Jarrett v. Headley

MEMORANDUM AND ORDER EDELSTEIN, District Judge: Patsy Kelly Jarrett (“Jarrett”) petitioned this court for a writ of habeus corpus. This matter was referred to Magistrate Michael H. Dolinger who issued a Report and Recommendation dated November 26, 1985. Magistrate Dolinger recommended that the petition be granted and that respondent be directed to release petitioner unless she is retried on the charges against her within sixty days. Objections were filed by respondent and petitioner filed a response. Respondent, by letter, submitted a reply. The Report and Recommendation of the Magistrate is hereby adopted by this court and the petition is granted. DISCUSSION 1 Respondent objects to the Magistrate’s Report and Recommendation on four grounds. First, that the Magistrate did not afford the proper deference to the factual *1404 findings of the state court; second, that the Magistrate’s finding that the identification procedure was unfair was incorrect; third, that the reliability of the identification was strengthened by other corroborating evidence; and finally, that the claim that pretrial publicity affected the identification was not raised in state court and therefore the petition must be denied. The court hereby adopts the Magistrate’s Report and Recommendation. Respondent’s objections are addressed herein. Deference to State Court Findings Respondent asserts that pursuant to Title 28, Section 2254(d) of the United States Code, the findings of the state court are entitled to a presumption of correctness. The facts underlying the state court’s decision on the constitutionality of the identification are governed by the statutory presumption. Sumner v. Mata, 455 U.S. 591, 597, 102 S.Ct. 1303, 1306, 71 L.Ed.2d 480 (1982) (per curiam). The ultimate question regarding the constitutionality of the identification, however, is a mixed question of law and fact that is not governed by Section 2254(d). Id. The court finds that the Magistrate gave the proper deference to the state court’s findings. The Magistrate did not make any factual findings that would be precluded by Section 2254. Rather the Magistrate relied on the facts established during the state court proceedings. Identification Procedure Respondent claims that the Magistrate incorrectly determined that the identification procedure was unfair. Respondent’s arguments are unpersuasive. The events leading up to the in-court identification 2 must be examined to determine if the identification was unconstitutionally tainted. In objecting to the Report, respondent assumes that the only events that affected the in-court identification were the events leading up to the photographic identification made by William Hyland (“Hyland”). This is incorrect. All of the events occurring after the photographic identification and prior to the testimony by Hyland at trial are also relevant. For example, discussions Hyland had with the police following the photographic identification and his discussion with the District Attorney before taking the witness stand at trial are clearly factors that the court must consider. Further, the respondent examines each event affecting the photographic identification in isolation. This is also incorrect. Thus, the Magistrate correctly considered the effect of the combination of the pre-trial factors in making his determination that the in-court identification was the result of unduly suggestive procedures. See Dickerson v. Fogg, 692 F.2d 238, 245 (2d Cir.1982). For the reasons set forth in the Report and Recommendation, the court finds that the pre-trial occurences were unduly suggestive. Further, the court agrees with the Magistrate that there was no independent and reliable basis for the in-court identification by Hyland applying the factors set forth in Neil v. Biggers, 409 U.S. 188, 199-200, 93 S.Ct. 375, 382, 34 L.Ed.2d 401 (1972). Corroborating Evidence Respondent asserts that the reliability of Hyland’s identification is strengthened by corroborating evidence and was therefore reliable. There are two possible interpretations of respondent’s assertion. The respondent may be asserting that even if the court finds that the in-court identification was flawed, there was sufficient additional evidence to support a conviction so that any error was harmless error, see People v. MacKay 98 A.D.2d 732, 469 N.Y.S.2d 146, 147 (2d Dep’t 1983) (per curiam). This is not the case here. The crucial evidence against the petitioner was the identification testimony of Hyland. See People v. Sapp, *1405 98 A.D.2d 784, 469 N.Y.S.2d 803, 804 (2d Dep’t 1983) (per curiam). If respondent is asking the court to consider this additional evidence in making a determination regarding the reliability of the identification itself this would be improper. Such “corroborating” evidence is clearly inconsistent with the factors traditionally examined to determine reliability. See Manson v. Brathwaite, 432 U.S. 98, 114-16, 97 S.Ct. 2243, 2253, 53 L.Ed.2d 140 (1977). Pre-trial Publicity Finally, respondent contends that the Magistrate’s consideration of pretrial publicity in determining whether the in-court identification was tainted was incorrect because petitioner failed to raise this issue in the state court. 3 It does not appear, however, that the Magistrate considered the pretrial publicity in making his determination. Rather the Magistrate simply indicated that Hyland was aware of the publicity. The Magistrate also indicated that the publicity was not the fault of the police indicating that this would not be considered in his determination although it may have affected the strength of the in-court identification. The court agrees with the Magistrate. Remaining Evidence As the Magistrate noted “[t]he prosecutor’s case plainly was far from overwhelming.” Report at 58. However, additional evidence was presented. The court agrees with the Magistrate that this remaining evidence is “sufficient to pass constitutional muster.” Id. Therefore, the respondent is ordered to release the petitioner unless she is retried on the charges against her within ninety days in conformity with this Memorandum and the Magistrate’s Report and Recommendation adopted herein. CONCLUSION The petition is hereby granted. Respondent is directed to release the petitioner unless petitioner is retried on the charges against her within ninety days of the filing of this Memorandum in conformity with this Memorandum and the Magistrate’s Report and Recommendation which is adopted herein. SO ORDERED. APPENDIX REPORT AND RECOMMENDATION MICHAEL H. DOLINGER, United States Magistrate. A. Introduction On August 11, 1973 near Utica, New York, a seventeen year old gas station attendant was murdered at his place of work and the gas station robbed of approximately $279.00 in cash receipts. More than three and a half years later, on March 29, 1977, two individuals were convicted of these crimes following a jury trial. Based upon their convictions on two counts of murder and two counts of robbery, the defendants were each sentenced to prison terms of twenty-five years to life. 1 One of those two defendants, Patsy Kelly Jarrett, has now petitioned for a writ of habeas corpus. She asserts principally *1406 that the key evidence against her—the testimony of the one witness who specifically identified her as being at the gas station at about the time of the killing—was the result of impermissibly suggestive police procedures and was otherwise so unreliable that its introduction denied her due process. For the reasons that follow, I recommend that the petition be granted. The Nature of the Petitioner’s Claims By her original petition, filed January 23, 1984, petitioner raised three claims. First, she asserted that her conviction had been based upon the eyewitness identification testimony at trial of one William Hyland, and that this testimony was purely the product of a series of efforts by the local police to encourage Mr. Hyland, long after the incident, to conclude that he could positively identify her as being present at the gas station. According to the petition, absent these suggestive procedures, the witness would not have so testified, as evidenced by his repeated statements prior to trial, and his testimony was entirely unreliable. Second, petitioner argued that, even if admissible, Hyland’s testimony was incredible as a matter of law and, as a result, the evidence in the record was constitutionally insufficient to sustain her conviction. Petitioner’s third claim—now withdrawn— was that the denial of her pre-trial motion for a severance denied her due process. In his response, respondent asserted that the petition must be dismissed because petitioner had failed to exhaust available state remedies with respect to her severance claim. He further argued that the pre-trial and trial record fully justified the introduction of the in-court identification of petitioner as reliable and not the product of police suggestion and that the denial of her motion to sever was proper. Apparently as a result of respondent’s exhaustion argument, in her reply papers petitioner limited her severance argument, 2 and thereafter withdrew the claim. 3 Accordingly the only issues to be addressed relate to the identification question and the sufficiency of the evidence. B. The Identification Process The state court pre-trial and trial record reflects a lengthy sequence of events leading ultimately to the eyewitness identification of petitioner at trial. Because the state court did not make detailed findings of historical fact, there are few if any findings to which deference is owed under 28 U.S.C. § 2254(d) and accordingly I rely directly on the state evidentiary record. 4 On August 13, 1973, two days after the murder in question, William Hyland was interviewed by the police concerning the events he had observed the prior date at the Seaway gas station. (Tr. 688-90.) 