243 for further
discussion of this point.
113 This was included in the argument of Matthews, above n. 11 at 364–7, as was Hall
VC’s use of the word ‘incorporated’, above n. 107 at 608. This unfortunate choice
of words cannot, when the judgment is read as a whole, be taken literally to mean
that half-secret trusts are incorporated into the will by reference.
114 See above n. 12 and n. 13 for examples.
115 Although note the comments of Kay J in Re Boyes, above n. 80 at 535–6. He
appeared to endorse the view that fully secret trusts are enforced to prevent fraud
and half-secret trusts are enforced for other reasons, although it is not stated or
even hinted what those reasons are. This is not consistent with other judgments
on fully secret trusts from the period.
116 See Re Stead [1900] 1 Ch 237 at 240 per Farwell J; Re Falkiner [1924] 1 Ch 88 at 96
per Tomlin J; Re Williams [1933] Ch 244 at 250–2 per Farwell J; Ottaway v Norman,
above n. 7 at 709–11 per Brightman J.
117 See Re Pitt-Rivers [1902] 1 Ch 403 at 407 per Vaughan Williams LJ, Stirling LJ and
Cozens-Hardy LJ concurred; Re Maddock [1902] 2 Ch 220 at 225 per Collins MR
and at 227 per Stirling LJ; Re Gardner (No 1) [1920] 2 Ch 523 at 530 per
Warrington LJ and at 534–5 per Younger LJ.
118 Above n. 116.
119 Above n. 25.
120 Above n. 116 at 241.
121 Ibid. at 240.
122 [1916] 1 Ch 142.
323
of secret trusts to an analogous situation,123 explaining that ‘a sub-

sequent action by [the secret trustee] in contradiction of what he
promised is a fraud’ and that ‘he should not, either for himself or for
anybody else, take advantage of the fraud that he had committed’.124
Perhaps the seeds of uncertainty were sown in two of the Court of
Appeal cases. In Re Maddock,125 Cozens-Hardy LJ, although in the
minority on this point, professed uncertainty as to whether the doc-
trine of secret trusts was based on ‘trust, or contract, or estoppel’.126
In Re Gardner (No 1),127 Lord Sterndale, similarly in the minority,
stated that the doctrine is based on what ‘has been called in some of
the cases a fraud. I do not think it matters which you call it’.128 When it
is recalled that the reason why the statutory formality requirements
do not apply to secret trusts is the instrument of fraud principle, then
the terminology does matter, however. To enforce secret trusts for
reasons other than the prevention of fraud would be unconstitutional.
It must be stressed, however, that in both judgments, the majority
subscribed to the fraud on the testator explanation.
In two relatively recent cases, misconceptions regarding Lord
Westbury’s judgment in McCormick129 led to unnecessary questions
being asked regarding the standard of proof required in secret trusts
cases. In Ottaway v Norman,130 Brightman J, while accepting that the
fraud in question is fraud on the testator,131 interpreted Lord West-
bury’s comments regarding the need for clear evidence of fraud132 as
meaning that, in secret trusts cases, the standard of proof is ‘perhaps
analogous to the standard of proof which this court requires before it
will rectify a written instrument’.133 He enforced the secret trust with-
out further reference to the standard of proof, however, so his com-
ments should not be taken to be an accurate reflection of the law.134
123 At the insistence of the defendant, the testator destroyed a codicil that had
revoked a power of appointment. He did this in reliance on the defendant’s
assurance that he would not exercise the reinstated power of appointment to the
prejudice of the defendant. Neville J found in favour of the plaintiff, although the
parties eventually settled.
124 Above n. 122 at 151–2.
125 Above n. 117.
126 Ibid. at 232 per Cozens-Hardy J. Curiously, Matthews does not cite this case when
he proposes an estoppel-based justification for the doctrine. See Matthews, above
n. 11 at 361.
127 Above n. 117.
128 Ibid. at 529 per Lord Sterndale MR.
129 Above n. 25.
130 Above n. 7.
131 Ibid. at 709–11.
132 Above n. 25 at 97.
133 Above n. 7 at 712.
134 Brightman J’s comments regarding the adoption into English law of the ‘floating
trust’, based on Birmingham v Renfrew (1937) 57 CLR 666 ought similarly to be
disregarded, as they were, again, speculative and have not been followed.
324
In Re Snowden,135 Megarry VC based his judgment on the orthodox

interpretation of Lord Westbury’s speech in McCormick.136 Having
stated that secret trusts can be enforced in the absence of fraud,137 he
concluded, on the basis of Lord Westbury’s comments, that in secret
trust cases where fraudulent enrichment is possible, a high standard of
proof ought to be applied, and that in all other secret trust cases, the
ordinary civil standard should apply.138 Megarry VC’s comments on
fraud are incongruous with all other authorities and ought to be disre-
garded, and his assertions regarding the standard of proof are ground-
less, especially in light of the interpretation of Lord Westbury’s speech
that has been submitted here. If Megarry VC’s take on the doctrine of
secret trusts is to be accepted, it must also be accepted that in 1869, not
only did Lord Westbury propose a new definition of fraud, he also
introduced an especially high standard of proof in cases where this
fraud was present, and that both of these changes went unnoticed by
the courts, even the House of Lords,139 for over a century. It is thus
submitted, therefore, that academic arguments140 or judgments141 rely-
ing on Megarry VC’s comments ought to be reassessed.
(ii) Half-secret Trust Cases

The first significant case142 is the leading case, Blackwell v Blackwell,143
the most complete House of Lords analysis of the principles govern-
ing secret trusts. A half-secret trust, which had been properly com-
municated and accepted144 was unanimously upheld for the
prevention of fraud. The view that fraud must involve the secret trus-
tee gaining personally was rejected outright. Lord Buckmaster ob-
served that if either a fully or a half-secret trust is not performed, ‘the
real beneficiaries are equally defrauded in both cases, and the faith on
which the testator relied is equally betrayed’.145 Lord Warrington
135 Above n. 19.

136 Above n. 25.
137 Above n. 19 at 536.
138 Ibid.
139 See discussion of Blackwell v Blackwell, above n. 1 and below, text to nn. 142–9.
Their Lordships did not consider or apply an extraordinary standard of proof.
140 See above n. 22 for examples.
141 For English cases, see above n. 22. In fact, this problem is greater in some other
common law jurisdictions. See above n. 28 for examples.
142 The first two cases from this period, Re Huxtable [1902] 1 Ch 214 and Re Hetely
[1902] 2 Ch 866, are both cases where the validity of the doctrine of half-secret
trusts and the decision in Fleetwood, above n. 107, were inexplicably called into
question. In neither judgment was Fleetwood expressly overruled, however, and
in neither case were any reasons offered as to why Fleetwood might be incorrect.
Thus, neither has any effect on the submissions that have been put forward here.
143 Above n. 1.
144 I.e. before the execution of the will.
145 Above n. 1 at 328. Note the second ‘equally’ in Lord Buckmaster’s quote. Wilson,
above n. 10 at 261, suggests that the argument that the fraud in question is fraud
on the beneficiary is ‘circular’. The answer to this is that while any failure to
perform the secret trust would amount to a fraud on the testator, the secret
beneficiaries would be injured by this fraud as well.
325
explained the fraud in similar terms, describing a secret trust as ‘aris-

ing 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, as the case might be’,146 so that ‘it would be a
fraud on the part of the legatees to refuse to carry out the trust’.147
Finally, Viscount Sumner was unequivocal in endorsing the fraud
theory, stating that the enforcement of half-secret trusts is ‘justified by
the same considerations as in the cases of fraud and absolute gifts’.148
It is strange that doubts as to why half-secret trusts, and also fully
secret trusts, are enforced have persisted since Blackwell.149 Indeed,
the enforcement of half-secret trusts has been expressly attributed to
the prevention of fraud three times150 since 1929. It has to be asked
why a unanimous House of Lords’ judgment which is consistent with
the overwhelming majority of other authorities has been so frequently
called into question, and why the myth that equitable fraud must
involve personal gain has proven so persistent.
(d) The Twenty-first-century Judgments

