[CHANCERY DIVISION]

In re MONTAGU'S SETTLEMENT TRUSTS

DUKE OF MANCHESTER v. NATIONAL WESTMINSTER

RANK LTD.

[1979 M. No. 2412]

1985 March 29

Trusts - Constructive trust - Beneficiary - Trustees releasing property to beneficiary absolutely in breach of trust - Beneficiary having notice of trust - Whether beneficiary having knowledge of terms of trust - Whether beneficiary constructive trustee of property

By clause 14 of a family re-settlement made in December 1923, the future tenth Duke of Manchester assigned to trustees "all articles of furniture plate pictures and other chattels" to which he was absolutely entitled in remainder after the death of the ninth Duke, and by paragraph (B) the trustees, after the death of the ninth Duke or, if and so far as practicable and convenient, during his lifetime, were to select and make an inventory or inventories of such chattels as the trustees in their absolute discretion might consider suitable for inclusion in the settlement, and were to hold the residue in trust for the future tenth Duke absolutely. The ninth Duke died in February 1947 but the trustees made no inventory of the chattels and by the end of 1948 had released them to the tenth Duke. The Duke's solicitor, who knew of the settlement and at an earlier stage had known of the effect of clause 14, informed the Duke in a letter dated 15 November 1948 that he was free to sell the items released. The Duke disposed of a number of items during his lifetime. After his death in 1977, the plaintiff, the eleventh Duke, brought an action for, inter alia, inquiries as to the proceeds of items sold and alleging, inter alia, that the trustees had been in breach of trust in failing to make any selection or inventory of the chattels and in releasing chattels to the tenth Duke. He further alleged that in taking possession of, and in dealing with the items delivered to him, knowing they were subject the trusts of the settlement, the tenth Duke held them as a constructive trustee:-

Held, that the trustees, along with the Duke's solicitor, had sufficiently conveyed to him in the letter of 15 November 1948 that they were treating clause 14 (B) of the settlement as allowing them to release the chattels to him as being his absolute property, but there was no reason why the solicitor's accurate knowledge of the meaning of clause 14 (B) at some earlier date should be imputed to the Duke so as to affect his conscience or result in his failure to make inquiries imposing a constructive trust on him; that even supposing the Duke did once understand the true meaning of clause 14 (B) of the settlement, there was nothing to suggest that he remembered its terms so as to know, at the time when he received the chattels, that they were trust property; that, accordingly, although the trustees had been in breach of their fiduciary duty in transferring the assigned chattels to the Duke in 1948, he had not received

them as a constructive trustee (post, pp. 282F-G, 284E, G - 285A,286A-B).

Carl Zeiss Stiftung v. Herbert Smith & Co. (No. 2) [1969] 2 Ch. 276, C.A. applied.

Baden, Delvaux and Lecuit v. SociŽtŽ General pour Favoriser le DŽveloppement du Commerce et de l'Industrie en France S.A. [1983] B.C.L.C. 325 considered.

Per curiam. It is at least doubtful whether there is a general doctrine of "imputed knowledge" that corresponds to imputed notice (post, p. 285G-H).

The following cases are referred to in the judgment:

Amalgamated Investment & Property Co. Ltd. v. Texas Commerce International Bank Ltd. [1982] Q.B. 84; [1981] 2 W.L.R. 554; [1981] 1 All E.R. 923; [1982] Q.B. 84; [1981] 3 W.L.R. 565; [1981] 3 All E.R. 577, C.A.

Baden, Delvaux and Lecuit v. SociŽtŽ General pour Favoriser le DŽveloppement du Commerce et de l'industrie en France S.A. [1983] B.C.L.C. 325

Barnes v. Addy (1874) 9 Ch.App. 244

Belmont Finance Corporation Ltd. v. Williams Furniture Ltd. [1979] Ch. 250; [1978] 3 W.L.R. 712; [1979] 1 All E.R. 118, C.A.

Belmont Finance Corporation Ltd. v. Williams Furniture Ltd. (No. 2) [1980] 1 All E.R. 393, C.A.

Blundell, In re (1888) 40 Ch.D. 370

Boursot v. Savage (1866) L.R. 2 Eq. 134

Carl Zeiss Stiftung v. Herbert Smith & Co. (No. 2) [1969] 2 Ch. 276; [1969] 2 W.L.R. 427; [1969] 2 All E.R. 367, C.A.

Competitive Insurance Co. Ltd. v. Davies Investments Ltd. [1975] 1 W.L.R. 1240; [1975] 3 All E.R. 254

Consul Development Pty. Ltd. v. D.P.C. Estates Pty. Ltd. (1975) 132 C.L.R. 373

Dearle v. Hall (1828) 3 Russ. 1

Diplock, In re [1948] Ch. 465; [1948] 2 All E.R. 318, C.A.; affirmed sub nom. Ministry of Health v. Simpson [1951] A.C. 251; [1950] 2 All E.R. 1137, H.L.(E.)

Habib Bank Ltd. v. Habib Bank A.G. Zurich [1981] 1 W.L.R. 1265; [1981] 2 All E.R. 650, C.A.

International Sales and Agencies Ltd. v. Marcus [1982] 3 All E.R. 551

Ipswich Permanent Money Club Ltd. v. Arthy [1920] 2 Ch. 257

Karak Rubber Co. Ltd. v. Burden (No. 2) [1972] 1 W.L.R. 602; [1972] 1 All E.R. 1210

Selangor United Rubber Estates Ltd. v. Cradock (No. 3) [1968] 1 W.L.R. 1555; [1968] 2 All E.R. 1073

Taylors Fashions Ltd. v. Liverpool Victoria Trustees Co. Ltd. (Note) [1982] Q.B. 133; [1981] 2 W.L.R. 576; [1981] 1 All E.R. 897

Vane v. Vane (1873) 8 Ch.App. 383

Williams v. Williams (1881) 17 Ch.D. 437

Willmott v. Barber (1880) 15 Ch.D. 96

The following additional cases were, among others, cited in argument, either at the hearing or at previous hearings:

Allen-Meyrick's Will Trusts, In re [1966] 1 W.L.R. 499; [1966] 1 All E.R. 740

Baden's Deed Trusts, In re [1967] 1 W.L.R. 1457; [1967] 3 All E.R. 159; [1969] 2 Ch. 388; [1969] 3 W.L.R. 12; [1969] 1 All E.R. 1016, C.A.; [1971] A.C. 424; [1970] 2 W.L.R. 1110; [1970] 2 All E.R. 228, H.L.(E.)

Costabadie v. Costabadie (1847) 6 Hare 410

Leek, decd., In re [1967] Ch. 1061; [1967] 3 W.L.R. 576; [1967] 2 All E.R. 1160; [1969] 1 Ch. 563; [1968] 2 W.L.R. 1385; [1968] 1 All E.R. 793, C.A.

Locker's Settlement In re [1977] 1 W.L.R. 1323; [1978] 1 All E.R. 216

Nelson v. Larholt [1948] 1 K.B. 339; [1947] 2 All E.R. 751

Swan, In re [1915] 1 Ch. 829

Tempest v. Lord Camoys (1882) 21 Ch.D. 571, C.A.

Turner v. Turner [1984] Ch. 100; [1983] 3 W.L.R. 896; [1983] 2 All E.R. 745

ACTION

By clause 14 of a family re-settlement dated 20 December 1923, the future tenth Duke of Manchester, then known as Viscount Mandeville, assigned to trustees certain chattels to which he was entitled in remainder expectant on the death of his father, the ninth Duke, and he directed by paragraph (B) that they should be held

"Upon trust after the death of the present Duke or (if and so far as may be found practicable and convenient) during his lifetime to select and make an inventory or inventories of such of the chattels hereby assigned as the trustees in their absolute discretion may consider suitable for inclusion in the settlement hereby made (which selected chattels are hereinafter called 'the selected chattels') and to hold the residue (if any) of the said assigned chattels in trust for Viscount Mandeville absolutely."

