MEMORANDUM AND ORDER RE: MOTION TO SUPPRESS

GERTNER, District Judge.

I. INTRODUCTION

The defendant, Ramses Smith, has moved to suppress testimony by Agent Christopher Demlein of the Bureau of Alcohol, Tobacco, and Firearms ("ATF") identifying him as a participant in the illegal sale of a firearm and ammunition. Specifically, Smith seeks to exclude an out-of-court photo identification made by Agent Demlein, as well as Agent Demlein's identifications of Smith during his detention hearing and a hearing on his motion to suppress. Smith also moves to suppress any future in-court identification that Agent Demlein might make.

Courts have long acknowledged the centrality of identification evidence in the government's case in chief, as well as the due process dangers it presents. As a result, identification procedures must be carefully reviewed to see if they are suggestive, and if subsequent identifications are tainted by earlier, distorted processes. Plainly, the counterpart of courts' obligation is the obligation of the government, and the law enforcement agents it uses, to monitor identification procedures, to keep accurate records of them, and obviously, to testify truthfully.

Sadly, I find that the ATF agent in this case failed to meet any of these obligations — including the obligation to tell the truth.

I held a multi-part suppression hearing on January 27, 2005 and February 16, 2006. As a result of this hearing, and the fundamental, troubling problems in the government's case it revealed, I hereby GRANT defendant's motion to suppress [docket entry #25] the out-of-court photo identification, the identifications during the detention and suppression hearings, and any future in-court identification by Agent Demlein of Ramses Smith.

II. FACTUAL BACKGROUND PROCEDURAL HISTORY A. The Gun Transaction

Defendant Ramses Smith was indicted on April 2, 2003 for possession of a firearm and ammunition by a convicted felon. The allegations against him stem from a midday sting operation on January 28, 2002, in which ATF Agent Demlein, working undercover, and a cooperating witness ("CW") purchased a firearm and ammunition from three or four men.

Agent Demlein and the CW had made arrangements with one of the men, John Ivory, to purchase the gun, and drove in the CW's car to a McDonald's parking lot in the Roxbury neighborhood of Boston to complete the transaction. Agent Demlein wore a video recorder on the lapel of his jacket, and the CW's car was wired for audio. In addition, multiple Boston Police officers, members of the Youth Violence Strike Force, were stationed in and around the parking lot, listening in on the gun sale.

Two men entered the CW's car and negotiated the sale. Demlein was familiar with one of the men, Paul Cato, but did not recognize the other. The government alleges that the defendant, Ramses Smith, was that unknown man, ultimately identified by Agent Demlein from a photograph two months later under what can be most charitably described as troubling circumstances.

Despite two police reports by Agent Demlein, his grand jury testimony, his testimony at the defendant's detention hearing, and his testimony during the hearing on the defendant's motion to suppress, the sequence of events between the gun transaction and the photo identification of Smith remains obscure. Indeed, over the course of these proceedings, Agent Demlein has offered four different accounts of his out-of-court photo identification of Ramses Smith. His credibility is doubtful. B. Police Reports

The first account of the photo identification is found in two police reports that Agent Demlein wrote in the months following the gun sale.

The report, written by Demlein on January 30, just days after the transaction, and approved by his supervisor on February 4, 2002, gives a chronological narrative of the January 28, 2002 gun sale and the subsequent activities of Agent Demlein, the CW, and the police officers who monitored the sale. He writes that he and the CW met three men in the McDonald's parking lot: John Ivory, Paul Cato, and an unknown man. The unknown man is described only as a "black male, medium-dark complexion, approximately 24 [years of age], 6'3", 200 pounds. . . . called "ribs or rizz. . . ." (Suppr. Hr'g, Gov't Exh. 2, Jan. 27, 2005). The report ends with a description of the audio and video equipment used to record the gun transaction and a notation that the firearm and ammunition purchased by Agent Demlein would be stored in the police vault. It contains no reference to any efforts by the Boston Police, the CW, or Agent Demlein to identify the unknown man, nor any of the additional details of the man's physical description that surfaced in later accounts.

