I concur in the judgment with respect to defendant Rivas but respectfully dissent with respect to defendant Trevino.1

*700Though I am in complete agreement with the majority’s holding that Hispanics are a cognizable class and that the exclusion of Spanish surnamed jurors is sufficiently indicative of exclusion of Hispanics so as to shift to the People the burden to rebut the inference of impermissible group bias, I cannot concur with the standard of review employed by the majority to determine whether the inference of group bias was rebutted in this case. Instead, I believe we should adhere to the approach set forth in People v. Wheeler (1978) 22 Cal.3d 258, 282 [148 Cal.Rptr. 890, 583 P.2d 748] and reiterated in People v. Hall (1983) 35 Cal.3d 161, 168 [197 Cal.Rptr. 71, 672 P.2d 854]—“we rely on the good judgment of the trial courts to distinguish bona fide reasons for such peremptories from sham excuses belatedly contrived to avoid admitting acts of group discrimination.” It is in this light that I review in some detail the specific reasons which the prosecutor gave for the peremptory challenges at issue in this case.

First, however, some general observations:

(1) This particular trial judge did precisely what the trial court in Hall failed to do: he made a “sincere and reasoned attempt to evaluate the prosecutor’s explanation in light of the circumstances as then known, his knowledge of trial techniques, and his observations of the manner in which the prosecutor has examined members of the venire and has exercised challenges for cause or peremptorily . . . .” (Hall, supra, 35 Cal.3d at pp. 167-168.) In Hall the trial court made it known that it was abdicating its function and “apparently considered itself bound to accept all of the prosecutor’s explanations at face value . . . .” {Id.., at p. 169.) Nothing of the sort can be said here. Yet the majority gives no weight to the trial court’s assessment of the prosecutor’s veracity and instead assumes the crucial fact-finding role itself. So much for reliance on the “good judgment” of the trial court.

(2) The majority focuses on the six Hispanics who were peremptorily excused by the prosecution and pretty well ignores the characteristics of the fifteen non-Hispanics who were similarly challenged and of the jurors and alternates who were eventually seated. This inevitably distorts the picture. For example, the majority discounts Faller’s reason that he dismissed certain Hispanics because they had no strong views on the death penalty, pointing out that seven of the jurors eventually seated professed similar attitudes. Not discussed, however, are other characteristics of the seated jurors which might properly have neutralized this “minus” from the prosecution’s point of view. Nor does the majority point out that Faller challenged four non-Hispanics—Opal Jennings, Bonnie Goodwin, Patricia Olds, Laquita Brown—who had no views on the death penalty.

(3) In this connection I note that the majority apparently objects to this dissent’s comparing of the prosecution’s claimed specific bias of challenged *701Hispanics with similar characteristics of non-Hispanics who were also challenged by the People—such as the non-views on the death penalty of prospective jurors Jennings, Goodwin, Olds and Brown. The majority says that “such a comparison must be presented to the trial court at the time the prosecution attempts to justify the challenges made to members of the cognizable class.” (Ante, p. 692.) One cannot quarrel with that statement, but unless we are to assume that the trial judge had left the bench at the relevant times, he was of course aware of the voir dire examination which preceded the challenges to Jennings, Goodwin, Olds and Brown. Thus the majority’s mandate that the comparison be before the trial court was amply fulfilled. While it was the prosecution’s burden to show that its challenges were triggered by specific bias, I know of no rule which prohibits it from letting the record speak for itself.

(4) The majority opinion states that “ [t]he district attorney exercised peremptory challenges to remove six Hispanics from the jury after asking them few if any questions on voir dire.” (Ante, p. 687.) In so stating, the majority overlooks the fact that the Witherspoon examination is an integral part of voir dire. During the Witherspoon examination, Faller questioned Alex Fació, Robert Guerrero, Justina Gonzalez, and Gloria Longoria. The only Spanish sumamed prospective juror that he did not question was Mary Silvas. The majority’s observation is, therefore, incorrect.

