Larry Huskins was charged in eight counts with sex offenses against his three natural daughters, ages six, three, and two, and in a non jury trial was convicted on one charge and acquitted on the others. Sexual psychopathy proceedings were initiated, but a jury found the defendant was not a sexual psychopath. (Welf. & Inst. Code, § 5512.) Thereafter, defendant, whose only prior offense was a 1948 conviction for resorting (Mun. Code 41.07), was sentenced to state prison. At all times, both in court and out of court, Huskins vigorously denied the charges against him. The principal issue is whether his motion for a new trial because of newly discovered evidence should have been granted.
*861In April 1962, Huskins and his wife, the mother of his three children, had reconciled after a separation. Their children continued to live in separate foster homes, 6-year-old Cathy living with foster parents named Jeanette and William White.
On behalf of the prosecution Mrs. White testified that on two occasions, once in February and once in April, Huskins had taken Cathy out and returned her in a dishevelled and upset condition, after which the child complained of sexual mistreatment. On both occasions Mrs. White took the child to a doctor three days after the visits, and the doctor found inflammation of the vaginal orifice caused by some foreign object. Mrs. White also testified that Cathy had been playing with herself during this period, which she said, “is very common in foster children. ’ ’ Mr. White did not testify.
The child had great difficulty in qualifying to testify, and there were signs she had been coached.1 One assignment of error is that she was not competent to testify (Code Civ. Proc., § 1880, subd. 2; People v. Burton, 55 Cal.2d 328, 341-342 [11 Cal.Rptr. 65, 359 P.2d 433]; People v. Banos, 209 Cal.App.2d 754, 756-757 [26 Cal.Rptr. 127]), but the trier of fact properly qualified her to testify, and he could reasonably have inferred from the child’s testimony that her natural father had molested her during the April visit.
In his defense Husldns testified he never molested any of his children, that on the day in question he and his wife worked around the house while their children played. Huskins’ wife testified she had been with her husband and the children during the entire day of the visit and Cathy was undisturbed and in good physical condition when she left the house with her father to return to her foster home.
After the criminal conviction and during the sexual psychopathy proceedings, the defense discovered that Mrs. White had instituted civil commitment proceedings against her own husband in 1951, accusing him of being a sex pervert who had attacked his own daughter and had performed sex acts with animals. These accusations were never proved. Rather Mrs. White herself was found to be suffering from paranoid schizo*862phrenia, and after her attempt at suicide was committed to Camarillo State Hospital, where she remained for a year and a half. The defense moved for a new trial, contending this newly-discovered evidence cast serious doubt on the credibility and motives of Mrs. White, the main prosecution witness, and suggested either that Mrs. White was suffering from delusions and no crime had ever been committed, or that Mrs. White had fabricated the accusations and coached the child in order to keep the child in her family, or that another person, the foster father, might have molested the child.
Penal Code, section 1181, authorizes a new trial on the discovery of material evidence which a defendant could not with reasonable diligence have discovered and produced at the first trial. “ To entitle a party to a new trial on the ground of newly discovered evidence, it must appear, '1. That the evidence, and not merely its materiality, be newly discovered ; 2. That the evidence be not cumulative merely; 3. That it be such as to render a different result probable on a retrial of the cause; 4. That the party could not with reasonable diligence have discovered and produced it at the trial; and 5. That these facts be shown by the best evidence of which the case admits. ’ ” (People v. Sutton, 73 Cal. 243, 247-248 [15 P. 86].) Here, undoubtedly the evidence was newly discovered, was not cumulative, and was the best available evidence. Reasonable diligence is conceded, for it appears that no one connected with the case had any inkling that Mrs. White had made comparable charges in the past. Was a different result on retrial probable ? The trial court thought not and denied the motion for a new trial, stating it had tried the case without a jury and did not think a different result on retrial probable. However, the test is not a subjective one whether a particular trier of fact would be persuaded by the new evidence to reach a different conclusion, but rather is an objective one based on all the evidence, old and new, whether any second trier of fact, court or jury, would probably reach a different result. We therefore consider the evidence on the basis of objective probabilities.
Ordinarily, evidence which merely impeaches a witness is not significant enough to make a different result probable, and respondent relies on the rule that “. . . newly-discovered evidence which would tend merely to impeach a witness is not of itself sufficient ground for granting a new trial.” (People v. Long, 15 Cal.2d 590, 607-608 [103 P.2d 969].) But the new evidence in this ease does more than merely impeach *863the main prosecution witness—it tends to destroy her testimony by raising grave doubts about her veracity and credibility. Since Mrs. White was the sole adult witness connecting the defendant with the charged acts, her credibility is central to the proof of the crime. As the Supreme Court said in another child molestation matter, "In this type of case the broadest latitude in cross-examination and production of rebuttal evidence is required. . . . [E] vidence contradicting the testimony of a witness, even if it consists of proof of other wrongful acts, is proper if it is relevant to an issue in the case.” (People v. Clark, 63 Cal.2d 503, 504-505 [47 Cal.Rptr. 382, 407 P.2d 294].) The defense, because of lack of knowledge, never had an opportunity at the criminal trial to cross-examine Mrs. White about the charges she had made against her husband, her history of mental illness, and her commitment to a mental hospital, or to develop the theory that Mrs. White to serve her own purposes had concocted the charges against Huskins in order to keep his children in her family. Every experienced trial attorney knows the devastating effect which pertinent cross-examination on a vulnerable subject can produce on a witness. On occasion the skillful use against a fabricating witness of ammunition such as that newly-discovered here may even cause the witness to break down on the stand and admit perjury in open court. Conversely, a witness who stands up well against such an assault tends to bring strengthened credibility to his evidence-in-ehief. These are some of the factors Wigmore had in mind when he called cross-examination the greatest legal engine ever invented for the discovery of the truth. (Wigmore on Evidence, § 1367 [3d ed.].)