In a second police report written by Agent Demlein a month later, on March 19 and approved on March 27, 2002, Demlein describes a meeting with Boston Police Detectives Susan Antonucci and Marvin Wright. He relates that on March 13, 2002, he and the CW who participated in the January 28 gun transaction "were shown a photograph of Ramses M. Smith . . . [and both] identified the individual as the subject who sold [Agent] Demlein a .357 Smith and Wesson Revolver one month prior." (Suppr. Hr'g, Gov't Exh. 3, Jan. 27, 2005).

From these two police reports, the following picture emerges of the January 28, 2002 transaction and subsequent events: 1) three men were present in the McDonald's parking lot; 2) one of them, called Ribs or Rizz, was unknown to Agent Demlein; 3) Agent Demlein identified that man as Ramses Smith from a single photo, two months after the encounter, on March 13, 2002.

C. Grand Jury

When he testified before the grand jury on April 2, 2003, however, Agent Demlein's story changed. This time, he testified that four men, not three, were present in the McDonald's parking lot, now including John Ivory's cousin Shon Jakes. (Grand Jury Tr. 7, 9-17, Apr. 2, 2003). The unknown fourth man, called "Ribs" or "Riz," was "very tall, six, bigger than six-feet tall. For the first time, nearly four months after the fact, Demlein described this man as having " a black and blue leather jacket on, short cropped, maybe corn rows, medium complexion, medium to dark complexion, over 200 pounds, a black male, early twenties or mid twenties." Id. at 7, 20-24; 10, 2-3 (emphasis added).

In addition, in this testimony Agent Demlein placed the photo showup just after the initial transaction, rather than over a month later. He testified that, upon returning to the police station after the gun transaction on January 28, one of the Boston Police officers "pulled up a photo of who he thought the [unknown fourth man] was, and instantly we recognized who the individual was." Id. at 13, 22-24. Agent Demlein stated that the individual in this photo was Ramses Smith.

Agent Demlein's grand jury testimony, then, provides a second version of events: 1) four, instead of three, men were present in the McDonald's parking lot; 2) one of them, called Ribs or Rizz, was unknown to Agent Demlein; 3) Demlein, along with multiple Boston Police officers, identified that man as Ramses Smith from a single photograph hours, rather than months, after the transaction, on January 28, not March 13, 2002; 4) for the first time, Agent Demlein's physical description of the man contained a reference to the man's "black and blue leather jacket" and his "short cropped hair," perhaps in "corn rows."

D. Detention Hearing

At Ramses Smith's December 5, 2003 detention hearing, Agent Demlein testified to yet another version of his story. Again, he recounted seeing four men in the McDonald's parking lot, one of whom he did not know and whom Paul Cato called "Rizz" or "Ribs." (Det. Hr'g Tr. 7, 12-15; 8, 23-24, Dec. 5, 2003). Again, he characterized the man's jacket and gave a physical description: He was "wearing a distinctive . . . black and blue leather Avirex jacket." He was "[a] large 200 pounds — a tall, very tall, over 6'2" or so, over 200 pounds, medium/dark complexion, short hair."Id. at 29, 21-25.

What is new — and troubling — about the detention hearing version of Demlein's story is that he now testified, for the first time, that after the gun sale on January 28, 2002, he and Boston Police officers reviewed the video taken in the car but could not identify the unknown man. Id. at 15, 14-15. Demlein also testified that he participated in an elaborate photo identification proceeding, far more elaborate than anything described before. Officers showed Agent Demlein "two or three" photos, although the exact number was unclear — fewer than "a hundred," though perhaps only one — and Demlein was able to identify Ramses Smith's photo as the unknown man. Id. at 16, 2-19. He was "a hundred percent sure" about this identification. Id. at 16, 25.