Further, the majority’s reliance on the fact that the prosecutor did not engage many of the Hispanic jurors in extended questioning is misleading in another respect. Faller was, of course, not the only one to question the jurors on voir dire. In this case, each juror was first questioned by the court and each of the two defense counsel, before the voir dire questioning passed to Faller. There is obviously nothing improper in an attorney exercising a peremptory challenge on the basis of impressions gained from a potential juror’s responses to questions posed by the court or by other counsel. The record here discloses that Faller declined to pose additional questions to many of the potential jurors—both non-Hispanic and Hispanic—and that he exercised peremptory challenges with respect to a number of the non-Hispanic jurors whom he did not personally question. Thus, the trial court could reasonably have found that no inference of group bias arose from the fact that Faller excused some jurors without lengthy questioning.

With these preliminaries out of the way, I review the reasons Faller gave for peremptorily challenging the six Spanish surnamed jurors. I do not propose to rely on any bases for challenging the six prospective jurors other than those adduced by Faller.

*702 Alex Fació

The majority concedes that this prospective juror was challenged for specific bias. It is, however, important to note that the reasons for challenging Fació were consistent with those given with respect to other prospective jurors. In brief, Fació had relatives charged with serious crimes and he was “soft” on the death penalty. Several non-Hispanics who were challenged shared similar attributes.

Catalina Martinez

Fuller said he excused Martinez because he felt he did not have any strong opinions on anything and gave the impression that he would simply go along with the other jurors rather than form his own opinion and because Martinez went drinking with friends who steal. The majority apparently agrees with the Court of Appeal that the reasons given for his exclusion are “suspect” (ante, p. 692), but does not explain why.2 The reasons are legitimate.

Martinez admitted that he had friends who steal. He said he went drinking with them. The special circumstance charged in defendants’ murder trial was felony murder: it was alleged the murder was committed during the commission or attempted commission of a robbery. Martinez did not know whether his friendship with thieves would interfere with his ability to decide the case fairly. He conceded that it was “tough to project what you are going to do.”3 In so doing, he was admitting a specific bias.

Even though Martinez eventually responded to defense attorney questioning in such a way as to give the appearance that he could control his bias, the trial court could have determined that Fuller was justified in doubting Martinez’ neutrality. The fact that Martinez admitted he might be biased provides a valid ground for challenge.

*703Faller’s second reason for dismissing Martinez—that Martinez would simply go along with the other jurors rather than form his own opinion—is also a valid ground for challenge, and the trial court could legitimately have accepted this justification as well.

Robert Guerrero

Faller indicated that he excused Guerrero because he sat with his arms folded during the voir dire—which Faller felt indicated a closed relationship—and because Faller did not believe Guerrero was sincere in his answers. The majority holds that “excluding Robert Guerrero based on his body language and mode of answering certain questions, is particularly untenable in light of Wheeler’s requirement of a showing of specific bias.” (Ante, at p. 692.) In so holding, the majority has—apparently inadvertently—substantially changed the law surrounding peremptory challenges.

We recognized in Wheeler that “a prosecutor may fear bias on the part of one juror because he has a record of prior arrests or has complained of police harassment, and on the part of another simply because his clothes or hair suggest an unconventional lifestyle. . . . Indeed, even less tangible evidence of potential bias may bring forth a peremptory challenge: either party may feel a mistrust of a juror’s objectivity on no more than the ‘sudden impressions and unaccountable prejudices we are apt to conceive upon the bare looks and gestures of another’ (4 Blackstone, Commentaries *353) . . . .” (Wheeler, supra, 22 Cal.3d at p. 275, italics added).

We recently reaffirmed this proposition in People v. Hall (1983) 35 Cal.3d 161, 170 [197 Cal.Rptr. 71, 672 P.2d 854], where we held that “A prosecutor may act freely on the basis of ‘hunches,’ unless and until these acts create a prima facie case of group bias, and even then he may rebut the inference.” Clearly, therefore, an attorney may excuse a prospective juror because of a gut feeling. When an attorney does so, his concern is that the juror may be biased against his client.

