Cooper appeals from his conviction for robbery (18 U.S.C. Secs. 2113(a), 2113(d)). We reverse because the Government failed to carry its burden of proving beyond a reasonable doubt that Cooper was sane when he robbed the bank.

The Court appointed a psychiatrist, Dr. Polos, to examine Cooper to determine his mental competence to stand trial and his sanity when the offense was committed. In his initial report to the court, Dr. Polos concluded that Cooper was competent to stand trial. He further reported that it was difficult to decide whether Cooper was sane at the time of the offense without more testing and indepth psychiatric evaluation.1 At Dr. Polos' request, the United States Attorney's office gave him statements from Cooper's coworkers and from persons who observed him before and after commission of the crime, FBI reports and interviews, and medical reports. Cooper was referred to Dr. Paulson of the UCLA Neuropsychiatric Institute for testing to ascertain the degree of Cooper's mental illness and his intellectual level and to determine whether Cooper was lying or malingering. Dr. Polos reexamined Cooper after he had studied the reports of Dr. Paulson and the materials he had received from the United States Attorney. He made a second report in which he stated:

"In my report of June 18, 1970, I indicated that the only question I found impossible to answer at that time was his capacity to appreciate the nature of his conduct at the time of the crime. There is no doubt that he had been taking excessive amounts of Valium as well as suffering from an acute onset of Bell's palsy. The consumption of the tranquilizer Valium was primarily to control emotional symptoms that he was experiencing at that time and created drowsiness and confusion in him.

"Diagnostically Mr. Cooper has a chronic anxiety reaction with a moderately severe depression, as well as a schizoid personality.

"It is my psychiatric opinion that at the time of the crime he was under severe emotional stress as well as the effect of Valium, consequently his capacity function was impaired. His behavior was the result of an anxiety state rather than of a psychotic state. However, I do not believe that he was fully capable of controlling his impulses. I realize that this opinion is not exact or clearcut, but I do believe that Mr. Cooper definitely has a mental illness, and along with the excessive use of Valium caused him to act out in a manner that he ordinarily would not. Furthermore, I believe that he is treatable and should be placed in a psychiatric facility for complete rehabilitation. If this is possible, I believe he will be fully capable of eventually returning to society as a useful member rather than a severely and chronically mentally ill individual, or someone with bitterness and criminal tendencies."

Dr. Polos testified at the trial. In the course of his testimony, he said that Cooper had told him that he had been taking Valium which had been prescribed for him. The existence of the prescription "wasn't established as a fact until [Dr. Polos] got the reports from . . . the FBI concerning the two doctors that had given [Cooper] prescriptions [for Valium]." He stressed that his opinion about Cooper's mental condition at the time of the crime was not solely based on Cooper's ingesting Valium. "It's the pain and disturbance from [the paralysis of one half of the face] as well as his mental illness." He also testified that on the basis of his examination of Cooper and his evaluation of the psychological tests, he had "no doubt in his mind that [Cooper] was being honest and was not attempting to lie or distort any of the material that was asked of him."

Evidence in addition to Cooper's statements tended to substantiate Dr. Polos' view that Cooper had been taking large amounts of Valium. Cooper had access to Valium from two prescriptions, and he also had access to a friend's Valium. He was in pain from the palsy, and he had reason to seek relief using Valium. His employer saw Cooper taking pills the week before the robbery, and at that time Cooper did not appear to be wide awake. There was testimony that one of the effects of Valium was drowsiness. His employer, coworkers, and social friends testified to Cooper's abrupt personality change shortly before the robbery and to his peculiar behavior then.

The Government offered no expert testimony directed to the sanity issue. The prosecutor called two doctors, neither of whom was qualified to make a psychiatric diagnosis and neither of whom was asked to do so. Dr. Toomajian, a jail physician, saw Cooper for five minutes the day after the robbery. He diagnosed Cooper's facial paralysis as Bell's palsy and gave him some vitamins. Dr. Bruce, a general practitioner, had treated Cooper two years earlier for a cervical spine injury and had not thereafter examined him. He gave some opinions about the effects of excessive dosages of Valium. There was no substantial conflict between Dr. Polos' and Dr. Bruce's opinions about the effects of Valium.