Demlein testified to this level of certainty despite also admitting that his vantage point for observing the unknown man was restricted. He stated, "I could just kind of acknowledge him out of the side of my eye," and testified further that his attention was not trained on the unknown man, but rather that he "stayed focused on the man with the gun [Paul Cato] in the back seat." Id. at 14, 25-25; 34, 2.

During the detention hearing, Agent Demlein also made an in-court identification of Ramses Smith as "the person that [he] saw seated behind [him] during that transaction on January 28. . . ." Id. at 18, 13-23.

In effect, Agent Demlein's detention hearing testimony presents the third version of the January 28, 2002 gun sale and subsequent photo identification of Ramses Smith: 1) four men were present in the McDonald's parking lot; 2) one of them, called Ribs or Rizz, was unknown to Agent Demlein; 3) Agent Demlein identified that man on January 28, 2002 as Ramses Smith from what was now described as a multi-photo array following a review of video footage; 4) Agent Demlein's physical description of this man omitted his earlier reference to "short cropped hair" or "corn rows," but now included the detail that the man's jacket was "distinctive," and that it was made by "Avirex."

E. Suppression Hearing

At the January 27, 2005 hearing on the defendant's motion to suppress, I heard yet another version of Demlein's story. Like the previous versions presented to the grand jury and at the detention hearing, Agent Demlein's suppression hearing testimony was made under oath.

Agent Demlein again placed four individuals in the McDonald's parking lot, one of whom was called "Ribs" or "Riz." (Suppr. Hr'g Tr. 9, 8-9; 12, 8-12, Jan. 27, 2005). This unknown man was "very tall, 6-3 foot tall, medium dark complexion, black male, mid-20's . . . not heavy set but broad build, like 200 pounds plus." Id. at 18, 11-14.

Agent Demlein described his effort to identify this unknown man. He testified that, after the gun transaction on January 28, 2002, he returned to the police station. There, he was shown "several photos," but conceded that could not identify the unknown man. Id. at 19, 4-22. This account contradicts Demlein's detention hearing testimony, according to which he made an identification on January 28, 2002 and was "a hundred percent sure." (Det. Hr'g Tr. 16, 25, Dec. 5, 2003).

Now, mixing in the version of events contained in his police reports, Agent Demlein testified that he returned again to the police station on March 13, 2002, and was shown a single photo by Boston Police Detective Marvin Wright. He identified the man in the photo, Ramses Smith, as "the individual that sold me a gun about a month ago with Paul [Cato]." Id. at 22, 11-12. He testified that he was "100 percent certain" about this identification. Id. at 23, 14.

Agent Demlein also identified Ramses Smith during the suppression hearing as the person he had previously identified from a photograph. Upon pointing out Smith in court, he remarked that he was "100 percent [certain] that's him," "the same as I was then." Id. at 47, 18-24.

Boston Police Detective Marvin Wright testified at the suppression hearing. He explained that he had stopped Ramses Smith on the street on February 1, 2002 because he fit Agent Demlein's initial description of the unknown man: "a black male, dark skin, approximately 6'3" wearing an Avirex jacket." Id. at 73, 18-19. Detective Wright testified that he pulled Smith's photo and later showed it to Agent Demlein, who identified Smith as the unknown man from the January 28, 2002 gun sale. Id. at 76, 17-20; 77, 1-5, 14-21.

This, then, is version number four, returning to some of the themes from earlier versions: 1) four men were present in the McDonald's parking lot; 2) one of them, called Ribs or Rizz, was unknown to Agent Demlein; 3) Agent Demlein first failed to identify this unknown man from a multi-photo array shown to him on January 28, 2002; 4) Agent Demlein ultimately identified Ramses Smith from a single photo shown to him by Detective Wright on March 13, 2002.

F. Reconstruction of the Photo Array Continuation of the Suppression Hearing

Seriously troubled by what I had heard at the January 2005 suppression hearing, I ordered the government to produce Detective Wright's field interrogation/observation report ("FIO") from his February 1, 2002 stop of Ramses Smith and a copy of the photographs shown to Agent Demlein on January 28 and March 13, 2002.

