This controversy involves the rightful beneficiary of a National Service Life Insurance policy on the life of Harold R. Bentley issued pursuant to 38 U.S.C.A. § 701 et seq. Jurisdiction in the court below was predicated on 38 U.S.C.A. § 784.

The pertinent facts of this case begin in 1941 when the insurance policy in question was issued. Nena Bentley, Harold's wife, was then named as the principal beneficiary. Bentley divorced Nena in 1960. Two years later, on October 10, 1962, Bentley executed an insurance form which removed his divorced wife as principal beneficiary and named instead appellant herein, Elizabeth Seekatz Lyle.1 On October 30, 1962, Bentley executed another document designating Mrs. Lyle as the principal beneficiary of his insurance policy.2

Bentley died in December, 1964, and in 1965 Nena Bentley filed a claim with the Veterans Administration for payment of insurance benefits. The claim was disallowed by the Veterans Administration on the grounds that the proceeds belonged to Mrs. Lyle. Nena Bentley then filed this suit naming Elizabeth Lyle and the United States of America as defendants, and alleging that at the time the changes of beneficiary were made in Bentley's policy he was lacking in mental capacity or subject to the undue influence of Mrs. Lyle. The case was submitted to a jury with instructions to consider both mental incapacity and undue influence. The jury returned a general verdict for Nena Bentley.

Mrs. Lyle appeals to this court solely on the ground that there was insufficient evidence in the record to support a jury finding of undue influence. She does not challenge the sufficiency of the evidence on the issue of mental incapacity,3 but she invokes the rule that it is improper for a court to instruct on an issue of law about which there is no evidence. Jackson v. Southern Railway Co., 5 Cir. 1962, 317 F.2d 532, cert. denied, 375 U.S. 837, 84 S. Ct. 77, 11 L. Ed. 2d 65. She argues that since this case was submitted to the jury on the issues of both mental incapacity and undue influence, and since the jury returned a general verdict, it is impossible to determine upon which theory the jury verdict was based. Mitchell v. John R. Cowley & Bro., Inc., 5 Cir. 1961, 292 F.2d 105, 109; Superior Combustion Ind. v. Schollman Bros. Co., 8 Cir. 1959, 271 F.2d 357. She maintains that under the existing circumstances, the very real possibility that the jury based its verdict on a finding of undue influence makes the verdict fatally vulnerable if in fact the evidence of undue influence was insufficient to justify submission of that issue to the jury.

We find appellant's contentions to be well founded. If the jury's verdict was in fact based on undue influence, which it could have been, and if the evidence to sustain such a finding was insufficient, then the verdict must not be allowed to stand. Under such circumstances it matters not how much evidence there is in the record of mental incapacity. Since the general verdict speaks affirmatively with respect to either or both of the issues submitted to the jury, we have no way of knowing upon what basis the jury's verdict was founded. Therefore, while the subjectives of the jury's deliberations are not for us to examine, if those deliberations as directed by the court's instructions permit a verdict to be based on an issue not supported by sufficient evidence, the jury verdict must be set aside and a new trial held. Mosley v. Cia. Mar. Adra, S.A., 2 Cir. 1963, 314 F.2d 223, 229, cert. denied, 375 U.S. 835, 84 S. Ct. 52, 11 L. Ed. 2d 65 and 375 U.S. 829, 84 S. Ct. 73, 11 L. Ed. 2d 61.

With these considerations before us, we proceed to consider the sufficiency of the evidence in support of a possible jury verdict of undue influence. In evaluating this evidence, we are of course duty bound to accept all evidence in favor of the verdict as true and to give such evidence the benefit of all permissible inferences that would help to sustain the jury's decision. State Farm Mut. Automobile Ins. Co. v. Jackson, 8 Cir. 1965, 346 F.2d 484; Hughes v. Moran, Tex.Civ.App.1959, 325 S.W.2d 829 (writ. ref., n. r. e.).