Despite the authorities considered above, Megarry VC’s take on fraud
has recently been followed twice151 in the tribunal courts. More
seriously, in Kasperbauer v Griffith,152 in the Court of Appeal, Peter
Gibson LJ stated that, in secret trusts cases, ‘equity acts to prevent
fraud or other unconscionable conduct’.153 It is submitted that Peter
Gibson LJ failed to appreciate that, in equity, unconscionable conduct
is fraud. In Healey v Brown,154 however, a case on mutual wills,155 it
was stated that secret trusts are enforced for ‘avoidance of a fraud on
the beneficiary and the testator’.156 This was very recently reaffirmed
in the Court of Appeal in De Bruyne v De Bruyne.157 Although the case
did not actually concern a secret trust, in a unanimous decision, the
principles governing secret trusts were explained and applied to a
146 Above n. 1 at 341.

147 Ibid. Lord Warrington also reiterates this point at 341.
148 Ibid. at 335.
149 Above n. 1. See above nn. 9–18 and accompanying text for a summary and
examples of the various academic views on the doctrine.
150 Re Keen, above n. 100 at 244 per Lord Wright MR; Romer LJ and Greene LJ
concurred; Re Cooper [1939] Ch 811 at 815 per Green MR; Clauson LJ and
Goddard LJ concurred; Young, above n. 101 at 349 per Dankwerts J.
151 Taylor v Revenue & Customs Commissioners, above n. 4; Davies v Revenue and
Customs Commissioners, above n. 4.
152 [2000] 1 WTLR 333.
153 Ibid. at para. 27, official transcript.
154 [2002] WTLR 849.
155 Mutual wills were described by Nourse J in Re Cleaver [1981] 1 WLR 939 at 940 as
arising ‘where . . . two persons . . . made an enforceable agreement as to the
disposal of their property and executed wills in substantially identical terms in
pursuance thereof’.
156 Above n. 154 at 907 per Donaldson QC sitting as a Deputy High Court Judge.
157 Above n. 3.
326
situation that was held to be analogous.158 Patten LJ explained with

regard to secret trusts (tellingly, he did not distinguish between the
two types), and also mutual wills that:
In neither case does the intended beneficiary rely in any sense on the
agreement (he may not even be aware of it) but, in both cases, equity will
regard it as against conscience for the owner of the property to deny the
terms upon which he received it. It is not necessary in such cases to
show that the property was acquired by actual fraud (although the prin-
ciple would apply equally in such cases). The concept of fraud in equity
is much wider and can extend to unconscionable or inequitable conduct
in the form of a denial or refusal to carry out the agreement to hold the
property for the benefit of the third party which was the only basis upon
which the property was transferred. This is sufficient in itself to create
the fiduciary obligation and to require the imposition of a constructive
trust.159
Thus the consistency of the judiciary in relation to the nature of the
fraud continues until the present day. The very small number of cases
containing conflicting statements can only sensibly be dismissed as
being anomalous.160 As it is abundantly clear from the authorities that
equitable fraud need not involve personal gain, the idea that fraud can
be prevented by the imposition of a resulting trust in favour of the
testator’s estate, and that therefore the fraud theory cannot explain
why the secret trust should actually be enforced,161 ought to be dis-
regarded, as should concerns regarding the standard of proof to be
applied.162 It is equally clear that all secret trusts are justified on the
same principles. There is simply insufficient evidence from the case
law to support assertions that half-secret trusts are incorporated into
158 The appellant (Mrs De Bruyne) appealed against an ancillary order that certain
assets were not matrimonial assets. The respondent (Mr De Bruyne) had, along
with two others, been a beneficiary of a discretionary trust established by his
father. The trust property consisted of shares and other property. In 1991 the
three beneficiaries agreed to dissolve the trust. The agreement was that the other
two beneficiaries would take certain property and Mr De Bruyne would hold the
shares for the benefit of his children. Mr De Bruyne, however, upon obtaining the
shares, transferred them to Mrs De Bruyne, as his nominee, who sold them.
Certain assets were purchased with the proceeds of sale. These assets were the
subject of the appeal. The respondents maintained that the assets were held on
constructive trust for the children. The appeal was dismissed.
159 Above n. 3 at 51. This seriously undermines the arguments of McFarlane, above
n. 11 at 676 and Hudson, above n. 13 at 243. Both attempt to separate fraud from
unconscionability, arguing that secret trusts are enforced as constructive trusts to
prevent unconscionable conduct by the secret trustee, rather than fraud.
160 It is interesting to note that Australia’s highest court has also accepted the ‘fraud
on the testator’ argument. See Voges v Monaghan [1954] HCA 63 per Fullagar and
Kitto JJ. As in England, questions inexplicably continue to be raised regarding the
nature of the fraud and the standard of proof. See Howell v Hyde [2003] NSWSC
732 and Child Support Registrar & Kanavos & Ors [2009] FMCAfam 871 for
examples of this apparent unwillingness to accept the judgment of the highest
court. The fraud on the testator approach has also been followed in Ireland. In
fact, Fullagar and Kitto JJ directly quoted from the judgment of Lord Davey in
French v French (1902) 1 IR 172 at 241.
161 See, for example, Critchley, above n. 9 at 652; McFarlane, above n. 11.
162 See above n. 22 and n. 28 for examples.
327
the will by reference163 or are enforced for any other reasons.164 Sim-
ilarly, arguments that fraud on the testator is not a strong enough type
of fraud to explain equity’s intervention,165 and that there is a lack of
clarity in the judgments regarding that nature of fraud,166 are not
borne out by the authorities. The fraud which causes equity’s inter-
vention arises if the secret trustee’s promise to perform the trust,
upon which the testator relied when executing his testamentary dis-
positions, is not performed. Adoption of any other notion of fraud
would involve overturning not only several centuries of juris-
prudence, but also three House of Lords’ decisions.
iii. Is the ‘Instrument of Fraud’ Principle Unconstitutional?

Although the fraud theory is indisputably the underlying reason that
the courts have provided for the enforcement of secret trusts, it has
been claimed that the instrument of fraud principle is unconstitu-
tional.167 The judges involved in formulating the principle, however,
were acutely aware of their duty to follow Parliament’s will. In Reech v
Kennegal,168 Lord Hardwicke explained that ‘the statute should never
be understood to protect fraud; and therefore whenever a case is
infected with fraud . . . the court will not suffer the statute to protect
it’.169 Similarly, in Podmore v Gunning,170 Shadwell VC stated that ‘the
very worst method of construing the Statute of Frauds would be that
which would give rise to frauds instead of preventing them’.171 In
Wallgrave v Tebbs,172 much the same reasoning was applied by Sir
Page Wood, as he made clear that when enforcing secret trusts, the
court ‘does not violate the spirit of the statutes; but for the same end,
namely prevention of fraud, it engrafts the trusts on the devise by
admitting evidence which the statute would in terms exclude’.173 Per-
haps, then, the principle is better understood as that equity will not
allow a statute intended to prevent fraud to be used as an instrument
of fraud.174
163 See Sheridan, above n. 11; Matthews, above n. 11.