On February 14, 2005, the government produced Detective Wright's FIO, but explained in a letter that none of the photos shown to Agent Demlein, apart from that of Ramses Smith, can now be located. (Gov't Letter FIO, Feb. 14, 2005). According to the government, the photos were printed from a Boston Police database at the time they were shown to Agent Demlein. The government, however, cannot provide either the original photo printouts or a record of the photos that were selected from the database on either January 28 or March 13, 2002.

I allowed the suppression hearing to continue so that the parties could address this new information. By the end of the second day of the suppression hearing on February 16, 2006, however, the record was no clearer. The government had presented four different versions of Agent Demlein's out-of-court photo identification of Ramses Smith as the unknown third (or fourth) participant in the gun sale, to wit: 1) Agent Demlein's two police reports mention only the March 13, 2002 single-photo identification. 2) His grand jury testimony refers to a single-photo identification on January 28, 2002 and describes multiple Boston police officers' making the identification as well. 3) At the detention hearing, Agent Demlein testified that he made the identification of Ramses Smith, allegedly from a multi-photo array after also reviewing video footage, on January 28, 2002. 4) Finally, at the suppression hearing, Agent Demlein recounted an initial, failed attempt to identify the unknown man from a multi-photo array on January 28, 2002 and a subsequent "successful" identification of Ramses Smith from a single photo on March 13, 2002. The government cannot reconstruct any of the photo arrays about which Agent Demlein testified, and therefore cannot corroborate any of these four versions of events.

The constitutionality of an out-of-court identification pivots on the procedures the police employed. It is hard to imagine a moment in which careful record-keeping and accurate reconstruction could be more important. But instead of care and accuracy, the government has offered nothing but chaos and contradiction. Indeed, after picking my way through the underbrush of these contradictions, only one thing is clear: The procedure employed to produce the out-of-court identification of Ramses Smith was unconstitutionally suggestive and unreliable. Agent Demlein's accounts "evolved," to put it mildly, as it became clear the procedures would be challenged and that earlier versions were suspect. While the circumstances of Agent Demlein's identifications of Ramses Smith during the detention and suppression hearings — made on the record and in open court — are more clear, they are plainly a product of the earlier suggestive identification and no less flawed.

III. ANALYSIS A. Legal Standard

The standard for determining whether to suppress testimony concerning both out-of-court and in-court identifications is "that of fairness as required by the Due Process Clause of the Fourteenth Amendment." Manson v. Brathwaite, 432 U.S. 98, 113 (1977). If the defendant is able to show "a very substantial likelihood of irreparable misidentification," then the evidence should be excluded. Id. at 116; see also United States v. McLain, No. 94-3192, 1995 U.S. App. LEXIS 12955, at *9-*10 (6th Cir. 1995) (noting that the burden rests on the defendant to demonstrate that a police identification procedure was improper).

Courts evaluate the fairness of out-of-court and in-court identifications using a two-step inquiry. United States v. de Jesus-Rios, 990 F.2d 672, 677 (1st Cir. 1993). First, a court must determine whether the identification procedure was impermissibly suggestive.Id. at 677. If the procedure was not impermissibly suggestive, the analysis is over and defendant's motion to suppress is denied. If, however, the defendant establishes that the procedure was impermissibly suggestive, the court must decide whether the identification was nevertheless reliable under the totality of the circumstances. Id. In assessing reliability, courts traditionally consider five factors:

1) the witness's opportunity to view or observe the defendant;

2) the witness's degree of attention;

3) the accuracy of the witness's description of the defendant;

4) the level of certainty demonstrated by the witness at the identification;

5) the length of time between the crime and the identification.