The Government presented five lay witnesses who had seen Cooper the day of the robbery. The bank teller saw Cooper somewhere between three and five minutes. She testified that when he demanded the cash, he spoke "very slowly" and a "little mumbly." He appeared "overly calm." She did not notice any facial paralysis. Mrs. Alexson saw the robbery as she was waiting at the drivein window of the bank. She said he walked away from the teller's window after he received the money. As she drove out of the lot, she saw Cooper taking "real large steps, and running." Deputy Sheriff Leake stopped Cooper's car a few minutes after the robbery and arrested him. When Cooper got out of his car, Leake noticed that Cooper "appeared to be bent over to the right slightly and a slight shuffle or dragging of one leg." "[H]is eyes were not coordinated." Leake made no close observation of Cooper, but, he said, Cooper did not appear to him to be under the influence of drugs. Mr. Chance, who booked Cooper, did not notice anything unusual about him, except that there was something wrong with half of his face, "[l]ike maybe he had had a shot of Novocain or something that would have impaired" his face and "his right eye had difficulty closing." Cooper complained of pain, and Officer Chance gave him two aspirin tablets. He appeared alert to Officer Chance and Officer Chance made no effort to examine him for signs of drugs. Officer Chance had heard that Valium was a muscle relaxant; he otherwise knew nothing about the drug. Finally, the Government called the FBI agent who had interrogated Cooper about four hours after the robbery. He observed that Cooper "had difficulty in his speech," but "his speech was intelligible." He saw the paralysis of the right side of Cooper's face. Cooper was relaxed and friendly and did not appear to be drowsy.

We held in Buatte v. United States (9th Cir. 1964), 330 F.2d 342 (Buatte I) that when the defendant introduces expert psychiatric testimony that is sufficient to support a finding of insanity and when the Government relies entirely on lay witnesses to prove sanity, the Government has failed to carry its burden of proof as a matter of law.

To sustain Cooper's conviction, the Government must accomplish two feats: (1) It must find a way to destroy Dr. Polos' testimony to eliminate Buatte I, and (2) after razing Dr. Polos' testimony, it must find enough affirmative evidence of Cooper's sanity to prove that he was sane beyond a reasonable doubt. The Government did not succeed on either count.

The Government's arguments are directed to the first point. The Government contends that Dr. Polos' testimony can be disregarded because he relied, in part, on Cooper's taking large amounts of Valium. There was evidence, it says, from which the jury could have decided that Cooper did not take large amounts of Valium. Therefore, the jury could reject Dr. Polos' opinion, citing United States v. Ingman (9th Cir. 1970), 426 F.2d 973.2

The evidence that Cooper was taking excessive quantities of Valium came from several sources: Cooper's statement to Dr. Polos that he had done so; reports from the FBI stating the prescriptions (with prescription numbers) for Valium given to Cooper by two doctors, which reports were given to Dr. Polos by the FBI. There also was evidence that Cooper had access to a friend's supply of Valium. Independent tests of Cooper were made at Dr. Polos' request by Dr. Paulson, chief of the psychological outpatient department at U.C.L.A. and the Neuro-psychiatric Institute, to determine whether Cooper had told Dr. Polos the truth. Dr. Paulson reported to Dr. Polos that there was "no doubt that the defendant was being honest and was not attempting to lie or distort any of the material that was asked of him."

The Government's argument that Cooper was not taking large amounts of Valium rests on inferences that the Government draws from testimony by Dr. Bruce that one who took excessive amounts of Valium would be drowsy and would go to sleep and the testimony of lay witnesses who saw Cooper on the day of the robbery and who said that Cooper appeared alert, and not drowsy, when he robbed the bank. The Government argues that because Cooper was neither asleep nor drowsy, he could not have been taking large amounts of Valium as Dr. Polos believed.