164 See Perrins, above n. 11; Andrews, above n. 11.
165 See Critchley, above n. 9; Challinor, above n. 9.
166 See McFarlane, above n. 11 at 676; Hudson, above n. 13 at 224. It is also clear that
the constructive trust arises at the moment of the testator’s death due to the
danger of fraud. This undermines Hudson’s argument that fraud cannot be the
justification for the doctrine because fraud theory operates on the basis that a
trust obligation arises only when ‘a claim of fraud’ is brought against the secret
trustee, i.e. as a ‘form of restitution’. This, he surmises, is incorrect because, in
reality, the secret trustee becomes a trustee at the death of the testator, regardless
of any subsequent court proceedings. See also below, Part V, for further
explanation of this point.
167 See especially Critchley, above n. 9 at 653; Challinor, above n. 9 at 497.
168 Above n. 30.
169 Ibid. at 125.
170 Above n. 88.
171 Ibid. at 655.
172 Above n. 47.
173 Ibid. at 322.
174 This point is made by Wilson, above n. 10 at 13, but is not widely accepted.
328
Cases involving the Mortmain Act 1736175 are also instructive be-
cause, in such cases, testators generally attempted to hide behind the
formality requirements of the Statute of Frauds as a means of facilitat-
ing dispositions that would otherwise have been rendered illegal by
the Mortmain Act. This prompted much judicial agonizing about the
intentions of Parliament, as is exemplified by Lord Eldon’s approach
in Stickland v Aldridge:176
It would be a strong proposition, that the providence of the Legislature,
having attempted expressly to prevent a disposition of land for purposes
of this sort, was so short as to be baffled by such a transaction as is
stated by this Bill . . . It would be singular, if the Court would protect
individuals, and would not act, to prevent a fraud upon the Law
itself.177
This again demonstrates that the intention of Parliament was para-

mount in the minds of the judges when the instrument of fraud prin-
ciple was developed.178
It appears that the instrument of fraud principle was fully formu-
lated by the early nineteenth century for, although it has been applied
many times, it has not since been explained in such detail. This does
not, however, mean that judges ceased to pay heed to the intention of
Parliament. In Blackwell179 Viscount Sumner, pointing out that secret
trusts have been enforced for several centuries, commented that the
WA 1837:
. . . is an amending Act, of which it may be said in no merely theoretical
sense that the Legislature was acquainted with the existing state of the
law . . . for two Royal Commissions . . . after enquiring (inter alia) into
the subject of wills of real and personal property had reported before
the Wills Act came before Parliament as a Bill. The extent to which parol
evidence was admissible under existing practice for various purposes
and the evils thereout arising were known . . . [and] no . . . remedy is
attempted by the Statute of Wills for the mischiefs that might arise from
admitting evidence [in secret trusts cases].180
In fact, as Lord Nottingham referred to the enforcement of secret

trusts as ‘the constant course of this court’181 and regarded the doc-
trine ‘as established’,182 it seems very likely that the legislature was
175 Above n. 42.

176 See above n. 43.
177 Ibid. at 519.
178 See Boson v Statham (1760) 1 Cox 16 at 18–20; Wallgrave v Tebbs, above n. 47 at
321–8; Adlington v Cann (1744) 3 Atk 141 for further examples.
179 Above n. 1.
180 Ibid. at 338.
181 Chamberlaine v Chamberlaine, above n. 29 at 35 per Lord Nottingham.
182 Walpole v Orford (1797) 3 Ves Jun 402 at 410. The Solicitor General made this
comment in relation to an apparently unreported secret trust case from the time
of Lord Nottingham called Berenger v Berenger.
329
equally aware of the doctrine in 1677 when the first compulsory stat-
utory formality requirements were enacted.183
Therefore, although there is no express reference to the doctrine of
secret trusts in the WA 1837 or the Statute of Frauds, it seems very
likely that the continued existence of the doctrine of secret trusts was
tacitly endorsed by Parliament when the statutory formality require-
ments for testamentary dispositions were enacted and amended with-
out reference to secret trusts. So, if Viscount Sumner’s view of
Parliament’s intention regarding secret trusts is accepted, to state that
the doctrine of secret trusts is unconstitutional is, in fact, to disregard
not only the weight of authorities, but also the will of Parliament. It is
clear that the judiciary has taken the view that to fail to enforce secret
trusts and thus permit s. 9 to be used as an instrument of fraud would
be subversive to the doctrine of parliamentary sovereignty. As Cotton
LJ put it, in cases of correctly communicated and accepted secret
trusts, ‘the court is justified and bound to admit parol evidence’.184
iv. Do the Differing Communication Requirements for Fully and

Half-secret Trusts Undermine the Fraud Theory?
The rule that half-secret trusts must be communicated before or
contemporaneously with the will’s execution has generated much
controversy. Numerous justifications for its existence have been pro-
posed, and several commentators have used the rule as the corner-
stone of their assertions regarding the justifications for the doctrine
itself.185 In fact, the rule is simply the result of policy decisions.186 In
Blackwell,187 Viscount Sumner stated that a half-secret trust could not
be enforced if communicated after the will’s execution; this would be
tantamount to allowing a testator to ‘reserve to himself a power of
making future unwitnessed dispositions [and thus to] “give the go-by”
to the requirements of the Wills Act’.188 Put another way, whereas
Viscount Sumner considered the enforcement of properly communi-
cated secret trusts to be in accordance with the legislature’s intention,
183 This also undermines the argument to the effect that the legislature ought to
intervene and place the doctrine on a statutory footing. See Martin, above n. 9 at
175.
184 Re Spencer’s Will (1887) 57 LT 519 at 521.
185 See particularly Matthews, above n. 11, and Perrins, above n. 11. For an elaborate
justification of the differing communication requirements, see D. Wilde, ‘Secret
and Semi-Secret Trusts: Justifying the Distinctions between the Two’ [1995] Conv
366.
186 See above n. 8; Johnson v Ball, above n. 96; Re Hetely, above n. 142; Re Keen,
above n. 100; all of which provide the same explanation as that of Viscount
Sumner in the text to n. 188 below. The rule was also applied in Re Bateman’s WT
[1970] 1 WLR 1463.
187 Above n. 1.
188 Ibid. at 339.
330
he considered that Parliament did not intend half-secret trusts com-