This two-step analysis applies with equal force to identifications made, as here, by police officers or other government agents, who are presumably better trained than civilians in the processes of observation and recollection. To the extent that courts take into account a police officer or government agent witness' heightened powers of observation, they do so in connection with the second Manson factor: the witness' degree of attention. In fact, Manson itself concerned a police officer's out-of-court identification of a drug suspect from a single photo. 432 U.S. at 108 (commenting that the police officer, "as a specially trained, assigned, and experienced officer, . . . could be expected to pay scrupulous attention to detail").

In dissent, Justices Marshall and Brennan were less convinced of the reliability of police and government agent witness' observations. The dissenters remarked:

[B]oth common sense and scholarly study indicate that while a trained observer such as a police officer `is somewhat less likely to make an erroneous identification than the average untrained observer, the mere fact that he has been so trained is no guarantee that he is correct in a specific case. His identification testimony should be scrutinized just as carefully as that of the normal witness.' Moreover, `identifications made by policemen in highly competitive activities, such as undercover narcotic agents . . ., should be scrutinized with special care.'

In cases involving identifications by police officers, the First Circuit has treated officers' testimony as Justices Brennan and Marshall suggested, "as that of a normal witness." See, e.g., United States v. Lopez-Lopez, 282 F.3d 1, 10 (1st Cir. 2002) (upholding the admission of police officers' identification of drug distribution suspects without reference to the officers' heightened observation skills).

It falls to this Court, therefore, to examine both the suggestivity of the identification procedures employed in this case and, if suggestive, the reliability of the identifications that resulted. To the extent that Agent Demlein's arguably heightened observation skills are relevant, I will consider them in connection with the second Manson factor.

A final note: Though courts generally analyze both out-of-court and in-court identifications using the same two-step inquiry described above, the analysis differs slightly in cases where a witness' in-court identification follows an out-of-court identification achieved using an improperly suggestive procedure. In such cases, courts have expressed concern that the later in-court identification would be tainted by the earlier, improper, out-of-court identification. Simmons v. United States, 390 U.S. 377, 383-84 (1968) ("Regardless of how the initial misidentification comes about, the witness thereafter is apt to retain in his memory the image of the photograph rather than of the person actually seen, reducing the trustworthiness of subsequent lineup or courtroom identification."). An in-court identification of this type may nevertheless escape suppression if the court determines that the in-court testimony is somehow reliable despite the suggestiveness of the earlier identification procedure. In the First Circuit, the touchstone of this analysis is the question of whether the earlier out-of-court identification "has proven so impermissibly suggestive" as to vitiate the reliability of the later in-court identification and "give rise to a very substantial likelihood of misidentification." United States v. Maguire, 918 F.2d 254, 264-65 (1st Cir. 1990) (citing Simmons, 390 U.S. at 384).

Thus, in the case at bar, I will apply the two-step test to the three identifications at issue (out-of-court, detention hearing, and suppression hearing), inquiring into both suggestivity and reliability. When examining the in-court identifications Agent Demlein made at the detention and suppression hearings, I will consider whether their reliability has been undermined by the earlier, out-of-court photo identification procedure. I will then address the question of additional future in-court identifications.

B. The Out-of-Court Photo Identification

Despite the contradictions in Agent Demlein's four accounts of his photo identification of Ramses Smith, the common thread connecting them all is that he made the identification from a single photo. As for the timing of the photo identification, despite Agent Demlein's grand jury and detention hearing testimony that he identified Ramses Smith on January 28, 2002, right after the gun transaction, I find that the identification in fact took place two months later on March 13, 2002. This conclusion is supported by Detective Marvin Wright's testimony, as well as the description of the March 13 photo identification in Agent Demlein's March 19, 2002 police report and the absence in his January 30, 2002 report of any reference to identification attempts on that day.