5 Hyland’s observations proved to be crucial because the evidence at trial tended to establish that the gas station attendant had been murdered at around the time that he was at the station. *1407 During the course of Hyland’s interview, his version of the facts was encapsulated in a statement typed by an interviewing police investigator (Tr. 693-94) and signed under oath by Mr. Hyland. (Deft.Exh.A) According to the statement, Mr. Hyland drove into the station at approximately 12:50 P.M. on August 11, and parked next to the gas pumps on the side nearest the road. Since no attendant was in view, Hyland got out of his car and stood for approximately one-and-a-half minutes behind his car. At that point, an automobile pulled out from the west side of the station, backed up, and then drove forward and parked parallel to the gas pumps on the side nearest the station building. Hyland’s description of the driver was as follows: I am not sure, but I believe the operator was a white female. She had long black shoulder length hair and was wearing dark clothing. I saw no clothing or baggage inside of this car. Upon moving closer to the gas pumps I observed this female going through items in a brown hexagon type pattern pocketbook. I noticed this pocketbook had gold trim on it. The material that it was made of appeared to be of an alligator type imitation. The type of hair style that this person had did not allow me to see her face. The girl did have a tan. I had never seen the girl or the car before. According to Hyland, “[a]fter the car had parked and a few seconds had elapsed,” he observed a white male emerge from the station and approach Hyland’s car. The man, whom Hyland described in his statement, walked to a position between the two cars and asked “which one is first[?].” Hyland told the man that he could pump either car, and the man apparently indicated that he would first take care of Hyland. Hyland then asked for $5.00 worth of gasoline and the man pumped the gas for him, at all times holding the nozzle of the hose. While this occurred, Hyland and the man engaged in “general conversation about the weather.” When the gasoline had been pumped, Hyland got in his car and left. According to the statement, Hyland “did not pay any attention as to what service [the man] offered the other car the girl was sitting [in].” By Hyland’s estimate, he left the station at 12:55 P.M., approximately five minutes after arriving. At the time that he first talked to the police Hyland was able to provide a relatively detailed description of the man who had pumped the gasoline. As a result of these discussions—which occurred on four occasions—the police subsequently prepared and circulated a sketch of the man. (Tr. 242-45, Gov’t.Exh. 28.) No such sketch was made of the person who had driven the other car at the Seaway station. (Tr. 277.) Following the preparation of this sketch, Hyland was not contacted again by the police until December 1975, approximately 28 months after the murder. (Tr. 246.) At that time a police officer came to Hyland’s home with two photo spreads to show him. He first displayed a set of ten to twelve photographs of males. Hyland narrowed the choice to two photographs, and when the officer told Hyland that the photos were of the same man (Tr. 316-17), he selected one as the man who had pumped gas for him on August 11, 1973. (Tr. 316, Gov’t.Exh. 7.) The photograph selected was of Billy Ronald Kelly, petitioner’s alleged paramour. (Wade Tr. 82). The officer then gave Hyland a set of twelve photographs of women. (Wade Tr. 57-58, Gov’t.Exhs. 16-27.) Of these, eight were unmarked photographs (Gov’t.Exhs. 19-26), 6 three were marked “New York State Police” (Gov’t.Exhs. 17,18 & 27), and one bore the notation “Sheriff’s Department.” (Gov’tExh. 16.) Hyland, who observed the markings at the time (Wade Tr. 73), initially selected two photographs, each of a different woman (Wade Tr. 74-75 & Gov’tExhs. 16 & 17), and both of which displayed law enforcement notations. He *1408 ultimately signed the back of one (Gov’t. Exh. 16), according to his Wade testimony, because it was “possibly” the person who had driven the other car at the gas station (Wade Tr. 61), although the second photograph also “look[ed] like” the driver. (Wade Tr. 75.) In explaining his choice to the grand jury, he stated that he had selected the photograph from the array because “[i]t was the same style hair, long hair,” as the driver had worn on the day of the murder (GJ 37), and confirmed that the selected photograph “could be” the driver “but [he couldn’t] say for sure.” (GJ 38.) The photograph that Hyland chose was of petitioner. Before leaving Hyland after he had made his selection from the photo array, the police officer engaged him in conversation about his choice. Specifically the officer told Hyland “who they might be,” their names, and “where they were from.” (Tr. 286-87.) Such disclosure meant that Hyland was informed that the two individuals were both from North Carolina and, quite probably, that they were believed to have been living together in Utica at the time of the murder. 7 Three months later, in March 1976, Hyland appeared before a grand jury. At that time he testified—in terms similar to his original statement—that the driver “looked like a female.” (GJ 30.) When asked whether he could positively identify the photograph of petitioner as depicting the driver, he stated: “Well, I can’t say positive about this, about the way—it was the same style hair, long hair.” (GJ 38.) As noted above, he agreed with the prosecutor’s characterization: “Is it safe to say then that the best you can say is that it could be the girl but you can’t say for sure?” (GJ 38.) The grand jury returned an indictment naming both Kelly and petitioner. Subsequently petitioner was arrested in her native North Carolina, and returned to Utica. The State never placed her in a line-up. Rather, in the course of pre-trial proceedings, the prosecutor informed petitioner’s counsel that the State intended to rely at trial on an identification by Hyland of a photograph of petitioner “as possibly being the girl seen seated in blue car at [the] station—” (See “Notice Pursuant to Section 710.30 in Petitioner’s Appendix II at 32.) Following this notification, petitioner moved to suppress any identification testimony by Hyland. 8 At the Wade hearing, which was commenced on January 31,1977, Hyland recounted his selection of photographs from the two photo arrays presented to him in December 1975. He denied that the police officer had given him any information prior to his making the selections (Wade Tr. 54), but admitted that he had noted the “Sheriff” and police markings on some of the photos. (Wade Tr. 73.) With respect to his choice from the female photo array, he testified that he had narrowed the selection to two photographs and that he had picked one because he “thought that was the one” although he also thought that both of the photographs looked like the person. (Wade Tr. 59, 75.) In explanation, Hyland agreed with the prosecutor that he had selected the photograph “as possibly being the female that was in the gas station.” (Wade Tr. 61.) Although petitioner was present in court—apparently at counsel table—Hyland was not asked to state either whether she was the person in the photograph that he had selected or whether she was the person he had seen in the car at the Seaway station. Indeed, Hyland’s only apparent reference to petitioner was to contrast her appearance with that of the person in the car. Thus, when asked if he could describe the person in the car, he said: A. Yes. It was a question whether I was going to be served first with the gasoline or the girl. And this girl had *1409 long, darker hair, it seemed, than the one that’s right there, now. Q. Darker than the one right there, right now, your’re referring to who? A. That girl, right, there. Q. At the Defense table? A. Yes. (Wade Tr. 79.) Apart from the description of the hair, Hyland did not refer to any other physical characteristics of the driver until asked by defense counsel whether he had “any idea how tall the girl was____” (Wade Tr. 79.) After noting that he had never seen the driver stand up but that “[s]he could see into the car mirror____” (ibid.), Hyland said, under further questioning, that he “wouldn’t be surprised if she was as tall as I am; maybe a little bit taller.” (Wade Tr. 80.) 9 Again, at no time did he refer to petitioner, who was seated in front of him. After the conclusion of the Wade hearing, the presiding justice denied the motion to suppress. 10 In so ruling the Court recounted the testimony, and appeared to assume, quite erroneously, that Hyland had identified petitioner at the hearing, rather than simply confirming that he had previously selected one photograph from an array. (Memorandum at 1404.) In any event, the Court held that the only question to be determined for the purpose of weighing the validity of an in-court identification in this case was whether the photo identification procedure was so suggestive as to create a very substantial likelihood of irreparable misidentification. (Id. at 1404.) In rejecting this possibility, the Court found the photo arrays to have been fair in that they presented a number of persons of generally similar physical appearance (id. at 1404), and it further found that the police had not made any suggestive comments to Hyland before he made his selections. (Ibid.) As for the police markings on some of the photographs, the Justice indicated that this was insignificant because he recalled Hyland having testified that he had believed at the time that he was being shown only police photographs (ibid.); in fact, the transcript does not bear out this finding since Hyland testified that he did not know the origin of the unmarked photographs. (Wade Tr. 72.) 