Paragraph 14 (C) then provided that the selected chattels should devolve with the settled hereditaments.

After the death of his father in February 1947, the trustees during 1948 and 1949 delivered to the tenth Duke certain of the chattels. Many of these were intended by him to be sold and were in fact sold, and others were retained by him unsold. No selection of chattels or any inventory had been made by the trustees in accordance with clause 14(B) of the 1923 settlement. In November 1977 the tenth Duke died and all the chattels subject to the trust came into the possession or under the control of his widow, the Dowager Duchess of Manchester.

By writ and statement of claim dated 29 June 1979, as amended, reissued on amendment and re-amended, the plaintiff, the eleventh Duke, alleged against the remaining trustees of the 1923 settlement, inter alia, that in breach of trust they had failed to make any selection or inventory of the assigned chattels as required by the tenth Duke under clause 14(B) of the settlement; further, that they were in breach of trust in delivering certain of the chattels to the tenth Duke knowing that he intended to sell them and did sell them without accounting to the

trustees for any part of the proceeds; further, that in taking possession of and dealing with the chattels delivered to him, knowing they were subject to the trusts of the 1923 settlement, the tenth Duke held them as a constructive trustee, and in causing them to be sold he was in breach of trust; further that the Dowager Duchess now held the chattels as a constructive trustee, alternatively, as a volunteer. The plaintiff claimed a declaration that, upon the true construction of clause 14 of the settlement, it was the duty of the trustees to select and make an inventory of such of the heirlooms, i.e. the chattels, as they in their discretion considered suitable for inclusion in the settlement and that in making such selection, they were bound to treat as suitable all articles other than those which, by virtue of their defective condition, were of no or trivial value. The plaintiff claimed an alternative declaration that the trustees had a duty to make an inventory within a reasonable time after the determination of the life interest of the ninth Duke, and that by their failure so to do, all the heirlooms (other than those of no or trivial value) were to be deemed to have been selected. Inquiries were asked for relating, inter alia, to the identity of the heirlooms at the date of the death of the ninth Duke and the identity of articles which might have been sold; and if sold, an inquiry to determine to whom the proceeds of sale were paid, which property now represented the same and who had possession or control of the same.

Of the eight defendants to the writ, only the eighth, the Dowager Duchess of Manchester, took part in the proceedings. She was sued as executor of the will of the tenth Duke in relation to his assets not situated in the United States of America or Kenya. An order was sought that, as executor, she should deliver up all articles in her possession or under her control which on inquiry might be found to be subject to the trusts of clause 14. A stay of proceedings was ordered against the first defendant, National Westminster Bank Ltd., and the other defendants except the fourth and fifth, the remaining trustees of the 1923 settlement, but they took no part in the proceedings; they considered they would be sufficiently covered by the contentions advanced on behalf of the eighth defendant.

On 21 December 1983 after a ten-day hearing, which started on 4 October 1983, Sir Robert Megarry V.-C., in a reserved judgment, said that clause 14 of the 1923 settlement had made elaborate provision for the future of the chattels which depended on the trustees carrying out their duty of selection - qualified by the words "practicable and convenient" - while the ninth Duke lived and thereafter unqualified. They had an absolute discretion in the process of selection but no discretionary power to refrain from making a selection. From the wording of clause 14(B) it was clear that a fiduciary duty was imposed: they were not merely given a power. The duty was to make a selection of those chattels considered suitable for inclusion in the settlement, with an inventory or inventories of those chattels; only when that duty had been performed could it be said what chattels were to be held on trust to devolve with the land.

The Vice-Chancellor said that on, if not before, the ninth Duke's death on 9 February 1947 the trustees were required by clause 14 of the

settlement to perform their task of selecting the chattels and having an inventory made of them. On the evidence, that had never been done; instead there was a process of treating all the chattels as being the absolute property of the tenth Duke, he being left to take what he wanted. It had been said that there had been a decision by the trustees that none of the chattels was "suitable for inclusion" in the settlement. The necessity of assuming that decision was a stumbling-block, for a number of inventories had been made from time to time. The conclusion therefore was that the trustees had never carried out the trust imposed on them by clause 14(B) of the settlement. The effect was that no "selected chattels" had come into existence and so no selected chattels which could be held by them on trust under clause 14(C); and there was no provision for any destination for the chattels or residuary gift in default of failure to operate clause 14(B). The difficulties of making a selection at the present stage were very great and the best course now was to direct an inquiry to determine ownership of the chattels with directions as to the basis on which the inquiry would be held.

In the course of his judgment, Sir Robert Megarry V.-C. said (see pp. 28-37 of the transcript):

"I turn now to the liability of the tenth Duke, and, now, his executor, the eighth defendant [the Dowager Duchess of Manchester], in respect of the chattels received by him and treated as being his absolute property, or disposed of by him, whether by sale or otherwise, and any proceeds of sale treated by him as being his. Until the necessary inquiry has been held it cannot, of course, be known which chattels are to be treated as being his absolutely, and so free from any claim, and which are to be treated as selected chattels in respect of which a claim may lie. What I must consider is the position of the tenth Duke (and, through him, the defendant) in respect of whatever chattels emerge as being in the latter category.

The two basic questions are (1) whether the tenth Duke received any or all of the settled chattels with sufficient knowledge that the trustees were in breach of duty in allowing him to treat them as his own absolute property so as to make him a constructive trustee of them; and (2) if not, whether the tenth Duke held them on trust by reason of being tenant for life for the purposes of the Settled Land Act 1925 under the 1923 settlement, or beneficiary for life of those chattels. I shall consider these questions in turn.

The starting point, I think, is that it is plain from the evidence that the tenth Duke at all material times had full notice of the terms of the 1923 resettlement. He had a copy of it, and his solicitor, Mr. Lickfold, showed in his correspondence that he knew the effect of clause 14. From this it follows that the tenth Duke had notice, both actual and imputed, of the terms of clause 14(B). Plainly he knew that the agreement of the trustees was needed before any of the chattels would become his, but I am far from sure that he in fact appreciated just how the sub-clause was to work, or that before any of the chattels were to become his absolute property the trustees were required to select the chattels which they considered suitable for inclusion in the 1923 resettlement, and make an inventory of them. As a matter of construction, it may be that when the

trustees had completed the process of selection any chattels not then selected would become the tenth Duke's absolute property, even before any inventory was prepared, and even if no inventory was ever prepared: the making of the inventory, it might be said, was a provision for providing indisputable evidence of the result of the process of selection, but was not a condition precedent to chattels becoming the tenth Duke's absolute property. This point was not argued, and I shall not pursue it. For present purposes I need only say that the mere absence of any such inventory does not seem to me necessarily to establish, by itself, that the process laid down by clause 14(B) had not been carried out so as to make any chattels the tenth Duke's absolute property, and so if he knew that no such inventory had been made, that does not establish his knowledge that none of the chattels had ever properly been made his. At the same time, the absence of any such inventory, in contrast with all the other inventories that had been made over the years, is at least a matter to put on inquiry anyone who knew of it and understood what clause 14(B) meant, and that must include Mr. Lickfold, the tenth Duke's solicitor, even if it did not include the tenth Duke, who for much of the time was in Kenya.

