delivered the opinion of the court:
This is an action to set aside a document dated January 10, 1950, which had been admitted to probate as the last will and testament of Charles Avery Hilliard. The plaintiff is Marian Heideman, the only child and sole heir of the testator, who was bequeathed one-sixth of the testator’s personal estate under the will. The defendants are Willard V. Kelsey, the executor, and Frank Wyman Hilliard, Leah Hilliard and Gladys H. Gates, the surviving brother and sisters of the testator, who are also beneficiaries under his will. The jury found that the contested document was not the testator’s will and the trial court entered a decree on the verdict. The defendants have appealed directly to this court. The testator owned a farm which he devised to his brother, and a freehold is involved.
*260On two previous occasions, decrees for the plaintiff in this case have been reversed and the cause has been remanded for a new trial. (Heideman v. Kelsey, 414 Ill. 453; 7 Ill.2d 601.) Originally the plaintiff charged undue influence and lack of testamentary capacity. At the first trial, a verdict was directed for the defendants on the issue of undue influence, and only the issue of testamentary capacity was submitted to the Jury, which found for the plaintiff.
On appeal from the decree entered on that verdict, we held that the testimony of Dr. Clark, one of the witnesses for the plaintiff, warranted a submission of the cause to the jury but that the trial court erred in denying the defendants’ motion for a new trial. (414 Ill. 453.) After a detailed review of the evidence presented on both sides, we concluded, “It is our opinion that the proof in this case demonstrates that Hilliard was a person with adequate testamentary capacity on January 10, 1950, and the jury’s verdict finding to the contrary is against the manifest weight of the evidence.” 414 Ill. at 466.
At the second trial the jury was unable to agree. At the third trial the jury again found for the plaintiff. On appeal we again reversed and remanded for a new trial. (7 Ill.2d 601.) The basis of this reversal was the admission of testimony of a handwriting expert to the effect that the testator’s signature on the contested document had not been executed on January 10, 1950. We held that this testimony raised an irrelevant and highly prejudicial issue not encompassed by the pleadings, which had presented only the issue of testamentary capacity. Although we found it unnecessary to set forth all the evidence presented at the third trial in view of this error, we noted that both sides relied on the same testimony used in the first trial and also produced six additional witnesses apiece, and we observed that the “contestant’s position is not greatly improved by their added proof.” 7 Ill.2d 601, 604.
*261Thereafter the plaintiff was given leave to file an amended complaint which included a new count alleging that the contested document was not executed on January 10, 1950, but instead was signed by the testator after April 1, 1950, at which time the testator was insane and without testamentary capacity. The defendants’ answer denied the allegations concerning the date of execution of the will, but admitted that “sometime after March 1, 1950, and before some time in April, 1950, the mental power, capacity and ability of said deceased became impaired as alleged.”
At the trial now before us the issue raised by the new count and the issue of testamentary capacity were submitted to the jury which returned a general verdict for the plaintiff. In addition, in response to three special interrogatories, the jury found that Charles Avery Hilliard was not possessed of sufficient mental capacity to transact ordinary business on January 10, 1950, and that the document offered as his last will and testament was not executed by him on that date, but instead was signed by Hilliard after April 16, 1950, at a time when it was conceded he was without testamentary capacity.
The testimony of most of the plaintiff’s witnesses who testified at the first trial on the issue of testamentary capacity was put before the jury at the present trial, either from the witness stand or by stipulation that their testimony might be read to the jury. Their earlier testimony is summarized in 414 Ill. at 456-459. The plaintiff also offered the testimony of five new witnesses on this issue. Several of them added nothing of substance to the plaintiff’s case.
Dr. Joseph Ernst examined the testator on October n, 1949, when he was admitted to Deaconess Hospital for treatment of persistent constipation. He had no present recollection of the occurrence, but from the record that he then made he testified on direct examination that the testator was vague as to his past medical history. On cross-*262examination he testified that the testator was alert and appeared to be a sane man. Dr. Claude Pickrell’s testimony at the second and third trials was read to the jury by stipulation. He admitted the testator to Deaconess Hospital in October, 1949, and had also spoken with him several times around January 31, 1950, about the advisability of a hernia operation. He was of the opinion, based on these conversations, that the testator was of sound mind and memory on January 31, 1950. Ben Maride, the testator’s barber, stated his opinion, based on his observation of the testator’s declining activity and increasing reticence, that the testator was not “of as sound a mind” in 1949 as he was previously. Anne Albro, the plaintiff’s only child, testified at length as to the relationship that existed between the testator, her mother and herself, which was one of normal affection. She also testified that she visited the testator in 1946 and that when she next returned from her home in Seattle and visited him in June of 1949 he was more feeble and aged, thin and more stooped, red-eyed and haggard. He did not recognize her 3 J4-year-old son, whom he had not seen for three years. She “didn’t think that he was of sound mind and memory that day.”