municated after the will to be enforceable,189 apparently because the
will in such cases serves as an overt statement by the testator that
although he has made his will, he still intends to alter by parol the final
destination of legacies bequeathed under it.
Viscount Sumner saw no contradiction between interpreting the
effect of the WA 1837 thus and attributing the enforcement of secret
trusts to the prevention of fraud, nor is any such contradiction sug-
gested in any of the other authorities.190 This is unsurprising because
even if the non-performance of a half-secret trust communicated and
accepted after the execution of the will amounts to a fraud on the
testator, it has been held on several occasions that the enforcement of
such half-secret trusts is against the policy of the WA 1837. Thus, from
a constitutional standpoint, equity is powerless to intercede to prevent
any such fraud. In jurisdictions where the same policy decision has
not been made, equity may act.191 An analogy may be drawn with the
Mortmain cases. In many such cases,192 the court accepted that the
secret trustees took subject to a mandatory obligation to perform
the secret trust because any non-performance would amount to a
fraud on the testator. This mandatory obligation upon the secret
trustee also amounted to an illegal devise, however, and the property
was returned to the testator’s estate by resulting trust. Again, because
of the manner in which legislation was interpreted by the judges,
equity was powerless to intercede to prevent the fraud on the
testator.
It is also worth noting that a will which refers to a half-secret trust
to be communicated in the future also serves as an overt and legally
significant193 statement that the testator has executed his will subject
to a secret trust without relying on any undertaking by the secret
trustee to perform it. In stark terms, the testator, having executed the
will referring to the secret trust, has accepted the possibility that he
may die without having procured acceptance of the secret trust obli-
gation. Therefore, it is in fact arguable that no fraud on the testator
occurs if the secret trust is not performed, whether or not commun-
ication and acceptance eventually took place. The same cannot be said
in the case of a fully secret trust communicated after the execution of
189 It must be said that judges in several common law jurisdictions have reached a
different conclusion regarding the communication requirement. For Ireland, see
Re Browne [1944] IrR 90; for New South Wales, Ledgerwood v Perpetual Trustee
Co Ltd (1997) 41 NSWLR 532; and for the United States of America, see American
Law Institute, Restatement of Trusts 55, 2nd edn 1959, comment (c), (h). All of
these references are cited in Martin, above n. 9 at 167, nn. 75–77.
190 In particular, in Re Keen, above n. 100, the fraud theory and the communication
requirement were accepted by the Court of Appeal.
191 See above n. 189 for examples of jurisdictions where the communication rule for
half-secret trusts is the same as for fully secret trusts.
192 E.g. Boson v Statham, above n. 178; Russell v Jackson, above n. 48.
193 I.e. a statement or act capable of contributing towards the creation of a valid
enforceable secret trust.
331
the will. In this latter case, there was no secret trust when the will was
executed because it was not mentioned in the will, nor had it been
communicated to anyone. Even if the secret trust existed in the mind
or the written notes of the testator, he had, at this point, taken no
steps to put this plan into operation. Hence, unlike in the situation
described above, it cannot be said that the testator executed his will
subject to any secret trust at all. If, after the execution of the will, the
testator communicates a secret trust to the relevant legatee, this is
the first overt and legally significant evidence of the existence of the
secret trust. If the testator does not amend his will after receiving an
express or implied assurance that the secret trust will be performed, it
is reasonable to presume that he is relying on this assurance. Thus,
the assertion that the differing communication requirements do not
impugn upon the fraud theory can be supported by the application of
logic, regardless of policy considerations. This avenue is yet to be
explored in the English courts, possibly due to the policy reasons
outlined above, but may warrant further consideration, especially in
jurisdictions where half-secret trusts communicated after the will’s
execution are enforceable, apparently on the ground of fraud.
III. The Dehors the Will Theory

Although it is beyond doubt that the secret trusts are enforced due to
the instrument of fraud principle, it is also apparent from the author-
ities that secret trusts are dehors the will and enforced in a way that
does not conflict with the WA 1837. This point was made by Viscount
Sumner in Blackwell v Blackwell,194 who, in addition to being un-
equivocal in attributing the enforcement of secret trusts to the preven-
tion of fraud, said that he could ‘not see how the statute-law relating
to the form of a valid will is concerned at all’195 with the enforcement
of secret trusts. In fact, there are numerous examples throughout the
case law of express or implied references to secret trusts being dehors
the will by judges endorsing the fraud theory.196 As this is at odds with
the majority of academic opinion,197 it is necessary to explore the
development of the theory in the courts. If the sections on fraud are
seen as explaining why secret trusts are enforced, the following sec-
tions can be considered as identifying how this is done.198
Many of the debates regarding the relevance of s. 9 to the doctrine
revolve around the question of whether secret trusts are testamentary
194 Above n. 1.
195 Ibid. at 334.
196 See, for example, Sweeting v Sweeting (1863) 33 LJ Ch 211; Re Spencer’s Will,
above n. 184; Re Keen, above n. 100; Re Young, above n. 101; Ottaway v Norman,
above n. 7.
197 See above n. 9 for examples.
198 This point is made by Viscount Sumner in Blackwell v Blackwell, above n. 1 at 337
and also by Wilson, above n. 10 at 261.
332
dispositions, in which case they are exceptions to the WA 1837,199

express inter vivos dispositions, in which case the WA 1837 is irrelev-
ant to questions regarding their enforcement,200 or constructive
trusts.201 Before proceeding, therefore, the meaning of s. 9 should be
reconsidered. The section states that ‘no will shall be valid’ unless it is
in writing and signed by the testator in the presence of two or more
attesting witnesses. The meaning of the term ‘will’ is described in s. 1
as extending to ‘a testament, and to a codicil . . . and to any other
testamentary disposition’. Thus, the term ‘testamentary disposition’ is
synonymous with the term ‘will’. The meaning of s. 9 is that all valid
testamentary dispositions must be executed in accordance with the
formality requirements. Any attempt to dispose of property upon
death that does not comply with s. 9 shall not be valid and therefore
cannot correctly be referred to as a testamentary disposition. In other
words, s. 9 regulates attempts at making testamentary dispositions.
Actual testamentary dispositions by their very nature satisfy s. 9. So,
to state that a secret trust is not a testamentary disposition is a state-
ment of the obvious. This does not necessarily mean that the WA 1837
is irrelevant to questions regarding its enforcement.
This narrow definition of ‘testamentary disposition’ is not widely
accepted.202 A wider definition that has been proposed is that any
ambulatory and revocable disposition that is enforced is a testament-
ary disposition, regardless of whether it appears on the document
commonly referred to as the ‘will’,203 and regardless of whether it is
admitted to probate. Such ‘testamentary dispositions’ are regulated
by s. 9. By this definition, a secret trust is a testamentary disposition
which is enforced as an exception to the Wills Act.
In order to answer the difficult questions204 surrounding the posi-
tion of the secret trust vis-à-vis the statutory formality requirements
of the Wills Act, the creation and enforcement of the secret trust will
be examined in detail.
i. Is the Testator’s Declaration of Trust a Testamentary

Disposition?
There is a strong judicial consensus that the testator’s declaration of
trust is not a testamentary disposition. In Chamberlain v Agar,205 in
which a fully secret trust expressed in writing was upheld, the Vice
199 See especially Critchley, above n. 9.

200 See especially Pearce and Stevens, above n. 9 at 224–9.
201 See especially Sheridan, above n. 11; Andrews, above n. 11; McFarlane, above n.
11; Hudson, above n. 13.
202 See especially Critchley, above n. 9. Also, for example, Martin, above n. 9 at 155,
describes secret trusts as ‘Testamentary Gifts not complying with the Wills Act
1837’.
203 See especially Critchley, above n. 9 at 634.
204 For examples of debates on this, see especially W. Holdsworth, ‘Secret Trusts’
(1937) 53 LQR 501; Holdsworth, above n. 9; Critchley, above n. 9; Kincaid, above
n. 16; Challinor, above n. 9.
205 Above n. 44.
333
Chancellor stated that ‘no Paper exists, that can be properly described
as a Will, Codicil, or testamentary Paper’.206 Similarly, in Smith v
Attersoll,207 it was held that the paper containing the details of the
half-secret trust was ‘not to be considered testamentary’,208 and in
Briggs v Penny,209 of four papers containing the details of purported
half-secret trusts, Knight-Bruce VC said that ‘not one can . . . be
treated or considered as testamentary. If any one of them is valid at
all, it can only be deemed valid in some other character than as a
testamentary instrument.’210 This is consistent with the view of Stuart
VC in Lomax v Ripley,211 who made it clear that papers containing
details of purported secret trusts did not ‘have a testamentary char-
acter’212 because they could not have been ‘admitted to probate’.213
Likewise, in Re Maddock,214 Collins MR referred to the written terms
of the secret trusts as ‘collateral non-testamentary document[s]’,215
and in Re Bateman’s WT,216 Pennycuick VC described a clause in the
testator’s will purporting to create a half-secret trust as ‘an attempt to
dispose of the estate by a non-testamentary instrument’.217 Finally, in
Re Cooper,218 Greene MR dismissed the argument that half-secret
trusts should be incorporated into the will as an argument for ‘incor-
porating by reference into the testamentary dispositions of the test-
ator the trusts actually declared to his trustees’.219 Clearly, he did not
regard the declarations of secret trusts as testamentary dispositions.
Significantly, there is not one case in which the testator’s declaration
of trust is described as being a testamentary disposition.220 What is
also apparent from these authorities is that by ‘testamentary disposi-
tion’, the judges concerned were referring to the narrow definition of
the term, i.e. the testator’s will and anything incorporated therein, as
opposed to the wider definition.
206 Ibid. at 263.