11 Finally, in support of the reliability of the identification, the Court found that Hyland had had a good opportunity to view the driver from a short distance. (Memorandum at 1404.) Trial began March 14,1977, approximately three and one-half years after the murder. At trial the only identification of petitioner as having been present at the Seaway gas station on the day of the murder was offered by Hyland. Just prior to Hyland taking the stand, the prosecutor conferred briefly with him, showing him the photographs of Kelly and petitioner that he had signed in December 1975 and urging him, in Hyland’s words, to “stick to your guns.” (Tr. 248-49, 304-05.) Hyland then took the stand and, for the first time, positively identified petitioner as the person who had driven the other car. When pressed on cross-examination he insisted that, regardless of his sworn statement two days after the murder, he was certain that the driver was female because of the way the driver used a comb. (Tr. 267.) As for his identification of petitioner, he indicated that on the day in question he could see the driver’s cheeks, nose, eyes and forehead, and that he had based his prior identification on the similarity of these features in the photograph. (Tr. 254-55.) Notwithstanding this testimony, he conceded having told the police two days after the murder that he had been unable to see the driver’s face because of the long *1410 hair. (Tr. 256.) He also appeared to concede that at the Wade hearing he had not recognized petitioner as the driver. (Tr. 256-57.) He further conceded that he had spent perhaps five minutes at the station (Tr. 264); that the man appeared to pump his gas within perhaps thirty seconds after his arrival (Tr. 260); that he had focussed more on the man than on the woman while at the gas station (Tr. 266); that, to the extent he looked at the other car, he was concentrating on the outside of the vehicle because he was impressed with its finish (Tr. 299-300); that he could not see much of the woman’s face except once, when she turned her head for a few seconds (Tr. 311); and that when the man came out and pumped the gasoline he looked mostly at the man. (Tr. 312.) As for his August 13, 1973 signed statement, he indicated that the police has drafted it and that he had signed it without reading it, although he might have “scanned” it. (Tr. 279-80.) 12 C. The Trial 1. The State’s Case At trial the prosecution called twenty-six witnesses. The gist of their testimony is briefly summarized. In the late spring or early summer of 1973, Billy Kelly and petitioner drove from North Carolina to Utica in a blue sedan owned by petitioner. They initially stayed for a few days at the home of Linda and Martin Comstock, whom they knew through a mutual friend. (Tr. 65, 67.) They then moved into an apartment in Utica (Tr. 114-15), where they remained until mid-August, when they departed in the same automobile, approximately one or two days after the Seaway murder. (Tr. 118-19.) While in Utica, Kelly worked on a construction site (Tr. 114), and was seen a number of times driving the car by himself. (Tr. 69, 75.) During that period both Kelly and petitioner frequented a local gay bar, and Kelly entered into a sexual relationship with a male patron of the bar named Billie Sullivan. (Tr. 95, 105.) During the morning of August 11, 1973, the Seaway gas station attendant, Paul Hatch, was alive and carrying out his job. (Tr. 126-27.) At 12:50 P.M., a customer named William C. Wheeler drove into the station and was waited on by Hatch at the gas pump. While standing there, the driver noticed another car pull into the far end of the station and stop. A female emerged from the passenger side, went to the driver’s side and talked to the male driver. She then walked to the side of the building and apparently entered. The witness was unable to identify the car (Tr. 172-76) and gave a tentative “impression” of the woman as “maybe” 5'5" or 5'6", medium build, twenty to twenty-five years old, with light brown hair and buxom. (Tr. 168, 171-72.) The witness further testified that he received his change from Hatch and then departed, with the other car still apparently parked at the station. (Tr. 167.) The witness also confirmed that despite his having provided his information to the police after the killing, they subsequently announced that they were looking for two males as their suspects. (Tr. 177.) At approximately 12:55 P.M., another witness, Theresa J. Law, drove along the road past the Seaway gas station. Driving at approximately thirty miles per hour, she observed a Volkswagen at the gas pump and an attendant coming out of the building, presumably to pump gas. She also saw a blue car parked at the west side of the station with no one near it. She thought it might be a Plymouth and, when shown a photograph of petitioner's car, said it was similar. (Tr. 182-83.) Another witness, Frank Mathalia, testified that he had driven past the Seaway station some time between 12:45 and 1:15 P.M. Although he was driving at thirty miles per hour and looking principally at the road, he reported seeing three people in *1411 the office of the station. He described one as a “girl” of medium height with long black hair, who was looking out of the window. The other two, he stated, were males engaged in a conversation. He further stated that no one else was visible and that there was no car parked at station. (Tr. 192-95.) Still another witness, Sharon Scribner, reported that shortly before 1:00 P.M. she had pulled into the station and seen no one around. (Tr. 198.) She observed a maroon Ford parked at the side of the building, but this car she had frequently seen at the station. As she departed, she observed in her rear-view mirror the taillight of another car near the maroon Ford. Although she saw less than six inches of the rear of the car (Tr. 208-09), she deduced that it was a blue Ford (Tr. 201-02) although she ultimately was unsure of the differences in appearance between a Ford and a Plymouth. Hyland testified to having entered the station at about 12:55 P.M. (Tr. 215.) Unlike his testimony at the Wade hearing— when he stated that he had seen no one at the station when he entered—he now testified that he had seen a man and woman between the gas pumps and the building. (Tr. 216.) According to Hyland, the man entered the building and he did not notice where the woman went. (Tr. 218.) He then recounted the appearance of the blue car from around the corner of the building and the emergence of a man from the building (Tr. 220-21) and specifically identified Kelly as the man who had pumped his gas and petitioner as the person driving the blue car. In all, he estimated that he had remained at the gas station for between four and five minutes. (Tr. 227.) At 1:05 P.M. two men drove into the station and found no one there. They also did not notice any car parked there. (Tr. 344-48, 354-56, 361.) At 1:10 P.M., Glen Adams dropped his daughter off at the station, as she liked to help Hatch pump gas. (Tr. 366-67.) She saw no one outside and went into a back room, where she saw Hatch lying on the floor. (Tr. 371-72, 377.) At the girl’s request, a postal driver went into the back room and observed Hatch's body with his hands tied behind his back. (Tr. 386-87.) The postal driver called the police after 1:15 P.M. (Tr. 387.) Hatch had been stabbed repeatedly and his throat had been slashed. (Tr. 436-38.) His mouth was taped over, and the tape yielded a fingerprint that was ultimately found to match one of Kelly’s. (Tr. 392, 416, 461, 504-11.) 13 The murder was apparently committed in the course of a robbery. Thus, the district manager of Seaway reported that $279.00 was found to be missing from the station. (Tr. 445-48.) As its last witness, the prosecution called a North Carolina police officer, who reported that in July 1976 he had executed an arrest warrant for petitioner. According to the officer, when questioned prior to her arrest petitioner said that she and Kelly had been living as man and wife in Utica in July and August 1973 and had left in late August. When shown a photograph of the Seaway station, petitioner was said to have stated “It looks familiar” (Tr. 564), and to have acknowledged that she and Kelly frequently stopped at gas stations to use their restrooms, but she stated that she could not recall whether they had ever been at that particular station. (Tr. 566.) She also denied any knowledge of the incident in question. (Tr. 569.) After her arrest, she was questioned again and gave substantially the same answers. (Tr. 572-76.) 2. Defendants’ Case Kelly rested without introducing any evidence. Petitioner, however, called seven witnesses. Petitioner herself testified that she had met Kelly in 1972 and had left North Carolina with him in July 1973. (Tr. 597-98.) They drove in her car to Herkimer, New York, where several friends of Kelly lived. *1412 (Tr. 598-602.) After staying with those friends—Mr. and Mrs. Comstock—for approximately one week, they rented an apartment in Utica. (Tr. 605.) According to petitioner, Kelly had driven the car most of the way from North Carolina, and prior to that time may have used the car in North Carolina without her. (Tr. 606-07.) She testified that, when they settled in Utica, Kelly got a construction job, and used her car each day to drive to and from work while she remained at home. (Tr. 606.) In Utica he also used the car on his own to date other women and at least one man, the government’s witness Billie Sullivan. (Tr. 610-11, 636-37.) She further stated that on one occasion he had used the car overnight on a date with a man. (Tr. 639.) Petitioner testified that she and Kelly had remained in Utica for approximately six weeks but that when her money ran out she had her father wire her $40.00 and then returned to North Carolina in late August 1973. (Tr. 607.) Petitioner admitted that she had been in Utica on the day of the murder but denied that she had been involved in the crime. (Tr. 616-17, 623-24.) Although she could not recall specifically what she had been doing at the time, she speculated that she had either been sleeping, drinking coffee or playing softball. (Tr. 616-17.) Petitioner also called a New York State Police investigator, Ronald Hojnacki, who testified that he had taken Hyland’s statement on August 13, 1973. According to Hojnacki, the statement contained only what Hyland had told him; he further testified that the two of them had prepared it together line by line, and that Hyland had in fact read it when it was completed. (Tr. 691, 693-97.) 