To illustrate the problem and not to imply that the effect of Valium is medically similar to alcohol, we put the testimony hypothetically in the more familiar context of alcohol consumption. Dr. Polos testifies that Cooper had been drinking heavily, suffered from Bell's palsy, and had a schizoid personality. At the time of the robbery, the combination of his physical disease, mental illness, and over-consumption of alcohol rendered him incapable of controlling his impulses. Dr. Bruce testifies that people who consume excessive amounts of alcohol are not alert, become drowsy, go to sleep, and lose consciousness. Government lay witnesses testify that Cooper appeared alert, not sleepy, and was not unconscious. From this evidence, the Government argues, the jury could find that Cooper had not been drinking heavily, and, therefore, it could completely discredit Dr. Polos' opinion.

The Government's assumption in the hypothetical is that from proof that one who consumes excessive amounts of alcohol becomes drowsy, goes to sleep, and loses consciousness and from proof that the defendant did not exhibit those symptoms, it follows that defendant was not drinking heavily. The difficulty is that there is not sufficient information in the proof to permit the inference to be drawn. It is common knowledge that the effects of alcohol vary depending upon tolerance of the individual to alcohol, the amount consumed, the timing of the consumption, and other factors. It is also common knowledge that consumption of alcohol affects behavior before the consumer exhibits more acute symptoms.

In this case, neither Drs. Polos nor Bruce tried to define excessive consumption of Valium. Unlike alcohol, there is no common knowledge about the effects of Valium consumption. Dr. Bruce's testimony indicated that there were some variables in symptomology because he described pain relief, drowsiness, and sleep, obviously different conditions, but he did not explain the factors that controlled the incidence of any of them. Drs. Polos and Bruce did not disagree that at some point in time or quantity of ingestion the consumption of Valium would make the patient sleepy and he would go to sleep. However, there was no proof either about the timing of Cooper's Valium consumption or about the quantity he took. Without greater precision in the testimony, no inference can be drawn that Cooper did not take excessive amounts of Valium.

The Government contends that our conclusion that it did not escape Buatte I is in conflict with United States v. Ingman, supra, 426 F.2d 973. It is not. Ingman was charged with federal marihuana offenses. His primary deense was insanity. Two psychiatrists testified on Ingman's behalf that Ingman suffered an acute psychiatric episode following the taking of LSD. "[N]either psychiatrist based his opinion on objective symptoms . . . tests, or treatment. Rather, each relied on subjective symptoms revealed only through statements made to them by Ingman." (426 F.2d at 977.) The Government introduced no expert testimony. The lay witnesses testified that they did not notice anything unusual about him. Two of the witnesses who had seen many persons under the influence of narcotics testified that Ingman was not under such influence when they observed him at the time of the offenses. Another witness testified that Ingman had told him that he did not use "acid" (LSD). The sole basis of the psychiatrists' opinions was their belief in the truth of Ingman's statement that he had been taking LSD. The truth of that report was directly impeached. We reaffirm the principle stated in Ingman: "A jury may, of course, reject expert opinion if it finds that the opinion was based on an incorrect view of the facts."3

In our case, Cooper's taking large amounts of Valium was not impeached. Consumption of Valium was not the sole basis of Dr. Polos' opinion. Dr. Polos did not rely only upon Cooper's statements to him. Cooper was subjected to extensive observation, examination, tests, and treatment.

Even if we assumed, arguendo, that the Government avoided Buatte I, it did not produce enough evidence to carry its burden of proving that Cooper was sane beyond a reasonable doubt.4 Cooper's lay witnesses alone amply eradicated the presumption of sanity. All of Cooper's lay witnesses attested to his erratic and peculiar behavior shortly before the robbery.

The Government cannot carry its burden by attacking the strength of Cooper's case. To meet its burden, it must point to affirmative evidence which is adequate to prove sanity without the benefit of a sanity presumption. We thus ignore the testimony of the Government's witnesses that Cooper did not appear to be under the influence of drugs. Proof of nondrug use does not prove sanity. All of the Government's lay witnesses, except Mrs. Alexson, who saw Cooper only at a distance, noticed that something was wrong with him. The affirmative evidence is limited to this: Cooper could go through the motions to rob a bank, he could run, he appeared alert, his speech was intelligible, and he was relaxed and friendly. Grossly psychotic patients can perform all of these acts and maintain all of these appearances. Although this testimony may constitute some evidence bearing on sanity, it cannot be escalated to proof beyond a reasonable doubt that Cooper was sane.

