(dissenting):
Again we have a holding of this Court reversing a conviction which was not only *1216fairly obtained in 1963 but affirmed on direct appeal to this Court, with certiorari denied by the Supreme Court, Levin v. United States, 119 U.S.App.D.C. 156, 338 F.2d 265 (1964), cert. denied 379 U.S. 999, 85 S.Ct. 719, 13 L.Ed.2d 701 (1965). This is but another of the long line of cases demonstrating this Court’s chronic aversion to finality in criminal cases. This holding is a grave abuse of the Great Writ of habeas corpus which was intended to correct injustice, not frustrate justice ; it lays down an unworkable and totally specious requirement.
After failing to persuade this Court and the Supreme Court to disturb his conviction, Appellant sought release on a petition for habeas corpus claiming “newly discovered evidence,” and arguing the Government had knowingly used perjured testimony; later he changed his petition to claim that the Government had concealed certain evidence. Both these claims were totally without basis and the majority agrees that this is so; the first claim was abandoned and this division of this Court rejected the second claim.1 The majority now hints that perhaps we did not reject the second contention when we remanded the habeas corpus petition to the District Court, but even a cursory reading of our opinion, Levin v. Katzenbach, 124 U.S.App.D.C. 158, 162, 363 F.2d 287, 291 (1966), will show that at most the remand found only a possibility that the Government may have “negligently” failed to disclose what had been stated by a potential witness interviewed by both sides.
This Court’s remand to the District Court was for two purposes: first, to determine whether the Government was negligent in not advising the defense that a bank officer had said he could not remember the exchange of large bills into $20.00 bills and second, if there was negligence, to determine whether this non-recall “might have led the jury to entertain a reasonable doubt about appellant’s guilt.”
After a hearing and extended consideration on remand the trial judge found as a fact that the Government had not been negligent and that even had the jury been told of the bank officer’s non-recall of the events, it would not have affected the result. Rule 52, Fed.R.Civ.P. limits our review narrowly to determining whether the District Court findings were clearly erroneous.
Before reaching analysis of what was done and what the majority now does, it is important to make clear what is not in dispute in this case. The prosecutor’s duty to disclose evidence favorable to the defense was defined by the Supreme Court in Brady v. State of Maryland, 373 U.S. 83, 83 S.Ct. 1194, 10 L.Ed.2d 215 (1963), but the issues here presented are not governed by that holding.2
But that rule is not the issue in this case, notwithstanding the belated and the tortured effort of the majority to make it appear so. What is involved is this: in preparing the case for trial each side interviewed one McCeney, a bank officer, concerning the $35,000 check drawn on union funds; this was the $35,000 paid *1217to Levin on his fraudulent claim that he could “fix” a case against one James G. Cross, then under indictment for embezzlement of union funds.
In light of this factual background the not-very-subtle quotation in the majority opinion from other cases which are inap-posite, using such inflammatory terms as “lawless” and “shocking” to describe conduct of a prosecutor, is an affront to the facts of this case. That these lurid allusions purport to be part of tracing the development of the law of the subject is a very thin excuse for this tactic and I suggest is a poorly veiled device to hint darkly at some nefarious act of Government which cannot be supported by reference to facts. The remand decision suggested only that the Government may have negligently failed to disclose information it had received; and even that claim rested on the fragile reed of an assumption that the Government had a duty to say in effect:
Now look, Mr. Defense Counsel, we interviewed Mr. McCeney and here is what he said to us: He remembers cashing the $35,000 check but cannot recall changing the bills into small denominations. We want to make sure he gives you the same information he gave us.
To apply any such remarkable standard, the Government must first surely have some reason to believe that the information of non-recall is relevant in some way so that it rises to the level of “evidence.” To recite this proposition, which is simply to apply the majority’s thesis, is to demonstrate that it has no basis in the realities of litigation.
It is, of course, too elementary to require citation of authority that there are two predicates of negligence: first, the existence of a duty and second, failure to meet that duty. Can it be possible — rationally possible — that the prosecutor has a duty to monitor or guide and oversee the defense counsel’s preparation and conduct of his case to the extent of requiring an exchange of information on a statement of non-recall derived from an interview with a common witness? Can it be that defense lawyers want or expect this kind of “Big Brother” treatment so long as they know of the existence and whereabouts of the witness and have access to him, as the defense did here? Even were civil pre-trial processes available, it is most unlikely that the bank officer’s non-recall would have been noted.
Having previously registered my dissent to the nebulous and novel “negligence” concept relied on by the majority, but being bound by it as the law of this case, I shall try to demonstrate that, even assuming the validity of the wrongly conceived and undefined law-for-this-case guidelines of the majority, the District Court was not “clearly erroneous” in finding that the “new evidence” in question would not “have led the jury to entertain a reasonable doubt about Appellant’s guilt.”