We begin by noting the general contours of undue influence and the considerations and kinds of evidence which courts have considered probative in establishing its existence. Our point of departure is the Texas Supreme Court case of Long v. Long, 1939, 133 Tex. 96, 125 S.W.2d 1034, wherein the following helpful discussion appears:

"It is not possible to frame a definition of undue influence which embraces all forms and phases of the term. Every case is different from every other case, and must depend largely on its own facts and circumstances. Generally speaking, undue influence is such influence or dominion as exercised at the time, under the facts and circumstances of the case, which destroys the free agency of the testator, and substitutes in the place thereof the will of another. Undue influence has also been defined as that dominion acquired by one person over the mind of another which prevents the latter from exercising his discretion, and which destroys his free agency. Also, undue influence has been defined as `that which compels the testator to do that which is against his will from fear, the desire of peace, or some feeling which he is unable to resist.' 42 Tex.Jur., p. 792, sec. 2, and authorities there cited.

"It cannot be said that every influence exerted by one person over the mind of another is undue. The influence is not undue unless the free agency of the testator has been destroyed, and a will produced that such testator did not desire to make. 42 Tex.Jur., p. 793, sec. 4.

* * * * * *

"It is rarely possible to prove undue influence by what is generally known as direct testimony. Undue influence is usually a subtle thing, and by its very nature it usually involves an extended course of dealings and circumstances. Usually a person charging undue influence must substantiate such charges by circumstances extending over a considerable length of time. It is therefore the settled rule that undue influence can be established by what is known as circumstantial, as well as direct, evidence. Besteiro v. Besteiro, Tex.Com.App., 65 S.W.2d 759; Bergstedt v. Bender, Tex.Com. App., 222 S.W. 547.

"Undue influence and mental incapacity are two distinct grounds for avoiding a will. Undue influence in its essential elements has no real relation to mental incapacity. Mental incapacity implies the lack of intelligent mental power; while undue influence implies within itself the existence of a mind of sufficient mental capacity to make a will, if not hindered by the dominant or overriding influence of another in such a way as to make the instrument speak the will of the person exercising undue influence, and not that of the testator. Scott v. Townsend, 106 Tex. 322, 166 S.W. 1138.

"In spite of the rule just announced, weakness of mind and body, whether produced by infirmities of age or by disease or otherwise, may be considered as a material circumstance in determining whether or not a person was in a condition to be susceptible to undue influence. 42 Tex.Jur., p. 796." 125 S.W.2d at 1035-36.

More recent judicial pronouncements set forth somewhat more specific criteria of undue influence than those discussed in the Long case. For example, in Rothermel v. Duncan, Tex.1963, 369 S.W.2d 917, the Supreme Court of Texas observed:

"* * * before a testament may be set aside on the grounds of undue influence the contestant must prove: (1) the existence and exertion of an influence; (2) the effective operation of such influence so as to subvert or overpower the mind of the testator at the time of the execution of the testament; and (3) the execution of a testament which the maker thereof would not have executed but for such influence. See: Stewart v. Miller, Tex.Civ. App. (1925), 271 S.W. 311, wr. refused; Olds v. Traylor, Tex.Civ.App. (1944), 180 S.W.2d 511, wr. refused." 369 S.W.2d at 922.

"* * * The exertion of influence that was undue cannot be inferred alone from opportunity, but there must be some testimony, direct or circumstantial, to show that influence was not only present but that it was in fact exerted with respect to the making of the testament itself." 369 S.W.2d at 923.

Measured by the above rules, we must now consider whether the evidence in the record before us would sustain a jury finding of undue influence. Indulging in all inferences favorable to such a verdict, we find the following: For many years prior to the changes of beneficiary herein contention, Harold Bentley was psychologically disturbed and alcohol dependent. Psychiatric testimony established that he had a schizoid personality and paranoid trends and that he suffered from alcoholism. Familial instability and heavy drinking interfered with his on-the-job performance as an engineer with the Texas Water Rights Commission and eventually led to his dismissal on April 7, 1962. The record shows that from at least 1960 until Bentley's death in 1964, he was in a weakened mental and physical condition.

Bentley's friendship with the appellant, Mrs. Lyle, began in June of 1960 following Bentley's divorce from Nena Bentley. At the time Mrs. Lyle was single. During the next few years the relationship between Bentley and Mrs. Lyle grew progressively more intimate, and in fact there is evidence in the record that their relationship was sexually illicit. However, the intimacy of this relationship is further emphasized by the fact that Bentley manifested great affection for Mrs. Lyle and on several occasions asked her to marry him. Only Bentley's heavy drinking seems to have frustrated this possibility.