Long before this case was tried the Government was aware that the sanity issue would be presented, and it had ample time within which to prepare and present further evidence on the issue if any were available. The record appears to have been developed as thoroughly as the Government was able to do. Under these circumstances we decline to remand for a new trial.

The judgment is reversed with directions to dismiss the indictment.

I respectfully dissent. I have no quarrel with the proposition that the burden of proving Cooper's sanity rested with the prosecution. The evidence for the defense successfully dispelled the presumption of sanity. The prosecution must prove a defendant's sanity beyond a reasonable doubt, once the defense presents enough evidence of insanity to raise the issue. Davis v. United States, 165 U.S. 373, 17 S. Ct. 360, 41 L. Ed. 750 (1897). The meaning of that well-settled abstract principle must necessarily evolve from its application to the facts of specific cases. After carefully examining the trial transcript, I find myself unable to agree with the majority's assessment of the evidence on the insanity issue.

The majority interprets Buatte v. United States, 330 F.2d 342 (9th Cir. 1964) (Buatte I) to have held

"that when the defendant introduces expert psychiatric testimony that is sufficient to support a finding of insanity and when the Government relies entirely on lay witnesses to prove sanity, the Government has failed to carry its burden of proof as a matter of law." [Majority opinion, p. 453.]

This is an incorrect statement of Buatte I. No rule of law requires the government to offer expert medical testimony in every case in which a defendant has a psychiatrist testify for him. The need for expert testimony depends upon the facts of the case. United States v. Ingman,1 426 F.2d 973 (9th Cir. 1970); Mims v. United States, 375 F.2d 135 (5th Cir. 1967); Dusky v. United States, 295 F.2d 743 (8th Cir. 1961), cert. denied 368 U.S. 998, 82 S. Ct. 625, 7 L. Ed. 2d 536 (1962). As United States v. Ingman, supra, makes evident, Buatte I held only that the prosecutor's lay evidence failed, on that record, to produce sufficient evidence to support a jury finding of sanity. It did not hold that the prosecution must produce an expert whenever one testifies for the defense.

United States v. Handy, 454 F.2d 885 (9th Cir. 1971), cert. denied - U.S. -, 93 S. Ct. 49, 34 L. Ed. 2d 86 (1972), does not hold to the contrary. In Handy, Judge Weick wrote, "While the testimony of lay witnesses, standing alone, cannot rebut the testimony of Handy's psychiatrists and psychologist, Buatte v. United States, 330 F.2d 342 (9th Cir. 1964), it is important evidence . . . ." 454 F.2d at 888. As the citation to Buatte indicates, Judge Weick meant only that on the record before the court the lay testimony in that case could not alone adequately answer the expert opinion offered by the defendant.

A broader reading of Judge Weick's language would put his opinion in conflict with the Ingman precedent. And, if read to mean that the prosecution must always meet expert testimony with expert testimony, the language is merely dictum, since the government in Handy had presented expert psychiatric evidence.

The question before us, therefore, is simply whether the government presented sufficient evidence to take the sanity issue to the jury. To answer it, we must view the evidence in the light most favorable to the government. United States v. Handy, supra at 888. The point bears emphasis. This court does not sit to make its own judgments upon a defendant's sanity. Our function is limited to determining whether, as a matter of law, the prosecution presented enough evidence of sanity to raise a jury question. A motion for acquittal should be granted only if any reasonable man would necessarily retain a reasonable doubt about the defendant's sanity at the time of the offense. United States v. Julian, 440 F.2d 779 (9th Cir. 1971).

The defense did not claim that Cooper suffered from a mental illness sufficiently severe to satisfy the test for legal insanity. Instead, it sought to develop the theory that a number of factors, each insufficient in itself, combined on the day of the bank robbery to rob Cooper temporarily of the capacity to control his impulses. Under the standard adopted in Wade v. United States, 426 F.2d 64 (9th Cir. 1970) (en banc), inability to conform his conduct to the requirements of law would render Cooper legally insane.