To an utterly absurd legal standard of prosecutorial negligence-without-duty, the majority now applies a review standard which is ridiculous and^anticlimatical^ in the extreme. The second aspect of the remand was to have the District Judge, after passing on the “negligence” aspect, decide whether the alleged newly discovered detail would have had an impact on the jury had the defense called McCeney and developed his lack of recollection. We must remember this is a case tried more than three years ago on events now nine years in the past.
Having sent the case back for a factual determination the majority — perhaps because they do not relish the result — now makes the discovery that this is not a factual issue after all but a legal matter. I suggest this is a transparent device to avoid the impossible task of demonstrating that the District Judge was “clearly erroneous” under Rule 52. In a most remarkable piece of judicial legerdemain, what was once a factual issue for the trial judge as fact trier, now emerges as a legal question for appellate judges! Of course, when this case is tried again — as *1218it must — the issue will again be emeshed in the jury’s fact-finding deliberations.
Just how a jury could have been “influenced,” as the majority now decides, by a piece of peripheral non-recall “evidence” available to and brushed aside by defense counsel is left dangling in midair. The action of the majority, ignoring firmly established concepts of appellate review and the Federal Rules of Civil Procedure as well, demonstrates the wisdom of the historic limitations imposed on reviewing courts. It was to hold in check undisciplined judicial action by remotely situated appellate judges that these rules were framed, but those concepts are cast aside today even if only to make a Rule-for-Levin’s-Case.
The District Judge, whose trial experience vastly exceeds that of all three members of this panel and who lived with this case for many days, observing witnesses and jurors alike, was the best, if not the only, person qualified to make the appraisal for which we remanded. Since that appraisal cannot be improved upon by paraphrasing, I quote:
If, as indicated by the Court of Appeals, significance attaches to the fact that neither of two Bank officers remembered exchanging the $1,000 bills for twenties but did recall the cashing of the Union check, such significance dwindles to the vanishing point in light of (1) the failure of the officers to remember the cashing of the Union check two years after the event when they were first asked about it; (2) the reconstruction or reviving of their recollections in this regard from bank records ; (3) the time lapse of one and one-half years between inquiry of the officers as to the cashing of the check and inquiry of them as to the exchange of the bills; (4) the fact that this second inquiry was made nearly four yéars after the event in question; (5) the lack of bank records to disprove the exchange of the $1,000 bills for smaller bills; (6) the possibility that one of the several tellers other than Hooper exchanged the thirty-five $1,-000 bills into $20 bills for Ashby; (7) the fact that Hooper did not know Ash-by; (8) the fact that two witnesses— Ashby and Landriscina — testified unequivocally that the $1,000 bills were exchanged for $20 bills; and (9) the obvious memory deficit of McCeney and Hooper.
Levin v. Katzenbach, 262 F.Supp. 951, 960 (D.D.C.1966).
The Government case, presented to the jury in 1964, satisfied 12 jurors — beyond a reasonable doubt — that Levin told Cross and Landriscina, officers of the Bakery and Confectionery Workers’ Union, that he could “fix” Cross’ perjury trial at a cost of $35,000; that $35,000 was raised by Peter Olson, Secretary-Treasurer of the Union, by embezzling Union funds; that Landriscina delivered the money to Appellant in Washington and that Appellant kept the money; in short, the jury found that Levin’s whole story was simply a confidence scheme concocted by him to bilk his victims out of $35,000.
A review of the evidence before the jury is called for by the majority’s action since the central issue at Appellant’s jury trial was whether Levin had received the $35,000. Landriscina testified that he gave Levin $10,000 at 11:00 o’clock on the morning of February 12, 1959, and $25,000 at 5:00 o’clock on the evening of Friday, February 13. Olson and his subordinate Ashby, disagreeing with Lan-driscina only in detail, also testified that they had given Landriscina the $35,000 in two installments — $10,000 on the morning of February 13 and $25,000 that same afternoon.
Although Landriscina’s version as to the date of the $10,000 payment was contradicted by Ashby and Olson as to the particular day, the transaction was corroborated by them in all other details. Landriscina testified that he received from Olson an envelope containing ten $1000 bills and gave them to Levin at the first meeting. Shortly thereafter Lan-driscina said he was contacted by Levin, who reported that the “fellow who was to take care of the jury” would not ac*1219cept bills of such large denominations. Landriscina then took the ten $1000 bills back to the Union office and arranged to have the money changed. He thereafter delivered $10,000 to Levin in smaller bills. Then, at the second meeting, he transferred the remaining $25,000 in small bills.