The tenth Duke was not, of course a lawyer himself, and it seems reasonable to infer, in the absence of evidence to the contrary, that on legal matters he simply accepted what the lawyers told him. In a letter dated 15 November 1948 to the tenth Duke by Mr. Lickfold, a paragraph under the heading 'Heirlooms' runs as follows:

'I had a long conversation with Nicholl on Thursday last, and the trustees have agreed that the heirlooms should be released, except the pictures. Under the resettlement executed by you on 20 December 1923 there is a clause by which the trustees can in their discretion release a large quantity of heirlooms and make a new list of such articles as are to remain as heirlooms. The amount obtained for the sale of any articles will be your personal property and the proceeds of sale will not have to be considered as capital trust money.'

This letter seems to me to indicate clearly enough that Mr. Lickfold, despite various other correct references to the terms of clause 14(B), had by this time become infected by the idea that the sub-clause worked by giving the trustees a power simply to release chattels from the 1923 resettlement to the tenth Duke, and that they then merely had to list those that were left. In a sense, of course, the trustees had such a power, in conjunction with the tenth Duke, by means of an overriding power in clause 3(A). But that required an appointment by deed, and it could not be suggested that what happened was an exercise of that overriding power. Yet as I have said, possibly that overriding power had in some way become confused with clause 14(B), or affected recollection of what the trustees and the tenth Duke could do in concert; and the general attitude at the time seems to have been one of the trustees leaving it to the tenth Duke to say what chattels he wanted, and then concurring, expressly or tacitly, with his taking or disposing of those that he had decided upon.

I do not for a moment say that any of this is clearly established; but on what evidence is available this seems to me to be a reasonable inference to draw, considering not merely those parts of the evidence that I have mentioned, but the evidence as a whole. It seems to me (1) that the tenth duke, although he knew of clause 14(B) of the 1923 settlement, did not himself know that the release of the settled chattels to him amounted to a breach by the trustees of the terms of the sub-clause; and (2) that his solicitor, Mr. Lickfold, though knowing the terms of clause 14(B), had at least for the time being come to regard it as operating in an inverted form, so that all that was required was for the trustees to decide to release the chattels to the tenth Duke, and that the trustees had done this. How this happened I cannot say on the evidence that has survived; but I cannot reach the conclusion that Mr. Lickfold, Col. Nicholl or the trustees were acting in conscious disregard of the terms of clause 14(B). Sir Arthur Collins, who acted as solicitor to the plaintiff, the eleventh Duke, from 1950 to about 1965, said that Col. Nicholl was a careful solicitor who was not likely to be overawed by an earl or marquis, but that he dealt with a very large number of trusts, and he expected that he would be very busy. Plainly there was a muddle, but however careless it was, I think that it was an honest muddle. Mr. Green, a trustee who knew Mr. Lickfold pretty well, said that he was quite sure that Mr. Lickfold would not have let the tenth Duke deal with any property unless he was quite satisfied that he was entitled to it.

It is against that background of fact that I have to consider whether the tenth Duke became a trustee of any of the chattels which the inquiry determines should have been settled chattels. Mr. Taylor for the plaintiff contended that In re Swan [1915] 1 Ch. 829 showed that the tenth Duke took as an express trustee. That was an unusual case, but I do not think that it supports Mr. Taylor's submission of express trust. I think that Mr. Chadwick was right when he classified it as being a case of a constructive trust, as was argued at p. 830. I do not think that pp. 834, 835 are inconsistent with this.

The question, then, is whether the tenth Duke was a constructive trustee. Various formulations of constructive trust are to be found in the cases and in the books. I can at the outset discard one category, namely, that of a person, who has not himself received the trust property, knowingly assisting the trustees in some dishonest breach of trust: I put it shortly. That plainly is not applicable to this case. Accordingly, I turn to the other main head of constructive trust, a head which (if I may say so) is usually put forward in two limbs. The first is that of a person who receives trust property with notice that it is trust property which is being transferred in breach of trust. The second limb is that of a person who, not being a bona fide purchaser for value without notice of the trust, deals with the property in a manner inconsistent with the trust after he has received notice of it. I think that for present purposes these statements are sufficiently accurate and supported by authority: see Underhill's Law of Trusts and Trustees, 13th ed. (1979), pp. 331, 332, and Snell's Principles of Equity, 28th ed. (1982), p. 194, and the authorities there cited.

Now there is modern authority for the proposition that if a client pays money to his solicitor, that solicitor is not liable for it as a constructive trustee merely because he has express notice that there is a disputed claim to that money by another: Carl Zeiss Stiftung v. Herbert Smith & Co. (No. 2) [1969] 2 Ch. 276. The case was much discussed during argument: and one point which I think is of some importance is considered by Sachs L.J., at p. 296, namely, whether in stating one of the essential requirements of a constructive trust it is right to refer to 'notice,' with its threefold subdivision into actual, constructive or imputed, or whether 'knowledge' is the right word; for the two are by no means necessarily the same.

I feel considerable hesitation in saying that the rules about what constitutes 'notice' for the ancient doctrine of purchaser without notice apply, without more, in determining whether a person holds as a constructive trustee. I doubt whether one can state the rules about what creates a constructive trust in terms of the doctrine of purchaser without notice. Suppose, for instance, a trustee transfers trust property to a person who takes it in all innocence, believing that he is entitled to it as a beneficiary: he may, perhaps, believe that he is a beneficiary within a testamentary discretionary trust, whereas in fact he is not. He cannot claim to be a purchaser for value without notice, for he is a mere volunteer. If when the truth emerges he still has the property he must restore it, whereas if he no longer has either the property or its traceable proceeds, he is under no liability, unless he has become a constructive trustee. If it then appears that his solicitor who has been acting for him in respect of the will ought to have discovered that his client was not entitled to receive anything, I would hesitate to say that the solicitor's notice must be imputed to his client, or that the client was a constructive trustee ab initio and is liable for what he has disposed of in all innocence, believing that it was truly his. In other words, though the doctrine of notice is applicable in determining whether or not a purchaser should take free of an undiscovered equity, and whether he has discharged the duties of investigation which are needed to establish him as a purchaser without notice, I have considerable doubts about whether it should be applied without at least some modification to cases where the issue is whether liability as a constructive trustee is to be imposed, cases in which the investigations apt for a purchaser are normally not required. After all, a purchase is an occasion for suspicion and investigation as between parties who are at arm's length, and for this the doctrine of notice is entirely apt, whereas the receipt of bounty under a trust is entirely different. In short, for the creation of a constructive trust, I would ask whether the right word is not 'notice' but 'knowledge,' or, to borrow from Sachs L.J., 'cognizance,' a word which he regarded as being neutral (see the Carl Zeiss case at p. 296) but which has a flavour of awareness which I would regard as making it lean towards 'knowledge' rather than 'notice.'

The relevance of this to the present case is obvious. I do not think that the tenth Duke had any knowledge at any material time that the chattels that he was receiving or dealing with were chattels that were still subject to any trust. I think that he believed that they had been

lawfully and properly released to him by the trustees. If liability as a constructive trustee depended on his knowledge, then he was not liable as a constructive trustee, and his estate is not liable for any chattels that have been disposed of, as distinct from any traceable proceeds of them. Even if he was not a constructive trustee and was a mere volunteer, his estate is liable to yield up any chattels that remain, or the traceable proceeds of any that have gone, subject always to the chattels emerging from the inquiry as 'selected chattels.' But unless he was a constructive trustee, there appears to be no liability if the chattels have gone and there are no traceable proceeds.

If on the other hand the test is one of notice rather than knowledge, any actual or constructive notice that Mr. Lickfold had in the transaction will prima facie be imputed to the tenth Duke; and this, if sufficient, would make him a constructive trustee, and so subject him to liability as such, even if no chattels remain and there are no traceable proceeds. Plainly there were times when Mr. Lickfold knew the meaning and effect of clause 14(B) accurately, and yet at other times he seems to have remembered only the transmogrified version that he set out in his letter of 15 November 1948.

