At a jury trial in the state court, petitioner was found guilty of violating two sections of the California Penal Code; section 187 (murder in the second degree) and section 12022.5 (use of a firearm in the commission of the named offense). After exhausting his rights of direct appeal, petitioner sought a writ of habeas corpus in the district court. Therein, petitioner argued that he had been denied constitutional due process through the admission of an in-court identification by a prosecution witness who had been involved in an encounter with petitioner prior to trial. The district court granted the writ, and the state now appeals.

There are two essential issues presently before the Court. First, can an accidental, pre-trial encounter between an accused and a prosecution witness ever support a federal habeas corpus claim? If so, second, did the district court herein properly find that petitioner had been denied constitutional due process by the admission of the challenged in-court identification?

FACTS

The victim, Eddie Hunter, was managing a restaurant on July 13, 1975, at approximately 2:30 a. m., when he was fatally shot by an assailant with a pistol. Although petitioner admitted at trial that he had been inside the restaurant at the time of the shooting, petitioner argued in his defense that he had been simply an innocent bystander in the shooting episode.

At trial, the prosecution called four witnesses, all of whom had been present at the scene of the crime. Nonetheless, the prosecution’s case rested primarily on the testimony of one witness, an eyewitness to the shooting, David Terry. In his testimony, Terry detailed the following scenario. In the early morning hours of July 13,1975, in the restaurant which the victim managed, the victim and his assailant became involved in an argument over a dice game. A few minutes after the argument, the assailant left the restaurant, went out to his car, and removed from the car a .22-caliber pistol. (At this time, the assailant was approximately thirty feet away from the witness Terry.) After he had procured the pistol, the assailant returned to the restaurant and confronted the victim with his weapon. During the confrontation, the victim drew his own weapon, a .357-magnum revolver, and struck the assailant on the left side of his head with it. Immediately thereafter, two or more rounds of shots were fired. In his trial testimony, Terry identified petitioner as the man he had seen shoot the victim.

The events which occurred immediately following the shooting are not disputed by the parties. Within a moment or two after the shots had been fired, the victim, then fatally wounded, came out of the restaurant and walked toward the parking lot. A few seconds later, petitioner, who had been shot in the armpit area, also came out of the restaurant and stumbled across the street. A police officer, who had been dispatched to the scene of the shooting, approached petitioner, discovered his wound, and drove him to an emergency room of a nearby hospital. Although petitioner had initially stated that he had been shot while walking down the street, petitioner maintained at trial that he had been caught in the cross-fire in the restaurant.

The appeal now before the Court focuses on a series of incidents which began approximately four hours after the fatal shooting. At that time, the witness Terry viewed a photo array that had been prepared by police officials. Although the photo array included two photographs of petitioner, Terry selected a man named Hightower as the man who had fatally shot the victim earlier that morning. A few days after the shooting, Terry disappeared without explanation for a period of more than three months. During this period, the state considered dismissal of its case against petitioner due to the absence of its key witness.

Terry was eventually relocated on October 17, 1975. A short time thereafter, police scheduled a line-up in order to resolve the questions that had been raised by Terry’s previous “photo array” identification of Hightower. The line-up was scheduled for Monday, October 20, 1975, and Terry was warned not to have any contact with police prior to that time. Nonetheless, and in spite of the warnings to the contrary, Terry went to a local police station on Sunday, October 19, and asked police officials to provide him with protective custody. Terry smelled of alcohol, and the police agreed to allow him to “sleep off” his intoxication in a holding cell at the station house.

As chance would have it, petitioner was being held by police at the same station house where Terry had gone to seek protective custody. Moreover, petitioner was conferring with his attorney at the exact time that Terry was in the station house on October 19. The conference between petitioner and his attorney concluded during the lunch hour, and consistent with the jail policy that no inmate was to be returned to his cell during a meal period, petitioner was placed in a holding cell — the same holding cell in which Terry was sleeping. Terry awoke a short time after petitioner had been placed in the cell, and he remained there with petitioner for approximately one hour. Terry did not appear to recognize petitioner until, near the end of the hour period, a booking officer asked both Terry and petitioner to identify themselves by name.

