implied-in-fact trust, is enforced for the very same fundamental reason that a constructive trust is, namely, to prevent the unjust enrichment of the fraudulent retainer of the property. We are too apt to forget that the sole reason why chancery took jurisdiction to enforce uses, the earliest trusts, was to prevent the unjust enrichment by feoffees through their fraudulent retention of land conveyed to them in use or confidence, and that to-day there is no other reason for chancery's enforcement of any kind of a trust.34

In the case of express trusts the trust will everywhere be enforced regardless of whether the fraud consists of fraudulent acquisition and fraudulent retention combined or only of fraudulent retention. In the case of resulting trusts the same thing is true. But some courts hesitate to raise constructive trusts in the Statute-of-Frauds cases on fraudulent retention alone, though it is fraudulent retention and not fraudulent acquisition that injures the defrauded party and enriches the fraudulent party, and though the same courts in effect recognize the truth of that statement whenever they enforce a resulting trust in favor of the payer of the purchase money against a grantee who took innocently but afterwards decided to retain the trust res for himself through the aid of the parolevidence rule or of the Statute of Frauds or of both. However inconsistent in granting and in withholding relief in the trusts cases some courts may be, the proper principle for their guidance is clear the moment the essential reason for the recognition and enforcement of trusts - the rectification of unjust enrichment is seen to be the same for all trusts, whether they are express or implied, and whether, if not express, they are resulting or constructive. With this explanation, we may now proceed to consider the classification problems in more detail.

I.

Express Trusts.

Express trusts heretofore have needed practically no subdivision. By reason, however, of the somewhat old use in the books of the word "implied" to describe what are really, in a fair sense, only express trusts, it will be well to divide express trusts into (1) Ex

34 This is common historical knowledge. See Ames, Lectures on Legal History, 237-238; 21 HARV. L. REV. 265.

press trusts manifest without interpretation, and (2) Express trusts ascertained, clarified, and defined by construction.35

The first class of express trusts needs no elaboration, but the second class of express trusts requires further subdivision.

One subdivision of that second class of express trusts will consist of precatory trusts, i. e., trusts "created by certain words, which are more like words of entreaty and permission than of command or certainty." 36

Another subdivision, it seems, will be comprised of one of the trusts commonly regarded as resulting, namely, the kind of trust which Pomeroy splits into two subdivisions, as noted on p. 446, ante. Where property is conveyed by deed without consideration, or is disposed of by will, in trust, but the trusts specified fail in whole or in part, it is customary to say that there is a resulting trust to the grantor or to those who have succeeded either to his rights or to those of the testator. Is this a real resulting trust, i. e., a genuine implied-in-fact trust, or is it something else? While at first sight the idea seems strange, it is probable that it is an express trust ascertained by construction. The current assumption that the trust is resulting is made in forgetfulness of the fact that construction involves inferences of fact. The question, then, is whether in the case of this kind of a trust inferences of fact have to be used to an extent not reasonably to be regarded as justifiable in cases of mere interpretation. In the trust supposed, the clearly express trust is the trust which fails. Then the question is whether the trust which is enforced for the grantor or his heir, or for the residuary devisee or the heir of the testator, is also express or is implied in fact. It seems to be a case of what may be called indirect or negative construction, i. e., construction by elimination. Knowing

35 The word "construction" is here meant to include both the ascertainment of the actual intention of the creator of the trust, as revealed in the express words of the trust, and the adoption of what the court believes from the revealed actual intention would have been the intention of the creator of the trust if the situation calling for definition of the trustee's powers and duties under the express trust, or for a determination of the equitable interests which the trust gives rise to, had presented itself to his mind. Compare Gray's Nature and Sources of the Law, § 370 (interpretation of statutes), and §§ 702-705 (interpretation of wills). Although some day it may be well to discriminate in terminology these two very different exercises of the judicial function, necessity has not yet forced that discrimination on us in the law of trusts.