What, then, is the result of this? A solicitor knows that the right meaning of a clause is A, but he then writes to his client and through some error tells him that the meaning is B, which is quite different. Is the court then to say that although the solicitor has told his client that the meaning is B and all is well, nevertheless the client is to have imputed to him notice of meaning A, which shows that all is far from well, with the result that liability as a constructive trustee is to be imposed on him? Is it to be said that this shows the 'want of probity' that in the Carl Zeiss case, at p. 301, Edmund Davies L.J. required? If the answer is Yes, I can only say that I find it surprising, and at some remove from my concept of equity. The question is not merely whether a person is to take a property subject to an equity but whether the full burden of trusteeship is to be imposed on him by construction of equity. Nor am I clear how far the true construction of a document is something that falls within the doctrine of notice.

I do not think that much help is to be obtained from most of the authorities put before me. Many of them deal with cases in the category that at the outset I discarded, that of knowingly assisting trustees in a dishonest breach. In such cases, mere constructive notice of the dishonesty, as distinct from actual knowledge or a reckless failure to make the inquiries an honest or reasonable man would make, is not enough: see, e.g., Belmont Finance Corporation Ltd. v. Williams Furniture Ltd. [1979] Ch. 250, 267, 275, where, however, the point was not decided. If in this category constructive notice of the fact which will impose the trusts is not enough, it may be asked why constructive notice of that fact should suffice for the other category.

It should also be remembered that the doctrines of purchaser without notice and constructive trusts are concerned with matters which differ in important respects. The former is concerned with the question whether a person takes property subject to or free from some equity. The latter is concerned with whether or not a person is to have imposed

upon him the personal burdens and obligations of trusteeship. I do not see why one of the touchstones for determining the burdens on property should be the same as that for deciding whether to impose a personal obligation on a man. The cold calculus of constructive and imputed notice does not seem to me to be an appropriate instrument for deciding whether a man's conscience is sufficiently affected for it to be right to bind him by the obligations of a constructive trustee.

I am conscious that this part of my judgment does not mirror the way in which the argument proceeded. It will be seen that the provisional view towards which I have been tending is that the tenth Duke never became a constructive trustee of any chattels. But I do not think it would be right for me to decide the point without hearing further submissions on either side that the parties may wish to make in the light of this judgment (if light be the right word). Accordingly, I propose to adjourn this part of the case for further argument which will take into account the views that I have provisionally expressed; for I am not satisfied that I have got to the bottom of the matter. Mr. Taylor may be able to disabuse my mind, and Mr. Chadwick may be able to support my tentative conclusion, either on the grounds that I have suggested or upon other grounds. I am well aware that there are dicta which may be said to point against that conclusion, as distinct from decisions; yet where the distinction between knowledge and notice is not in issue, a judge may speak indifferently of either. I think that the cases deserve further consideration against the facts of this case in relation to my provisional conclusion. The stage at which this further argument should take place is for consideration, though plainly this should not be before counsel have been able to consider a transcript of this judgment."

The court made a declaration that the trustees of the 1923 settlement ought, within a reasonable period after the death of the ninth Duke, to have selected out of the chattels assigned to them by the tenth Duke under clause 14 of the settlement, such chattels as they in their absolute discretion considered suitable for inclusion in the settlement pursuant to clause 14(B); and that neither those trustees nor their successors had ever made any such selection. The court ordered that, for the purpose of an inquiry to be directed and of further proceedings in the action relating to the identity of the assigned chattels at the date of the death of the ninth Duke, the Dowager Duchess by her solicitors should deliver to the plaintiff's solicitors a list of such of the assigned chattels as, in her contention, had not, before the death of the ninth Duke, been lost, disposed of or dealt with so as to cease to be comprised in the settlement. The court further ordered that on delivery of the said list the plaintiff by his solicitors should deliver to her solicitors a copy of the said list showing, inter alia, items which the plaintiff contended should be added or deleted.

Among other declarations and orders for inquiries, it was ordered that the question whether the tenth Duke became a constructive trustee of the settled chattels should be adjourned for further consideration. The action was adjourned generally, with liberty to the parties to apply.

At the hearing beginning on 18 June 1984, counsel adduced further argument on the question whether the tenth Duke became a constructive

trustee of the settled chattels transferred to him by the remaining trustees of the 1923 settlement.

P. W. E. Taylor Q.C. and Peter Rawson for the plaintiff, the eleventh Duke of Manchester. In 1948 and 1949 when the tenth Duke received from the then trustees certain of the settled chattels which he had assigned to them as part of the 1923 settlement, he received those chattels with sufficient knowledge that the trustees were in breach of duty in allowing him to treat them as his own absolute property, with the result that he became an express trustee of them; alternatively, he became a constructive trustee.

John Chadwick Q.C. and Lynton A. Tucker for the defendant, the Dowager Duchess of Manchester. The circumstances involving the tenth Duke's receipt of the settled chattels were not such as to constitute him either an express trustee or a constructive trustee of them.

The submissions addressed to the court on the issue are fully considered and dealt with in the judgment, particularly at pp. 270D-E, 276D-F, 279C-D, G - 280D, 281E - 282F, 283B-D, 284F.

Additional cases were cited in argument either at the hearing or earlier hearings as follows: On the duty of the trustees as a matter of law to keep a proper record of any decisions made in respect of the chattels, Costabadie v. Costabadie (1847) 6 Hare 410. On the position of one holding chattels, whether he be an express or constructive trustee, In re Swan [1915] 1 Ch. 829.

On imputation of knowledge of a trust, Nelson v. Larholt [1948] 1 K.B. 339. The knowledge of the trust by the Duke's solicitor must be imputed to the Duke.

On the failure of the trustees to make a selection of the chattels, leading to the court's intervention, Tempest v. Lord Camoys (1882) 21 Ch.D. 571; In re Allen-Meyrick's Will Trusts [1966] 1 W.L.R. 499; In re Locker's Settlement [1977] 1 W.L.R. 1323 and Turner v. Turner [1984] Ch. 100.

On the discretionary power of trustees to make a selection: In re Leek, decd. [1967] Ch. 1061 and In re Baden's Deed Trusts [1967] 1 W.L.R. 1457; [1971] A.C. 424.

29 March 1985. SIR ROBERT MEGARRY V.-C. read the following judgment. This is a further stage in massive litigation. On 21 December 1983 I gave judgment on a number of points; but on one of them I refrained from reaching a firm conclusion so as to give the parties an opportunity of adducing further arguments. The parties took this opportunity, and for most of five days from 18 June 1984 I heard their submissions. The contentions were advanced by Mr. Taylor on behalf of the plaintiff, the eleventh Duke of Manchester, and by Mr. Chadwick, on behalf of the eighth defendant, the Dowager Duchess of Manchester. She is executor of the tenth Duke's will, and is the plaintiff's stepmother. What is in issue is the result of the receipt by the tenth Duke of a large number of settled chattels, and his disposal of them during his lifetime.

On many matters the evidence is slender or obscure, since the tenth Duke (whom I shall call simply "the Duke") received the chattels in the late 1940s and disposed of many of them then; he died in 1977 and others concerned have also died. What has to be resolved was whether the Duke held the chattels as a constructive trustee so that his estate is accountable for them or their proceeds.