ACCIDENTAL NATURE OF THE ENCOUNTER

The state courts and the district court herein all agreed that the jailhouse encounter between petitioner and the witness Terry had been an accidental, inadvertent encounter. The state argues, with reference to this finding, that an accidental, pre-trial confrontation between an accused and a prosecution witness can never support a state prisoner’s claim of a due process violation. We disagree.

The state’s argument is premised on the incorrect conclusion that the fundamental purpose of judicial review in the context of a pre-trial confrontation is the deterrence of culpable police conduct. The state argues, based on this conclusion, that an accidental encounter should never give rise to habeas corpus relief because the exclusion of the challenged in-court identification will not serve any deterrent purpose. Although we recognize that there is no intentionally wrongful police conduct involved in an accidental encounter, we also recognize that the deterrence of such conduct is not the primary purpose behind judicial review of tainted identification testimony. Rather, a court reviews a challenged in-court identification essentially to determine whether the witness’ testimony retains sufficient indicia of reliability. In Manson v. Brathwaite, 432 U.S. 98, 114 , 97 S.Ct. 2243, 2253 , 53 L.Ed.2d 140 (1976), the Supreme Court stated clearly that “. . . reliability is the linchpin in determining the admissibility of identification testimony . . ” which may have been affected by a pre-trial confrontation. It is certainly possible that an in-court identification by a prosecution witness may prove to be unreliable, even though the pre-trial encounter in question has not involved any culpable police conduct.

A trial court’s focus on the reliability of identification testimony is consistent with the fact that the accidental nature of a pre-trial encounter is often an important factor in a court’s decision to admit a witness’ subsequent testimony. In this regard, the state again fails to understand that the accidental nature of a pre-trial encounter is important, not because there has been an absence of wrongful police activity, but only because the circumstances surrounding an accidental encounter are often directly relevant to the question of the reliability of the in-court identification. More specifically, the accidental nature of a pre-trial encounter plays a role in a court’s consideration of identification testimony because the majority of accidental encounters do not involve any significant degree of suggestiveness. A witness who has been involved in an accidental encounter has generally seen a defendant in a seemingly innocent light, or as stated by the district court herein, “. .an accidental confrontation usually gives no indication to the witness that ‘this is the man.’ ” Accordingly, the witness’ subsequent identification is not influenced by the encounter to a significant degree. A typical case is United States v. Massaro, 544 F.2d 547 (1st Cir. 1976), cert. denied, 429 U.S. 1052 , 97 S.Ct. 766 , 50 L.Ed.2d 769 (1977), wherein the First Circuit concluded that a defendant’s due process rights had not been violated by an accidental encounter which had been “. pure happenstance without suggestive circumstance.” The encounter in Massaro had occurred in a hallway of a courthouse, and the First Circuit described the scene as follows:

544 F.2d at 550 . See also United States v. Colclough, 549 F.2d 937, 941-42 (4th Cir. 1977).

However, the fact that most accidental encounters do not involve a significant degree of suggestiveness does not mean that every accidental encounter is automatically above constitutional scrutiny. Rather, a court is obligated to review every pre-trial encounter, accidental or otherwise, in order to insure that the circumstances of the particular encounter have not been so suggestive as to undermine the reliability of the witness’ subsequent identification. The Supreme Court has instructed explicitly that “it is always necessary to ‘scrutinize any pre-trial confrontation.’ ” Kirby v. Illinois, 406 U.S. 682, 690 , 92 S.Ct. 1877, 1883 , 32 L.Ed.2d 411 (1972), quoting United States v. Wade, 388 U.S. 218, 227 , 87 S.Ct. 1926, 1932 , 18 L.Ed.2d 1149 (1967). (Emphasis added by Kirby Court.) Further, this obligation of the courts to scrutinize all pre-trial encounters is applicable both in the context of direct appeal, as well as in the context of petitions for writs of habeas corpus. See, e. g., Baker v. Hocker, 496 F.2d 615, 617 (9th Cir. 1974); Cassasa v. Nelson, 452 F.2d 1083, 1084 (9th Cir. 1971). Thus, the district court herein was correct in reviewing the jailhouse encounter between petitioner and the witness Terry for the purpose of determining whether the encounter had affected Terry’s in-court identification in such a way that the admission of the identification denied petitioner constitutional due process.