I also find that, despite testimony about multiple identification attempts, the only time Agent Demlein viewed photos and attempted to identify the unknown participant in the gun transaction was on March 13, 2002, when he was shown a single photograph. This conclusion is consistent with Detective Wright's suppression hearing testimony that he did not recall showing any photos to Agent Demlein on January 28, 2002. (Suppr. Hr'g Tr. 80, 13-19, Jan. 27, 2005). It is also bolstered by the lack of any record of photos printed on January 28, 2002 by the detectives working on the gun sale investigation and the absence of any reference to a January 28 identification attempt in Agent Demlein's January 30 and March 19, 2002 police reports.

Having stilled the moving target of Agent Demlein's identification-related testimony, I now proceed to the two-step suggestivity and reliability analysis.

1. Suggestivity

It is axiomatic that identifications achieved through the use of a single photo are highly problematic. A single photo shown to an eyewitness (the proverbial "Is this the man you saw?" question) plainly suggests the guilt of the person pictured. As the First Circuit has commented, "[s]ingle photo identifications do, indeed, present so serious a danger of suggestiveness as to require that they be given extremely careful scrutiny. . . ."Nassar v. Vinzant, 519 F.2d 798, 801 (1st Cir. 1975); see also United States v. Gomez Pizarro, 364 F. Supp. 2d 56, 67 (D.P.R. 2005) (suppressing testimony concerning a single-photo identification on the ground of its suggestivity); United States v. Shay, No. 92-10396, 1993 U.S. Dist. LEXIS 9169 (D. Mass. June 29, 1993) (same). The Supreme Court has held similarly that "identifications arising from single-photograph displays may be viewed in general with suspicion." Manson, 432 U.S. at 116 (citing Simmons, 390 U.S. at 383). The Manson court ultimately determined that the single-photo identification procedure employed in that case was not impermissibly suggestive, noting that the police officer witness was given an opportunity to view the single photo "at his leisure," alone and outside the presence of other officers. 432 U.S. at 116.

In the case at bar, the single-photo identification procedure used to achieve Agent Demlein's identification of Ramses Smith on March 13, 2002 was impermissibly suggestive. In contrast toManson, Agent Demlein was not alone when viewing the photo. He was in Detective Wright's presence and, if his March 19, 2002 police report is to be believed, in the presence of Detective Antonucci as well. Moreover, though Detective Wright testified at the suppression hearing that he merely asked Agent Demlein, "Does this guy look familiar to you?", Agent Demlein had discussed the gun investigation and described the unknown man to Detective Wright prior to March 13. (Suppr. Hr'g Tr. 77, 14-18; 73, 3-19, Jan. 27, 2005). He was clearly aware that Detective Wright had been looking for the unknown man he had generally described. He surely knew who the "guy" in Detective Wright's single photo might be.

Agent Demlein's March 13, 2002 identification of Ramses Smith from a single photo was therefore the result of an impermissibly suggestive procedure. The next step in the analysis requires that I examine, under the totality of the circumstances, whether the identification was nevertheless reliable.

2. Reliability

Though it is incongruous to discuss "reliability" when the government's main witness has contradicted himself to such an extent, this Court will nevertheless proceed through the five reliability factors set out by the Manson Court.

a. Opportunity to Observe

At the suppression hearing, Agent Demlein testified that the gun transaction occurred sometime shortly after noon on January 28, 2002. (Suppr. Hr'g Tr. 7, 5-7, Jan. 27, 2005). He described the day as "bright." Id. at 20, 1-4.

Agent Demlein's position in the car, however, severely limited his opportunity to observe the unknown man, who was sitting directly behind him in the rear passenger's seat. Id. at 12, 13-16. In fact, Agent Demlein testified at the detention hearing that he could only "kind of acknowledge" this man "out of the side of [his] eye." (Det. Hr'g Tr. 14, 24-25, Dec. 5, 2003). Finally, Agent Demlein testified at the suppression hearing that the unknown man was in the car for a mere forty-five seconds. (Suppr. Hr'g Tr. 59, 4-20, Jan. 27, 2005).