14 3. The State’s Rebuttal Case On rebuttal the prosecution called two police witnesses. One testified to having questioned petitioner in December 1975. The other described very briefly how he had presented the two photo arrays to Hyland in December 1975. In this regard he stated that, “prior to [Hyland] looking at these photographs” he had not told the witness “that there was an association between one of the males and one of the females that was in the pile____” (Tr. 738.) D. Verdict and Post-Trial Proceedings Following summations, the jury returned verdicts of guilty on all counts against both defendants. (Tr. 847-52.) Each was sentenced to concurrent terms of twenty-five years to life. Petitioner appealed pro se from the judgment of conviction and on her appeal raised the claims that she is now asserting. (Habeas Petition ¶ 7.) That judgment was affirmed without opinion by the Appellate Division, Fourth Department on April 8, 1980. 15 Thereafter she sought leave to appeal to the New York Court of Appeals. 16 That application was denied on May 27, 1980. 17 Analysis A. The Introduction of the In-Court Identification of Petitioner As the Second Circuit has aptly noted: Centuries of experience in the administration of criminal justice have shown that convictions based solely on testimony that identifies a defendant previously unknown to the witness is highly suspect. Of all the various kinds of evidence it is the least reliable, especially *1413 where unsupported by corroborating evidence. Jackson v. Fogg, 589 F.2d 108, 112 (2d Cir.1978). Because of concern about the multifarious ways in which a witness can become persuaded that a particular individual was in fact the perpetrator of a crime and because such eyewitness identification can have a devastating impact at trial, see, e.g., Kampshoff v. Smith, 698 F.2d 581, 585-86 (2d Cir.1983), 18 the federal courts have developed a set of general guidelines to limit the use of such evidence to those cases in which it appears likely to be reliable. Application of those standards to the particular circumstances of this case demonstrates that the in-court identification of petitioner should have been suppressed. 1. The General Criteria Commencing with Stovall v. Denno, 388 U.S. 293, 302, 87 S.Ct. 1967, 1972, 18 L.Ed.2d 1199 (1967), the Supreme Court has repeatedly recognized that pretrial identification procedures may be “so unnecessarily suggestive and conducive to irreparable mistaken identification” that the introduction of evidence derived from that process—including an in-court identification—would be inconsistent with due process. In such circumstances, the Court has held, the suggestiveness of the procedures is not merely a matter for the jury to consider in evaluating the testimony of the witness. See Neil v. Biggers, 409 U.S. 188, 196-98, 93 S.Ct. 375, 380-81, 34 L.Ed.2d 401 (1972); Coleman v. Alabama, 399 U.S. 1, 5-6, 90 S.Ct. 1999, 2001, 26 L.Ed.2d 387 (1970); Foster v. California, 394 U.S. 440, 443, 89 S.Ct. 1127, 1128, 22 L.Ed.2d 402 (1969). In designing a standard to deal with this problem, the Court has emphasized “that reliability is the linchpin in determining the admissibility of identification testimony____” Manson v. Brathwaite, 432 U.S. 98, 114, 97 S.Ct. 2243, 2253, 53 L.Ed.2d 140 (1977). As articulated in Manson the court must apply a two-step analysis to determine whether to exclude either an in-court identification or testimony about a pre-trial identification. 19 First, the defendant must demonstrate that the procedures utilized by the police prior to trial in obtaining an identification by the witness were not unduly or unnecessarily suggestive. See Manson v. Brathwaite, 432 U.S. at 105-14, 97 S.Ct. at 2248-53. Second, if he succeeds in making that showing, the court must determine whether, “on the totality of the surrounding circumstances,” Coleman v. Alabama, 399 U.S. at 4, 90 S.Ct. at 2000, the proposed identification testimony at trial is independently reliable, notwithstanding the suggestive procedures; that is, it must decide whether or not the manner in which the identification was derived suggests that there is “a very substantial likelihood of irreparable misidentification.” Manson v. Brathwaite, 432 U.S. at 116, 97 S.Ct. at 2254 (quoting Simmons v. United States, 390 U.S. 377, 384, 88 S.Ct. 967, 971, 19 L.Ed.2d 1247 (1968)). See, e.g., Dickerson v. Fogg, 692 F.2d 238, 244 (2d Cir.1982); Sales v. Harris, 675 F.2d at 538; Styers v. Smith, 659 F.2d 293, 297 (2d Cir.1981); United States v. Evans, 484 F.2d 1178, 1184 (2d Cir.1973). In making this two-stage analysis the court is to engage in a balancing process, weighing the degree of suggestiveness against whatever indicia of reliability may be invoked. As the Supreme Court has noted, the central purpose of this analysis is to weed out unreliable identifications. See, e.g., Manson v. Brathwaite, 432 U.S. *1414 at 114, 97 S.Ct. at 2253; Neil v. Biggers, 409 U.S. at 198 (quoting Simmons v. United States, 390 U.S. at 384, 88 S.Ct. at 971). Accordingly, to reach the second stage of the analysis, the court need not find that the police procedures were extraordinarily egregious; all that is required is sufficient suggestiveness in those procedures to justify looking to whatever countervailing indicators of reliability may exist. Dickerson v. Fogg, 692 F.2d at 248 n. 3 (Newman, J., concurring.) See, e.g., Velez v. Schmer, 724 F.2d 249, 251 (1st Cir.1984). 2. Undue Suggestiveness Most commonly the evaluation of the suggestiveness of the police procedures involves an analysis of only one event, such as a line-up, a show-up, or a photo array. See, e.g., United States v. Damsky, 740 F.2d 134, 140 (2d Cir.), cert. denied, — U.S. —, 105 S.Ct. 298, 83 L.Ed.2d 233 (1984); United States v. Archibald, 734 F.2d 938, 940-41 (2d Cir.1984); United States v. Bubar, 567 F.2d 192, 197-98 (2d Cir.), cert. denied, 434 U.S. 872, 98 S.Ct. 217, 54 L.Ed.2d 151 (1977). Nonetheless, a claim of undue suggestiveness may be premised upon a series of events that, collectively and in the particular circumstances of the case, are deemed unnecessarily suggestive. See, e.g., Foster v. California, 394 U.S. at 442-44, 89 S.Ct. at 1128-29. Even if none of the individual components of the pretrial procedure are separately unduly suggestive, the combination of them may result in unnecessary suggestiveness, and in such a case the defendant has met his burden with respect to the first stage of the analysis. See, e.g., Dickerson v. Fogg, 692 F.2d at 245. 20 In this case, a series of steps taken and not taken by the police from the time of the crime until Hyland took the witness stand were sufficiently suggestive in their cumulative impact to satisfy the first stage of the two-part analysis and thus to require the Court to review the independent basis, if any, for Hyland’s in-court identification. 21 Immediately following the commission of the murder, Hyland was able to give an apparently very detailed description of the man who had pumped gasoline for him. He was plainly unable, however, to describe the second suspect—the driver of the other car—as other than a person of uncertain sex and race who had a tan and long hair, a situation explainable by the fact that he had not been able to see the driver’s face while the driver sat in the *1415 other car. 22 Indeed, it was plainly Hyland’s inability to describe the second suspect that led the police to circulate only one composite drawing, the sketch of Ronald Kelly. This lack of information by the eyewitness about the second suspect must be borne in mind in evaluating the subsequent steps taken by the police. The first contact by the police with the witness after his initial efforts at describing the two suspects occurred more than two years later, when he was given the two photo arrays. This long delay, together with Hyland’s initial inability to describe the second suspect, already suggests that the witness was potentially vulnerable to suggestion and that the police were therefore required to exercise particular care to avoid manufacturing an identification by him. See, e.g., United States ex rel. Cannon v. Smith, 527 F.2d 702, 704 (2d Cir. 1975); United States v. Fedorenko, 455 F.Supp. 893, 906 (S.D.Fla.1978), rev’d on other gds., 597 F.2d 946, reh. denied, 601 F.2d 1195 (5th Cir.1979), aff’d, 449 U.S. 490, 101 S.Ct. 737, 66 L.Ed.2d 686 (1981). Although the array they showed him was not fatally defective in itself, the manner in which it was handled contributed significantly to the process of suggestion. As testified without contradiction, the police officer did not make any suggestive remarks before handing Hyland the photos in two batches. Moreover, the twelve female photographs that comprised the second batch represent a fair range of physical types. The problems with the two arrays were of a different nature. First, the fact that the police initially gave the witness a set of pictures of males, and, after his selection of one of them, gave him a set only of females may have contributed to reinforcing in his mind the notion that the second suspect was a female, particularly since he had earlier expressed some uncertainty about the sex of the second suspect. Presumably this effect could have been avoided by simply giving the witness a second set of photos composed of both men and women with long hair since he had initially been able to identify the second suspect solely as having long hair and a tan. Second, and more significantly, four of the twelve photographs of women had plainly visible law enforcement markings, and only one—that of petitioner—had a “Sheriff’s Department” identification. The use of photographs so marked has been strongly criticized, see, e.g., United States v. Archibald, 734 F.2d at 940, and with good reason. If police markings are plainly visible on one or several but not all of the photographs displayed, it is quite likely that the witness—whether consciously or not—will focus on the marked photographs both because they are distinctive and because the markings suggest that the person portrayed, unlike the others in the array, is under arrest or has been arrested on earlier occasions. Indeed the Second Circuit has criticized the use of marked photographs even when all of the photographs were marked, simply because one of the markings differed from the others and the display of the police labels was plainly unnecessary. See United States v. Archibald, 734 F.2d at 940. In this case the use of law enforcement markings was more suggestive than in Archibald. 