More important is the testimony of Dr. Groves B. Smith, a physician and surgeon specializing in nervous and mental diseases. In response to a hypothetical question the doctor gave his opinion that the man described in the hypothetical question had advanced arteriosclerotic degeneration, both physical and mental, and was not of sound and disposing mind and memory on January 10, 1950. His testimony was in some measure based upon the behavior and medical condition of the hypothetical man, particularly his longstanding and advanced arteriosclerosis. His opinion was also influenced, however, by his conclusion that the man who signed his-name in a cramped and jerky manner to the contested document of January 10, 1950, could not have been of sound mind.
*263This new testimony does not strengthen plaintiff’s case upon the issue of testamentary capacity. It was Dr. Ernst’s opinion that the testator was alert and was a sane man in October of 1949. Dr. Pickrell testified that the testator was of sound mind on January 31, 1950. The factual foundations upon which Maride and Anne Albro based their contrary opinions were meager, and so their opinions were of little probative value. Indeed, Maricle’s conclusion that in 1949 the testator was not “of as sound a mind as he was before” was of no value in determining the testator’s capacity. Dr. Smith testified on cross-examination that he regarded himself as an advocate. His conclusions were based upon a hypothetical question that omitted many facts that had already been developed, as well as uncontroverted facts later developed by the defendants’ evidence which tended to establish the testator’s testamentary capacity during the period in which he signed the will. And his assertion that the testator’s mental capacity could be determined by an examination of his signature on the contested document must be weighed against the subsequent testimony of Dr. Val Satterfield, a distinguished psychiatrist, that there is no relation, in medicine and psychiatry, between the handwriting of a person and his mental condition.
Dr. Clark was the witness whose testimony at the first trial was held to be sufficient, standing alone, to prevent a directed verdict for defendant. His testimony then is difficult to reconcile with his testimony at the present trial. At the first trial he testified that the testator suffered from brain disease caused by arteriosclerosis, or hardening of the arteries. At the present trial he testified that there had been a change in medical opinion during the intervening years; and although he testified that the testator “had a mental condition,” he was unable to say that it was caused by arteriosclerosis. At the first trial he testified that the testator did not have senile dementia. (414 Ill. at 456.) At the present trial he testified that in 1948 the testator *264suffered from early senile dementia. On cross-examination he enumerated the “classical symptoms” of senile dementia, but he was not able to associate these classical symptoms with significant conduct of the testator.
For reasons that will be stated, it is not necessary to determine whether or not the testimony of Dr. Smith and Dr. Clark at the present trial was sufficient to establish a prima facie case for the plaintiff on the issue of testamentary capacity.
While the plaintiff did not notably strengthen her case in this retrial of the testamentary capacity issue, the defendants did. They introduced in evidence additional letters regarding his will which the testator wrote to his lawyer during the latter part of 1949 and in January, 1950. These letters, with those previously introduced at the earlier trial, show that the testator at that time was able to and did in fact deliberate at length and with understanding upon the problems of drafting a will. The defendants also offered testimony of the lawyer’s secretary to show that the contested document incorporated the plan developed in this correspondence. In addition, they offered additional testimony tending to show that the testator retained the capacity to understand and manage his business affairs when the will was signed. Although we are reluctant to disturb findings in which three juries have concurred, we must conclude that the jury’s determination that the testator lacked testamentary capacity is against the manifest weight of the evidence and can not stand.
The only evidence supporting the plaintiff’s contention that the contested document was not signed on January 10, 1950, but rather at some time after April 1, 1950, when the testator was admittedly incompetent, is the testimony of George Swett, a handwriting expert, and of Dr. Smith. Swett testified that in his opinion the testator’s signatures on page one and page two of the contested document were executed at a time closer to April 25, 1950, *265than to January 10, 1950. He reached this conclusion (1) by a comparison of the signature on page one, first with seven checks written by the testator in January and then with a letter written on April 25 by the testator, and (2) by a comparison of the signature on the second page with the letter of April 25 and with the signature on the first page. Dr. Smith testified that from an inspection of signatures he was able to state that the signatures on the will were made by a person who was physically and mentally incapable of writing a legible signature, and by one who was not competent to understand what he was signing. He also testified that in his opinion the signatures on the will were placed there at a time much closer to April 25 than to January 10.
