(dissenting):
Levin’s conviction of grand larceny was substantially based on testimony of James Landriscina that he met Levin in Washington, D. C., at 5:00 p. m., Friday, February 13, 1959, and gave him $25,-000.1 Levin testified that he was in Washington that day but that he left in the early afternoon. Levin attempted to support his alibi by establishing his religious habit of always being home with his family in New York on the Jewish Sabbath — sundown Friday through sundown Saturday.
The majority here accepts the rule generally followed by courts and commentators that habit evidence, if specific enough, may be admitted. Such evidence is logically probative and no rule excludes it. Whether its probative value is worth the excursion into collateral issues that proof of the habit may require depends on the particular facts of a case. Such questions are generally confided to the trial court’s discretion. In this case, the trial court apparently had in mind neither the probative value of habit evidence nor the inconvenience of an excursion into collateral issues. It took the view that the evidence of habit may not be used to prove alibi in a criminal case. This view has no support in reason or authority and its application here was highly prejudicial.
I think that in the circumstances of this case — where the central issue concerned an event four years before the trial and the testimony of the sole alleged eyewitness was open to suspicion2 — the defendant was clearly entitled to support his alibi with evidence of his religious habit. The question remains whether he was effectively denied the opportunity to do so, either because his introduction of such evidence was too sharply curtailed or because the trial court’s rulings and charge dissipated the value of the habit evidence he did introduce.
Exclusion of the rabbi’s testimony of habit was probably not prejudicial in view of the similar testimony presented by other witnesses. But I think the value of their testimony was vitiated by the court’s erroneous rulings and charge. Because the question of guilt was close, I think these errors require reversal of the conviction.
The court told counsel that habit evidence was irrelevant. See TR 627-634. It also told counsel it would instruct the jury that evidence of the defendant’s habit of doing something is no proof that he did so on a particular occasion, TR 634, and that all matters with respect to habit and custom were to be excluded and stricken. TR 633, 925. On one occasion the court cut off questioning when defense counsel began inquiring into appellant’s habit, saying in the presence of the jury:
“Mr. Stein, we are not concerned here with what Mr. Levin has done all the rest of his life.” [TR 627]
All this may well have discouraged defense counsel from attempting fully to explore the matter with any witness, and may well have led the jury to believe it could not consider appellant’s habit in determining whether he had or had not *277met Landriscina and received the money at 5:00 on a Friday afternoon.
If there was any question about this in the jurors’ minds, it must have been resolved by the court’s charge. The court told the jury:
“It is my recollection that Mr. Levin testified, as did other witnesses, that he has a habit or practice of a religious nature whereby he stays at home from sundown Fridays to sundown Saturdays. You are told that proof of such a habit or practice is not proof that the defendant is never elsewhere at that particular time.”
'The majority, in finding that this charge was not erroneous in the circumstances here, says:
“Proof of habit, for religious reasons, to be at home on Friday evenings clearly is not conclusive proof that the person having the habit may not be elsewhere on occasion and particularly on the specific Friday evening in issue.” [Emphasis supplied.]
;Sueh a charge would have been correct. But the charge given did not use the word ■“conclusive”; it said proof of the habit is “not proof”. In light of all that had .gone before, and especially the court’s .statement of how it expected to charge the jury, I cannot assume the omission of the word “conclusive” was unlikely to mislead the jury.3 Bather, it fits the pattern, consistently maintained throughout the trial, of not allowing the jury to consider habit evidence in relation to Levin’s alibi.
The trial lasted eight days; the transcript is 1076 pages long. Despite the effort and expense involved in a new trial, I think one is required. In determining whether error was prejudicial, the closeness of the case is a crucial consideration. Edwards v. United States, 265 F.2d 302 (9th Cir. 1959). Here the only witness directly implicating appellant was Landriscina, who claimed to be the other party to the transaction. Landriscina had already pleaded guilty to conspiracy to obstruct justice under another count of the indictment involved here. His testimony was thus subject to impeachment and was in fact impeached. His testimony as to dates and times was contradicted by other prosecution witnesses.
A jury in this jurisdiction is permitted to believe the uncorroborated testimony of an accomplice, or testimony corroborated circumstantially by other accomplices ; 4 and the Government may establish a prima facie case with testimony that is contradictory in several material respects. But I think we cannot hold, as *278a matter of law, that evidence tending to exculpate the defendant would have been unlikely to influence the jury’s decision if the jury had been allowed to consider it. I would grant a new trial.
. Landriscina also testified lie gave Levin $10,000 on Thursday morning, February 12. But other union officials testified Landriscina did not receive the $10,000 until Friday morning. See note 4 infra, and majority opinion, note 4 supra.
. At least where the proffered habit evidence tends to show that an event never took place, I see no reason to exclude the habit evidence simply because there is a purported eyewitness to the event. McCormick, Evidence 342 n. 9 (1954); Morgan, Maguire & Weinstein, Cases and Materials on Evidence 390 (1957). That direct evidence is more weighty than habit evidence is adequately conveyed by an instruction; it does not call for exclusion of the “inferior” evidence.
. That defense counsel’s requested instruction was erroneous means that denying it without more would not have been error. The error here was not refusing to give the requested instruction; it was giving the prejudicial instruction.
. Several union officials testified that a cheek for $35,000 was cashed on Friday, June 13. But only Vincent Belloni, who the court charged might be an accomplice, gave testimony tending to connect Levin with the money. Belloni testified that in January 1959 he and Landriscina visited Levin’s New York office where they discussed Cross’ legal representation at his pending perjury trial. Nothing was said about fixing the ■ case. Belloni had a subsequent conversation with Levin at the latter’s New York -office:
“I told him that Jimmy had called me and had told me that Mr. Levin wanted $35,000 for the case, and Jimmy bad told me to speak to Mr. Levin to see if he could bring the price down and where the money was going.
*****
“Mr. Levin says no, that would be the price and what’s the difference where the money would go.
“Q. Did he say anything more?
“A. What’s the difference where the money would go as long as the job is done.”
Belloni did not think there was anything wrong or irregular about the questions he propounded to Levin at Landriscina’s request. A few days after Cross’ perjury trial had terminated, Levin called Belloni on the telephone and told him “when I say something will be done it is usually done.” In several conversations with Belloni, Levin never said whether or not he received the $35,000.