23 Only four of the twelve photographs had law enforcement markings, thus directing the witness’s attention to those four; in contrast, in Archibald all of the photos had police markings. 734 F.2d at 940. Moreover, the photograph chosen by the witness had a different type of *1416 marking from the three others, a difference that was deemed in Archibald to have some unnecessary attention-getting effect. Ibid. Finally, unlike Archibald, id. at 741, it was evident to the police in this case, based upon Hyland’s original statement, that he had only a minimal amount of information about the physical appearance of the second suspect, and was therefore particularly susceptible to such extraneous influences, a vulnerability that could only have been enhanced by the passage of more than two years since Hyland had seen the suspects. See, e.g., United States ex rel. Cannon v. Smith, 527 F.2d at 704. Cf. Foster v. California, 394 U.S. at 442-43, 89 S.Ct. at 1128. Not surprisingly, although Hyland signed the back of the photograph of petitioner, his choice was so tentative that the prosecution itself described the selection as having been of a photograph that “could possibly be” the suspect. Hyland subsequently confirmed his substantial uncertainty by testifying that he had initially selected the photographs of two different women, both of whom resembled the suspect (Wade Tr. 74-75); that his selection had been based on the fact that the photograph portrayed “the same style hair, long hair____” (GJ 37); and that his choice reflected only his belief that the person in the photograph could “possibly” be the second suspect. (Wade Tr. 61.) In view of this evident uncertainty by the witness at the time that he selected the photograph of petitioner, the next step taken by the police in the identification process must be deemed particularly significant. According to Hyland, after he had made his selection, the police officer discussed the two individuals whom he had selected, telling Hyland “who they might be” and “where they were from.” (Tr. 286-87.) 24 In this case the provision of such information was tantamount to telling the witness that he had selected the correct photograph of the second suspect, a practice repeatedly recognized as suggestive in the context of a photographic array. See, e.g., United States v. Leonardi, 623 F.2d 746, 755 (2d Cir.), cert. denied, 447 U.S. 928, 100 S.Ct. 3027, 65 L.Ed.2d 1123 (1980); United States v. Field, 625 F.2d 862, 869 (9th Cir.1980); United States v. Jarvis, 560 F.2d 494, 500 (2d Cir.1977), cert. denied, 435 U.S. 934, 98 S.Ct. 1511, 55 L.Ed.2d 532 (1978). Cf. United States v. Russell, 532 F.2d 1063, 1068 (6th Cir.1976). The suggestiveness of the police officer’s comments are apparent from the circumstances. Hyland was, from the very outset, quite firm and detailed in his description of the man who had pumped his gasoline. Indeed, he participated in the preparation of a composite sketch of the man that proved to be a reasonable likeness of Kelly. Moreover, at the photo array he selected two male photographs, both of which were of Kelly, and apparently expressed no uncertainty about his identification, in sharp contrast to his evident uncertainty about the appearance of the other suspect. The officer's immediate disclosure to Hyland that both of the people he had chosen were from North Carolina and his apparent disclosure that they had been travelling together in New York 25 plainly would be expected to have a suggestive impact on an uncertain witness, suggesting to him that, notwithstanding his doubts about the second suspect, he had chosen correctly since she was a companion of the man whom he confidently believed had *1417 been at the Seaway station at the crucial time. As the Second Circuit has observed, such post-array comments “may tend to reinforce an otherwise weak or tentative identification” and should suffice to invalidate a later identification unless “the [original] identification is strong and unequivocal.” United States v. Leonardi, 623 F.2d at 755. Since both the initial and the second identifications by Hyland of the second suspect were extremely tentative, the police comments must be viewed as highly suggestive. Hyland’s next contact with law enforcement authorities was his testimony at the grand jury. The principal significance of this event for the identification process was that Hyland’s testimony confirmed that his selection of petitioner’s photograph had been extremely tentative, and, indeed, that he was not entirely certain that the person in the other car was a woman. The significant actions, or inactions, of the police in connection with the identification process resumed after Kelly and petitioner had been indicted. As noted, Hyland had been very uncertain about the physical appearance of the second suspect from the very outset, and his appearance before the grand jury served to confirm the tentative nature of his selection of petitioner’s photograph. Following the return of the. indictment, both defendants were returned to Utica for trial amid substantial newspaper coverage of their purported roles in the crime and their link to each other. (See Petitioner’s Appendix II at 33-37.) While obviously not the fault of the police, this coverage—which Hyland apparently followed (Tr. 286-87)—was likely to add to the effect of the police officer’s post-array comments, which, as we have seen, encouraged Hyland to view his photographic selection of petitioner as correct. The effect of this combination of the initial uncertainty of the witness, the passage of time, the suggestive comments after the photo array and the post-indictment publicity was then further enhanced by the occurrence of one event and the non-occurrence of another event. The event that did occur was the Wade hearing, at which petitioner was seated at the defense table while Hyland testified. In view of Hyland’s inability to specify at any time until then any permanent features of the second suspect, 26 the presence of the petitioner at the hearing may be assumed to have had a significant potential for suggestability. Indeed, several courts have noted the problems inherent in a criminal defendant’s being seated in a manner that plainly identifies him as the suspect when the reliability of a witness’s identification is at issue. See, e.g., United States v. Archibald, 734 F.2d at 941-43; People v. James, 100 A.D.2d 552, 553, 473 N.Y.S.2d 252, 254 (2d Dep’t 1984). Cf. Moore v. Illinois, 434 U.S. 220, 229-30, 98 S.Ct. 458, 465, 54 L.Ed.2d 424 (1977); Boyd v. Henderson, 555 F.2d 56, 61-62 (2d Cir.), cert. denied, 434 U.S. 927, 98 S.Ct. 410, 54 L.Ed.2d 286 (1977). See also People v. Moger, 85 A.D.2d 610, 611, 444 N.Y.S.2d 676, 677 (2d Dep’t 1981). The fact that petitioner underwent a Wade hearing at which Hyland appeared, and the manner in which she was displayed in the courtroom are presumably not the fault of the State; a Wade hearing is conducted at the option of the defendant, and in this case petitioner did not request either not to be present or to be seated somewhere other than at counsel table. Compare United States v. Archibald, 734 F.2d at 941-43; People v. James, 100 A.D.2d at 553, 473 N.Y.S.2d at 254. Accordingly, the Wade hearing cannot itself be deemed an “unduly suggestive” identification procedure, but it is surely one of the “surrounding circumstances” that properly affect the court’s evaluation of the appropriateness of what the police and prosecutors chose to do in obtaining the witness’s in-court identification testimony. *1418 As for the event that did not occur, the determination of the State not to conduct a line-up following petitioner’s extradition from North Carolina takes on particular significance in the chain of events leading to Hyland’s in-court identification of her. The Supreme Court has recognized that line-ups are generally more reliable than photographic arrays, see Simmons v. United States, 390 U.S. 377, 386 n. 6, 88 S.Ct. 967, 972 n. 6, 19 L.Ed.2d 1247 (1968), although the unsuggestive use of photographs is of course permitted. Id. at 384, 88 S.Ct. at 971. Moreover, in this case even the State conceded that the witness’s photo identification had been very tentative. Under these circumstances, by not conducting a line-up either before or after petitioner’s appearance at the Wade hearing, the State in effect avoided any further test of the witness’s independent ability to identify the suspect, and thus left him open to the suggestive effects of the post-array police comments, the publicity in the press, and the appearance of petitioner at the Wade hearing. It is to be assumed that this handling of Hyland prior to trial in all likelihood served to lessen, if not eliminate, the uncertainty that he had displayed at the photo array and even thereafter. 