The chief issue presented on this appeal is whether the prosecution's acquiescence in a witness's erroneous assertion that he had not been offered any deal for his testimony constituted a denial of due process under Napue v. Illinois, 360 U.S. 264, 79 S. Ct. 1173, 3 L. Ed. 2d 1217 (1959), and Giglio v. United States, 405 U.S. 150, 92 S. Ct. 763, 31 L. Ed. 2d 104 (1972). Finding that it did not, we affirm the decision of the district court, 479 F. Supp. 495, and deny the petition for habeas corpus relief.

On the night of Saturday, February 27, 1971, an eighteen year old woman named Susan Randall left a Manchester, New Hampshire, restaurant with a young woman friend. Wearing a two-tone, brown fur coat, and a brown floppy hat with a dark rim, Ms. Randall parted company with her friend shortly before midnight and walked in the direction of Granite Square, Manchester. This was the last time that anyone who knew her saw her alive. Approximately three days later her partially nude and brutally beaten body was discovered on the ice twenty-five feet beneath the Merrimack River bridge in Concord, New Hampshire.

Numerous witnesses, some of whom had been in a Granite Square restaurant on the night of February 27, testified that shortly after midnight there was a girl wearing a "floppy hat, brown hat, and a fur coat" hitchhiking from Granite Square, looking for a ride east on Granite Street, in the direction where Susan Randall lived. Several of these witnesses stated that at approximately 12:45 a. m. a white car with blue upholstery stopped and picked up the girl. One witness, an auto mechanic, stated that the car was a 1964 Chevrolet Impala. Another, who drove a two-door 1964 Chevrolet Impala, stated that the car was not an Impala. One witness also testified that the driver of the car, who remained seated in the car, was a "big man", "around six feet tall; very broad shoulders", with "a big head for a man of his size".

Petitioner, who is approximately six feet tall and weighs two hundred pounds, owned and drove a white four-door 1964 Ford with blue upholstery. During the last few days of February, 1971, he was moving furniture from a former residence in Manchester to a new apartment in Lowell, Massachusetts. A Lowell neighbor testified that he assisted petitioner in unloading furniture on the night of February 27 and that shortly before 11 p. m. petitioner departed in the direction of Manchester, telling the neighbor that he was returning for another load of furniture. The neighbor also stated that petitioner had not returned by the time that he went to bed, shortly after midnight.

A Manchester woman testified that petitioner, whom she described as "disturbed, angry and belligerent" at the time, arrived at her house at approximately 11:45 p.m. February 27, saying that he wanted to speak to one of her children. She stated that she told petitioner to leave her house, since she did not want him seeing her teenage children, and that he finally left shortly after midnight, getting into his car and driving in the direction of Granite Square. The woman called the police the following Monday to complain about petitioner, telling them roughly the same story set forth in her testimony.

When questioned by the police two weeks after the homicide, petitioner told them that after unloading the furniture in Lowell late on the night of February 27, he never returned to Manchester, either that evening, or the following day. However, another witness who had been at the home of the Manchester woman corroborated her testimony that petitioner had been at her house at midnight.

On April 2, 1971, state police searched petitioner's car pursuant to a warrant. From among the millions of minute particles picked up in a vacuuming of the car, the police isolated two small horse hairs and numerous small bits of paint-like material. They also performed on the interior of the car certain chemical tests that indicated the presence of possible blood stains on the passenger side of the front seat, and they removed from the floor mat what appeared to be a dried crust of blood. Similar tests indicated the possible presence of blood stains on one of petitioner's boots. The crust, hairs and particles, along with hairs and particles removed from the victim's coat, were forwarded to the United States Alcohol, Tobacco and Firearms Bureau where they were subjected to microscopic examination, and, in the case of the hairs and particles, a process known as neutron activation analysis.

