Circuit Judge:
In voting to deny rehearing en banc in this case, I remain equally unpersuaded of the accuracy of (1) the dissenting judge’s characterization of the information in question as “non-evidence”, (2) the majority’s generalization that there is a constitutional duty resting upon the prosecution to disclose voluntarily any and all “evidence which would be helpful to defendant”, or (3) the Government’s representation in its petition for rehearing en banc that the panel’s decision gives it no alternative but to turn over to the defense its complete investigation file in every criminal case.
I do not think it is possible in any one case to write a definitive demarcation between the Federal Criminal Rules relating to discovery, on the one hand, and those particular situations where a fair trial may have been significantly blurred by the Government’s failure to comply with a request for information. Judge Frankel, in a singularly perceptive, useful, and recent discussion of this problem, United States v. Gleason et al., 265 F.Supp. 880 (S.D.N.Y.1967), has said that, although “the dimensions of due process are not limited or fully defined by the Rules of Criminal Procedure * * * it makes obvious practical and doctrinal sense to consider the directly pertinent rules promulgated by the highest Court when appraising the meaning of the Constitution which is ultimately for that Court to expound.” Dealing with a motion to compel disclosure by the Government of any evidence favorable to the defense, he said that “it seems safe to take as a starting point the proposition that the prosecution need not deliver up, either before or during trial, all that would be literally embraced” by such a demand. But he goes on to conclude that:
It seems doubtful, however, .that there should be a blanket rule postponing to the trial all disclosures of the type in question [that is to say, disclosures beyond the scope of the Federal Rules], For example, where the prosecutor knows of witnesses potentially useful to the defense, does not intend to call such witnesses himself, and knows — or should reasonably be expected to suppose — that his knowledge is not shared by defense counsel, the information may come too late for effective preparation if it is not delivered until the case is on trial. [Citations omitted] Other kinds of instances will undoubtedly arise where the Government “has in its exclusive possession specific, concrete evidence” of a nature requiring pretrial disclosure to allow for full exploration ánd exploitation by the defense.
In other words, the prosecutor, like the rest of us, is going to have to learn how to live in the shadowy world between the Rules and the Fifth Amendment. I sympathize with his plight, but, contrary to the assumptions of his petition for rehearing, we cannot release him from it by a broad pronouncement in this or any other particular ease. The panel decision may have been wrong, but we do not en banc every panel decision which might well have gone the other way. On the other hand, it may have been right. Its important characteristic for present purposes is not its individual rightness or wrongness, but whether it is simply one application of a general legal doctrine which existed before this case was decided and which, after that event,. continues to exist in essentially the same form.
The following chronology of this prosecution will, I believe, be helpful in weighing the question of whether the decision *1223is so far out on its facts as to constitute a departure from the doctrinal channel. These facts (as to who knew what, and when) are essential in my appraisal of the need to consider this case en bane.
The alleged theft occurred in 1959. In March of 1961, the Government investigators went to the bank to inquire about the cashing of the union cheek which generated the bribe money. They were first told by Mr. McCeney, an Assistant Treasurer, only that transactions of the kind about which they were inquiring were normally handled by Mr. Hooper, an Assistant Head Teller. Hooper told them that he remembered nothing about the cashing of the check, but he apparently agreed to take a further look because, when the investigators came back the next day, he said that Olson had cashed a $35,000 check and received $1000 bills.
The matter was presented to the grand jury in June of 1962. It was in this proceeding, and by means of the testimony of Ashby, that the Government apparently first learned of the exchange of the $1000 bills into $20 bills. It recognized the significance of this testimony of its own witness, because it made an earnest effort to corroborate it. In September of 1962, its investigators went to the bank again where they interviewed McCeney. They asked him if he could confirm what had been said to the grand jury about the $20 bills; and they took from him on this occasion a written statement in which he said in substance that (1) he did not recall any exchange for smaller bills, (2) Hooper would, in any event, have handled it, and (3) Hooper had told him that he did not recall any such exchange. Further emphasizing the importance which the Government attached to this matter, particularly in the light of this initial failure to corroborate its own witness, the Assistant United States Attorney in charge of the prosecution telephoned McCeney promptly after he read the written statement brought back to him by his investigators. This telephone call was a seeming further effort to elicit some corroboration. It failed because Mc-Ceney apparently said in response to this inquiry that he would not say that it did not happen but that he could not remember any such thing.