The defense claimed that Cooper suffered from a moderately severe neurosis, a painful organic disease called Bell's palsy, and the toxic effects of a sustained overdose of a tranquilizer, Valium.

As the majority points out, Cooper's employer, co-workers, and social friends testified to Cooper's abrupt personality change shortly before the robbery and to his peculiar behavior then. The defense also presented the medical testimony of Dr. Theodore Polos, a well qualified and experienced psychiatrist.

I have no argument with the majority's assertion that Dr. Polos' testimony adequately dispelled the presumption of sanity, and I also agree with their reaffirmation of Ingman: "A jury may, of course, reject expert opinion if it finds that the opinion was based on an incorrect view of the facts." United States v. Ingman, supra, 426 F.2d at 977. An expert's opinion may be disputed on either of two levels. One may dispute his version of the facts supporting the diagnosis, or one may disagree with his medical reasoning and conclusions from those facts. Certainly the prosecution would need expert evidence if it were to challenge the medical reasoning or conclusions.

If, however, the prosecution attempts to prove that the expert opinion is premised upon false factual assumptions about the defendant's behavior, then in some cases at least no expert opinion will be needed. The government in this case chose the latter course.

In order to sustain Cooper's conviction, the government must: (1) demonstrate enough affirmative evidence to support a jury finding that Dr. Polos' factual assumptions were incorrect; and (2) demonstrate enough affirmative evidence of Cooper's sanity to prove that he was sane beyond a reasonable doubt. The government has met its burden on both counts.

Dr. Polos first examined Cooper, pursuant to a district court request, to determine whether he was mentally competent to stand trial. After spending about an hour with Cooper, Dr. Polos reported to the court:

"It is my psychiatric opinion that the defendant is presently sane. . . . . The prognosis of his current mental competency is good.

"The question as to whether the defendant was legally insane at the time of the commission of the offense charged against him is difficult to answer at this time. It is my impression that there is a good chance had the defendant been taking as much as 100 to 150 milligrams of Valium per day he could have well have had a toxic reaction that would have rendered him incapable of having the capacity to appreciate the criminality of his conduct at the time of the crime."

At Dr. Polos' request, Dr. Morris Paulson, a qualified psychologist, administered a psychological testing battery to Cooper. The prosecution made available to Dr. Polos a number of F.B.I. reports on interviews with people who had seen Cooper both before and after the robbery. Dr. Polos also interviewed Cooper a second time.

On the basis of this information, Dr. Polos concluded that Cooper had, on the day of the robbery, temporarily lost his capacity for self-control:

"It is my psychiatric opinion that at the time of the crime he was under severe emotional stress as well as the effect of Valium, consequently his capacity function was impaired. His behavior was the result of an anxiety state rather than of a psychotic state. However, I do not believe that he was fully capable of controlling his impulses. I realize that this opinion is not exact or clear cut, but I do believe that Mr. Cooper definitely has a mental illness, and along with the excessive use of Valium caused him to act out in a manner that he ordinarily would not."

Dr. Polos thought that three separate and indispensable elements coalesced to make Cooper temporarily insane:

"I would like to clarify one point. It isn't the Valium alone. If you and I or most of us here took Valium and took an overdose, it isn't the Valium alone that is going to make him go out and try suicide or do something self-destructive or rob a bank. It isn't the neurosis alone that is going to do it. And it isn't the fact that he had a very painful organic or probably a virus infection in his nerve and the paralysis and the pain and-

"But I think it is a combination of all of these things." (emphasis supplied)

The gist of Cooper's story was that he had taken large doses of the tranquilizing drug Valium during the weeks prior to the robbery, as well as on the robbery day itself. He took it to alleviate the pain associated with his attack of Bell's palsy. The Valium, he said, was left over from prescriptions given him by two doctors.

This story came into evidence through the testimony of Dr. Polos, who related what Cooper had told him. Cooper himself did not take the witness stand, so the jury had no opportunity to evaluate his credibility. Dr. Polos thought Cooper told the truth, but the court and the jury need not have agreed.