This is a question that I dealt with at some length in my judgment, ante, pp. 268C - 273E. My provisional conclusion was that the Duke had not become a constructive trustee of the chattels, but that as that part of my judgment had not mirrored the way in which the argument had proceeded, I should not decide the point then but should adjourn the case for further argument on it. I do not wish to repeat in this judgment more of my previous judgment than is necessary: but although I shall treat the relevant part of that judgment as being incorporated by reference, I must say enough to make this judgment intelligible on its own.

The issue centred on clause 14(B) of the 1923 settlement. That settlement assigned a large number of chattels to the trustees of the settlement upon certain trusts. In the events which happened, the trustees had a fiduciary duty on the death of the ninth Duke to select and make an inventory of such of the chattels as they considered suitable for inclusion in the settlement, and to hold the residue of the chattels in trust for the Duke absolutely. In the event, the trustees made no selection or inventory but instead treated all the chattels as being the absolute property of the Duke. The Duke's solicitor, Mr. Lickfold, undoubtedly knew the terms of clause 14(B) and understood its effect, and the Duke had a copy of the settlement: but the time came when Mr. Lickfold, Col. Nicholl (a solicitor advising the trustees), the trustees themselves and Mr. Gilchrist, an American lawyer advising the Duke, all seemed to have treated clause 14(B) as allowing the trustees to assent to the Duke taking any of the chattels that he wished and either keeping them or else selling them and keeping the proceeds of sale. In 1949 there were two large sales of the chattels; and many chattels were shipped out to Kenya, where the Duke was living.

There is no suggestion that anyone concerned in the matter was dishonest. There was a muddle, but however careless it was, it was an honest muddle. Further, I do not think that the Duke was at any relevant time conscious of the fact that he was not entitled to receive the chattels and deal with them as beneficial owner. Of course, if clause 14(B) is singled out for attention and read carefully it could be seen by a reasonably intelligent layman not to empower the trustees simply to release chattels to the Duke, but to require them first to select chattels for inclusion in the settlement and to provide that only when they had done that would the chattels not selected become the Duke's property. But clause 14(B) was deeply embedded in a long and complex document, and in view of the advice and information that the Duke received from his solicitor I can see no reason why the Duke should be expected to attempt to construe the settlement himself. If expressed in terms of the doctrine of notice, I have held, ante, p. 268F-G, that the Duke "had

notice, both actual and imputed, of the terms of clause 14 (B)." But I have also said, ante, pp. 271H - 272A:

"I do not think that the tenth Duke had any knowledge at any material time that the chattels that he was receiving or dealing with were chattels that were still subject to any trust. I think that he believed that they had been lawfully and properly released to him by the trustees."

That brings me to the essential question for decision. The core of the question (and I put it very broadly) is what suffices to constitute a recipient of trust property a constructive trustee of it. I can leave on one side the equitable doctrine of tracing: if the recipient of trust property still has the property or its traceable proceeds in his possession, he is liable to restore it unless he is a purchaser without notice. But liability as a constructive trustee is wider, and does not depend upon the recipient still having the property or its traceable proceeds. Does it suffice if the recipient had "notice" that the property he was receiving was trust property, or must he have not merely notice of this, but knowledge, or "cognizance," as it has been put?

In my previous judgment I provisionally took the view that mere notice was not enough, and that what was required was knowledge or cognizance. In saying this, I very much had in mind what was said in the Court of Appeal in Carl Zeiss Stiftung v. Herbert Smith & Co. (No. 2) [1969] 2 Ch. 276; and I shall not repeat what I have already said about that case. It is that question which Mr. Taylor and Mr. Chadwick have now explored before me, with an ample and helpful citation of authority, most of which had not been cited previously. It was common ground that it was impossible to contend that the law to be found in the cases was clear and not in something of a muddle. Part of the difficulty arises from the fact that in cases on constructive trusts in which there is clearly knowledge the term "notice" is often convenient to use, without any distinction between notice and knowledge being intended.

At the outset, I think that I should refer to Baden, Delvaux and Lecuit v. SociŽtŽ General pour Favoriser le DŽveloppement du Commerce et de l'industrie en France S.A. [1983] B.C.L.C. 325, a case which for obvious reasons I shall call "the Baden case." That case took 105 days to hear, spread over 7 months, and the judgment of Peter Gibson J. is over 120 pages long. It was a "knowing assistance" type of constructive trust, as distinct from the "knowing receipt or dealing" type which is in issue before me. I use these terms as a convenient shorthand for two of the principal types of constructive trust. Put shortly, under the first of these heads a person becomes liable as a constructive trustee if he knowingly assists in some fraudulent design on the part of a trustee. Under the second head, a person also becomes liable as a constructive trustee if he either receives trust property with knowledge that the transfer is a breach of trust, or else deals with the property in a manner inconsistent with the trust after acquiring knowledge of the trust. It will be seen that the word "knowledge" occurs under each head; and in the Baden case, at p. 407, the judge in effect said that "knowledge" had the same meaning under each head.

I pause at that point. In the books and the authorities the word "notice" is often used in place of the word "knowledge," usually without any real explanation of its meaning. This seems to me to be a fertile source of confusion; for whatever meaning the layman may attach to those words, centuries of equity jurisprudence have attached a detailed and technical meaning to the term "notice," without doing the same for "knowledge." The classification of "notice" into actual notice, constructive notice and imputed notice has been developed in relation to the doctrine that a bona fide purchaser for value of a legal estate takes free from any equitable interests of which he has no notice. I need not discuss this classification beyond saying that I use the term "imputed notice" as meaning any actual or constructive notice that a solicitor or other agent for the purchaser acquires in the course of the transaction in question, such notice being imputed to the purchaser. Some of the cases describe any constructive notice that a purchaser himself obtains as being "imputed" to him; but I confine "imputed" to notice obtained by another which equity imputes to the purchaser.

Now until recently I do not think there had been any classification of "knowledge" which corresponded with the classification of "notice." However, in the Baden case, at p. 407, the judgment sets out five categories of knowledge, or of the circumstances in which the court may treat a person as having knowledge. Counsel in that case were substantially in agreement in treating all five types as being relevant for the purpose of a constructive trust; and the judge agreed with them: p. 415. These categories are (i) actual knowledge; (ii) wilfully shutting one's eyes to the obvious; (iii) wilfully and recklessly failing to make such inquiries as an honest and reasonable man would make; (iv) knowledge of circumstances which would indicate the facts to an honest and reasonable man; and (v) knowledge of circumstances which would put an honest and reasonable man on inquiry. If I pause there, it can be said that these categories of knowledge correspond to two categories of notice: Type (i) corresponds to actual notice, and types (ii), (iii), (iv) and (v) correspond to constructive notice. Nothing, however, is said (at least in terms) about imputed knowledge. This is important, because in the case before me Mr. Taylor strongly contended that Mr. Lickfold's knowledge must be imputed to the Duke, and that this was of the essence of his case.

It seems to me that one must be very careful about applying to constructive trusts either the accepted concepts of notice or any analogy to them. In determining whether a constructive trust has been created, the fundamental question is whether the conscience of the recipient is bound in such a way as to justify equity in imposing a trust on him. The rules concerning a purchaser without notice seem to me to provide little guidance on this and to be liable to be misleading. First, they are irrelevant unless there is a purchase. A volunteer is bound by an equitable interest even if he has no notice of it; but in many cases of alleged constructive trusts the disposition has been voluntary and not for value, and yet notice or knowledge is plainly relevant. Second, although a purchaser normally employs solicitors, and so questions of imputed notice may arise, it is unusual for a volunteer to employ solicitors when

about to receive bounty. Even if he does, he is unlikely to employ them in order to investigate the right of the donor to make the gift or of the trustees or personal representatives to make the distribution; and until this case came before me I had never heard it suggested that a volunteer would be fixed with imputed notice of all that his solicitors would have discovered had he employed solicitors and had instructed them to investigate his right to receive the property.