In June of 1962, Bentley became an out-patient at a hospital in Houston, where he participated in group therapy sessions at least once a day. He was at this point thoroughly estranged from his former wife and all of his children with the exception of his nine year old daughter. He remained under the care of hospital personnel until at least February of 1962, and it was during this period that Bentley executed the changes of beneficiary.

It appears that at this time Mrs. Lyle was residing in Austin and the record indicates that she visited the hospital only twice — once during the period of Bentley's hospitalization and once to take him home after he was discharged. Members of Mrs. Lyle's family also visited Bentley.

The record is singularly bare as to the circumstances surrounding the changes of beneficiary. There is no indication as to where the changes were made, whether in the hospital or outside it. One of the record forms changing the beneficiary which Bentley executed and sent to the Veterans Administration contains the signature of a Mr. Flippen as "witness," but who Mr. Flippen was or in what capacity he acted is again unclear. He never testified. The second change of beneficiary was witnessed by a Mr. Wilson. He also failed to testify.

On the other hand, the testimony of Mrs. Lyle that she did not even know she had been named beneficiary under Bentley's insurance policy until after the change had already been made remains uncontradicted. She stated that Bentley first discussed the change with her sometime in late November 1962, and prior to that she was not even aware of what he had done. The jury may have disbelieved this testimony, but there was no other testimony relative to the circumstances or preparation of the instruments in question.

Bentley was discharged from the hospital in Houston sometime between February and April of 1963. In the spring of 1964, he left for Mexico, where he died shortly thereafter. Before leaving this country Bentley executed a power of attorney in Mrs. Lyle's favor and authorized her to draw upon his checking account. His apparent purpose was to permit Mrs. Lyle to wind up certain unsettled business he had left behind. The power of attorney does not seem to have been used, and the checking account was drawn upon for only small sums to settle a few of Bentley's unpaid bills.

A fair summary of this evidence would appear to justify two elements of undue influence in favor of the jury's verdict. There is first of all ample evidence of Bentley's weakened mental and physical condition prior to and during the period when the changes of beneficiary were made. There is secondly a reasonable amount of evidence that Mrs. Lyle had the opportunity to exert undue influence on Bentley. But are these two elements enough? We think that standing alone they are insufficient to establish undue influence.

As previously stated, there is a complete absence of testimony either direct or circumstantial on the events leading up to and surrounding the execution of the changes of beneficiary. Moreover, there is ample testimony that Bentley was at odds with his wife and children and evidence of his affection for Mrs. Lyle. Such conditions do not justify, as appellee contends, a jury determination that the changes of beneficiary were themselves "unnatural." As said by the Court in Rothermel v. Duncan:

"* * * it is only where all reasonable explanation in affection for the devise is lacking that the trier of facts may take this circumstance [an unnatural disposition of property] as a badge of disorder or lapsed mentality or of its subjugation. See: Craycroft v. Crawford, Tex.Com.App. (1926), 285 S.W. 275 * * *." 369 S.W.2d at 924.

The fact that Bentley's affection for Mrs. Lyle may have existed within the context of an illicit relationship does not change this result. Affection for a paramour is not probative evidence of undue influence. Courts do not rewrite wills or redesignate beneficiaries in insurance policies in order to have them conform to "natural dispositions." Nor are courts empowered to impose their own moral standards or familial preferences upon litigants and to derive from them a test of undue influence.

In our view the evidence in this case which would support a jury verdict of undue influence merely "sets the stage" for such a finding, but is ultimately insufficient. We find the following language in Boyer v. Pool, 1955, 154 Tex. 586, 280 S.W.2d 564, corroborative of this position:

"There is testimony that the testator told his grandchildren (contestants) that he would provide for them while the will does not do that. The contestants established a confidential relationship, the opportunity, and a motive for undue influence and that during his last years the testator was virtually dependent upon his daughters, two of the principal beneficiaries. There is proof that the daughters made a daily charge for nursing their brother during his last last illness, which indicates a grasping attitude by the daughters and a harsh and unnattural family relationship. Proof of this type simply sets the stage. Contestants must go forward and prove in some fashion that the will as written resulted from the daughters substituting their mind and will for that of the testator. Here the will and the circumstances raise suspicion, but it does not supply proof of the vital facts of undue influence — the substitution of a plan of testamentary disposition by another as the will of the testator." 280 S.W.2d at 565.