207 Above n. 90.
208 Ibid. at 270–1.
209 Above n. 103.
210 Ibid. at 527.
211 Above n. 69.
212 Ibid. at 76.
213 Ibid.
214 Above n. 117.
215 Ibid. at 224. This contradicts what is stated about the case by Challinor, above n. 9
at 497, who suggests that Maddock is an authority that suggests that a secret trust
is not dehors the will because ‘it was treated as if it had been made by will’. In
Maddock, however, it was emphasized that although the secret trust was, for the
purposes of the litigation, being treated as part of the will, it was in fact not part
of the will. This authority supports the assertion that a secret trust is dehors the
will.
216 See above n. 186.
217 Ibid. at 1468.
218 Above n. 150.
219 Ibid. at 819.
220 Re Maddock, above n. 117, has been cited as authority that secret trusts are
testamentary dispositions (e.g. by Challinor, above n. 9). In fact, the secret trusts
were merely treated as part of the testamentary disposition for the purposes of the
litigation.
334
The next question is whether s. 9 applies to the testator’s declara-

tion of trust. The most instructive cases are those which deal with
uncommunicated secret trusts which are, of course, void. The reason
for this is that the testator’s declaration of trust does not comply with
the WA 1837. The leading case on this rule is Wallgrave v Tebbs,221 in
which Page Wood VC refused to enforce the uncommunicated secret
trust on the grounds that it is ‘impossible for the Court to look upon a
document which is excluded by the statute’.222 Similarly, in Moss v
Cooper,223 the same judge explained that ‘if you attempt to raise a
trust out of some uncommunicated intention, you contravene the ex-
press provisions of the statute by varying the dispositions of the will
by parol evidence’.224 In Briggs v Penny,225 it was also held by Knight-
Bruce VC that the testator’s papers were not admissible as evidence
as ‘the statute of 1837 seems to me to exclude them’.226 The require-
ments of communication and acceptance have frequently been re-
affirmed, including by the House of Lords.227
These authorities reaffirm that the narrow definition of testament-
ary disposition is the correct one, as they are direct authorities that
the Wills Act does apply to the declaration of secret trust, even though
it has been repeatedly held that such a declaration is not a testament-
ary disposition. Thus, assertions that secret trusts are enforceable as
express inter vivos trusts, to which the Wills Act has no application,228
are unsustainable. Express inter vivos trusts are unilateral229 and need
not comply with statutory formality requirements for valid wills.230
The testator’s declaration of secret trust is affected by the statutory
formality requirements, and the secret trust can only be given effect if
the testator’s declaration is communicated to and accepted by the
secret trustee. Classifying a secret trust as an express inter vivos trust
is simply contrary to the authorities. Perhaps the reason why the
classification of secret trusts as express inter vivos trusts has endured
is that it seems to be presumed that, if the declaration of secret trust is
not a testamentary disposition, then it must be an inter vivos disposi-
tion.231 In fact, the declaration of secret trust has no effect due to non-
221 Above n. 47.

222 Ibid. at 326. See also Tee v Ferris, above n. 61 at 366 per Sir Page Wood VC;
Whitton v Russell, above n. 35 at 449 per Lord Hardwicke.
223 (1861) 1 J & H 352.
224 Ibid. at 366.
225 Above n. 103.
226 Ibid. at 547.
227 See McCormick v Grogan, above n. 25; Blackwell v Blackwell, above n. 1.
228 For examples, see above n. 16.
229 For a comprehensive list of theoretical reasons why secret trusts are not express,
see R. Burgess, ‘The Juridical Nature of Secret Trusts’ (1972) 263 Northern Ireland
Legal Quarterly 23.
230 They should, of course, comply with statutory formality requirements for inter
vivos trusts. This is another controversial point, although one which is addressed
by the classification of secret trusts as constructive trusts (see III.iii below).
231 See especially Pearce and Stevens, above n. 9 at 224–9, for a summary of this line
of thinking.
335
compliance with the Wills Act. It is plainly inaccurate to describe as an

inter vivos disposition that which is not a disposition at all.
Although not a disposition, the declaration of secret trust by the
testator has been described as being ‘in furtherance of the testament-
ary dispositions’,232 and, in Blackwell v Blackwell,233 Lord Buckmaster
referred to the declarations of secret trusts as ‘testamentary inten-
tions’,234 as did Vaughan Williams LJ in Re Pitt-Rivers.235 It is therefore
submitted that the testator’s declaration of trust is best described as
an expression of his testamentary intentions.
ii. Is the Secret Beneficiary’s Interest Acquired by Testamentary

Disposition?
It would seem self-evident that if the testator’s declaration of secret
trust is not a testamentary disposition, then the interest that the
secret beneficiary takes cannot be part of a testamentary disposition
either. This is indeed the case, as several authorities demonstrate. The
starting point is the House of Lords’ judgment of Cullen v Attorney-
General for Ireland,236 in which it was held that a secret trust was not a
testamentary disposition and did not therefore qualify for a certain
taxation exemption.237 Although this case has been dismissed as ‘a
policy decision’,238 it deserves closer attention. Lord Westbury de-
scribed the title claimed by a secret beneficiary as ‘a title dehors the
will, and which cannot be correctly termed testamentary’,239 and Lord
Chelmsford reached a similar conclusion.240 In Re Young,241 it was
held that a half-secret beneficiary could take his interest, even though
he had witnessed the testator’s will.242 Although s. 15 of the WA 1837
renders void any disposition by will to a person who has been a
witness to that will, it was held that ‘a beneficiary under a secret trust
does not take under the will, and that he is not, therefore, affected by
s. 15’.243 This is further authority that the secret beneficiary’s interest
is not bestowed upon him by way of a testamentary disposition. These
232 Johnson v Ball, above n. 96 at 91 per Parker VC.

233 Above n. 1.
234 Ibid. at 325.
235 Above n. 117 at 407.
236 Above n. 25.
237 The wording of the Stamp Act in question (5 & 6 Vict. c. 82 s. 38 and 8 & 9 Vict. c.
76 s. 4) stated that the exemption from duty applied to ‘a gift by any will or
testamentary instrument’.
238 See Critchley, above n. 9 at 641.
239 Cullen, above n. 25 at 198.
240 Ibid. at 197–8.
241 Above n. 101.
242 Section 15 of the Wills Act renders void any disposition by will to a person who
has been a witness to that will.
243 Above n. 101 at 351.
336
authorities are entirely consistent with the narrow definition of testa-

mentary disposition;244 by this definition, it is obvious, despite numer-
ous assertions to the contrary,245 that a secret trust is not enforced as a
testamentary disposition for the simple reason that it does not appear
in full in the will. Thus, as well as it being impossible that a secret trust
is enforced as an express inter vivos trust, it is also plainly the case
that a secret trust is not an express testamentary trust either.246
iii. How is the Secret Beneficiary’s Interest Acquired?