Third, there seems to me to be a fundamental difference between the questions that arise in respect of the doctrine of purchaser without notice and constructive trusts. As I said in my previous judgment, ante, pp. 272H - 273B:

"The former is concerned with the question whether a person takes property subject to or free from some equity. The latter is concerned with whether or not a person is to have imposed upon him the personal burdens and obligations of trusteeship. I do not see why one of the touchstones for determining the burdens on property should be the same as that for deciding whether to impose a personal obligation on a man. The cold calculus of constructive and imputed notice does not seem to me to be an appropriate instrument for deciding whether a man's conscience is sufficiently affected for it to be right to bind him by the obligations of a constructive trustee."

I can see no reason to resile from that statement, save that to meet possible susceptibilities I would alter "man" to "person." I would only add that there is more to being made a trustee than merely taking property subject to an equity.

There is a further consideration. There is today something of a tendency in equity to put less emphasis on detailed rules that have emerged from the cases and more weight on the underlying principles that engendered those rules, treating the rules less as rules requiring complete compliance, and more as guidelines to assist the court in applying the principles. A good illustration of this approach is to be found in the judgment of Oliver J. in Taylors Fashions Ltd. v. Liverpool Victoria Trustees Co. Ltd. (Note) [1982] Q.B. 133, 145-155. This view was adopted by Robert Goff J. in Amalgamated Investment & Property Co. Ltd. v. Texas Commerce International Bank Ltd. [1982] Q.B. 84, 104, 105, and it was, I think, accepted, though not cited, by the Court of Appeal in the latter case: see at pp. 116-132. Certainly it was approved in terms by the Court of Appeal in Habib Bank Ltd. v. Habib Bank A.G. Zurich [1981] 1 W.L.R. 1265, 1285, 1287. The Taylors Fashions case [1982] Q.B. 133 concerned equitable estoppel and the five probanda to be found in the judgment of Fry J. in Willmott v. Barber (1880) 15 Ch.D. 96, 105; and on the facts of the case before him Oliver J. in the Taylors Fashions case concluded that the question was not whether each of those probanda had been satisfied but whether it would be unconscionable for the defendants to take advantage of the mistake there in question. Accordingly, although I readily approach the five categories of knowledge set out in the Baden case [1983] B.C.L.C. 325 as useful guides, I regard them primarily as aids in determining whether or not the Duke's conscience was affected in such a way as to

require him to hold any or all of the chattels that he received on a constructive trust.

There is one further general consideration that I should mention, and that is that "the court should not be astute to impute knowledge where no actual knowledge exists": see the Baden case at p. 415, perPeter Gibson J. This approach goes back at least as far as Barnes v. Addy (1874) 9 Ch.App. 244, 251, 252. The view of James L.J., at p. 256, was that the court had in some cases

"gone to the very verge of justice in making good to cestuis que trust the consequences of the breaches of trust of their trustees at the expense of persons perfectly honest, but who have been, in some more or less degree, injudicious."

Of the five categories of knowledge set out in the Baden case [1983] B.C.L.C. 325, Mr. Chadwick, as well as Mr. Taylor, accepted the first three. What was in issue was nos. (iv) and (v), namely, knowledge of circumstances which "would indicate the facts to an honest and reasonable man" or "would put an honest and reasonable man on inquiry." On the view that I take of the present case I do not think that it really matters whether or not categories (iv) and (v) are included, but as the matter has been argued at length, and further questions on it may arise, I think I should say something about it.

First, as I have already indicated, I think that one has to be careful to distinguish the notice that is relevant in the doctrine of purchaser without notice from the knowledge that suffices for the imposition of a constructive trust. This is shown by a short passage in the long judgment of the Court of Appeal in In re Diplock [1948] Ch. 465, 478, 479. There, it was pointed out that on the facts of that case persons unversed in the law were entitled to assume that the executors were properly administering the estate, and that if those persons received money bona fide believing themselves to be entitled to it, "they should not have imposed upon them the heavy obligations of trusteeship." The judgment then pointed out:

"the principles applicable to such cases are not the same as the principles in regard to notice of defects in title applicable to transfers of land where regular machinery has long since been established for inquiry and investigation.

To that I may add the obvious point that the provisions about constructive notice in section 199 of the Law of Property Act 1925 apply only to purchasers (as defined in section 205 (1) (xxi)) and are not in point in relation to a beneficiary who receives trust property from the trustees.

With that, I turn to the cases on constructive knowledge. Mr. Taylor relied strongly on Selangor United Rubber Estates Ltd. v. Cradock (No. 3) [1968] 1 W.L.R. 1555 and Karak Rubber Co. Ltd. v. Burden (No. 2) [1972] 1 W.L.R. 602. Each was a knowing assistance case. In the Selangor case at p. 1582, Ungoed-Thomas J., immediately after speaking of tracing property into the hands of a volunteer, said that equity

"will hold a purchaser for value liable as constructive trustee if he had actual or constructive notice that the transfer to him was of trust property in breach of trust ...";

and at p. 1583 he went on to refer to equitable rights and to say that in general "it is equitable that a person with actual notice or constructive notice of those rights should be fixed with knowledge of them." I find this view hard to reconcile with the passage in In re Diplock [1948] Ch. 465 (a case not cited to the judge) which I have just mentioned; and with all respect, it also seems to me to tend to confuse the absence of notice which shields a purchaser from liability under the doctrine of tracing, with the absence of knowledge of the trust which will prevent the imposition of a constructive trust. The judge went on to consider the meaning of "knowledge" in various judgments, and reached a conclusion that knowledge was not confined to actual knowledge; and with this, as such, Mr. Chadwick had no quarrel. But he strongly contended that the cases cited on the extended meaning of "knowledge" were cases within the "wilful and reckless" head in the classification in the Baden case [1983] B.C.L.C. 325 (i.e. type (iii)), and that there was nothing to justify the inclusion of types (iv) and (v). The essential difference, of course, is that types (ii) and (iii) are governed by the words "wilfully" or "wilfully and recklessly," whereas types (iv) and (v) have no such adverbs. Instead, they are cases of carelessness or negligence being tested by what an honest and reasonable man would have realised or would have inquired about, even if the person concerned was, for instance, not at all reasonable. Yet Ungoed-Thomas J. in his conclusion, at p. 1590, applied the standard of what would have been indicated to an honest and reasonable man, or would have put him on inquiry, and so, I think, included all five of the Baden types of knowledge, and not only the first three.

In the Karak case [1972] 1 W.L.R 602, Brightman J. considered this conclusion. Again In re Diplock [1948] Ch. 465 was not cited, but Williams v. Williams (1881) 17 Ch.D. 437, another case not cited to Ungoed-Thomas J. in the Selangor case [1968] 1 W.L.R. 1555, was duly examined. Brightman J. distinguished that case by pointing out that it was a knowing receipt case, whereas the case before him was a knowing assistance case; and he said, at p. 638, that Williams v. Williams had "no relevance at all to the case before me." In Williams v. Williams Kay J. had held that the recipient of the trust property, a solicitor, was not liable as a constructive trustee, and, at p. 445, the judge said that the case would be "very different" if the recipient had "wilfully shut his eyes." He then referred to the "very great negligence" of the solicitor, qua solicitor, in ignoring the trust, though holding that he was not liable as a constructive trustee. I do not see how Kay J. could have reached that conclusion if he had thought that knowledge of Baden types (iv) and (v) had sufficed; and, of course, what is before me is a case of knowing receipt, like Williams v. Williams and unlike the Karak case.