FINDINGS AND CONCLUSIONS

The Supreme Court has set down a two-step test to determine whether an in-court identification has been irreparably tainted by a pre-trial encounter between a witness and an accused. First, a court must determine whether the challenged pre-trial encounter was unnecessarily or impermissibly suggestive. Second, the court must examine the totality of the circumstances in order to decide whether the witness’ in-court identification is nonetheless reliable. See Manson v. Brathwaite, supra, 432 U.S. 98, 114 , 97 S.Ct. 2243, 2253 , 53 L.Ed.2d 140 (1976); Neil v. Biggers, 409 U.S. 188, 199-200 , 93 S.Ct. 375, 382-383 , 34 L.Ed.2d 401 (1972); Simmons v. United States, 390 U.S. 377, 384 , 88 S.Ct. 967, 971 , 19 L.Ed.2d 1247 (1968); Stovall v. Denno, 388 U.S. 293, 302 , 87 S.Ct. 1967, 1972 , 18 L.Ed.2d 1199 (1967).

As to the first step of the above test, the district court correctly concluded that the encounter between petitioner and the witness Terry, although accidental, had been unnecessarily and impermissibly suggestive. There are at least three aspects of the encounter in the present case which distinguish it from those accidental encounters that do not involve highly suggestive circumstances. First, the setting of the encounter between petitioner and Terry strongly suggested to Terry that petitioner had been accused of a crime. Unlike the situation in Massaro, supra, where there had been “no evidence of custody,” the confrontation in the jailhouse cell clearly labeled petitioner as a criminal defendant. Second, the mention of petitioner’s name by the booking officer identified petitioner specifically as the state’s suspect in the victim’s killing. In other words, at the moment when petitioner identified himself in the holding cell, Terry realized that he was looking at the man who had been accused of the Hunter murder. Finally, the encounter in the present case was a result of the state’s negligent exercise of its control over both the witness and the accused.

After the district court had determined that the encounter between petitioner and Terry had been impermissibly suggestive, the court then proceeded to examine the overall reliability of Terry’s in-court identification. After a detailed analysis, the court concluded that Terry’s testimony had lacked the requisite indicia of reliability. On appeal, the state raises a variety of objections to the court’s conclusion on this issue.

First, the state argues that the district court acted contrary to 28 U.S.C. § 2254 when it examined the trial record with respect to the issue of reliability. In short, the state contends that the district court improperly ignored the state court’s determination, made at the pre-trial suppression hearing, that Terry’s testimony was admissible. The district court, in its order granting the state’s motion for a stay pending appeal, correctly responded to this argument as follows:

Moreover, section 2254(d) provides that, in a habeas corpus proceeding,

However, the state court in the present case essentially made only one factual finding in regard to the pre-trial encounter between petitioner and the witness Terry; namely, that the encounter had been accidental in its nature. Based on this finding, the state court concluded that the encounter, because of its accidental nature, could not have been unduly suggestive. (Note that the district court agreed with this factual finding of the state court, but disagreed with its legal conclusion.) In other words, the state court, after it had made a factual determination relevant to the issue of the suggestiveness of the encounter, never proceeded to consider any facts relating to the second issue of the reliability of Terry’s identification, given the totality of the circumstances. Section 2254(d)(1) provides that a determination of a state court need not be presumed correct if “the merits of the factual dispute were not resolved in the State court hearing.” 28 U.S.C. § 2254 (d)(1). As noted above, the district court properly concluded that the pre-trial encounter between petitioner and Terry, even though accidental, had been unnecessarily and impermissibly suggestive. Once it had arrived at this conclusion, the district court, was then obligated to examine the overall reliability of Terry’s subsequent identification. Because the state court had made no factual determinations in this regard, the district court properly proceeded to conduct its own investigation, in accordance with section 2254(d)(1).