In Jones v. Jones, 406 Ill. 448, 450, this court held that expert testimony can not be allowed to prevail over the uncontradicted and unimpeached testimony of two disinterested attesting witnesses. This rule of priority for one form of testimony over another has been criticized. (McCormick on Evidence, sec. 173, n. 15.) It is unnecessary to reiterate that broad holding in this case. The propositions advanced by the witnesses are sufficiently startling to warrant scrutiny of the methods they employed, and their testimony can not survive that scrutiny.
The choice of signature specimens to be used for comparison was made by Swett. He used only seven out of 28 checks written by the testator between October, 1949, and February, 1950, which the plaintiff had submitted to him and which were placed in evidence. In arriving at his opinion he took no account of other uncontroverted samples of the testator’s handwriting executed during the period in controversy. If his theory was one of progressive deterioration in handwriting, as it seems to have been, a fair analysis of those earlier writings which showed characteristics similar to those manifested in the letter of April 25 was obviously required. A conclusion based on only a portion of the speci*266mens available is at best meaningless and at worst highly prejudicial. We conclude that Swett’s testimony was not competent evidence and was improperly admitted.
What we have said applies equally to the testimony of Dr. Smith that the contested document was executed nearer to April 25 than to January 10. He, too, failed to take into account handwriting specimens that were inconsistent with his hypothesis. He advanced no medical authority to support his proposition that it is possible to determine the mental competence of one who signs a will, or the date of its execution, by inspecting comparative signatures years after the event. Without the testimony of Swett and Dr. Smith, there was no evidence to support the plaintiff’s charge that the contested document was not executed on January 10, 1950, and a verdict should have been directed for the defendants on that issue.
The defendants urge us to reverse without remanding for a new trial. We rejected the same contention on review of the first trial on the ground that Dr. Clark’s testimony established a prima facie case which warranted submission to the jury. Dr. Clark’s testimony at the present trial is difficult to reconcile with his earlier testimony, and he seems to have disclaimed his earlier opinion as to the testator’s lack of ability to handle his business affairs when he testified, “I do not know of any basis for me to determine how he was doing with respect to his business affairs.” His testimony alone is no longer sufficient, but even if it could be regarded, when coupled with that of Dr. Smith, as establishing a prima facie case, we are of the opinion that the cause should not be remanded.
We are reviewing not the first trial, but the fourth. To be sure, this fact has made us increasingly reluctant to set our view of the evidence against that of three juries. But it also makes us increasingly reluctant to remand the cause for another trial with the prospect that we will again be called upon to reverse a jury finding and remand for *267another trial. “This court recognizes the tendency of juries to look for an excuse to hold invalid a will making an unequal division of the testator’s property among his children. (Turnball v. Butterfield, 304 Ill. 454; Freeman v. Easly, 117 Ill. 317.)” DeMarco v. McGill, 402 Ill. 46, 51.
The plaintiff has had four trials and over eight years in which to produce a convincing case. The resourcefulness of her attorneys in exploring new theories has produced only the unsatisfactory expert testimony of Dr. Smith and of Swett. She has failed to meet or surmount the defendants’ strong evidence which shows that the testator could and did handle business affairs and plan the disposition of this property when his will was executed.
This court faced a similar situation in Graham v. Deuterman, 244 Ill. 124. Three juries in three successive trials had there found that the contested document was not the testator’s last will and testament because he lacked testamentary capacity. This court reversed decrees entered on the first two verdicts and remanded for new trials on the grounds that the verdicts were against the weight of the evidence. The plaintiff’s evidence consisted primarily of nonexpert opinions that the testator was not of sound mind and could not transact ordinary business affairs. At the third trial, this evidence was repeated along with new testimony of a doctor, who had seen the testator frequently and had also treated him professionally eight years before his death, that the testator was suffering from senile dementia, or softening of the brain due to old age. After considering all the evidence, the court concluded that the plaintiff could not be expected to produce any substantial new evidence and that the interests of neither party would be served by again remanding the cause for another new trial.
The same is true of the case now before us. It has been demonstrated that the plaintiff cannot produce sufficient evidence to sustain her claim that the testator lacked testa*268mentary capacity, and there is no competent evidence in the record supporting her claim that the contested document was not signed on January 10, 1950, but rather at a later date when the testator was admittedly insane. Another trial would be fruitless and a waste of the court’s resources as well as those of the parties. The decree will therefore be reversed.
Decree reversed.