There is also In re Blundell (1888) 40 Ch.D. 370, a case not cited in the Selangor case [1968] 1 W.L.R. 1555. It was cited but not mentioned in the judgment in the Karak case [1972] 1 W.L.R. 602, but like

Williams v. Williams, 17 Ch.D. 437 and In re Diplock [1948] Ch. 465 it does not appear in the Baden case [1983] B.C.L.C. 325: all three, I may say, were duly cited in the Carl Zeiss case [1969] 2 Ch. 276. In In re Blundell, Stirling J. refused to hold that a solicitor was a constructive trustee of costs that a trustee of property had allowed him to take out of the estate, even though he knew that the trustee was guilty of a breach of trust, "unless there are facts brought home to him which show that to his knowledge the money is being applied in a manner which is inconsistent with the trust": see p. 381. Both Williams v. Williams and In re Blundell figure prominently in the judgments of Sachs and Edmund Davies L.JJ. in the Carl Zeiss case, in support of their conclusion that negligence is not enough and that there must be dishonesty, a conscious impropriety or a lack of probity before liability as a constructive trustee is imposed.

I should also mention certain other cases. Consul Development Pty. Ltd. v. D.P.C. Estates Pty. Ltd. (1975) 132 C.L.R. 373 is a case where from the judgments of Barwick C.J. and Gibbs and Stephen JJ. may be collected a somewhat tentative acceptance of type (iv) knowledge, but not of type (v). Belmont Finance Corporation Ltd. v. Williams Furniture Ltd. [1979] Ch. 250, 267, 275, shows Buckley and Goff L.JJ. taking the view (though not as a matter of decision) that what is required in a knowing assistance case is actual knowledge, wilful shutting the eyes to dishonesty, or wilful and reckless failure to inquire, but that other forms of constructive notice are not enough; and Orr L.J. concurred. Goff L.J. at p. 275 reaffirmed the view that he had expressed in Competitive Insurance Co. Ltd. v. Davies Investments Ltd. [1975] 1 W.L.R. 1240 that "constructive notice of the section 199 type would not be sufficient." I do not read the judgments of Buckley and Goff L.JJ. in Belmont Finance Corporation Ltd. v. Williams Furniture Ltd. (No. 2) [1980] 1 All E.R. 393, 405, 412, as resiling from the views that they had expressed in the earlier stages of the case.

I have by no means exhausted the wealth of authorities on the subject, but although I have examined them all, I do not think that I need cite any more of them. Mr. Taylor was critical of certain of the authorities that I have cited. In re Diplock [1948] Ch. 465, he said, was a case of institutions which received cheques; and an ordinary reasonable institution would not investigate the right of the executors to send a mere cheque. The result was that In re Diplock did no harm to his contentions in a case where settled chattels rather than cheques were in issue; and the Duke's solicitor was thus under a duty to see that the Duke received the chattels free of all obligations to others. This view seemed to me to have strange results. I suggested to Mr. Taylor that this appeared to mean that whether you had to look a gift horse in the teeth depended on whether it was a simple or a complicated horse; and with this I think he agreed, at any rate, he said, if the horse was valuable. I cannot accept this. Short of a wilful or reckless disregard of the obvious, it seems to me to be quite unreal to suggest that a donee's liability to be subjected to a constructive trust may depend upon his correctly estimating whether the gift is sufficiently complex and valuable to require him to employ a solicitor to investigate the donor's right to

make the gift. As for Williams v. Williams, 17 Ch.D. 437, that was a special case, said Mr. Taylor. The solicitor had received the property not beneficially but on behalf of his client, and so was a mere conduit pipe; yet I do not see how that disposes of what appears to have been the judge's distinction between recklessness and negligence. In re Blundell, 40 Ch.D. 370 was another case of a solicitor, but in what Mr. Taylor said about it I could not perceive any ground for disregarding it in relation to this case.

Mr. Taylor's submissions on what should have been done by the Duke and Mr. Lickfold if the Duke was to escape being held a constructive trustee were along the following lines. Mr. Taylor disclaimed any suggestion that the Duke himself should have made any inquiries about the chattels on receiving Mr. Lickfold's letter of 15 November 1948. As mentioned in my former judgment, ante, p. 269D-F, that letter contained a paragraph headed "Heirlooms" which ran as follows:

"I had a long conversation with Nicholl on Thursday last, and the trustees have agreed that the heirlooms should be released, except the pictures. Under the resettlement executed by you on 20 December 1923 there is a clause by which the trustees can in their discretion release a large quantity of heirlooms and make a new list of such articles as are to remain as heirlooms. The amount obtained for the sale of any articles will be your personal property and the proceeds of sale will not have to be considered as capital trust money."

Mr. Taylor also disclaimed any suggestion that on receiving this letter the Duke should have dug out earlier letters that he had received, which had enclosed a copy of the settlement and had set out a correct prŽcis of it. What Mr. Taylor did say was that this letter did not sufficiently convey to the Duke that (apart from the pictures) the Duke could have all the chattels. If the Duke did read it in that way, then he ought to have written back to Mr. Lickfold making inquiries. Further, Mr. Lickfold was engaged in checking what the Duke would get, and whether he would take it free from the trust, so that he was in a position similar to a solicitor to a purchaser. As I have already indicated, Mr. Taylor said that the essence of his case was the imputation of Mr. Lickfold's knowledge to the Duke.

In my judgment, the letter is perfectly capable of conveying to the Duke that (pictures apart) he could have all the chattels, and I think that it did convey this to him. Furthermore, everyone concerned acted at the time as if that were the position. I cannot see why the Duke, reading the letter that way, was under any obligation to write back to Mr. Lickfold to make inquiries, in default of which he would become liable as a constructive trustee. Furthermore, I cannot see why the correct knowledge of clause 14 (B) that Mr. Lickfold certainly had at one time should be imputed to the Duke. Mr. Lickfold was plainly acting for the Duke at all material times, and he had much to do, since the Duke was in Kenya most of the time. But I cannot see that Mr. Lickfold was in any way employed by the Duke to ascertain whether the Duke would take the chattels free from the trusts of the settlement or subject to them. Mr. Lickfold was acting generally for

his mainly absent client, and was naturally concerned to see that he received no less than he was entitled to; but that is quite different from being employed to investigate title. I cannot see that by acting generally for the Duke he became subject to any duty or expectation of making any such investigation, or at least any duty that is at all similar to the duty owed by a purchaser's solicitor to his client. Even under the ordinary doctrine of purchaser without notice, a purchaser does not have imputed to him notice of everything of which his solicitor has notice, but only the notice that the solicitor has acquired in the particular transaction: see section 199(1)(ii)(b) of the Law of Property Act 1925.

I am not saying that there can never be any such thing as imputed knowledge; but I have considerable doubts about it generally, and especially in relation to imposing a constructive trust. Mr. Taylor cited Vane v. Vane (1873) 8 Ch.App. 383, 399, 400, where there are strong dicta that the actual knowledge of an agent would be treated as the actual knowledge of the principal. Yet this was said not generally but in relation to the "actual knowledge of the agent through whom an estate is acquired," and "the knowledge of the purchaser's agent acquired in the course of the transaction." That case, I should say, had nothing to do with constructive trusts, but concerned a statutory defence under the Statutes of Limitation for a purchaser for value who "did not know and had no reason to believe" that a fraud had been committed.

