Although I concur with the majority opinion that the lower court did not err in denying the request of counsel to argue cross-racial identification to the jury, I write separately to express my profound concern that, in a proper case, counsel should be allowed latitude by the trial judge with respect to argument concerning matters legitimately a part of a judicial proceeding. At the outset, I agree wholeheartedly with the *709dissenting opinion of my colleague who points out that “appellant clearly had the right to attack the credibility of this witness, and to do so by arguments directed at proving a weakness in her capacity to perceive or remember the facial characteristics of her attacker.” I further recognize that, as the dissent points out, the eyewitness identification by the victim of the appellant was not only the “critical evidence,” it was the only evidence which was contested at trial.
Significantly I further acknowledge that “there is a substantial body of empirical study suggesting that cross racial identification, particularly by whites of blacks, is more difficult than identification of a person within one’s own race.” I view cross-racial identification as a subset of the more universal problem of the unreliability of eyewitness identification, generally. Historically, as a precursor to addressing the problem of cross-racial identification, many studies have shown that eyewitness identification is often unreliable, particularly when the opportunity to observe is short and the critical moment is stressful. The factors which contribute to the unreliability of eyewitness identification, generally, are merely exacerbate when the witness is called upon to distinguish features unfamiliar to him or her. Thirty seven years ago, the Supreme Court attempted to address factors known to result in mis-identification of criminal defendants.12
Counsel should be allowed, in my judgment, with proper safeguards, to argue to a jury, that it may consider the greater difficulty of recalling the features and identifying characteristics of a member of a group with whom one has had *710little or no contact as opposed to a member of one’s own ethnic or racial group.
That said, I part company with my dissenting colleague in equating an argument regarding cross-racial identification based on “the common sense view that whites have greater difficulty identifying blacks” with the principle, well-established in Maryland, that counsel may draw reasonable inferences from evidence before the court or jury. As the dissenting opinion notes, appellant made a motion in limine, requesting the court to approve a jury instruction on cross-racial identification in contemplation of raising the issue in his opening statement. It was incumbent upon counsel, at that juncture or during the trial, to proffer an evidentiary basis for the theory he wished to argue to the jury.
Cross examination remains the most effective means of challenging the accuracy of the witness’ in-court identification. In the instant case, as the majority points out, appellant was allowed to argue that which had been explored on cross-examination, e.g., lack of prior knowledge of assailants, conditions bearing on opportunity to observe and the complexions and hair styles of the persons shown in photo array vis-a-vis those of appellant. It should be noted that the trial judge permitted appellant to point out during closing argument to the jury the racial difference between the victim and the appellant. What was left unsaid was that in the experience of some (or many), it is more difficult to identify members of a different race than members of one’s own. I do not believe it would have been improper had the trial judge allowed counsel to make that comment. In the absence of circumstances indicating the victim in this case lacked familiarity with African Americans or that there were other race-based circumstances bearing on her ability to identify appellant, the trial judge, in my judgment, did not abuse her discretion.
In many of the court decisions 13 which consider cross-racial identification, empirical data has been introduced into evi*711dence, providing a sound common basis for each of the jurors to apply the principle. The majority discusses the certitude of the testimony of Christine Randall, the victim, who stated she was “good with faces” attributing that quality to her background in art and “painting people.” Although I believe that it is preferable that the jury be aided in what to consider when judging the accuracy of cross-racial identification, i.e. the results of empirical studies or research, at the very least, there should be some evidentiary basis to believe a witness’ identification may be impaired because of unfamiliarity with members of another race or ethnic background. The witness’ testimony, in my judgment, could properly provide that evi-dentiary basis. In the case, sub judice, however, I discern no indication in the testimony of the victim that the fact that appellant was of a difference race impaired her ability to identify him.
. Recognizing the vagaries of eyewitness identification, the Supreme Court, set forth elaborate procedures in the Wade (United States v. Wade, 388 U.S. 218, 87 S.Ct. 1926, 18 L.Ed.2d 1149), Gilbert (Gilbert v. State of California, 388 U.S. 263, 87 S.Ct. 1951, 18 L.Ed.2d 1178), Stovall (Stovall v. Denno, 388 U.S. 293, 87 S.Ct. 1967, 18 L.Ed.2d 1199) trilogy (decided on June 12, 1967) to ensure that judicial identifications upon which convictions are obtained are not the result of observations made at any time other than during the criminal event or that such identifications are not "impermissibly suggestive” because of improper police conduct. These procedures, however, address external influences rather than questions of perception and memory of the witness.
. See e.g., State v. Cromedy, supra.