It is well established that secret trusts must be communicated to and
accepted by the secret trustee in order for the secret beneficiaries to
acquire an interest on the death of the testator.247 It is as a result of the
communication and acceptance of the secret trust obligation that, at
the time of the testator’s death, the secret trustee’s conscience is af-
fected.248 This is entirely consistent with the classification of a secret
trust as a constructive trust which is imposed on the secret trustee
because any deviation from his promise would amount to a fraud,249
rather than as an express trust. Again, there are a number of in-
structive authorities. In Stickland v Aldridge,250 Lord Eldon stated that,
in cases of secret trusts, ‘though within the intention [of the legis-
lature] it cannot be said a trust is declared under these circumstances,
it is clear, a trust would be created, upon the principle, on which this
Court acts, as to fraud’.251 Put another way, the testator’s declaration
of trust cannot be said to create an enforceable express trust due to
the statutory formality requirements, but because failure to adhere to
the agreement would constitute a fraud on the part of the secret
trustee, a constructive trust arises. Direct authority that a secret trust
is not an express trust can be found in Lomax v Ripley,252 in which
Stuart VC stated that secret trusts may only be enforced if it is possi-
ble to ‘prove by evidence a trust expressed, or such an engagement by
words or by silence as would authorize the Court to say that [the
secret trustee] undertook to do that which prevented the devisor from
244 See also Re Keen, above n. 100 at 244 per Lord Wright.
245 See, for example, Hodge, above n. 9; Wilde, above n. 185; Critchley, above n. 9;
Hayton and Mitchell, above n. 14.
246 Both Hodge, above n. 9, and Wilde, above n. 185, apparently classify secret trusts
as express testamentary trusts.
247 See, however, Re Gardner (No 2) [1923] 2 Ch 230, which appears to indicate that
some sort of beneficial interest may pass prior to the testator’s death. There are,
however, inconsistencies in Romer J’s judgment (especially at 232–3), which is
contrary to other authorities and appears to expressly contradict the earlier Court
of Appeal judgment of Re Maddock, above n. 117.
248 See above nn. 221–7 and accompanying text for authorities regarding the
requirements for communication and acceptance.
249 Although note that constructive trusts are notoriously difficult to define. In, for
example, Yeoman’s Row Management Ltd v Cobbe [2008] 1 WLR 1752 at 1769,
Lord Scott said that ‘It is impossible to prescribe exhaustively the circumstances
sufficient to create a constructive trust . . .’.
250 Above n. 43.
251 Ibid. at 519.
252 Above n. 69.
337
imposing upon her an express trust’.253 Lord Westbury used language

consistent with the constructive trust classification in McCormick v
Grogan,254 stating that equity ‘imposes upon [the secret trustee] a
personal obligation, because he applies the Act as an instrument for
accomplishing a fraud’.255 The same view was echoed in Re Spencer’s
Will.256 Cotton LJ stated that in alleged secret trust cases:
. . . parol evidence is to be produced for the purpose of showing that
there were circumstances which induced the testator to make this be-
quest, and which would enable the court to fasten upon the legatees an
obligation or trust . . . [and that] without creating an express declaration
of trust . . . under the circumstances, implies an obligation of perform-
ing the wishes of the testator which the testator had relied upon their
performing as a ground for giving the legacy.257
Another illuminating judgment is that of Kay J in Re Boyes,258 who

said that in cases of valid secret trusts, ‘the Court has compelled
discovery and performance of the promise, treating it as a trust bind-
ing the conscience of the donee, on the ground that otherwise a fraud
would be committed’.259 Similarly, in Re Pitt-Rivers,260 Vaughan Wil-
liams LJ described the secret beneficiary’s interest as being ‘created
by the giving by the [secret trustee] of a promise which it would be
unconscientious for [the secret trustee] not to perform’261 (as opposed
to being created by the testator’s declaration of trust). In Blackwell,262
Lord Warrington described a secret trust in similar terms, as ‘arising
from the acceptance by the legatee of a trust, communicated to him by
253 Ibid. at 73. Note also that Stuart VC was paraphrasing the comments of Lord
Eldon in Paine v Hall (1812) 18 Ves Jun 475 at 475.
254 Above n. 25.
255 Ibid. at 97.
256 Above n. 184.
257 Ibid. at 521–2. It ought to be pointed out here that what are now commonly called
constructive trusts have not always necessarily been described as so. Two broadly
contemporary definitions are provided here. Smith states that ‘implied trusts and
constructive trusts . . . are frequently confounded or classed together; and the
same trusts are sometimes designated by the name of implied trusts, and at other
times by that of constructive trusts’, and that a constructive trust ‘may be defined
to be a trust which is raised by construction of equity, in order to satisfy the
demands of justice, without reference to any presumable intention of the parties’:
J. Smith, Smith’s Equity Jurisprudence, 5th edn (Stevens & Norton: London, 1856)
151. Similarly, Story states that ‘implied trusts may be divided into two general
classes: first, those which stand on the presumed intention of the parties; secondly
those which are independent of any such intention, and are forced upon the
conscience of the party by operation of law’: J. Story, Commentaries on Equity
Jurisprudence as administered in England and America, 9th edn (Little, Brown and
Company: Boston, 1866) 414. Therefore, Cotton LJ’s use of the term ‘implied’
assumes extra significance.
258 Above n. 80.
259 Ibid. at 535.
260 Above n. 117.
261 Ibid. at 408.
262 Above n. 1.
338
the testator, on the faith of which acceptance the will was made or left
unrevoked . . .’.263
In Blackwell,264 Viscount Sumner emphasized the similarities be-
tween secret trusts and resulting trusts, pointing out that it would be
illogical for the court to refuse to give effect to a half-secret trust on
the ground that it does not appear in the will but then to impose
unquestioningly a resulting trust in favour of the estate ‘[as if] the will
gives the fund to the legatee in trust for the residuary legatee, as if the
document, signed and witnessed, had said so in words’.265 Neither
resulting nor constructive trusts appear in the will, but both are im-
posed when appropriate as a result of the ‘exercise of general equit-
able jurisdiction’266 with which the legislature has never sought to
interfere. Resulting trusts, which are routinely and uncontroversially
imposed on legacies,267 are dehors the will in exactly the same way as
secret trusts are. This also goes a long way towards answering those
who claim that a half-secret trust, because it is identified on the face of
the will, must be an express trust.268 A half-secret trust is no more an
express trust than is a resulting trust arising when property is be-
queathed by will to a trustee to hold subject to a trust appearing in the
will but whose objects are uncertain. There is no reason why a person
named in the will as a trustee may not, in appropriate circumstances,
be compelled to hold on constructive trust for a party not named in
the will.
More recently, in Kasperbauer v Griffith,269 Paragon Finance v
Thakerar,270 Healey v Brown,271 and De Bruyne v De Bruyne,272 secret
trusts were described as constructive trusts, and in Re Tyler,273 secret
trusts were expressly distinguished from ‘trusts created inter vivos’.274
Thus, in the face of so many authorities, the classification of secret
trusts as constructive trusts, imposed for the prevention of fraud, can
scarcely be in doubt.275 As secret trusts are constructive trusts,
263 Ibid. at 341. See also Re Falkiner, above n. 116 at 95 per Tomlin J.
264 Above n. 1 at 338.
265 Ibid.
266 Ibid. at 339.
267 For example, in the case of a half-secret trust communicated after the execution of
the will (e.g. Briggs v Penny, above n. 103), or in a case where a party takes as a
trustee but the objects or beneficial interests are uncertain (see Boyce v Boyce
(1849) 16 Sim 476).
268 See especially Sheridan, above n. 11; Andrews, above n. 11; Martin, above n. 9.
269 Above n. 152 at para. 27, official transcript per Peter Gibson LJ.
270 [1999] 1 All ER 400 at 409 per Millett LJ.
271 Above n. 154 at 907 per Donaldson QC sitting as a Deputy High Court Judge.
272 See above n. 3.
273 [1967] 1 WLR 1269.
274 Ibid. at 1275.
275 Note that, in Brown v Pourau, above n. 28 at 368, Hammond J stated that the
imposition of a remedial constructive trust is the reason why the doctrine of
secret trusts need not be attributed to the prevention of fraud. Hammond J also
claimed that the same was true in the USA and Canada. If this statement is
339
s. 53(1)(b) LPA 1925 is not relevant to questions regarding their en-