I shall not go through all the other cases cited by Mr. Taylor, but I think that I should say something about three of them. In Boursot v. Savage (1866) L.R. 2 Eq. 134, 142 Sir Richard Kindersley V.-C., speaking of a solicitor's knowledge in relation to himself as purchaser, said "his knowledge is my knowledge." But from the context it is plain that what was being discussed was notice, not knowledge, and that the judge was not intending to lay down any rule for knowledge as distinct from notice. Belmont Finance Corporation v. Williams Furniture Ltd. (No. 2) [1980] 1 All E.R. 393 concerned a constructive trust of the "knowing receipt" type, and actual or constructive knowledge. But the reference to "imputed knowledge" in the judgment of Buckley L.J., at p. 404, is to the imputation to a company of the knowledge of its directors and secretary; and that is a very different matter. A company is an artificial person, and so the knowledge of the natural persons who control and manage it must be treated as being the knowledge of the company; but that is no ground for saying that the knowledge of one natural person must be treated as being the knowledge of another. The word "imputed" is being used in an entirely different sense. In International Sales and Agencies Ltd. v. Marcus [1982] 3 All E.R. 551, 557, "imputed knowledge" is mentioned in relation to constructive trusts; but that too was a company case, and in any event it was held to be beyond argument that there was actual notice of the trust.

Even if it is accepted that on a purchase there may for certain purposes be "imputed knowledge" that corresponds to "imputed notice," that still leaves untouched cases where there is no purchase and no employment of a solicitor to investigate the right of a donee or a beneficiary to receive the gift or beneficial interest in question. In

any case, in relation to constructive trusts I cannot see why equity should say to that donee or beneficiary, "True, your solicitor, who had not been employed to investigate title, told you that the property was yours, and you believed him; but he was wrong, and as he had actual knowledge that the property was subject to a trust, you too had actual knowledge of that trust. This bound your conscience, and therefore you took the property as a constructive trustee." This would in no way be equity as I know it.

There is a further question that I should consider, and that is forgetfulness. Little was said about this in argument; but in a case in which at one time the true position was known to Mr. Lickfold, and possibly to the Duke, I must say something about it. If a person once has clear and distinct knowledge of some fact, is he to be treated as knowing that fact for the rest of his life, even after he has genuinely forgotten all about it? To me, such a question almost answers itself. I suppose that there may be some remarkable beings for whom once known is never forgotten; but apart from them, the generality of mankind probably forgets far more than is remembered. So far as the doctrine of notice is concerned, there is authority, in relation to the rule in Dearle v. Hall (1828) 3 Russ. 1, for saying that the question is whether at the time in question notice previously obtained continues to operate on the mind of the recipient: see Ipswich Permanent Money Club Ltd. v. Arthy [1920] 2 Ch. 257. Of course, since 1925 there is a statutory scheme under section 137 of the Law of Property Act 1925 for regulating priority by means of the receipt of written notices; and nothing I say is intended to suggest that such notices might lose their effect if the recipient or anyone else forgets them. But apart from such statutory provisions, it seems to me that a person should not be said to have knowledge of a fact that he once knew if at the time in question he has genuinely forgotten all about it, so that it could not be said to operate on his mind any longer. This is emphasised in relation to constructive trusts in that, in my view, it would be wrong to hold that a person's conscience is affected by something that he does not know about. Even if section 199 of the Law of Property Act 1925 had any application, and notice were in issue, I cannot accept Mr. Taylor's contention that the section shows that what a person once knew he is conclusively presumed still to have notice of; for the section is framed in terms of what "is" within the purchaser's own knowledge, and not "is or ever has been" within his knowledge.

Of course, the court may be slow to conclude that what was once known has been forgotten; but such a conclusion is likely to be aided if the person has received professional advice which treats the fact forgotten as if it did not exist. The relevance of this is that even if the Duke did once understand the true meaning of clause 14(B), there is nothing to suggest that he still knew it when he received the chattels, especially in the light of what Mr. Lickfold had written to him. The fact that he had a copy of the 1923 settlement and could have read clause H 14(B) and questioned Mr. Lickfold about it does not, in my judgment, show that the Duke at any relevant time had anything that can be called "knowledge" that the transfer of the chattels to him was a breach of

trust or in any way improper. The question must be determined by what knowledge the Duke had at the time of the transfer, and not by his state of knowledge at any previous time.

I shall attempt to summarise my conclusions. In doing this, I make no attempt to reconcile all the authorities and dicta, for such a task is beyond me; and in this I suspect I am not alone. Some of the difficulty seems to arise from judgments that have been given without all the relevant authorities having been put before the judges. All I need do is to find a path through the wood that will suffice for the determination of the case before me, and to assist those who have to read this judgment.

(1) The equitable doctrine of tracing and the imposition of a constructive trust by reason of the knowing receipt of trust property are governed by different rules and must be kept distinct. Tracing is primarily a means of determining the rights of property, whereas the imposition of a constructive trust creates personal obligations that go beyond mere property rights.

(2) In considering whether a constructive trust has arisen in a case of the knowing receipt of trust property, the basic question is whether the conscience of the recipient is sufficiently affected to justify the imposition of such a trust.

(3) Whether a constructive trust arises in such a case primarily depends on the knowledge of the recipient, and not on notice to him; and for clarity it is desirable to use the word "knowledge" and avoid the word "notice" in such cases.

(4) For this purpose, knowledge is not confined to actual knowledge, but includes at least knowledge of types (ii) and (iii) in the Baden case [1983] B.C.L.C. 325, 407, i.e. actual knowledge that would have been acquired but for shutting one's eyes to the obvious, or wilfully and recklessly failing to make such inquiries as a reasonable and honest man would make; for in such cases there is a want of probity which justifies imposing a constructive trust.

(5) Whether knowledge of the Baden types (iv) and (v) suffices for this purpose is at best doubtful; in my view, it does not, for I cannot see that the carelessness involved will normally amount to a want of probity.

(6) For these purposes, a person is not to be taken to have knowledge of a fact that he once knew but has genuinely forgotten: the test (or a test) is whether the knowledge continues to operate on that person's mind at the time in question.

(7) (a) It is at least doubtful whether there is a general doctrine of "imputed knowledge" that corresponds to "imputed notice." (b) Even if there is such a doctrine, for the purposes of creating a constructive trust of the "knowing receipt" type the doctrine will not apply so as to fix a donee or beneficiary with all the knowledge that his solicitor has, at all events if the donee or beneficiary has not employed the solicitor to investigate his right to the bounty, and has done nothing else that can be treated as accepting that the solicitor's knowledge should be treated as his own. (c) Any such doctrine should be distinguished from the process whereby, under the name "imputed knowledge," a company is treated as having the knowledge that its directors and secretary have.

(8) Where an alleged constructive trust is based not on "knowing receipt" but on "knowing assistance," some at least of these considerations probably apply; but I need not decide anything on that, and I do not do so.

From what I have said, it must be plain that in my judgment the Duke did not become a constructive trustee of any of the chattels. I can see nothing that affected his conscience sufficiently to impose a constructive trust on him: and even if, contrary to my opinion, all of the five Baden types of knowledge are in point, instead of only the first three, I do not think that he had any such knowledge. He was a layman, and he accepted and acted on what he was told by his solicitor and was acted on by the trustees and the solicitor to the trustees. Furthermore, as stated in my earlier judgment, the plaintiff himself had executed a deed of appointment of new trustees dated 25 June 1953. This deed contained a full release of the then trustees in respect of all acts and things done or omitted to be done by them; and yet there has been no suggestion that at that time the plaintiff had made any complaint of the massive breaches of trust that the trustees had committed a few years earlier by failing to make the selection of chattels required by clause 14(B) and allowing the Duke to take any chattels he wanted. However, quite apart from this, I would reach the same conclusion. Accordingly, I hold that the Duke never became a constructive trustee of any of the chattels.

Solicitors: Beachcrofts; Allen & Overy.

K. N. B.