was held to have been erroneous. The evidence should have been admitted not as directly assertive, but as permitting an inference that lack of intent on the decedent's part had prevented a legally effective delivery of the deeds.

These cases invite the following comments:

(a) They do not involve declarations of intention, and therefore show that the matter now under discussion demands acceptance or rejection of the Hillmon doctrine in a form broadened to include all declarations susceptible of inferential or circumstantial use.

(b) Neither case involves an inference back from present state of mind to past act. The inference permitted is simply from present state of mind to past state of mind. This is clear enough in the first case, where the necessary physical acts for a valid delivery had occurred and the mental element alone was found lacking. The second case is not quite so clear. But there, after all, the administrator had two lines of attack: (1) no physical act and (2) no conveying intent. Since he might win on either line, his evidence was acceptable although it bore on (2) only. The opinion touches this squarely when it says "delivery is essentially a matter of intent." Here, then, we deal with evidence involving no risk of faulty perception and little risk of misstatement. The main danger is that the decedent's memory, unconsciously or subconsciously appealed to, may have played him false.

9 59

(c) It is highly doubtful whether courts would thus respect subjective intent as distinguished from manifest meaning in any but gratuitous transactions. Suppose the defendant in either case had paid hard cash for the deed. Would the decedent's representative or successors have been allowed to negative the conveyance by showing a secret retentive state of mind contradicting the outward physical act of delivery? Williams v. Kidd and Mower v. Mower may never carry over into the majority of the world's transactions. We have much more buying and selling than we do free giving. Incidentally, is it wise in this respect to treat a grantee who takes by gift in a radically less favorable way than we do a grantee who pays?

59 228 Pac. at 914.

A more general comment upon this entire subdivision of the article is also invited. Observe the interplay of the first and the second limiting suggestions. A declaration offered as inferential proof of a past event not involving an act of the declarant's encounters a double objection, and has comparatively small chance of favorable reception. The more interesting and doubtful issue refers only to the proof of the declarant's past acts, and success for the proponent of the evidence in this narrowed aspect would hardly threaten the structure of our law of evidence.

2. But we cannot, in Mr. Seligman's opinion, stop here. His next step is to discuss the admissibility of explicit statements of memory to prove a past act or event. He believes that we must accept these too, if we accept the Hillmon evidence of intention for the purposes of that case. I think not. The argument for exclusion stated in the preceding subdivision again applies. The risk of faulty memory is now in full force. No longer is the risk of misstatement whittled thin. The indirect and inferential nature of the evidence is dangerously disguised

it "would be liable to be given the same weight as actual testimony by the jury, who can easily give declarations of intention their proper value." 60 Existing law manifests a plain distrust of the state of mind called memory when expressed in direct form. Even when past physical sensation or past mental condition is directly in issue, the weight of authority refuses to permit proof by extrajudicial declarations couched in the past tense. All the more will such evidence be excluded if offered to prove a past act or event by inference from a past state of

60 Quoted from Professor Chafee's valuable review of Wigmore's second edition. 37 HARV. L. REV. 513, 519.

See 3

61 The cases on physical sensation or condition are fairly numerous. WIGMORE, EVIDENCE, 2 ed., § 1722 (b). As to mental state, the authorities are much befogged by use of the pernicious term res gestae. 4 CHAMBERLAYNE, EVIDENCE, § 3660, n. 3, lists an imposing string of cases, but the list does not bear up well under scrutiny. Perhaps the best citations in it are Groff v. Rohrer, 35 Md. 327, 340 (1872), and Mayfield v. State, 101 Tenn. 673, 676, 49 S. W. 742 (1899). 3 WIGMORE, EVIDENCE, 2 ed., § 1732, n. 8, cites some better cases. In this connection, see 3 ibid., §§ 1750 (c) and 1754, indicating a partially similar limitation in connection with the spontaneous exclamation exception to the hearsay rule. See, however, Chapman v. Bimel-Ashcroft Mfg. Co., 263 S. W. 993 (Mo., 1924). Compare Mr. Seligman's reasoning in 26 HARV. L. REV. at 154-155.

mind reflected only through conscious exercise of memory. This cumulation of objections is important. Hearsay may struggle into a case despite a metaphorical ball and chain on one leg. But when we also fetter the other leg and manacle the wrists for good measure, the aggregate resistance is too much to overcome.

It would be reckless to say that every precedent accords with the contention just made. The helter-skelter methods of digesting evidence cases make it practically impossible to find them all, let alone discover exactly what they hold. But it will perhaps be fair to close the discussion of this point with comment upon a decision likely to be cited as contradicting the contention.6

62

A gentleman by will gave one of his servants a legacy "provided [she is] . . . in my employ at the time of my death." The testator died. Payment of the legacy was demanded and refused. The servant sued the estate. At the trial the defence offered a memorandum written on the envelope in which the will was found. The memorandum was in the testator's writing, was dated September 17, 1917, and stated that the plaintiff "left my employ this day." The lower court admitted the evidence and directed a verdict against the plaintiff. On appeal, the decision was affirmed.