The support for Cooper's story came primarily from his narrative to Dr. Polos, plus some ambiguous circumstantial evidence. No one ever saw Cooper taking Valium, although his girl friend had been with him on the night before and the morning of the robbery. She was not called as a witness. Cooper had no Valium when arrested, and F.B.I. agents were unable to discover any that had belonged to him. Cooper's attorney told Dr. Polos that Cooper had access to a friend's supply of Valium, but this was unsubstantiated by any admissible evidence.

Two physicians supposedly supplied Cooper with prescriptions for Valium. One of these, Dr. Bruce, testified that, although he treated Cooper for a back injury, he had never prescribed Valium for him. The other doctor's prescription had been written more than a year before the robbery. Cooper told Dr. Polos he had obtained an illegal refill of the prescriptions, but we have only that hearsay evidence, Cooper's unsworn word, in support of this.

The defense case showing Cooper's possession of a Valium supply was weak and rested almost entirely on the defendant's own tale to the doctor. The explanation of the source of supply was implausible.

Even assuming, arguendo, that Cooper's evidence showed he had access to some amount of Valium, only his own completely unsupported statements provided evidence of any consumption on the day of the crime. Dr. Polos opined that a toxic reaction would likely have occurred if Cooper had taken 100 to 150 milligrams of Valium per day. Only Cooper himself, through Dr. Polos, produced any account of his consumption.2

* * *

The prosecution challenged Cooper's defense by seeking to prove that his actions during and after the robbery were inconsistent with the behavior of a man suffering from a toxic reaction to Valium. Both sides offered expert opinion about the side effects of an excessive Valium intake.

Dr. Bruce opined that after a 100 milligram dosage a person would not be able to drive and would probably fall asleep within an hour and a half. Drowsiness and confusion, he said, are the toxic effects.

Dr. Bruce admittedly did not have experience or information concerning patients who had taken large doses over an extended period of time. Since Cooper claimed to have taken an overdose for several weeks, Dr. Bruce's testimony alone does not adequately predict the behavior pattern Cooper should have displayed.

Dr. Polos filled in this gap.3 He explained the tolerance effect of a prolonged overdose. He agreed with Dr. Bruce that confusion and extreme drowsiness are the toxic effects of a dosage above 40 milligrams a day. He thought a patient might be able to develop enough tolerance to drive an automobile after 100 milligrams. But he said that one would have difficulty in running quickly; the ability to run at all would be impaired.

Finally, in his report, Dr. Polos said the Valium intake produced drowsiness and confusion in Cooper. This is the most relevant statement about side effects, since it is directed specifically to Cooper's condition.

The prosecution presented five witnesses who observed Cooper on the day of the robbery. Not one recalled seeing in him the least sign of drowsiness or confusion.

The bank teller related Cooper's words to her:

"I said, 'May I help you, sir?'

"And he said, 'Well, yes,' and he started to open-put his hand under his shirt . . .

"He reached up and then he went down and he come out with a gun and pointed it at me and he said, 'Give me all your paper money. Don't make any noises, don't press any buttons. Give me five minutes to get out or somebody will get hurt."'

The government argued that the coherence of this robbery demand indicated a lack of Valium influence. The teller thought Cooper seemed unusually slow moving and relaxed, but she said nothing indicating he was confused or drowsy.

The bank teller's testimony arguably could be consistent with a toxic Valium reaction, but that of the next witness was not. Mrs. Alexson, a bank customer, saw Cooper making his getaway. She saw him run rapidly for a distance of 200-250 feet, taking "real large steps." Cooper's own expert had said that a man suffering a toxic reaction to Valium would show impaired running ability.

Cooper was apprehended a few minutes later. The arresting officer did not observe him closely, but did not believe that Cooper was intoxicated or under the influence of drugs.

The booking officer, Chance, spent about 25 minutes with Cooper. He saw nothing unusual or abnormal in him, except the partial facial paralysis from the palsy. Chance said he was alert throughout the booking process:

"For one thing, when I advised him of his rights he stated that he wanted an attorney. For another thing, when I asked him questions to fill the booking slip [name, age, height, weight, address, place of employment, birth date, birthplace, who his employer was, occupation, social security number, special medical problems, next of kin] he answered them coherently. There was no pause where he would have to search for an answer or search for a reason."