27 See generally United States v. Fernandez, 456 F.2d 638, 641 n. 1 (2d Cir.1972), cited in United States v. Danzey, 594 F.2d 905, 909-10 (2d Cir.), cert. denied, 441 U.S. 951, 99 S.Ct. 2179, 60 L.Ed.2d 1056 (1979). The final stage of the identification process occurred at trial. Just before Hyland took the witness stand, and before he had ever positively identified petitioner as the second suspect, the prosecutor showed him the two photographs that he had previously signed—one of Kelly and the other of petitioner—and, in Hyland’s words, told him to “stick to your guns.” (Tr. 305.) This form of pre-testimonial encouragement is frowned on, see Davis v. United States, 425 F.2d 673, 674 (9th Cir.1970), and even though in itself it would not justify setting aside the conviction, see United States v. Danzey, 594 F.2d at 916, in context it surely contributed to the cumulative suggestiveness of the State identification process. Hyland had consistently expressed doubt and uncertainty about the details of the second suspect’s appearance, and his selection of the petitioner’s photograph was consistent with this uncertainty. In the three and a half years since the crime, the only additions to Hyland’s store of knowledge about the second suspect had come from a series of suggestive acts, some by law enforcement authorities and some by others, and accordingly there is no reason to believe, on the present record, that at the time of trial Hyland’s independent basis for identifying petitioner was any stronger than on the day that he signed his original statement expressing the tentative belief that the suspect was a white female. For the prosecutor to say to such an uncertain witness—just before he takes the stand—that he should, in effect, stand fast on his identification must be viewed as a further contribution to the suggestive process by which Hyland eventually became certain that he could identify petitioner as the driver of the second car at the Seaway station. In sum, it is apparent that the procedures utilized by the State to obtain Hyland’s in-court identification of petitioner were, in their collective effect, sufficiently suggestive, under the circumstances of this case, to require a review of the independent basis, if any, for the witness’s identification testimony. As will be seen, the pertinent criteria for this evaluation demonstrate little justification for the introduction of his testimony. 3. Independent Reliability of Identification Upon a determination that identification procedures have been utilized that are unduly or unnecessarily suggestive, the court *1419 must look to “whether, under the ‘totality of the circumstances,’ the identification was reliable even though the confrontation procedure was suggestive.” See Manson v. Brathwaite, 432 U.S. at 106, 97 S.Ct. at 2249 (quoting Neil v. Biggers, 409 U.S. at 199, 93 S.Ct. at 382.) As described by the Court in Manson and Neil, this analysis involves a balancing of six factors: 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. Against these factors is to be weighed the corrupting effect of the suggestive identification itself. Manson v. Brathwaite, 432 U.S. at 114, 97 S.Ct. at 2253. Application of these criteria to this case strongly suggests that there was little or no independent basis for Hyland’s in-court identification of petitioner and accordingly that it must be viewed as largely a product of the various suggestive events previously described. (a) Opportunity to view Hyland testified that he was at the station for only four to five minutes. During that time, he parked his car by the gas pumps, and stood behind his car at an angle from which he apparently could not see the face of the driver of the car that subsequently parked on the other side of the gas pumps. Moreover, he spent most of his few minutes at the station talking to and looking at the man who pumped his gasoline and only glanced at the other driver for a “few seconds” (Tr. 311, see Deft. Exh. A), presumably when the ostensible station attendant asked whom he should first serve. Such a brief glimpse of a person’s hair or of the back of his head or even of his face is scarcely likely to give a witness more than the vaguest impression of the other person’s appearance. See, e.g., Jackson v. Fogg, 589 F.2d at 109; United States ex rel. Cannon v. Smith, 527 F.2d at 703. See also Dickerson v. Fogg, 526 F.Supp. 1299, 1307 (S.D.N.Y.1981), aff’d, 692 F.2d 238 (2d Cir.1982). Under these circumstances it is not surprising that Hyland could not describe the person except by reference to long hair and a tan. (b) Witness’s degree of attention By his own testimony it is evident that Hyland was not paying any special attention to the driver of the other car, nor was there any reason why he should do so. He entered the Seaway station to obtain gasoline and, by his own estimation, spent most of the few minutes that he was there first looking around the empty station and then talking with and looking at the man pumping his gasoline. He had no reason to look elsewhere since the driver of the other car appeared only to be another customer, and Hyland was certainly unaware that a crime was taking place. Compare, e.g., United States v. Leonardi, 623 F.2d at 756; United States ex rel. John v. Casscles, 489 F.2d 20, 25 (2d Cir.1973), cert. denied, 416 U.S. 959, 94 S.Ct. 1976, 40 L.Ed.2d 310 (1974). Since Hyland was not a trained observer, such as a policeman, and since the events at the station appeared to be quite routine, there is no reason to assume that he was paying close attention to any of these events, and particularly not to the appearance of the other driver. See, e.g., Manson v. Brathwaite, 432 U.S. at 115, 97 S.Ct. at 2253; United States v. Sanchez, 603 F.2d 381, 385-86 (2d Cir.1979). (c) Accuracy of witness’s description Two days after his visit to the Seaway station, Hyland proved unable to describe the other driver except to note that he believed, although he was not sure, that the person was a white female, and that the driver had “long black shoulder length hair” and a tan. (Deft. Exh. A.) He was unable at any time until the trial to describe any facial features and gave a height size significantly different from petitioner’s and based solely on his recollection of how the person sat in the car. This description was “of no real value,” United States ex rel. Cannon v. Smith, 527 F.2d *1420 at 704, since it could be the description of millions of people. Hyland’s inability to describe the second suspect persisted until trial, when he referred in vague terms to certain of the petitioner’s features—a sequence of events that strongly suggests that his description at trial was the culmination of the various suggestive effects described, as well as seeing petitioner seated before him at trial, and was not the product of independent recollection. See, e.g., Dickerson v. Fogg, 692 F.2d at 246. 28 (d) Level of certainty demonstrated by witness at prior confrontation Hyland was plainly uncertain in his original selection of petitioner’s photograph, and he remained uncertain about her identity until the trial. When shown the array, he initially selected photographs of two women as resembling the driver, and although he ultimately signed petitioner’s photograph, he believed only that it was “possibly” the person in the car, and still thought that the other woman also looked like the second suspect. This uncertainty was plainly attributable to his having not seen or focussed on the driver’s face, and thus before the grand jury he said that he had picked the photo because it showed “the same style hair, long hair ...” (GJ 37) and that “the best [he could] say is that it could be the girl but [he could not] say for sure____” (GJ 38.) This same uncertainty persisted through the Wade hearing, when he agreed with the prosecutor that he had selected petitioner’s photograph because it could “possibly” be the driver of the other car. (Wade Tr. 61.) In view of this persistent uncertainty, it is difficult to avoid the conclusion that Hyland’s newly acquired certainty at trial was a product of suggestion rather than of independent memory. See, e.g., Dickerson v. Fogg, 692 F.2d at 246; Solomon v. Smith, 645 F.2d 1179, 1187 (2d Cir.1981). (e) Length of time between the crime and the confrontation Hyland did not view the photographs until December 1975, well over two years after the crime had been committed. The Wade hearing, where he first saw petitioner, did not occur until January 1977, nearly three and one-half years after the events at the Seaway station, and the trial did not take place until two months later. This lengthy span of time further weakens the reliability of Hyland’s identification, which was based upon a brief, unfocussed glance at the back or side of the head of another motorist sitting in a car more than three and a half years before the trial. Compare, e.g., Jackson v. Fogg, 589 F.2d at 111 (emphasizing passage of one month before confrontation). See also Neil v. Biggers, 409 U.S. at 201, 93 S.Ct. at 383 (passage of seven months is “seriously negative factor in most cases”). The mere passage of time is not an automatic determinant of unreliability. See, e.g. United States v. Williams, 596 F.2d 44, 49 (2d Cir.), cert. denied, 442 U.S. 946, 99 S.Ct. 2893, 61 L.Ed.2d 317 (1979) (32 months). Nonetheless, where all of the other pertinent criteria suggest a lack of any independent basis for Hyland’s identification of petitioner, the lapse of years before Hyland even saw a photograph of petitioner is surely consistent with the inference of unreliability. (f) The effect of the suggestive procedures As the foregoing suggests, the relevant criteria for determining the independent reliability of Hyland’s identification testimony all suggest that his in-court identification had little or no independent basis. The inference is therefore virtually unavoidable that his identification of petitioner was a product principally of the suggestive procedures and circumstances previously described. *1421 4. Conclusion as to identification testimony Two days after the incident at the Seaway