36 Black's Law Dictionary, 928. That precatory trusts are express trusts ascertained by construction is clear. See I Pomeroy, Equity Jurisprudence, 3 ed., p. 178, n.

Ι

the interest which the grantor or the testator had, the court construes the conveyance or will and finds that all not effectually conveyed remained at law in the grantor, or by analogy vested in the grantor in equity, immediately upon the conveyance, and that all not effectually devised to the cestuis vested in the testator's heir or in the residuary devisee. The direct or positive construction consists in ascertaining the property interest of the express cestuis; but as the property interest of the grantor, or of the successor in interest of the testator, is figured out by deducting the property interests of the express cestuis, ascertained by positive construction, from the total property interests to be disposed of, that deduction would seem to be still part of the construction of the express trust, even though it is only a by-product part. To be sure, a presumption that there is no trust for the grantor is raised by the stated consideration or by the use habendum in the deed or by both; and a presumption that there is no trust for the testator's successor in interest is raised because the testator, anticipating death, does not expressly provide for the successor in interest, but in each case that presumption is overcome by the "in-trust-nevertheless" clause in the deed and the "in-trust" clause in the will. Where the voluntary deed or devise is expressly "in trust" and the trusts fail, the "in-trust" clause keeps the deed or will from carrying a conclusive or other presumption of gift. The words "in trust" are words which raise a presumption that a gift to cestuis and not to the grantee was intended, and which therefore make rational the presumption of fact that the portions of the equitable interest which are unclaimable by the intended cestuis were intended to belong to, and so are unreflectingly but plausibly said "to result to," the creator of the trust or his successors in interest. Those presumptions are interpretation presumptions, it would appear. Surprising as it may be, this kind of a trust seems, then, fairly to be classified as express by construction.

II.

Resulting Trusts.

As we include under express trusts those trusts defined by construction which many writers have called implied,37 the term

37 A vendor against whom a bill for specific performance will lie is often spoken of as a trustee, and in Felch v. Hooper, 119 Mass. 52 (1875), it was deemed by the court

"implied can be utilized to cover, as already noted, resulting trusts. True resulting trusts are implied in fact, and they are the only implied trusts properly so called.

Implied-in-fact trusts give us our first serious classification problem. The typical resulting trust is implied in fact, but some socalled resulting trusts can be regarded as implied in fact only by a resort to a very artificial reasoning.

Take the case where A. pays B. the purchase price for property and has B. convey it to A.'s wife, C., who receives it on an oral trust for A. From the fact that C. is A.'s wife there is a presumption of a gift to her by A., but as this presumption is rebuttable the oral trust may be proved to rebut it. The presumption of gift having been rebutted, equity makes C. hold as trustee for A. despite C.'s plea of the Statute of Frauds and despite the fact that C. may have taken with innocent intent and have decided to keep for herself only after quarrels between her and her husband took place or some other occurrence influenced her.38 So chancery will make that the land contracted to be sold was "charged with an implied trust" within the meaning of a statute which gave the court power, on service by publication, to decree a conveyance by a non-resident person "seised of an estate upon a trust express or implied," and to appoint some suitable person to make the conveyance. This case was one where on a bond for conveyance given by the vendor the vendee had paid part of the purchase money, entered into possession of the land, made improvements thereon, and tendered the balance of the purchase price, and the court seemed to think that because the part payment and the tender of the balance of the purchase price had to be shown dehors the bond the trust was implied and not express. Colt, J., said for the court:

"This statute expressly includes implied trusts and cannot be confined in its application to trusts which are created by deed or will and do not depend upon the proof of facts which may be open to dispute."

It is believed, however, that the trust, if it really was a trust within the meaning of the statute, was an express trust found to be such by construction, and that by "a trust. . . implied" was probably meant a resulting trust and a constructive trust only. The bond for title was certainly an express contract and if equity says that the vendor was a trustee because of his contract obligations, he was of course an express trustee. This is none the less true although he would have been astonished to be told that he was an express trustee. That he was seised of an estate upon express trust within the meaning of the words in that Massachusetts statute, if he was a trustee at all within the intent of that statute, seems reasonably clear. On whether he was a trustee, see Merrill v. Beckwith, 163 Mass. 503, 40 N. E. 855 (1895).