The crust was established to be blood, but its type, including whether it was animal or human, could not be conclusively determined. The forensic chemist who performed the neutron activation analysis interpreted the results as establishing that the paint chips removed from the car were from the same manufacturer and batch as those found on the victim's coat, and that the horse hairs found in the car were similar to those taken from the coat. He concluded that the victim's coat "was in contact with (petitioner's) automobile." Another forensic expert who worked for the state testified that based on the results of the various analyses there was a "high degree of probability" that "we have had contact between (the victim's) clothing and (petitioner's) car."

The defense presented in rebuttal the testimony of two other experts, one of whom was a well regarded international authority in the field of neutron activation analysis. These witnesses testified that there was insufficient evidence and background scientific data to tell whether any of the perceived relationships between the chemical make-ups of the particles and horse hairs found in the car and on the coat were significant.

Against this background of circumstantial evidence was the testimony of one David Carita. Carita stated that when he and petitioner were fellow inmates in Massachusetts1 petitioner told him that he killed Ms. Randall. At the conclusion of Carita's direct testimony the prosecutor, in an apparent attempt to demonstrate that Carita had no motive for fabricating his testimony, asked him the following questions:

"Q: Were any promises or inducements made to you by myself or anyone else in the state of New Hampshire?

A: No.

Q: To get you to testify?

A: No."

Defense counsel followed with an extensive cross-examination, the chief thrust of which was to establish a basis for convincing the jury that Carita was lying. The cross-examination focused on the fact that Carita had a history of testifying against criminals, including an organized crime chief "enforcer", that he was considered a "rat" by many convicts, and that he feared for his life. Counsel also persisted at length in attempting to get the witness to contradict or retract his statement that no inducements had been offered for his testimony. The following series of questions and answers, precipitated by Carita's statement that he originally did not want to testify because he had a "pretty good deal" in Massachusetts, represents what counsel was able to achieve in this attempt:

"A: So I agreed that I would come (to New Hampshire), you know. But I wasn't too, you know, happy about it. In fact, I wasn't even sure about it at that time.

Q: You wanted to bargain a bit?

A: No, I didn't think I had to bargain.

Q: You thought they'd take care of you?

A: Well, I explained that if I came to New Hampshire, it would be because I was already released and in Massachusetts. That I didn't want my name in the papers anymore. That I was trying to make a life for myself, and it's rough if they find my whereabouts. It's dangerous to me.

Q: As a result you had to have something from them, didn't you?

A: Assurance of safety.

Q: And did you make any requirements in this direction? Did you make any requests of them?

A: Did I make any requests?

Q: Yes, sir.

A: Yes, I wanted to be assured that I just wouldn't be thrown in the State's Prison in the population and that who knows who is up here; who knows who may come here. The only people that I knew of that were interested in my safety were the people of Massachusetts.

Q: So, when you came to New Hampshire, you wanted some sort of assurances that you had something worthwhile here, didn't you?

A: Well, I wanted to make sure that I wasn't giving up what I had in Greenfield to come up here. Like I said before, I didn't want to leave where I was, you know, because of the hard work I put in. I didn't want to come up here, so I asked, 'Well, do you have work release program up here?' And I was informed that they do. And that along with that I wouldn't be just thrown into the, you know, anywhere just any old prison. I wanted to be assured of my safety. So, if you talk about bargaining or anything like that, yes."

During the closing arguments, defense counsel returned to this line of attack, asserting that Carita must have made a deal with the state. The prosecutor, in his final argument, briefly rebutted this assertion:

"Now if someone wanted to make a deal, I, honestly, believe he's got no deal to be made. . . ."

Several years after the trial and his conviction for first degree murder, petitioner discovered that the assurances of safety for which Carita had bargained included a promise made by a New Hampshire sheriff, with the assistance of the Attorney General's office, that Carita would be provided with a new name, location, and haircut.2 Petitioner filed a motion in state court for a new trial, alleging that the prosecution had violated his due process rights by failing to disclose at trial that a deal had in fact been struck with Carita in return for his testimony. The state trial court denied the motion, finding that disclosure of the sheriff's promise could not with any reasonable likelihood have affected the judgment of the jury. The New Hampshire Supreme Court affirmed, and the habeas petition followed. The district court, in a careful and thorough twenty-six page opinion, found that there was no reasonable likelihood that the disclosure of the promise could have affected the judgment of the jury.