Thus it was that the Government knew as of the fall of 1962 that its key witnesses were going to testify at the trial that the defendant had made them exchange the $1000 bills for $20 bills, but that the bank officials who allegedly effected this exchange could not corroborate this testimony. The Government could not have been unaware of the danger this involved to its case, particularly one of this kind where the jury would immediately recognize the prosecution witnesses as operating in an atmosphere where cash floating around is as likely to end up in someone else’s pocket as in that of the defendant. If any shadow of doubt could be thrown on the testimony of these men the Government said actually handled the cash eventually delivered to the defendant, the Government would be hurt. Certainly the failure of the bank officials to corroborate such a transaction as this one would have cast such a doubt.
In January, 1963, the defendant’s counsel made his motions for discovery and a bill of particulars. Item 7 of the latter was as follows:
State the denomination of the $35,-000 in money which was allegedly taken by the defendant, Levin. If the exact denominations are unknown, then give the approximate denominations.
The Government vigorously opposed the giving of this information. With regard to Item 7 in particular, the Government said:
Paragraph 7 is indicative of the nature of defendant’s motion — to secure the Government’s evidence in advance of trial. Proceeding from the fact that the denominations of the bills used in the larceny is evidentiary, defendant can suggest no good reason as to how such knowledge would aid him *1224in respect to the functions of a bill of particulars as enumerated above.
It is, of course, not surprising that the defendant could not, in the Government’s phrase, “suggest [any] good reason as to how such knowledge would aid him,” since as of that time the defense appears not to have known of the fact that the check was initially cashed in $1000 bills, much less that these were immediately changed into $20’s.1
Active opposition to an innocent request for information as to the bill denominations would surely, under these circumstances, create some sensations of discomfort in the more knowledgeable party. Such instincts are perhaps as good a guide as any to the need for voluntary disclosure. If, however, such feelings are successfully resisted but the defense comes upon the information by other means, no harm may be done; and this is essentially the basis upon which the Government seeks to sustain this conviction. To do so, in my view, it should, in the light of its earlier affirmative withholding, be able to demonstrate on the record that the defense did know of the $20 bill exchange in advance of trial.
Exactly when defense counsel first learned of these facts is far from clear. At the habeas corpus hearing, he testified that the Government did exhibit to him, apparently not long after the discovery hearing, the $35,000 check. His testimony also suggests that he may have first heard about the $1000 bills during the Cross trial, which took place in April, 1963. There is also a strong intimation that his first and only interview with McCeney came after the Cross trial, since the purpose of that interview was to inquire about records of “big bills.”2 *1225The Cross trial preceded the trial of this case by about five weeks.
The dissenting opinion states that defense counsel “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.” The actual answer was:
Q. When did you first learn from any source that the thirty-five $1,000 bills, in which the cheek was originally cashed on February 13th of 1959, might have been returned to the National Savings and Trust Company to be changed into bills of smaller denominations?
A. I think I heard testimony during the Cross trial along those lines, but I am not sure of that. As I was saying before, I was in a trial before Judge Walsh during the time that James Cross was being tried and I came into the courtroom intermittently as I could, and I believe that I heard testimony about the cashing of this check during the Cross trial which preceded the Levin trial.
The prosecutor then sought to establish by his questions what is beyond dispute, namely, that the $20 bill exchange came out in the testimony at the Levin trial; and the prosecutor conducting this examination summed it in this wise: “So you knew about it probably at the Cross trial, you certainly knew about it at the Levin trial * * *.” For me, that “probably” is, because of the uncertainties in this record, at least at the outside limits of accurate characterization; and it is certainly short of the demonstration needed to retrieve the effect of the Government’s initial denial of the request for the bill denominations.