38 Smithsonian Institution v. Meech, 169 U. S. 398, 18 Sup. Ct. 396 (1897); Duvale v. Duvale, 54 N. J. Eq. 581, 35 Atl. 750 (1896); Short v. Short, 62 Ore. 118, 123 Pac. 388 (1912). See Harden v. Darwin, 66 Ala. 55 (oral trust performed) (1880); Carpenter v. Gibson, 104 Ark. 32, 148 S. W. 508 (oral trust not proven) (1912). But see Kinley. v. Kinley, 37 Colo. 35, 86 Pac. 105 (1906); Murray v. Murray, 153 Ind. 14, 53 N. E.

C.'s heirs or devisees hold in trust for A. despite their plea of the Statute of Frauds, and despite the fact that C. never had a fraudulent intent but always acknowledged orally the existence and binding nature of the trust.39

Why does chancery do this? Were the history of the impliedin-fact and the so-called "implied-in-law" trusts to be forgotten, it would seem as if there could be but one logical answer, namely, that when the presumption of a gift was rebutted by showing that a gift was not intended and a trust was, all presumptions of fact were at an end, and the situation was then one of an oral trust of land, a repudiation of that trust by the trustee, and an attempted unjust enrichment of the trustee by the retention for his own use of the trust res at the expense of the buyer of the property. In other words, on principle, the trust, and, in the absence of a controlling historical development to the contrary, the only trust for chancery to declare and enforce would seem to be a constructive trust.

But the historical development of the trust doctrine has been against that analysis. If, as just suggested, this kind of a trust is different from the ordinary resulting trust, most of the courts have not been aware of it. The typical resulting trust has been where A. has paid for property and B. has conveyed it at A.'s request to C., who was not related to A. There chancery indulged the presumption of fact-rebuttable, as such presumptions normally are that C. was to hold in trust for A. But if C. was A.'s wife a presumption of gift was indulged. Without noticing that this presumption of gift was on principle the only presumption where C. was A.'s wife, the chancery judges regarded the presumption of a trust as the first one entertained and as rebutted by proof of the relationship of the parties, with the consequent presumption of fact of a gift. Then when that presumption of fact of a gift 946 (1899); Andrew v. Andrew, 114 Ia. 524, 87 N. W. 494 (1901); Mullong v. Schneider, 155 Ia. 12, 134 N. W. 957 (1912). Compare Johnson v. Ludwick, 58 W. Va. 464, 52 S. E. 489 (1906); Ludwick v. Johnson, 67 W. Va. 499 (1910).

39 Sherman v. Sherman, 20 D. C. 330 (1892); Bachseits v. Leichtweis, 256 Ill. 357, 100 N. E. 197 (1912); Cotton v. Wood, 25 Ia. 43 (1868); Price v. Kane, 112 Mo. 412, 20 S. W. 609 (1892); Bartlett v. Bartlett, 15 Neb. 593, 19 N. W. 691 (1884); Bailey v. Dobbins, 67 Neb. 548, 93 N. W. 687 (1903); Lahey v. Broderick, 72 N. H. 180, 55 Atl. 354 (1903); Yetman v. Hedgeman, N. J. Eq. (1913), 88 Atl. 206; Hickson v. Culbert, 19 So. Dak. 207, 102 N. W. 774 (1905); Bickford v. Bickford's Estate, 68 Vt. 525, 35 Atl. 471 (1896). See Corr's Appeal, 62 Conn. 403, 26 Atl. 478 (1892). Compare Livingston v. Livingston, 2 Johns. Ch. (N. Y.) 537 (1817). But see Chapman v. Chapman, 114 Mich. 144 (1897); Ryan v. Williams, 92 Minn. 506 (1904).