station, the one eyewitness who was apparently at the station near the time of the crime and saw one or two of the perpetrators was able to provide a detailed description of one of the two individuals but was unable to give any meaningful description of the other person. More than three and a half years later, for the first time he positively identified the petitioner as the second person at the station, notwithstanding his earlier inability to describe any permanent traits of the person and despite his having stated under oath—only two days after the incident—that he had not seen the face of the second suspect and was uncertain of the person’s sex and race. In context it is apparent that his substantial change of position is attributable to a series of events that occurred in the intervening years, none of which involved the refreshing of his memory of the incident. Rather, those events were the product, in large measure, of the manner in which the State went about the task of developing its case and eliciting identification testimony from this crucial witness. Under the circumstances, those procedures in all likelihood created in the witness’s mind a belief as to the identity of the suspect that was not supported by the sense impressions that he had acquired at the time of the event. Since the procedures utilized were unduly suggestive and since there is little reason to believe that Hyland independently could have identified petitioner with the requisite certainty that he displayed at trial, his identification testimony should have been suppressed. The failure to do so denied petitioner due process. 29 B. Sufficiency of the Evidence Petitioner argues, in the alternative, that even if the admission of Hyland’s in-court identification were not erroneous, the petition would have to be granted because his testimony was incredible “as a matter of law” and the balance of the evidence was insufficient to convict petitioner. This argument is not well-founded, and I therefore do not recommend issuing the writ on such a basis. Claims of insufficient evidence are frequently made and rarely upheld in habeas proceedings. The standard for evaluating such a claim is “whether, after viewing the evidence in the light most favorable to the prosecution, any rational trier of fact could have found the essential elements of the crime beyond a reasonable doubt.” Jackson v. Virginia, 443 U.S. 307, 319, 99 S.Ct. 2781, 2789, 61 L.Ed.2d 560 (1979) (emphasis in original). Cf. United States v. Brown, 776 F.2d 397, at 402 (2d Cir.1985). Petitioner’s argument in this case is premised upon the notion that Hyland’s testimony was so suspect that it must be disregarded and that the remainder of the evidence was insufficient to meet the Jackson standard. Neither proposition is sustainable. If we start from the premise that Hyland’s in-court identification was admissible, then there are elements of his testimony that the jury could have relied upon, even if a rational trier of fact would have had to reject his trial statement that he was certain that petitioner was the driver of the other car based on her facial features. First, he stated two days after the incident that he believed the second suspect to be a white female with long dark hair, an impression that—even if tentative—can be considered by a rational jury. Second, as emerged during cross-examination of Hyland, he had been given a set of twelve photographs of women without prior comment by the police, and he eventually selected petitioner’s photograph. If Hyland’s in-court identification testimony was properly admitted even though incredible, then a rational jury could at least consider the fact that Hyland had—whether by coin *1422 cidence or not—selected the photograph of the woman who was the paramour or “roommate,” of the man whom he had very reliably identified as the purported gas station attendant. 30 If this limited portion of Hyland’s testimony were all that the jury could consider, petitioner might well be correct in her assessment of the adequacy of the evidence. Although “the testimony of a single, uncorroborated eyewitness is generally sufficient to support a conviction,” United States v. Danzey, 594 F.2d at 916, the court “must scrutinize such evidence with care.” Ibid. In this case Hyland’s selection of petitioner’s photograph is somewhat suspect for the reasons already indicated, including his initial inability to describe the second suspect, the fact that he apparently could not see the person’s face for more than a few seconds, and the likelihood that he was not paying much attention. The difficulty with petitioner’s argument is that there was other evidence, some of it circumstantial, that supported Hyland’s selection of petitioner’s photograph. First, the evidence was overwhelming that Kelly was one of the killers, based upon his fingerprint being found on the tape over the victim’s mouth, as well as Hyland’s very specific identification of him. Second, the State presented strong evidence that petitioner's car had been used to carry out the crime, including the eyewitness testimony of a number of individuals who saw a car resembling petitioner’s at the Seaway station around the time of the murder and the evidence that Kelly had free use of the car. Third, the prosecution introduced evidence suggesting that a second person was involved, based upon the testimony of eyewitnesses that there had been two people in the blue car identified as petitioner’s and that three people had been seen standing in the gas station office around the time of the crime. 31 Fourth, the same testimony strongly suggested—as had Hyland—that the second person was a woman. Indeed, one witness testified to having seen a woman emerge from the blue car and approach the station building and another reported seeing a woman with long dark hair standing in the station office with two men during the time period in question. Fifth, the testimony—both from Hyland’s first statement and from the other witnesses—justified the conclusion that the woman had long hair, as did petitioner. Sixth, there was no dispute that petitioner had come north with Kelly—the unquestioned principal in the crime—and was living with him as a couple, thus permitting the inference that if Kelly committed a crime with a woman who resembled petitioner, the woman could well have been petitioner. Seventh, this inference was somewhat reinforced by the testimony of one of Kelly’s acquaintances that he had never seen Kelly with any woman other than petitioner. (Tr. 97-98.) Eighth, petitioner and Kelly departed together from Utica only one or two days after the Seaway murder and headed back towards North Carolina. Ninth, petitioner did not provide any verifiable alibi for the date and time of the murder, and, when questioned before her arrest in North Carolina, did not unequivocally deny having been at the Seaway station. Tenth, petitioner testified at trial, denying her involvement, and it was certainly within the province of the jury to disbelieve her testimony based on her demeanor or other factors not reflected in the printed record. If she did not appear credible to the jury, this could be deemed additional corroboration for the foregoing inculpatory evidence. The prosecutor’s case plainly was far from overwhelming. Nonetheless, it did not involve solely an incredible eyewitness, as petitioner suggests. If viewed “in the light most favorable to the prosecution,” *1423 Jackson v. Virginia, 443 U.S. at 319, 99 S.Ct. at 2789, as it must be, the evidence was sufficient to permit a rational juror to find guilt beyond a reasonable doubt even if this Court would not have so found. See United States v. Brown, 776 F.2d at 402. 32 Accordingly the evidence was sufficient to pass constitutional muster. CONCLUSION For the reasons stated, I recommend that the petition for a writ of habeas corpus be granted, and that respondent be directed to release petitioner unless she is retried on the charges against her within sixty days in conformity with the findings made in this Report and Recommendation. Pursuant to Rule 7 of the Rules of Proceedings Before a United States Magistrate, the parties shall have ten (10) days from this date to file written objections to this Report and Recommendation. Such objections shall be filed with the Clerk of the Court, with extra copies delivered to the chambers of the Honorable David N. Edelstein, Room 2104, and to the chambers of the undersigned, Room 503. Footnotes [1] The facts relating to this petition are set forth in the Magistrate’s thorough Report which appears as an appendix to this Opinion. They need not be restated here. [2] New York law prohibits the prosecution from introducing evidence of a pre-trial photographic identification on their direct case. N.Y.Crim. Proc.Law § 60.30 (McKinney 1986); see People v. Brewster, 100 A.D.2d 134, 473 N.Y.S.2d 984, 987-88 (2d Dep’t 1984). [3] The exact nature of respondent’s objection is unclear. If respondent is claiming that this matter must be dismissed pursuant to Rose v. Lundy, 455 U.S. 509, 102 S.Ct. 1198, 71 L.Ed.2d 379 (1982), because petitioner has raised claims in this court without exhausting her state court remedies this would be inaccurate. Jarrett did not base her petition on the issue of pretrial publicity. Rather, the Magistrate merely mentioned the fact of publicity in the Report. If respondent is arguing that pursuant to Wainwright v. Sykes, 433 U.S. 72, 97 S.Ct. 2497, 53 L.Ed.2d 594 (1977), there was a procedural default regarding the pretrial publicity claim without a showing by the petitioner of "cause” and "prejudice” this would also not require a dismissal. Based on the court’s finding that the Magistrate did not base his decision on pretrial publicity, the court need not reach this issue. [1] The convictions were for second degree murder (N.Y.Penal Law § 125.25(1)), felony murder (also defined as murder in the second degree) (Penal Law § 125.25(3)), first degree robbery (Penal Law § 160.15(1)) and first degree robbery with an instrument (Penal Law § 160.