forcement.276 Although the question of whether this provision should
affect secret trusts is a favourite topic of commentators,277 there is,
perhaps unsurprisingly, a notable scarcity of references to s. 53(1)(b)
or its predecessor in the case law.278 Simply put, the view that secret
trusts are express trusts,279 or that their classification as express or
constructive trusts is immaterial or impossible to ascertain,280 is not
supported by the authorities. That the secret beneficiary’s interest
arises dehors the will by way of a constructive trust, the imposition of
which is detached from the requirements of s. 9, also further under-
mines the argument that the doctrine of secret trusts is unconstitu-
tional. It is also submitted that much of the confusion surrounding the
dehors the will theory has stemmed from a fundamental misunder-
standing of the meaning of ‘testamentary disposition’ and ‘will’ and
thus s. 9. If the narrower meaning proposed here is accepted, all of the
authorities cited here can be reconciled with one another.
IV. Do Social Changes since the Doctrine was Developed

Mean that it is now Obsolete?
It is often assumed that the doctrine of secret trusts was developed in
order to facilitate secret bequests to mistresses or illegitimate fam-
ilies.281 This has led some to argue that, because illegitimacy and in-
fidelity are more socially acceptable today than they once were, the
doctrine of secret trusts is no longer of relevance.282 In fact, out of 60
correct, one must ponder why it is still necessary to rely on cases such as
McCormick v Grogan and Re Snowden in such jurisdictions in order to
demonstrate that fraud is no longer relevant.
276 Section 53(2) LPA 1925 expressly exempts constructive trusts from the
requirements of s. 53(1).
277 See, for example, McFarlane, above n. 11; Critchley, above n. 9; Martin, above
n. 9.
278 Most notoriously in Ottaway v Norman, above n. 7, an orally expressed and
communicated secret trust of land was enforced without mention of s. 53(1)(b).
The now overruled Re Baillie, see above n. 102 and accompanying text, should
also be noted as North J apparently considered that half-secret trusts must be
reduced to writing in order to comply with s. 7 of the Statute of Frauds (s.
53(1)(b)’s predecessor). This is a rather confused judgment though, and is the sole
authority on this.
279 See, for example, Hodge, above n. 9; Wilde, above n. 185; Pearce and Stevens,
above n. 9; Pettit, above n. 39. Also, Sheridan, above n. 11, insists that, while fully
secret trusts are constructive, half-secret trusts must be express, as do Andrews,
above n. 11 and Perrins, above n. 11.
280 Hayton and Mitchell, above n. 14; R. Clements and A. Abass, Equity and Trusts,
1st edn (Oxford University Press: Oxford, 2009) ch. 9.2; Martin, above n. 9;
Wilson, above n. 10.
281 See Sheridan, above n. 11 at 328; Critchley, above n. 9 at 645; Challinor, above n. 9
at 498; Martin, above n. 9 at 156; Pearce and Stevens, above n. 9 at 207; Wilson,
above n. 10 at 250.
282 See especially Challinor, above n. 9, also Critchley, above n. 9.
340
secret trust cases283 that have been identified, only six284 clearly con-
tain secret trusts in favour of mistresses or illegitimate children, and
two others285 possibly do. Of the others, 19 are Mortmain cases. The
remaining cases contain examples of secret trusts created for a multi-
tude of reasons; for example, there are cases where the testator was
persuaded to create the secret trust by a deceitful secret trustee,286
cases where the testator wanted to change his will at a very late stage
and it appeared easier to create a secret trust than to amend the
will,287 cases where the secret trust arose as a result of a poorly
drafted will,288 cases where the secret trust was created in an apparent
attempt to avoid tax liability,289 cases where the secret trust was cre-
ated as a result of legal advice,290 and cases where the secret trust
arose out of the testator’s indecisiveness.291 It is clear that the doctrine
was not developed to allow testators to provide secretly for illegit-
imate families, or indeed for reasons involving the Mortmain Acts.292
In fact, the motivation of the testator is rarely mentioned in the judg-
ments. The obvious conclusion to draw is that any failure to perform a
secret trust is a fraud, regardless of why the secret trust was created.
V. Conclusion
Despite the lack of academic consensus regarding the justifications
for the enforcement of secret trusts, analysis of the case law has
revealed remarkable judicial consistency regarding the doctrine. The
inescapable conclusion is that secret trusts are enforced, notwith-
standing the formality requirements, because equity will not allow s. 9
283 In addition to those cited so far, the following are Mortmain cases: Edwards v Pike
(1759) 1 Eden 267; A-G v Poulden (1837) 8 Sim 472; Johnstone v Hamilton (1865) 12
LT 823; Springett v Jennings (1870) LR 10 Eq 488; O’Brien v Tyssen (1884) 28 Ch D
372. Other examples of cases concerning secret trusts are: Burney v Macdonald
(1845) 15 Sim 6; Le Page v Gardom (1915) 84 LJ Ch 749; Re Wedgwood [1915] 1 Ch
113; Re Spence [1949] WN 237; Re Ree’s WT [1950] Ch 204; Margulies v Margulies
(1999–2000) 2 ITELR 641.
284 McCormick v Grogan, above n. 25; Re Boyes, above n. 80; Blackwell v Blackwell,
above n. 1; Re Cooper, above n. 150; Re Keen, above n. 100; Re Bateman’s WT,
above n. 186.
285 Smith v Attersoll, above n. 90; Re Spence, above n. 283.
286 E.g. Thynn v Thynn, above n. 24; Oldham v Litchford, above n. 32.
287 E.g. Reech v Kennegal, above n. 30; Chamberlain v Agar, above n. 44.
288 E.g. Re Williams, above n. 116; Re Gardner (No 1), above n. 117.
289 E.g. Kasperbauer v Griffith, above n. 152; Davies v Revenue and Customs
Commissioners, above n. 4.
290 E.g. Re Wedgwood, above n. 283; Whitton v Russell, above n. 35.
291 E.g. Proby v Landor, above n. 50; Re Snowden, above n. 19.
292 Challinor, above n. 9 at 492–3, also brings the Mortmain cases into her argument
that the doctrine of secret trusts is now ‘divorced from its original function and
context’. As demonstrated here, this argument is based on assumptions that are
not borne out by the authorities.
341
of the WA 1837, a statute enacted by Parliament to prevent fraud, to