Tyrone Miller, an F.B.I. agent, saw Cooper for one and a half to two hours. By the time Miller departed, nearly four hours had elapsed since the robbery. Cooper still showed no signs of confusion or drowsiness. Miller said he spoke coherently and intelligibly, and did not appear drowsy.

Although the jail medical staff examined Cooper's palsy, Cooper did not ask for Valium or any other pain killer. He said nothing to them about having taken Valium.

Two other witnesses said that Cooper had appeared depressed and disoriented in the weeks before the crime. His job performance declined markedly. The prosecution argued that these changes could easily have been caused by the pain and depression occasioned by the palsy attack. Moreover, even if one believes that this evidence establishes excessive Valium consumption beyond the possibility of rational dispute, it still says nothing about his condition at the time of the robbery.

To grant the motion for acquittal, the trial judge would have had to conclude that the evidence was such that any reasonable man would necessarily retain a reasonable doubt about each of the following propositions: (1) Cooper possessed a large supply of Valium; (2) he took 100 to 150 milligrams of Valium for several weeks prior to the robbery, thereby building up a tolerance; (3) he had taken a Valium overdose on the morning of the robbery; and (4) on that occasion the Valium had interacted with his underlying neurosis and his organic pain to deprive him temporarily of the capacity to resist his impulse to rob the bank.

Support for those assertions cannot be found in this record. We are not asking whether the jury could have believed Cooper's story or whether we believe it. The correct standard is whether, viewing the evidence in the light most favorable to the government, a reasonable man would necessarily have a reasonable doubt about Cooper's sanity at the time of the crime. Cooper's defense rested upon his unsworn narrative to the doctor and upon some ambiguous circumstantial evidence from his acquaintances' testimony about his behavior preceding the robbery.

The government's evidence, at the very least, created a jury question about the truth of Cooper's tale. The prosecution did a competent job of proving that Cooper did not behave as one would if suffering from a toxic reaction to Valium. Under our precedent in United States v. Ingman, supra, that proof sufficed to take the sanity issue to the jury.

The majority asserts that, even assuming Dr. Polos' testimony to be discredited, the government did not produce enough evidence to carry its burden of proving that Cooper was sane beyond a reasonable doubt. I respectfully disagree.

When Dr. Polos' testimony is taken out of the case, Cooper's evidence of insanity consists solely of testimony by his employer, co-workers, and social friends to the effect that Cooper underwent an abrupt personality change shortly before the robbery. Assuming, arguendo, that this testimony is sufficient to rebut the presumption of insanity, we still must recognize that

"The quantum and nature of proof the Government must offer to take the case to a jury varies in different situations and to some degree depends upon the quantum and nature of proof the defendant offers."4

Evidence of an abrupt personality change, coupled with peculiar behavior around the time of the robbery, is certainly flimsy evidence of insanity. To controvert it, the government offered the following evidence: Cooper made a coherent robbery demand at the bank; he appeared alert at the booking process; when advised of his rights, he stated that he wanted an attorney; he answered all questions on the booking slip concerning his background coherently ["[t]here was no pause where he would have to search for an answer or search for a reason,"]; he spoke coherently and intelligently to an F.B.I. officer some two and one-half hours after the robbery.

Based on this evidence the jury, after being properly instructed on the law regarding insanity, found Cooper to have been sane at the time he robbed the bank. The majority now holds that on this evidence a reasonable man must necessarily retain a reasonable doubt about Cooper's insanity at the time of the offense. I strenuously disagree. The jury saw the witnesses, evaluated their testimony, and determined beyond a reasonable doubt that Cooper was sane. The evidence supports that determination, and we should not disturb it on appeal.

The logical consequence of the majority position is that, as soon as the defense raises a doubt about the defendant's sanity, no matter how flimsy that doubt may be, the government is obligated to come forward with expert psychiatric evidence. That is not the law in our circuit, nor should it be. I would affirm.