The first question that we must face concerns the standard of review. Both parties agree that the district court articulated the appropriate substantive legal standard for determining the effects of the nondisclosure of the promise to Carita; that is, whether the nondisclosure "could . . . in any reasonable likelihood have affected the judgment of the jury. . . ." Napue v. Illinois, 360 U.S. 264, 271, 79 S. Ct. 1173, 1178 (1959); see also Giglio v. United States, 405 U.S. 150, 154, 92 S. Ct. 763, 766; United States v. Ramos Algarin, 584 F.2d 562 (1st Cir. 1978). Despite this fact, however, we find that we are required to make an independent review of the record and arrive at our own factual conclusion. The Supreme Court so proceeded in Napue, supra, finding contrary to Illinois state courts that certain false testimony could have affected the judgment of the jury. "The duty of this Court to make its own independent examination of the record when federal constitutional deprivations are alleged is clear. . . ." Id., 360 U.S. at 271, 79 S. Ct. at 1178. This duty exists even when a federal district court has already examined the state court record, at least in cases such as this in which no additional federal evidentiary hearing is held and the chief issue is the constitutional significance of the facts. See Neil v. Biggers, 409 U.S. 188, 193, n. 3, 93 S. Ct. 375, 379, n. 3, 34 L. Ed. 2d 401 (1972); see also Taylor v. Lombard, 606 F.2d 371 (2d Cir. 1979); but see United States v. Antone, 603 F.2d 566, 570-71 (5th Cir. 1979).

In making our own examination in this case we shall first assess the nature and strength of the incriminating evidence, compared to that of contradicting evidence. We shall then focus on the thrust of Carita's testimony, the extent to which he had been subject to impeachment at trial, and the extent to which he would have been rendered more vulnerable had the full story of his understanding with the sheriff been revealed. Finally, we shall have to determine whether any such additional vulnerability could in any reasonable likelihood have caused the jury to bring in a verdict of not guilty.

This is not a case where one could say that the evidence was so overwhelming that the new evidence could safely be ignored. It calls for closer analysis of each type of evidence. To begin, there was evidence, rather weak when standing alone, that a white car with blue upholstery and carrying a big man picked up the victim and that petitioner, a big man, owned a white car with blue upholstery. The former vehicle was described both as being and not being a Chevrolet Impala; the latter was a Ford. The state's case begins to gather strength with the testimony of a Manchester woman (and another witness) that petitioner (a) was in Manchester an hour (or less) before the victim was seen being given a ride, (b) was, when he was told to leave, in an angry state of mind, and (c) drove off in the direction of Granite Square. This incident becomes more important by providing a compelling basis for the jury to have concluded that petitioner had lied to the police in saying that he had not returned to Manchester on the night of the homicide after 11 p. m. Against this background of increasingly suggestive circumstantial evidence the jury had to choose among experts. As for the choice which had to be made concerning the possible blood stains on the passenger seat and boot, along with the blood crust on the floor mat whether it was human or animal blood we suspect that the jury could have conceived of nonincriminating explanations, yet would have been pushed by the growing force of the evidence to discount them in favor of inculpatory inferences. This leaves the disputed results from neutron activation analysis, two experts giving opinions positively linking the paint chips and loose hairs taken from the victim's coat to those taken from the car, and two others claiming that such a link could not be made.

The net effect of the evidence so far would, in our opinion, point strongly toward guilt, though not compellingly. Once Carita testified concerning the confession, we think the force of the evidence in terms of its effect on the jury became compelling. In this respect we agree with petitioner that Carita's testimony was very important, and not mere "icing on the cake" as claimed by the state. Had Carita's testimony been eliminated altogether, we think that there would be a reasonable likelihood that the judgment of the jury could have been affected. Carita, however, did testify; the narrower determination which we must make is whether, given his testimony and the other evidence, the judgment of the jury could have been affected by disclosure of the deal with the sheriff. We think not.