-15(3)). [2] See Affidavit of Claudia Angeles, Esq., sworn to May 25, 1984 at ¶¶ 18-19. [3] See Affidavit of Claudia Angeles, Esq., sworn to December 14, 1984 at ¶¶ 2-7. [4] As both parties have conceded, the state record is adequately developed and there is no dispute between the parties as to any material historical fact. (See letters to the Court from Assistant Attorney General Monica R. Jacobsen, dated December 19, 1984 and from Claudia Angeles, Esq., dated January 18, 1985.) Accordingly, there is no need to conduct an evidentiary hearing in connection with this matter. See, e.g., Johnson v. Fogg, 653 F.2d 750, 753 (2d Cir.1981), cert. denied, 456 U.S. 907, 102 S.Ct. 1755, 72 L.Ed.2d 164 (1982); Simmons v. Clemente, 552 F.2d 65, 68-69 (2d Cir.1977). [5] The facts described here are based upon the transcripts of the trial, the pre-trial Wade hearing, grand jury testimony, and the exhibits to those various proceedings. The trial transcript—which is found in petitioner’s Appendix I, Vols. 1-3—will be refered to by the designation “(Tr. ),” with the page numbers inserted. The Wade and grand jury transcripts—pertinent portions of which are found in petitioner’s Appendix II—will be referred to by the designations “(Wade Tr. )’’ and "(GJ )," with the page numbers inserted. The transcripts of Mr. Hyland’s testimony before the grand jury and at the Wade hearing were received in evidence at the trial as defendant Jarrett’s exhibits P and R respectively. Trial exhibits will be referred to as either “(Deft. Exh. —)” or "(Gov’t. Exh. —).” [6] One of these photographs, exhibit 25, bears a part of what may be a State police marking, but the only lettering that appears is "NY STA.” [7] That the discussion included a specific reference to the relationship between the two is indicated as well by the rebuttal trial testimony of the officer. (See Tr. 738.) [8] Petitioner’s motion papers are included in her Appendix II at 16-37. [9] Hyland was 5 foot, 6V2 inches tall. (Wade Tr. 80.) Petitioner is 5 feet 2 inches. (Gov’t. Exh. 40.) [10] The four-page memorandum of the Honorable John J. Walsh is found in Petitioner's Appendix II at 129-32. [11] At one point Hyland stated that he assumed that the four photographs with police markings on them came from the police. (Wade Tr. 73.) He certainly did not testify that he believed the unmarked photographs had come from the same source. [12] The police investigator before whom the statement was sworn also testified at trial and stated his belief that the statement accurately represented what Hyland had told him on that day. (Tr. 690-94.) [13] The match was apparently not done until the end of 1975 or sometime in 1976. (Tr. 514-15.) [14] Petitioner’s other witnesses included two court reporters, who identified the transcripts of Hyland's prior grand jury and Wade hearing testimony; petitioner’s father; her brother-in-law, who testified about the appearance of petitioner’s automobile; and a Utica-based friend of petitioner and Kelly. [15] The order of the Appellate Division is reproduced in Petitioner’s Appendix II at 190. [16] Petitioner’s application is found in Appendix II at 192-95. [17] The order of the New York Court of Appeals appears in Petitioner’s Appendix II at 197. [18] One noted commentator has observed: "In England and America most of the spectacular miscarriages have been due to wrong identification of the defendant as a culprit.” G. Williams, The Proof of Guilt 106 (3d ed. 1963). [19] The standard is the same for both situations. See Manson v. Brathwaite, supra, 432 at 106 n. 9, 97 S.Ct. at 2249 n. 9. Because New York law prohibits the prosecution from introducing evidence of a pre-trial photographic identification, N.Y.C.P.L. § 60.30, see Sales v. Harris, 675 F.2d 532, 535 n. 1 (2d Cir.), cert. denied, 459 U.S. 876, 103 S.Ct. 170, 74 L.Ed.2d 140 (1982), the only evidence introduced by the State on its case-in-chief was the in-court identification by Hyland. [20] The Court in Dickerson, after noting a series of police actions that contributed to the witness’s identification of the defendant, concluded: While none of these aspects of the confrontation alone necessarily would have invalidated the identification, the combination of them all comprised a highly suggestive identification procedure. Ibid. [21] Respondent urges that this Court should defer to the findings of the state trial court on the identification issues, including its finding of no undue suggestiveness. This argument fails for several reasons. First, the determination of lack of undue suggestiveness is not in itself a finding of historical fact and accordingly is not entitled to deference under 28 U.S.C. § 2254(d). See Sumner v. Mata, 455 U.S. 591, 596-97, 102 S.Ct. 1303, 1306, 71 L.Ed.2d 480 (1982); Dickerson v. Fogg, 692 F.2d at 243. Second, the factual findings of the state court related solely to the photo array, and, as will be seen below, the events relevant to the present claim encompass far more than this one step. Third, the Court’s few findings with respect to the array are in several respects entirely unsupported by the record and therefore not entitled to deference. See id. at 243 (referring to 28 U.S.C. § 2254(d)(8)). Specifically the Court erred in implying that Hyland had identified petitioner at the Wade hearing; quite to the contrary, he was not asked to identify her, and his testimony, fairly construed, suggests that he did not recognize her as a suspect. The Court also erred in concluding that Hyland had assumed that all the photographs had originated with the police, a finding on which the Court based its conclusion that the law enforcement markings on some of the photos were immaterial; Hyland in fact testified that he had no idea as to the source of the unmarked photographs. Finally, the court erred in finding that Hyland had testified that he had observed petitioner at close range. At most, he stated that he had been standing behind the rear end of his car on one side of the pumps and the driver of the other car had stopped on the other side of the pump and remained in her car at an unspecified distance from him. (Wade Tr. 78, 80.) [22] Although Hyland's statement refers to the driver of the other car in the feminine gender, it states: "I am not sure, but I believe the operator was a white female.” (Def. Exh. A.) This statement could be read as expressing doubt solely about the race of the driver, but this does not appear to be the case since, in his testimony before the grand jury, Hyland indicated some uncertainty about the sex of the driver. (GJ 30.) Moreover, even after speaking to Hyland the police apparently continued to look for two male suspects. (Tr. 177.) [23] The Court in Archibald held that the particular array used there were not sufficiently suggestive to require suppression of the identification testimony. 734 F.2d at 940. [24] I note that at the Wade hearing Hyland had testified that he did not recall any conversation with the police officer after making his selections. {Wade Tr. 60-61.) [25] Although Hyland’s testimony that the officer told him "who they were” was not further explained by him, it is reasonable to infer that this meant that the officer had informed him that the two individuals were related—an inference that was later in effect confirmed by the prosecutor’s rebuttal trial examination of the police officer: Q..... Did you at any time, prior to his looking at these photographs, tell him that there was an association between one of the males and one of the females that was in that pile? A. No, I did not. (Tr. 738)(emphasis added.) [26] Prior to trial, Hyland had mentioned only that the suspect had had long hair and a tan, both obviously impermanent attributes, and had testified that he believed the suspect was female. [27] This analysis is not intended to suggest that the State is invariably under an obligation to conduct a line-up for an uncertain eyewitness. Rather, under the particular circumstances of this case the lack of a line-up—which could easily have been held—simply enhanced the suggestive effect of a series of events, some within the control of the State and others not. [28] It should be noted that, in contrast to his inability to describe the second suspect, Hyland was able to provide a detailed description of Kelly—a description that led to the creation of a reasonably accurate composite drawing. Accordingly, Hyland’s failure to describe the second suspect cannot be attributed to an inability to provide physical descriptions. [29] There can be little question that the error was not harmless in view of the key role that Hyland’s identification of petitioner played in the State’s case and the customarily strong impact of identification testimony. See, e.g., Kampshoff v. Smith, 698 F.2d at 585. [30] Petitioner has quite properly challenged the handling the photo array because of the markings on some of the photographs. If, however, Hyland’s testimony was admissible—as we are assuming—then necessarily the jury could consider the photo array since it would not have been deemed unduly suggestive. [31] The State also presented the testimony of the victim’s wrestling coach to the effect that he had been physically very strong (Tr. 454), and from this fact and the manner in which he had been trussed and killed the prosecution urged as well the inference that more than one person had been involved. (Tr. 784.) [32] Petitioner asserts that the evidence other than Hyland’s in-court identification proved nothing more than that petitioner "showed poor judgment in the choice of roommates.” (Petitioner’s Memorandum at 5.) If the evidence were viewed favorably to petitioner, this characterization would not be unreasonable. Under Jackson, however, the habeas court is obliged to view the evidence favorably to the state, and so viewed it is surely far more inculpatory than petitioner suggests.