be used to perpetrate fraud. The case law overwhelmingly indicates
that any failure by the secret trustee to perform the secret trust
amounts to a fraud on the testator because it is taken that the testator
relied on the secret trustee’s promise to perform when determining
his testamentary dispositions. Of course, the secret beneficiaries will
also be injured by such fraud.
While the fraud theory explains why secret trusts are enforced, in
order to understand how this is done, it must be appreciated that
secret trusts are dehors the will. The testator’s declaration of trust
cannot have effect as an express trust. It does not comply with s. 9 and
thus it cannot be considered to be a testamentary disposition, if the
term ‘testamentary disposition’ is understood correctly. In fact, it is
not a valid disposition at all. It is an expression of testamentary inten-
tion. If the testator dies having relied on the secret trustee’s promise
to carry into effect this intention, however, the secret trustee’s con-
science is affected because it would, as explained above, be fraudulent
for him not to perform it. Therefore, at the moment of the testator’s
death, a constructive trust arises by operation of law, and the secret
trustee is bound to do what he promised to do. This constructive trust,
of course, arises dehors the will and cannot be termed testamentary
either. Thus, the fraud theory and the dehors theory are inexorably
linked. It is only in cases of fraud that equity is justified in imposing a
constructive trust because its imposition gives effect to the testator’s
testamentary intentions, even though those testamentary intentions
were not expressed in such a way as to satisfy s. 9 of the WA 1837.
Once this is understood, it is clear why the requirements for a valid
secret trust are intention, communication and acceptance. For the
secret trustee’s conscience to be bound, he must have accepted an
instruction sufficiently imperative for the court to hold that failure
to perform would amount to a fraud. If it is borne in mind that the
court is only justified in imposing the constructive trust because to do
so would be in accordance with the will of Parliament, it can be
appreciated why the differing communication requirement exists for
half-secret trusts. English judges have, whether rightly or wrongly,
reached the conclusion that it was not the intention of Parliament
that half-secret trusts communicated after the will’s execution should
be enforceable in any circumstances. The differing communication
requirement is the inevitable and proper consequence of this
conclusion.
This conclusion is based on thorough analysis of all relevant
English case law. The consistency in the judgments, especially from
the higher courts, is such that any academic arguments or legal devel-
opments based on considerations other than the long-established
doctrine of secret trusts revealed here ought to be reassessed.
342
Appendix
Cases where the Prevention of Fraud was held to be the

Underlying Reason for the Enforcement of Secret Trusts
I. Fully Secret Trusts Cases

Thynn v Thynn (1684) 1 Vern 296
Devenish v Baines (1689) Prec Ch 3
Oldham v Litchford (1705) 2 Freem Chy 284
Sellack v Harris (1708) 2 Eq. CA. ABR 46
Whitton v Russell (1739) 1 Ves 124
Drakeford v Wilks (1747) 3 Atk 539
Reech v Kennegal (1748) 1 Ves Sen 123
Boson v Statham (1760) 1 Cox 16
Barrow v Greenough (1796) 3 Ves Jun 152
Muckleston v Brown (1801) Ves Jun 53
Stickland v Aldridge (1804) 9 Ves Jun 517
Chamberlain v Agar (1813) 2 V & B 257
Russell v Jackson (1852) 10 Hare 204
Wallgrave v Tebbs (1855) 2 K & J 313
Lomax v Ripley (1855) 3 Sm & G 48
Tee v Ferris (1856) 2 K & J 357
Moss v Cooper (1861) 1 J & H 352
Sweeting v Sweeting (1863) 33 LJ Ch 211
Cullen v Attorney General for Ireland (1868) LR 3 Ch App 362
Jones v Badley (1868) LR 3 Ch App 362
McCormick v Grogan (1869) LR 4 HL 82
Norris v Frazer (1873) LR 15 Eq 318
Rowbotham v Dunnett (1878) 8 Ch D 430
Re Boyes (1884) 26 Ch D 531
Re Spencer’s Will (1887) 57 LT 519
Re Stead [1900] 1 Ch 237
Re Pitt-Rivers [1902] 1 Ch 403
Re Maddock [1902] 2 Ch 220
Re Gardner (No 1) [1920] 2 Ch 523
Re Falkiner [1924] 1 Ch 88
Re Williams [1933] Ch 244
Ottaway v Norman [1972] Ch 698
II. Half-secret Trusts Cases

Podmore v Gunning (1836) 7 Sim 643
Briggs v Penny (1851) 3 M & G 546
Re Fleetwood (1880) 15 Ch D 594
Blackwell v Blackwell [1929] AC 318
Re Keen [1937] Ch 236
Re Cooper [1939] Ch 811
Re Young [1951] Ch 344
343
III. Cases Where the Doctrine of Secret Trusts was Being

Discussed Generally
Tharp v Tharp [1916] 1 Ch 142
Healey v Brown [2002] WTLR 849
De Bruyne v De Bruyne [2010] EWCA Civ 519
344

The Secret Is Out There

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Secret Is Out There

Uploaded by

Copyright:

Available Formats

The Secret Is Out There:

Searching for the Legal

Abstract: The doctrine of secret trusts, which appears to operate in

Keywords: constructive trusts, equitable fraud, equity, express trusts,

* Senior Lecturer, University of Wolverhampton; e-mail: gregory.allan@wlv.ac.uk.

Common Law World Review 40 (2011) 311–344 311

raised in the courts, most recently in De Bruyne v De Bruyne3 in the

trusts are difficult to justify, especially half-secret trusts, as a half-

II. The Fraud Theory

the prevention of fraud is referred to as being the underlying justifi-

ii. What is the Nature of the Fraud?

(a) The Eighteenth-century Judgments29

25 McCormick v Grogan (1869) LR 4 HL 82 and Cullen v Attorney General for Ireland

‘a fraud upon the testator and the legatee’.33 It is therefore unsurpris-

(b) The Nineteenth-century Judgments

‘fraudulent inducement’56 by the secret trustee. Lord Westbury used

mere reinforcement of the idea that it would be dishonest for the

bed’.78 Tellingly, Bacon VC cited Lord Westbury’s speech in

(ii) Half-secret Trust Cases

In Smith v Attersoll,90 a half-secret trust was enforced, although

90 (1826) 1 Russ 266.

there is only one case, Re Baillie,102 in which the court declined to

He also described half-secret trusts as being created ‘by communica-

102 (1886) 2 TLR 660.

(c) The Twentieth-century Judgments

of secret trusts to an analogous situation,123 explaining that ‘a sub-

In Re Snowden,135 Megarry VC based his judgment on the orthodox

(ii) Half-secret Trust Cases

135 Above n. 19.

explained the fraud in similar terms, describing a secret trust as ‘aris-

(d) The Twenty-first-century Judgments

146 Above n. 1 at 341.

situation that was held to be analogous.158 Patten LJ explained with

iii. Is the ‘Instrument of Fraud’ Principle Unconstitutional?

163 See Sheridan, above n. 11; Matthews, above n. 11.

This again demonstrates that the intention of Parliament was para-

In fact, as Lord Nottingham referred to the enforcement of secret

175 Above n. 42.

iv. Do the Differing Communication Requirements for Fully and

he considered that Parliament did not intend half-secret trusts com-

III. The Dehors the Will Theory

dispositions, in which case they are exceptions to the WA 1837,199

i. Is the Testator’s Declaration of Trust a Testamentary

199 See especially Critchley, above n. 9.

206 Ibid. at 263.

The next question is whether s. 9 applies to the testator’s declara-

221 Above n. 47.

compliance with the Wills Act. It is plainly inaccurate to describe as an

ii. Is the Secret Beneficiary’s Interest Acquired by Testamentary

232 Johnson v Ball, above n. 96 at 91 per Parker VC.

authorities are entirely consistent with the narrow definition of testa-

iii. How is the Secret Beneficiary’s Interest Acquired?

imposing upon her an express trust’.253 Lord Westbury used language

Another illuminating judgment is that of Kay J in Re Boyes,258 who

s. 53(1)(b) LPA 1925 is not relevant to questions regarding their en-

IV. Do Social Changes since the Doctrine was Developed

of the WA 1837, a statute enacted by Parliament to prevent fraud, to

Cases where the Prevention of Fraud was held to be the

I. Fully Secret Trusts Cases

II. Half-secret Trusts Cases

III. Cases Where the Doctrine of Secret Trusts was Being

You might also like