In the first place, disclosure of the deal could not have significantly added to the ability of defense counsel to impeach Carita. Carita's criminal past and his history of testifying against criminal defendants were revealed to the jury. The only relevance of the deal would have been for purposes of suggesting that in testifying as he did, Carita had something to gain and thus a motive for telling the story he told.3 However it is not by any means clear that disclosure of all the facts surrounding the deal would have provided any basis for such a suggestion. Carita had already received a new identity from the F.B.I. while he was in Massachusetts, and his refusal to answer certain questions at trial concerning names he used in Massachusetts suggests that he was in fact operating under the alias before he testified. Had the New Hampshire promise of an identity change been disclosed at trial, the state thus could have argued rather persuasively that the promise was made to Carita not as a reward for testifying, but rather as an assurance that by testifying and drawing attention to himself he would not lose the benefits of the "pretty good deal in Massachusetts" to which he referred in his testimony.

More importantly, even if we assume that the jury could have perceived the promise of an identity change as a significant motive for Carita's testimony, we cannot see any reasonable likelihood that its judgment could have been affected. While, as we have said, Carita's testimony was by no means unimportant, it was, in terms of its impact on the jury, inherently credible within the context of the increasingly suggestive inferences raised by the circumstantial evidence. As a practical matter, testimony that fits like the final piece of a puzzle into an increasingly clear pattern needs a much stronger rebuttal than one that stands almost alone as a purported representation of the whole. In this case, successful impeachment of Carita in order to negate the impact of his judgment on the jury would have required much more than the additional evidence of the promised identity change.

Affirmed.

BOWNES, Circuit Judge (dissenting).

At the outset, I suggest that when we say that we are determining whether a nondisclosure "could . . . in any reasonable likelihood have affected the judgment of the jury," Napue v. Illinois, 360 U.S. 264, 271, 79 S. Ct. 1173, 1178, 3 L. Ed. 2d 1271 (1959), we are engaging in a legal fiction. We are really deciding whether it would have affected our own judgment. There is no way that we can determine how twelve jurors would have reacted if the state had lived up to its duty of conducting a fair prosecution by revealing the full details of the deal when Carita took the stand. The real question is whether the defendant was deprived of a fair trial.

The basic facts are undisputed. Prior to testifying, a New Hampshire sheriff, with the assistance of the state's Attorney General's office, promised Carita that, in return for testifying, certain steps would be taken to protect him. At the close of direct examination, the trial prosecutor deliberately asked Carita whether any promises or inducements had been made. His negative answer was a flat lie. At this point, the prosecutor had a duty to immediately disclose to the judge and jury the bargain that had been struck between the state and the witness. The failure to do so was compounded when the prosecutor in closing argument again stated that there was no deal.

The suggestion made on appeal that the trial prosecutor did not know of the promises made to Carita by the state cannot be taken seriously. If he did not know, then he was either grossly negligent or the members of his office who actively participated in the bargaining concealed it from him. Either way, the New Hampshire Attorney General's office presents a sorry picture.

I agree with my brethren that the cross-examination did inform the jury that some sort of a bargain had been made prior to the witness testifying. It is anomalous that what saves the verdict for the state is the cross-examination of defense counsel. Had defense counsel relied on the integrity of the state and not explored the issue of state promises in return for testimony, there is no doubt that a new trial would be required. If it were not for the prosecutor's remarks in closing argument, I would agree that, despite the palpable misconduct of the Attorney General's office, the defendant, due to the cross-examination of his attorney, received a fair trial. However, for me, the direct suggestion by the prosecutor to the jury at the close of the case that there was no deal tips the scale. In my opinion, defendant was deprived of his sixth amendment right to a fair trial because of prosecutorial misconduct.