In Boyer v. Pool, supra, the crucial failure of proof necessitating a reversal of the jury verdict as to undue influence was a failure to show the circumstances under which the will was executed. The Court in Boyer noted:

"In fact it is not shown that any of the children had any knowledge that the will was being executed until afterwards. There is no evidence of the direct intervention of the will and mind of the daughters or of the dominance of the free will of the testator in the preparation and execution of the will so as to bring this case within Long v. Long, 133 Tex. 96, 125 S.W.2d 1034." 280 S.W.2d at 566.

The Court in Boyer also noted that "Proof of the planning and preparation of the will, where the witnesses are available, is the heart of an undue influence case." 280 S.W.2d at 566. It may be that the mysterious Mr. Flippen and Mr. Wilson would have had no probative testimony to offer as to the circumstances surrounding Bentley's changes of beneficiary. But since they were the only known witnesses to the changes themselves and the record is otherwise bare of the circumstances in which the changes were executed, the absence of such witnesses cannot be entirely ignored.

On the other hand, the fact that we deal here with an insurance policy and not a will is of no great moment. Cf. Irons v. Smith, 4 Cir. 1933, 62 F.2d 644. As noted by the Court in Boyer, undue influence must be exercised at the time an instrument is executed, and this doctrine applies as much to deeds as to wills. Besteiro v. Besteiro, Tex.Com. App.1933, 65 S.W.2d 759, 761. This fact was recognized by the district court in this very case for in its instructions to the jury on the issue of undue influence, the Court properly noted:

"Improper influence cannot be inferred alone from opportunity, and there must be testimony, either direct or circumstantial, to show that undue influence not only existed, but that it was, in fact, exercised with respect to the making of the instrument itself."

Our careful examination of the record shows no evidence in support of this instruction.

We come finally to the case of Olds v. Traylor, Tex.Civ.App.1944, 180 S.W.2d 511 (writ ref.), recently cited with approval by the Texas Supreme Court in Rothermel v. Duncan, supra, and supportive of the view we take of the case before us. In Traylor, the evidence disclosed that the testator was old and ill, that he was irrational the day prior to the execution of the will, and that his mind and body had been impaired by the ravages of disease. His wife, the principal beneficiary under his will, had ample opportunity to influence the testator and the Court conceded for the sake of argument that the disposition of property may have been "unnatural." Despite these facts the Court concluded:

"These facts and circumstances, standing alone, are not sufficient to raise the issue of undue influence, in view of the record before us, because, in our opinion, there was no evidence, either direct or circumstantial, that any undue influence was exercised on the testator, or was operating on his will, at the time he executed the instrument under attack." 180 S.W.2d at 516.

In accord with the rationale of Olds v. Traylor, supra, we cannot permit a conjuration of inferences to substitute for evidentiary underpinnings. A mere suspicion or hunch that undue influence has been exerted is not enough. Here there is no indication that Mrs. Lyle took advantage of either Bentley's psychoses or his inebriation. There is instead the uncontradicted evidence that she was ignorant of the very document she is alleged to have subverted.

As we read the record before us, we do not perceive a case in which the excluded and the included claimants were competing for the mind, the largess and the affections of the deceased during his lifetime. Rather the very existence of domestic infelicity and post-marital bitterness between Bentley and his former wife argues against the possibility that Bentley's mind was directed contrary to its own inclinations. This is, after all, not a suit for alienation of affections. Therefore even if Mrs. Lyle is, because of her immoral conduct, a woman to be scorned, she does not thereby become a conqueror of Bentley's will or a despoiler of his dispositive volitions.

On the basis of the foregoing we conclude that where there is no evidence that undue influence was in fact exerted at the time the instrument was made, where the record is barren of any circumstances leading up to the preparation of such instrument, and where the disposition of property is not without all reasonable basis in affection, then a finding of undue influence is not permissible.

Reversed and remanded for a new trial.