It would be a reasonable evaluation of the habeas corpus record (and there is no express finding by the habeas corpus court to the contrary)3 that defense counsel did not know of the $20 bill transaction until it first came out in the Government’s case during the trial of his client. It is, of course, true that he made no request for a recess in order that he might investigate this matter, nor did he even telephone McCeney about it. He said at the habeas corpus trial that this was because he had already learned from McCeney that the bank had no records, but his testimony also was that his interview with McCeney, which arguably did not occur until after the Cross trial, was confined to asking whether the bank had any big bill transactions records. A teller at the bank might have remembered, even without records, an incident involving the changing of thirty-five $1000 bills into $20’s. And, of course, Hooper, at the habeas corpus hearing, said for the first time not only that he would have been the one to have handled such a transaction but that, if it had occurred, he would have remembered it.
The scope of a defense investigation during the days preceding the trial is one thing. What, under the pressure of trial itself, counsel decides to do about new information is quite another.4 For me, this is a case where, at the time the defense made a timely effort to get some *1226information about the denominations of the bills allegedly given to the defendant, the Government opposed it, even though it must be charged with knowledge that the bill-exchanging testimony, uncorroborated by the bank, was pregnant with significance for the defense. In the special setting of this case, it is perhaps not enough to avoid the danger of injustice that the Government may be said to have taken a technically accurate position with respect to a bill of particulars.
This can be regarded, therefore, as a situation where, within the meaning of Brady v. State of Maryland, 373 U.S. 83, 83 S.Ct. 1194, 10 L.Ed.2d 215 (1963), the defense made an “actual request” of the Government for information which the Government deliberately refused, conscious of its danger to the prosecution and of the inequality which then existed between prosecution and defense in terms of their relative knowledge. Falling back to the Government’s next line of defense, it is literally impossible to glean from this record a clear showing that no harm was done because the information became known in any event to the defense prior to trial.
In its brief upon this appeal, the Government refers to a deposition of Ashby in a civil case, used by defense counsel in cross-examining Ashby at the Levin trial. That deposition, in pertinent part, is as follows:
BY MR. DICKSTEIN:
Q. Mr. Ashby, we were speaking yesterday of the check for $35,000 drawn to Peter Olson as paying officer of Local 3. Do you remember that check?
A. Yes, sir.
Q. Do you know whether Mr. Olson cashed that check?
A. Yes, he did cash it.
Q. How do you know?
A. He gave me the cash to put in the vault in the safe.
Q. Did you go to the bank with him when he cashed it ?
A. I can’t recall. I may have.
Q. You don’t remember ?
A. It is vague. I may have gone with him to cash the check. I just can’t recall.
Q. What were the denominations of' the currency which Mr. Olson received?
A. That I don’t know.
Q. What were the denominations of the currency that he gave you to put in the vault?
A. I know they were small bills. I don’t recall the denominations. They were relatively small bills.
Q. By “small,” do you mean less than $100 bills?
A. As I recall, they were less than $100 bills.
Q. It was not 35 $1,000 bills, was it?
A. No, not to the best of my recollection.
Q. When, in relationship to the date of the check, did he give you this money to put in the vault?
A. I don’t know if it was the same day the check was drawn or a day later or two days later. It was in close proximity of the date of the check.
******
Q. And did you count the money before you put it in the vault?
A. I counted — Yes.
Q. What was the highest denomination of bill that you noted while you were counting the money?
A. I can’t recall the exact denominations of the bills.
Q. Approximately.
A. There were quite a few bills. There may have been hundred-dol*1227lar bills. I don’t think there was anything larger than hundreds.
Q. Were most of them 20’s and 10’s?
A. Some of them were smaller than that; 20’s I think. But I can’t recall what the breakdown was.
Apart from the fact that the deponent is, in the light of his later testimony before the grand jury and at trial, being egregiously evasive, it can scarcely be said that pre-trial examination of this deposition would acquaint the defense with the fact that 35 one-thousand dollar bills were exchanged, at Levin’s insistence, into $20 bills.