In People v. Williams, 57 Cal.2d 263 [18 Cal.Rptr. 729, 368 P.2d 353], evidence impeaching the testimony of the principal prosecution witness was discovered subsequent to the trial. That evidence consisted of affidavits of witnesses who quoted the principal prosecution witness as having said he was going to concoct a story of armed robbery for the purpose of fixing the defendant. The court granted a new trial, not only finding a different result on retrial probable but also stressing the duty of the courts to be particularly alert to review their processes wherever the possibility of false charges exists, While the impeaching evidence in the present case does not directly contradict testimony at the original trial, still, it bears an intimate relationship to the subject matter and raises the possibility of false accusation. In People v. Clark, 63 *864Cal.2d 503, 505 [47 Cal.Rptr. 382, 407 P.2d 294], the court reversed a child molestation conviction on the specific ground that the trial court had unduly restricted the cross-examination of the prosecuting witness and improperly refused to permit the use of defense witnesses to impeach her testimony on matters not directly related to the charge. A broad latitude in cross-examination was compelled, said the court, because of the defense contention that the prosecuting witness had a propensity to fabricate stories about sexual molestation.
We think it of utmost importance in this case for the trier of fact, in evaluating the credibility of Mrs. White’s current charges, to observe her demeanor and posture and reactions and behavior on the stand while she explains under cross-examination what brought about her prior unproved charges of child molestation against her husband. Her intimate involvement as accuser in two such matters could be the result of an unfortunate coincidence—or it could reflect a pattern of mind which predisposes her to jump to unwarranted conclusions on the subject. In our view this is one of those exceptional cases with unusual facts in which newly-discovered evidence impeaching the credibility of a prosecution witness makes a different result on retrial probable. (People v. Williams, 57 Cal.2d 263, 274-275 [18 Cal.Rptr. 729, 368 P.2d 353].) Our conclusion on the probability of a different result is fortified by the verdict of the jury in the sexual psychopathy trial, which, after hearing the evidence impeaching Mrs. White, concluded that defendant was not a sexual psychopath.
In denying the motion for a new trial and thereby determining the nonprejudicial effect of defendant’s ignorance of Mrs. White’s previous history, the trial court relied not only on the evidence in the case but also on the testimony of three psychiatrists at the sexual psychopathy trial as verification of the results of the criminal trial. The testimony of the psychiatrists illustrates the limitations of medical diagnoses for legal purposes and the dangers inherent in accepting legal conclusions from medical experts.
The first psychiatrist gave defendant a “very brief” examination at the time defendant was first admitted to Atascadero State Hospital and never saw Mm again. He did not attend the staff meeting which decided that defendant was a sexual psychopath, and he based his own conclusion that defendant was a sexual psychopath on a reading of other doctors’ reports, technicians’ reports, the probation report, *865and the transcript of the testimony at the defendant’s criminal trial. He considered defendant’s protestations of innocence as an indication of an introspective inverted type of thinking, and said the fact that the prosecution witness had made similar charges against other people and had herself been committed for mental illness would have no effect on his opinion that defendant was a sexual psychopath.
The second psychiatrist interviewed defendant for half an hour in the hospital unit of the Los Angeles County jail. He concluded that defendant was a sexual psychopath because he had been convicted of child molestation and stated that if he had not been so convicted he would not have classified him as a sexual psychopath. This psychiatrist confused his function with that of the court when he stated, “My conclusions would have to be, if he is guilty of the act, he will be considered a sexual psychopath.” Obviously, if everyone convicted of a sex offense were automatically classified as a sexual psychopath, there would be no need for a separate trial on that issue.
The third psychiatrist had read the transcript of the preliminary examination and interviewed defendant for an unspecified length of time. This psychiatrist found that Hus-kins was strongly inclined to religion and stated that this religious inclination “simply establishes the portion of the welfare and institutions code qualification of a sexual psychopath ; that is, he is either afflicted in some mental capacity or through a certain amount of mental aberration or mental retardation, something of that nature, predisposing him to the commission of sexual offenses. Q. Do you feel then that his strong religious feelings are some form of mental aberration? A. Yes, I think the manner in which he protested [his innocence] and the manner in which he expressed his religious ideas are not infrequently associated with a mental disorder. ’ ’ He, too, testified it would make no difference in his opinion what prior history of unproved accusations the principal witness had in her background.
The trial court relied in part on the testimony of these psychiatrists as evidence in support of defendant’s guilt which militated against the probability of a different result on retrial. The fact that the doctors’ testimony in turn was largely based on the outcome of the original trial was given no weight. The legal process thus found itself on a merry-go-round in which a new trial was denied because the psychiatrists were sure defendant was a sexual psychopath, and the *866psychiatrists were sure defendant was a sexual psychopath because he had been convicted in his original trial. In essence, the results of the trial were used to sustain the opinions of the psychiatrists, and the opinions of the psychiatrists were used to sustain the results of the trial. This circular reasoning created a vicious circle, which it was the duty of the trial court to break on the belated discovery of the shaky foundation on which the original structure was built.
We conclude that in a case of such slenderness as this the discovery of previously unknown evidence which would tend to destroy the credibility of the principal prosecution witness on cross-examination provides a solid basis for a new trial. (People v. Williams, 57 Cal.2d 263 [18 Cal.Rptr. 729, 368 P.2d 353].) The judgment is reversed.
Roth, P. J., and Herndon, J., concurred.