Olson testified that he cashed a $35,000 Union check on February 13 and gave Landriscina $10,000 in $1,000 bills to “fix the Cross trial.” Ashby testified that he saw the exchange of bills and that Olson then instructed Ashby to give Landris-cina the remaining $25,000 when he asked for it. Ashby confirmed that Landris-cina returned and said Levin claimed the bills were unacceptable because they were too large. Ashby then took the entire $35,000 to the National Savings and Trust Company, changed the bills into twenties and gave Landriscina $10,000 and put the rest in the safe. Later the same afternoon, Landriscina returned for the $25,000 which he gave Levin later on February 13.
The Government also introduced the check which Olson cashed to obtain the $35,000, dated February 12; bank markings on it indicate that it was cashed on February 13 thus supporting the view of those who said the currency passed on the 13th. Both Landriscina and Olson agreed that, whatever the day, the $35,000 was paid to Levin in two installments.
Levin denied receiving any of the money and claimed that he could not have been involved at all because as a diligent observer of the Jewish Sabbath he would have been home on Long Island by sundown on February 13, the time fixed by two witnesses of the delivery of the second payment of $25,000. Levin’s denial on this score was the essence of his defense.
Taking full advantage of the one-day confusion in dates in the Government’s case, i. e., whether the first delivery of money took place on February 12 or February 13, Levin sought to show that Ash-by and Olson were telling the truth about cashing the $35,000 since bank records on this were undisputed, but that Lan-driseina was lying and had kept the money for himself or passed it on to someone else. Defense counsel hinted repeatedly to the jury in his cross-examination of Landriscina and later in his argument, that Landriscina stood in line to become President of the Union if Cross were convicted of perjury, indicating that Lan-driscina had more to gain than the $35,-000 by not passing it on to Levin.3 Levin also sought to impeach Landriscina by demonstrating his memory faults with respect to the events of February, 1959, including the fact that Landriscina had previously stated to Government investigators that the first delivery had been on February 9, and by showing that Landri-scina had pled guilty to conspiracy to obstruct justice under another count of the same indictment on which Appellant was tried.4
Notwithstanding the great efforts of the defense to exploit the one-day discrepancy between the Landriscina and Olson-Ashby versions, the jury had little hesitancy about believing the essence of Landriscina’s account of transferring the money to Levin. After an eight-day trial, the jury promptly found Levin guilty. The verdict indicates the jurors considered the mistake of one day — a variance of a kind found in most lawsuits — was a natural result of the passage of time; the verdict also shows the jury rejected the various efforts made to impeach prosecution evidence.
The new “evidence” which the majority professes to believe might have changed the jury verdict is a statement by Benjamin MeCeney, Assistant Treasurer at the National Savings and Trust Company, *1220that while he remembered Olson cashing the $35,000 check, he did “not recall Mr. Ashby coming in to change the thousand dollar bills to smaller ones.” It must be emphasized that this piece of supposed “new evidence” is not evidence in the sense that it tends to prove any fact but is really non-evidence. McCeney’s statement was that he did not remember whether Ashby came in to exchange the bills; he did not say that he remembered that Ashby did not change the bills. It is simply a reflection of non-recall, made to Government investigators in September, 1962, three and one-half years after the transaction and long before the ha-beas corpus hearing on the “new evidence” claim.
It is inconceivable to me — as it was to the presiding trial judge — that this non-recall could now be said to have had any effect on the jury. A few reasons are immediately apparent:
(1) .The jury could reasonably have concluded that it is not surprising that a bank officer of a large and busy bank could not remember changing some money three and one-half years after the event even with $1,000 bills involved.5
(2) The majority totally fails to give weight to the fact that Appellant’s trial counsel, experienced in criminal matters, indicated that he did not regard McCeney’s testimony as important when trial counsel testified at the habeas corpus hearing. The attorney testified in the habeas corpus hearing that he had learned from the Cross trial, in advance of the Levin trial, that the large bills had been changed into smaller ones. He had previously spoken to McCeney to determine if there would be any records of the cashing of the $35,000 check and learned that they had not been retained. After learning of the exchange of bills at the Cross trial, the attorney testified: “I did not again go to the bank because I had no reason to. I had previously been told by two people that they had no records.” The logical inference is that he was aware that the bank officers could not remember and that without records he would not be inclined to want their testimony. Apparently trial counsel believed that any non-recollection would not seem significant to his defense tactics or to the jury. In other words, we now reverse for a new trial because “evidence” was not available that would apparently not have been used if it were!
That Levin’s trial counsel, when called at the habeas corpus hearing, believed that this so-called “evidence” would not have influenced the jury undoubtedly flows from the fact that the jury believed Landriscina’s testimony that he gave the money to Levin despite the fact that Landriscina was contradicted as to the details and time by his own associates and documentary evidence, and de*1221spite the fact that he was impeached by his role in the criminal conspiracy and by his possible motives for not delivering the money. Can there be any rational basis to believe that the jury might have changed the verdict if they had heard a statement of non-recall from a subsidiary participant about a tangential detail when the other, more important considerations did not sway them?