Neither is it clear beyond peradventure that the deposition was seen prior to the Levin trial. There was a day’s interval at that trial between the direct examination of Ashby and his cross-examination. The record shows that defense counsel, while conducting that cross-examination, had only some written notes in his hand and not the deposition itself. When objection was made to this circumstance, the court recessed the trial so that the deposition could be obtained from the files of this court, where it was then reposing. Our Clerk’s records do not show it as having been withdrawn earlier. It had doubtless been seen before by defense counsel, but no one can say when, or whether it might not have conceivably been after the Levin trial started. The Government did not introduce the deposition into evidence at the habeas corpus hearing, nor make any references to it in its effort to establish in that record exactly when trial counsel first learned of the $20 bill exchange.
Denial of rehearing en bane is not to be taken as indicating that the Government is required to honor a general request for any and all information helpful to a defendant — an approach expressly disavowed by Judge Frankel. I might have affirmed the habeas corpus judge in this instance, but on the facts the case would have been a close one. I think an en banc consideration would only embroil all of us in pondering the nuances of a record that can be read in more ways than one. In my view, the central inquiry here is whether affirmance on the facts of this case represents such a departure from established lines of legal doctrine as to justify en banc consideration. Applying that test here, I do not believe that rehearing en banc is warranted.
. The Government’s position vis-a-vis this request makes it useful to quote the following comment by Judge Frankel in the opinion referred to hereinabove (at p. 886 of 265 F.Supp.) :
Concluding that the Pitkin statements may be in part helpful to Karp, we have proceeded to inquire whether any “respectable interest” of the Government’s may be served by withholding this material. The prosecutor was asked to speak on this question. He responded with the observations that “the defendants are engaging in a fishing expedition” and “that anytime [sic] discovery of material is ordered in derogation of the Federal Rules of Criminal Procedure, the Government is injured.” We have not found enlightenment in these pronouncements.
. In answer to questions by the Government seeking to establish his knowledge of Mr. McCeney and the latter’s connection with the cashing of the $35,000 check, defense counsel testified as follows:
THE WITNESS: One. I had known Mr. McCeney before the trial and I knew that he held an office in the bank, I forget what it was, but I think it was an assistant vice president. Whether or not I referred to him in examining Mr. Olson is a matter of record and I do not recall it, because I have never read the record since the trial was completed. Thirdly, in the preparation of the case, I recall that I knew that this check was of some importance because I had heard the Cross trial which took place before the Levin trial and I believe this check was mentioned. It was my understanding that when a check is cashed where large bills are given, that some record is kept by the bank and, in my mind, I had some vague notion of an Internal Revenue regulation to catch people who are cashing big checks.
MR. ARNOLD: Will you identify what check you mean when you say “this check”?
THE WITNESS: I am thinking of the check bearing Mr. Olson’s name in some way or other, the check which Mr. Ash-by says he took to the bank, that is the check I have in mind.
BY MR. ALTSHULER:
Q. The $35,000 check?
A. Yes, the one that was introduced at trial, that is the cheek. One day in preparing myself for the trial, I went over to see Mr. McCeney and without telling him what I had in mind, I asked him, in what must have been to him a casual conversation, whether records are kept when large bills are given out by the bank and I believe he told me that some records are kept but that they are destroyed very shortly.
I then asked another fellow, Mr. Fred Loops, who unfortunately is now dead, who was working at the Bank of Commerce—
THE REPORTER: How do you spell that name?
*1225THE WITNESS: L-o-o-p-s, Mr. Loops, in another casual conversation with Mr. Loops, I was attempting to cheek the information that I got from Mr. Mc-Ceney and I asked Mr. Loops whether they, at the Bank of Commerce where I was then banking, kept records of the disbursement of large bills and he told me the same thing Mr. McCeney did, that the records they keep are destroyed very quickly, and that was the sum and substance of my relationship with Mr. McCeney in this case.
******
. The finding in terms is that trial counsel learned of the $20 bills “at or prior to trial.” But whether it was “at,” or “prior”, is critically important.
. The difference lies in the loss of the invaluable privilege of pondering these matters reflectively in the course of preparing for trial, and not to have his client judged on the basis of his reaction to new information under the tensions and pressures of trial.