Next the state contends that, contrary to the conclusion of the district court, there is sufficient evidence in the trial record to support the reliability of Terry’s in-court identification. The Supreme Court in Neil v. Biggers, supra, 409 U.S. 188, 199 , 93 S.Ct. 375, 382 , 34 L.Ed.2d 401 (1972), specified a number of factors that a court must consider in determining whether a witness’ identification is sufficiently reliable. These factors are . . the opportunity of the witness to view the criminal at the time of the crime, the witness’ degree of attention, the accuracy of his prior description of the criminal, the level of certainty demonstrated at the confrontation, and the time between the crime and the confrontation.” Manson v. Brathwaite, supra, 432 U.S. 98, 114 , 97 S.Ct. 2243, 2253 , 53 L.Ed.2d 140 (1976). We need not reiterate all of the detailed findings that were made by the district court as to these various factors. Suffice it to say that, in this regard, the court found as follows: (a) the witness Terry had viewed petitioner in the restaurant for at the most five minutes; (b) Terry had been merely a casual observer of the activity in the restaurant; he had not concentrated his attention on the assailant; (c) approximately four hours after the shooting incident, Terry had selected a photograph of a man other than petitioner as the assailant; further, Terry had made this selection despite the fact that he had seen two photographs of petitioner; (d) Terry gave no indication that he recognized petitioner during their jailhouse confrontation until after petitioner’s name had been mentioned; (e) more than three months elapsed between the shooting incident and Terry’s subsequent confrontation with petitioner on October 19. These findings of the district court adequately support the conclusion that Terry’s in-court identification lacked sufficient reliability, given the totality of the circumstances.

Finally, the state argues that the district court improperly refused to take into account other evidence of petitioner’s guilt when it considered the reliability of Terry’s identification testimony. We again disagree. The factors set out by the Supreme Court in Neil v. Biggers, supra, all relate to the quality of a witness’ personal observations and remembrances. Neither the incriminating testimony of other witnesses, nor any other evidence of a defendant’s guilt, add to, or detract from these personal characteristics of an identification. Although other evidence of guilt is relevant, such evidence is relevant only to a determination of whether an admission of a tainted in-court identification is error that is harmless beyond a reasonable doubt. Such evidence is not relevant to the issue of the reliability of an in-c'ourt identification. See Manson v. Brathwaite, supra, 432 U.S. at 118 , n *, 97 S.Ct. at 2255 (Stevens, J., concurring); Landry v. Alabama, 579 F.2d 353 (5th Cir. 1978); United States v. Jarvis, 560 F.2d 494 (2d Cir. 1977), cert. denied, 435 U.S. 934 , 98 S.Ct. 1511 , 55 L.Ed.2d 532 (1978).

As to the final issue of harmless error, we agree with the district court that the admission of Terry’s in-court identification was not error that was harmless beyond a reasonable doubt. See Chapman v. California, 386 U.S. 18, 24 , 87 S.Ct. 824, 828 , 17 L.Ed.2d 705 (1967). The other evidence of petitioner’s guilt was far short of overwhelming in its impact. For example, Terry was the only witness who had testified at trial that he had seen petitioner with a gun. Although the testimony of the other witnesses had connected petitioner to an earlier argument with the victim, Terry’s testimony was the only testimony that connected petitioner to the shooting itself. Moreover, it should be noted that any contention by the state that the admission of Terry’s testimony was harmless beyond a reasonable doubt flies in the face of the state’s own admission that it had considered dismissal of its case against petitioner during the three-month period of Terry’s absence.

The district court’s issuance of a writ of habeas corpus is affirmed.