At first blush this seems a realization of Mr. Seligman's worst fears. But we must fill in the situation. It was shown that on September 17, 1917, the testator closed his house, paid the plaintiff her wages to date, and went to live at a club. The plaintiff never again served him. If this were all, the suit might never have been brought. However, there is more to the story. After moving to the club the testator each month sent the plaintiff a cheque somewhat exceeding in amount her previous monthly wages. This gave ground for arguing that while the plaintiff's services had ceased her employment continued. The testator's cheque book supplied a counter-argument. On each of the stubs from which the plaintiff's cheques were torn the

62 Kelly v. Pitney, 98 N. J. L. 773, 121 Atl. 593 (1923), noted in 22 MICH. L. REV. 169.

testator had written either "Gratuity to date. Former cook" or simply "Gratuity.'

77 63

64

Now let us get back to the discussion of the memorandum or declaration on the envelope. The upper court starts off with the good old stock phrase "part of the res gestae," which leaves us fumbling in a Latin fog instead of an English one. From the fact that the declaration is in the past tense, it seems not to have accompanied any act. Perhaps it was made in an arm chair at the club, hours or even days after the house-closing was finished. But the court goes on to say that the memorandum is admissible "as characterizing and making plain testator's state of mind and purpose at the time he closed his house." This sounds more intelligible. Observe, though, that what we wish to prove is the termination of a contract or relation of employment. Suppose the continuance of employment to be a question of mutual consent, it is at least a question of expressed consent or dissent. State of mind bottled up in the testator's head discharges no servants. To be really serviceable, the memorandum must bear upon the case in one of the following manners:

(a) It may be said that no reasonable servant could have construed the closing of the house otherwise than as accomplishing her discharge, unless her employer gave her to understand to the contrary. The memorandum tends to prove that the testator neither did nor said anything to convey other than the natural impression.

(b) It may be said that the testator was here supplying a dictionary for the construction of the term "in my employ" as used by his will.66 He is saying: "The closing of my house always lay in my mind as something which would effect the discharge of my servants."

63 98 N. J. L. at 774, 775, 121 Atl. at 594.

64 98 N. J. L. at 776, 121 Atl. at 594.

65 Ibid.

66 In the Goods of Ashton, [1892] P. 83, 87. Here, however, the "dictionary was in the will itself and not in an independent document. See In re Mangan's Will, 200 N. W. 386, 392 (Wis., 1924), rejecting post-testamentary declarations which may perhaps have been an effort to alter the will, rather than to formulate the original interpretation.

(c) It may be said that the memorandum is evidence tending to prove that the testator himself, or possibly the plaintiff, did or said something sufficiently indicating termination of the employment.67

Of these three explanations, the first is fairly practical. As here applied, it does not show that the memorandum was properly admissible. It simply indicates that the plaintiff did not suffer because of the evidential ruling. The estate, having proved a situation which would make any reasonable servant appreciate that she was discharged, might rest on its oars until the plaintiff pulled up abreast by adducing adequate rebutting evidence. Her only evidence of this nature seems to have been proof of the continued monthly payments made to her after the house was closed. As pointed out above, it is clearly maintainable that these payments were colorless and did not help her side. Hence, having failed to overcome the estate's prima facie case, the plaintiff was not prejudiced even if her opponents were permitted to improve their position in an irregular manner. She could not have won the race with that handicap removed.

The second explanation is too thin and airy. It also runs us into various difficulties with the law of wills and the principles of construction.

The third explanation alone goes against my contention. The decision is not a convincing authority for the acceptance of this last explanation.

3. The reader will observe that I have not commented upon the possibility that a declaration may be used as tending to prove the doing of a contemporaneous act or the happening of

67 See, as having possible bearing on this point, Harvey v. Brown, 187 N. C. 362, 121 S. E. 626 (1924). Here it was material to prove that one of the defendants had no knowledge of a certain agreement until after he had taken a mortgage. A witness for the plaintiff was allowed to testify on cross-examination that the defendant in question told him that his (the defendant's) first knowledge of the agreement was after the institution of the suit. That statement obviously means: "Nobody previously informed me, nor did I otherwise find out, about the agreement." The opinion is so hazy on this point that the case cannot be taken very seriously. Besides, the plaintiff seemed to have no adequate affirmative evidence that this defendant was given notice of the agreement; thus the lower court's admission of the hearsay was immaterial error at

most.