Despite the hour and the brightness of the day, then, Agent Demlein's opportunity to observe the unknown man was limited by both his own position in the front seat and the length of time the man stayed in the back seat.

b. Degree of Attention

Though Agent Demlein is a trained ATF agent, and is presumably more skilled than the average observer in studying those around him, he clearly was not able to use those skills in the case at bar. Agent Demlein testified at the detention hearing that he "stayed focused on the man with the gun in the back seat," rather than on the unknown man. (Det. Hr'g Tr. 34, 2, Dec. 5, 2003). He was attuned to the danger of being shot, and therefore fixed his attention on Paul Cato, who had the gun and was negotiating its sale. (Suppr. Hr'g Tr. 20, 15-22, Jan. 27. 2005; Det. Hr'g Tr. 34, 1-2, Dec. 5, 2003). He hardly exercised the special training that so impressed theManson court, and surely did not pay close attention to the unknown man.

c. Accuracy of the Description

Though Agent Demlein's description of the unknown man changed slightly over time, particularly with respect to the man's hairstyle and jacket, it remained relatively consistent. It also fits Ramses Smith, at least as Detective Wright observed him on February 1, 2002 and as he was pictured in the single photo shown to Agent Demlein on March 13, 2002. (Suppr. Hr'g, Gov't Exh. 1, Jan. 27, 2005; Gov't Letter FIO, Feb. 14, 2005).

The greater concern here, however, is the description's extraordinary level of generality. The description was so broad as to encompass "hundreds of individuals." (Suppr. Hr'g Tr. 83, 3-6, Jan. 27, 2005). Indeed, Smith's attorney, at the suppression hearing, raised the possibility that Agent Demlein's description in fact fit Paul Cato more accurately than Ramses Smith. Id. at 60, 9-15; 64, 16-19. The nicknames "Rizz" or "Ribs," though they share the first letter "R" with "Ramses," are not distinctive enough to make this broad description more specific, and there has been no testimony regarding Smith's nickname, if he has one at all.

Thus, though Agent Demlein's description of the unknown man appears to fit Ramses Smith at the time he was stopped and later identified, its breadth limits its use in this analysis.

d. Level of Certainty

Agent Demlein testified at both the detention and suppression hearings that he was one hundred percent certain that the man he identified on March 13, 2002 was the unknown third (or fourth) participant in the gun sale. (Suppr. Hr'g Tr. 16, 25, Jan. 27, 2005; Det. Hr'g Tr. 23, 14, Dec. 5, 2003). He did not contradict himself on this point. The shifting sands of his testimony, however, in nearly every other respect, makes this statement less than credible.

e. Length of Time Between Crime and Identification

Almost two months elapsed between the date of the gun transaction and the photo identification. During that time, Agent Demlein testified that he was working at the Olympics in Salt Lake City, Utah. (Suppr. Hr'g Tr. 47, 10-17, Jan. 27, 2005), on scores of other cases.

The two month period between the gun transaction and Agent Demlein's viewing of the single photograph reduces the reliability of Agent Demlein's identification. Indeed, inManson, the Court found a photo identification reliable when it occurred "within minutes of the crime," commenting, "[w]e do not have here the passage of weeks or months between the crime and the viewing of the photograph." 432 U.S. at 115; but see Levasseur v. Pepe, 70 F.3d 187, 195 (1st Cir. 1995) (finding an identification reliable despite a six-month lapse between the crime and the identification because the strength of the other indicia outweighed the effect of the time lapse).

3. Conclusion: The Out-of-Court Photo Identification

Considering the totality of the circumstances, I conclude that Agent Demlein's single photo out-of-court identification of Ramses Smith was unreliable. The combination of Agent Demlein's limited opportunity to observe the unknown man, his relative inattention, his broad description of the man, and the two-month time lapse between the gun sale and the identification destroys the identification's reliability.