(3) There is nothing whatever, except hindsight, to suggest that the defense would have made any use of McCeney’s statement, and indeed I can hardly imagine that any experienced trial lawyer, such as Levin had, would have wanted to indicate to the jury the weakness of its case by calling McCeney as its witness only to have him say he “could not remember.”6 Moreover, the thrust of any defense utilization of McCeney’s non-recall would have been to argue that the exchange of bills never occurred. But to demonstrate that, Appellant would have had to argue that not only was Landriscina lying, but also Olson and Ashby, since all three testified that the $1,000 bills had been exchanged for smaller denominations. As I pointed out in my first dissent, it is more difficult to persuade a jury that three men are lying than, as Levin attempted to do at trial, that one is.
I regret the occasion to dissent in these terms, except that it becomes necessary to demonstrate the glib but fallacious assumptions which underlie the majority’s action. Even more I regret this Court’s repeated actions which plainly tell prisoners, “jail-house” lawyers and the bar generally that if they can find a way to continue the warfare with society long enough they may finally reap the natural rewards of lost evidence and fading memories. New trials, long after the events occurred, place enormous obstacles in the way of just results. It is on this very basis that courts dismiss indictments for lack of speedy trial. Compare Williams v. United States, 102 U.S.App.D.C. 51, 250 F.2d 19 (1957). Here the prosecution must now re-try a case concerning events of February, 1959, and by the time of the new trial nearly nine years will have elapsed. On this record, I feel fully warranted in charging the majority with another instance of appellate “nit picking.” This kind of perversion of the judicial process has gravely hampered speedy and certain justice in this jurisdiction.
Before BAZELON, Chief Judge, DAN-AHER, BURGER, WRIGHT, McGOW-AN, TAMM, LEVENTHAL and ROBINSON, Circuit Judges, in Chambers.
. Indeed, the remand order recited “the prosecution, no doubt in complete good faith, did not disclose to the defense at or before trial” the fact that a bank officer stated he “could not recall” changing large bills into $20 denominations. Levin v. Katzenbach, 124 U.S.App.D.C. 158, 159, 363 F.2d 287, 288 (1966) (emphasis added). Surely this made clear that no “suppression” or “concealment” of evidence was involved.
. Brady decided that “upon request” the prosecution must furnish evidence favorable to the accused, 373 U.S. at 87, 83 S.Ct. 1194, 10 L.Ed.2d 215. I would assume that where the government has positive exculpatory evidence which plainly would constitute a defense, the government has a duty to tender it without a request. But the majority here stretches the sound Brady concept to cover peripheral material which by no stretch of imagination could have been regarded as “evidence.” It is for this reason the majority must spin out its fanciful theory of government “negligence.”
. Levin admitted later receiving fees from Landriscina for lobbying on behalf of the union.
. Judge Bazelon in his dissent from the original affirmance of Levin’s conviction stated:
[Landriscina’s] testimony was thus subject to impeachment and was in fact impeached. His testimony as to date and times was contradicted by other prosecution witnesses.
119 U.S.App.D.C. at 158, 338 F.2d at 277.
. The majority seems to attach some significance to the fact that McOeney could remember cashing the $35,000 check but could not remember changing the bills. It was in March of 1961 that government investigators first asked McCeney and Hooper about cashing the $35,000 check; this was two years after the event. They cou,ld not remember it. When the investigators returned to the bank the next day, MeCeney and Hooper did recall cashing the check; the evidence adduced at the habeas corpus hearing showed that they recalled cashing the check upon consulting bank records. McCeney was not asked about changing bills until September of 1962 — a year and a half after being asked about cashing the check and three and a half years after the event. The Bank’s records of large bill transactions had been destroyed by that time. Is it any wonder, then, that three and one-half years after the event, the Bank officers could not remember the exchange when the only reason they could remember cashing the check two years after the event was that they had consulted their records? The majority’s suggestion, therefore, that it is significant that Mc-Oeney remembered the cashing but not the exchange is baseless.
The majority in note 28 states that “at the habeas hearing McOeney and Hooper recalled details of the transaction which would not be contained in records.” But the testimony which the majority relies upon is a description by bank officers of established bank procedures concerning large bill transactions. Of course, no records were needed to recall this.
. One other aspect of the majority opinion deserves comment. It is hinted that the statement of Hooper at the 1966 habeas corpus hearing — two years after the trial —is somehow relevant to whether the Government should have told the defense in 1964 of the McCeney statement. But the Government did not learn about that fact until the 1966 hearing. Can the Government be expected to give the defense information it did not have at the time of trial? The majority’s confusion is indicated by the statement in their first opinion that they “do not suggest that the government is required to * * * disclose all its evidence, however insignificant, to the defense,” Levin v. Katzenbach, 124 U.S.App.D.C. 158, 162, 363 F.2d 287, 291 (1966), but they now act to the contrary.