(concurring).
I find this a difficult case to analyze though, on the facts, I concur in the result. The court dwells on the inadequacy of the psychiatrist’s and psychologist’s diagnoses. While in certain respects I think it is overly critical, I agree that the jury was entitled to be skeptical of opinions of insanity based upon relatively brief examinations made several months after the crime and at a time when Dube had everything to gain from a finding of insanity. There were, besides, indications from which a jury might wonder if the experts were confusing insanity in the criminal sense with a less fundamental disorder.
Still it is not simple to identify the affirmative evidence from which the jury could find defendant sane beyond a reasonable doubt. Certain conclusions, could, it is true, be drawn from Dube’s girl' friend’s description of his conduct *254before and after the crime. She had lived with Dube for several weeks and was in his company when he fled. While the defense argues that by selecting a bank to rob on the spur of the moment, Dube behaved in a bizarre manner, this behavior does not necessarily compel an inference of mental abnormality; and his conduct during and after the robbery, including precautions to avoid detection such as discarding the gun and driving to a city where he felt the police were less likely to be on the lookout, seems rational enough. The two tellers, who saw him briefly during the robbery, observed nothing bizarre, and the jury was able to add to this evidence its own observations of Dube while in the courtroom. Thus, there was evidence that Dube at certain times had behaved in a way which, to the average eyes, might seem normal. Still one wonders by what standard the fleeting glimpses of behavior transmitted by Dube’s girl friend and the tellers allowed a finding of sanity beyond a reasonable doubt.* Dr. Voss, the psychiatrist, testified that the girl friend’s version of Dube’s behavior was consistent with a diagnosis of schizophrenia. Whether or not that is so, it is questionable whether her association with Dube was extensive enough, and her behavioral testimony detailed enough, to permit a positive diagnosis of sanity either by a layman or an expert.
Yet not without some hesitation I think the jury was entitled to receive help from another quarter. In Davis v. United States, 160 U.S. 469, 16 S.Ct. 353, 40 L.Ed. 499 (1895), the Supreme Court did not characterize the presumption of sanity as belonging to that category of presumption which vanishes once the defense shows evidence of insanity. Instead, it stated,
“If the whole evidence, including that supplied by the presumption of sanity, does not exclude beyond a reasonable doubt the hypothesis of insanity, of which some proof is adduced, the accused is entitled to an acquittal. tt
160 U.S. at 488, 16 S.Ct. at 358 [Emphasis supplied].
Except for the quoted reference in Davis (which was the case that established the federal rule requiring the prosecution to prove sanity beyond a reasonable doubt) there has been little attention paid in federal cases to whether the presumption of sanity, once questioned, continues to have evidentiary force. Some courts, like the court here, see it as a presumption that evaporates once evidence of insanity is introduced. Yet viewed as a common sense inference that a person without marked symptoms to the contrary is likely to be sane, I think the presumption is entitled to be given reasonable weight in determining whether on all the evidence the Government gets to a jury.
Massachusetts courts have for years relied upon the presumption of sanity as sufficient to take a case to the jury notwithstanding an absence of affirmative evidence of sanity. See, e. g., Commonwealth v. Masskow, Mass., 290 N.E.2d 154, 159 (1972). But cf. Commonwealth v. Mutina, Mass., 323 N.E.2d 294, 297 n. 2 (1975) (questioning but not deciding *255the continued viability of this doctrine in light of In re Winship, 397 U.S. 358, 90 S.Ct. 1068, 25 L.Ed.2d 368 (1970)). Cases such as Winship and, most recently, Mullaney v. Wilbur, 421 U.S. 684, 95 S.Ct. 1881, 44 L.Ed.2d 508 (1975), call into question the use of devices designed to shift the Government’s burden of proof. Still the presumption of sanity in the limited sense suggested states a commonly perceived probability and can be distinguished from a mere contrivance to undercut the Government’s burden.
Moreover, the approach falls well short of the rule adopted by England, Canada and many states that insanity is an affirmative defense, the burden of proving which is on the defense. That rule, while different from that applied in federal prosecutions, was held constitutional in Leland v. Oregon, 343 U.S. 790, 72 S.Ct. 1002, 96 L.Ed. 1302 (1952), and Leland was defended as being still good law by Mr. Justice Rehnquist in his concurring opinion in Mullaney v. Wilbur, supra.
In the present case, given the evidence of an ability to function normally, and an absence of evidence of abnormal behavior, I think the jury could summons assistance from the inference, or presumption, that Dube was sane. Evidence bearing upon insanity has never been restricted to expert evidence. Conduct, lay observations and even lay opinions have traditionally been given much weight. 2 J. Wigmore, Evidence, §§ 227 et seq. (3d ed. 1940). And the jury could add to factors such as the reasonableness of Dube’s conduct before and after the crime and his apparent lack of any history of mental disturbance, an inference of sanity drawn from its common experience that most people (at least those without marked outward symptoms) are sane. With the aid of this inference it could -reach the conclusion that he was sane beyond a reasonable doubt. I recognize that this rationale is not without its difficulties, but it seems more satisfying than to pretend that the Government’s meager evidence of Dube’s conduct established, or could establish, by itself, much of anything.
Had there been somewhat less evidence of ordinary behavior, or slightly stronger evidence of abnormality, reversal might be in order. But without condoning the Government’s failure to call an expert or otherwise bolster its case I think the issue was properly submitted to the jury.
Our approach is not easily reconciled with that taken in Beltran v. United States, 302 F.2d 48 (1st Cir. 1962), in which Judge Aldrich wrote, 302 F.2d at 52,
“The introduction of evidence of insanity places a burden on the government of proving sanity beyond a reasonable doubt. . . This burden cannot be spirited away by the simple method proposed by the government of the court’s saying it does not believe the evidence, therefore there is no evidence, therefore there is no burden [S]uch thinking would render the whole principle meaningless. Rather, the record must be looked at as a whole, with thé burden on the government to overcome any reasonable doubt.”
Beltran was, however, on its facts a stronger case for the defense. The diagnosis of paranoid schizophrenia had been made at about the time of the crime and was subsequently confirmed. It was accepted by the court with respect to Beltran’s competency to stand trial for a different offense. Nonetheless, I think it fair to say that our approach in the present case is closer to Judge Magruder’s dissenting opinion in Beltran than to the court’s.