C. The Detention and Suppression Hearing Identifications

My conclusion that Agent Demlein's out-of-court photo identification of Ramses Smith was both suggestive and unreliable does not necessarily require suppression of the two in-court identifications of Smith as well. Those identifications must only be suppressed if the suggestiveness of the earlier photo identification procedure vitiates the reliability of any later identification, if the previous identification "has proven so impermissibly suggestive" as to "give rise to a very substantial likelihood of misidentification." Maguire, 918 F.2d 254, 264-65 (1st Cir. 1990) (citing Simmons, 390 U.S. at 384). I may also consider the curative power of cross-examination. de Jesus-Rios, 990 F.2d at 678.

In the case at bar, I can find nothing in the record that would suggest that Agent Demlein's two in-court identifications were free of the taint of the earlier, suggestive photo identification. Agent Demlein provided no additional testimony showing that his in-court identifications were made on a more reliable basis than the earlier photo identification. In fact, Demlein admitted that he had not seen the unknown man before, and never saw him again after, the gun transaction. (Det. Hr'g Tr. 25, 1-7, Dec. 5, 2003). If anything, Demlein's ability to identify the unknown man was significantly weaker at the detention hearing, nearly two years after the gun sale, and even more so at the suppression hearing, almost three years after the gun transaction. In addition, though Agent Demlein testified at the suppression hearing that he was "100 percent [certain] that's him," again, this statement of certainty (particularly when it is itself suspect) does not outweigh the significant evidence of unreliability. (Suppr. Hr'g Tr. 47, 18-24, Jan. 27, 2005).

As for cross-examination as a cure, while questions by Smith's attorneys have effectively exposed the contradictions in Agent Demlein's testimony, those questions do not possess enough curative power to undo the "very substantial likelihood of misidentification" present in this case.

I therefore conclude that Agent Demlein's in-court identifications of Ramses Smith were both suggestive and unreliable.

D. Additional Future In-Court Identifications

The defendant has also moved to suppress any future in-court identification that Agent Demlein might make. At least one other court in the First Circuit has considered a prospective motion to suppress future in-court identifications. See Gomez Pizarro, 364 F. Supp. 2d at 67-68 (suppressing any future in-court identification when there was no evidence that such an identification would be reliable despite an earlier suggestive identification procedure). In evaluating this motion, I am mindful of the Manson Court's admonition that "[j]uries are not so susceptible that they cannot measure intelligently the weight of identification testimony that has some questionable feature."Manson, 432 U.S. at 116; see also de Jesus-Rios, 990 F.2d at 678 (referring to "a case of minimal suggestivity that could have been cured at trial").

I share the Manson Court's confidence in the ability of jurors to evaluate the evidence before them, flaws and all. I conclude, however, that any future in-court identification by Agent Demlein in this case would suffer from the same grave reliability problems that afflict the three identifications examined above. As with the detention and suppression hearing identifications, cross-examination is an insufficient cure for these problems. I therefore conclude that any future in-court identification by Agent Demlein is suppressed as well. IV. CONCLUSION

The First Circuit teaches that an identification should be suppressed only in "extraordinary circumstances." United States v. Watson, 76 F.3d 4, 6 (1st Cir. 1996). Agent Demlein's out-of court photo identification of Ramses Smith — all four versions of it — as well as his two previous and any future in-court identifications, however, present just such a circumstance.

In the case at bar, the Manson dissenters' misgivings appear to have been borne out. Justices Marshall and Brennan warned of the danger of identifications made by government agents "in highly competitive activities, such as undercover narcotics agents. . . ." 432 U.S. at 130. In Manson, the undercover operation was a drug sting; here, Agent Demlein was working to rid the streets of illegal firearms. I applaud the goals of this operation, but am troubled by what appears to be, at best, extreme sloppiness and, at worst, manipulation of this Court in an agent's zeal to obtain a conviction.

I therefore GRANT the defendant's motion to suppress [docket entry #25] as to the out-of-court photo identification, the in-court identifications at the detention and suppression hearings, and any future in-court identification of Ramses Smith by Agent Demlein.

SO ORDERED.