The majority’s rejection of body language as a basis for specific bias points up another reason why its evaluation of the trial court’s ruling is quite unfair: the prosecutor was, of course, never asked to explain why he challenged 15 prospective jurors who were not Spanish sur named. We have no way of knowing how many of those 15 exhibited unacceptable body language just as we have very few clues as to the other reasons why they were excused. The majority therefore expresses doubts about the sincerity of the prosecutor’s explanations without knowing whether his reasons for *704challenging non-Hispanics were perhaps entirely consistent with those he gave for challenging the Spanish surnamed jurors.4

Faller also felt that Guerrero’s answers were evasive and indicated that he had “something else on his mind regarding these subjects which he was not expressing, ...” The trial court, which found this explanation to be credible, had had an opportunity to observe the prospective juror during voir dire—an advantage not shared by the majority which nevertheless feels that it is in a better position to judge Faller’s credibility. The trial court could properly find that Faller had sustained his burden of justification for dismissing Guerrero.

Justina Gonzalez

Faller stated that he had excused Gonzalez because she was young, because she was defensive and lacked confidence, and because he did not think she could “cooly analyze the evidence.”

In explaining his challenge to several jurors, Faller expressed his desire to have a jury “a bit more mature in basic overall age” because he did not want to risk juror sympathy based on age identification with the defendants. The majority does not suggest that a juror’s potential age identification with a particular defendant cannot give rise to specific bias, but simply rejects the prosecutor’s professed reliance on this factor as insincere, in large part, apparently, because the prosecutor did not explicitly question the jurors as to their age. (Ante, at p. 691.) It is hardly unusual, however, to make a rough judgment about age on the basis of appearance, and the trial court— who viewed the jurors in question—evidently did not find the prosecutor’s explanation suspicious.

In addition, the majority disbelieves Faller’s explanation that he excused Gonzalez because she was defensive. The majority states that “the district attorney condemn[ed] Longoria for expressing herself ‘rather haltingly,’ then [took] the opposite approach and criticizefd] Gonzalez’ perceived verbosity.” (Ante, at p. 692.) This is not true.

Faller never criticized Gonzalez’ verbosity. On the contrary, he stated, apparently with approval, “[Gonzalez] was very verbose and very interested *705in expressing her views on a great variety of subjects, [f] Yet at the same time when approached even by defense counsel in regards to certain questions, she would say, ‘No, I don’t want to answer that,’ or T want to talk to you in private,’ and appeared to have a lack of confidence in her own ability to form opinions and to express herself.” (Italics added.) Characterizing this as a criticism of her verbosity is simply incorrect. From the record it appears that Faller could have concluded in good faith that Gonzalez, while outwardly verbose, was actually defensive and lacked confidence. The trial court, of course, could find that he sustained his burden of justification in dismissing her on this ground as well.

Mary Silvas

Faller explained that he had excused Silvas because she was female, had a child about the age of one of the defendants, and had not even thought about the death penalty—which Faller felt indicated a lack of commitment to consider seriously what she would do during sentencing. Although the majority simply rejects each of these reasons as disingenuous, the trial court could reasonably have found that at least the last two reasons demonstrated that Faller had not excluded her because she was Hispanic. As discussed above, Faller excluded other jurors who he thought might feel undue sympathy for the defendants because of age identification and he excused at least one non-Hispanic juror, Jill Moffitt, who—like Silvas—had children of about the defendants’ age. In addition, throughout the voir dire Faller utilized peremptory challenges frequently with respect to jurors who indicated either that they would not or might not vote for the death penalty.5 The trial court could have found that Faller did not want to take a chance that Silvas— who stated that she had not even thought about the death penalty—might decide that she could not impose such a penalty when called on to consider the issue during trial.

Gloria Longoria

Faller informed the court that he had excused Longoria because he felt she was an equivocal juror during the Witherspoon examination and because she expressed herself rather haltingly. Faller explained that he was concerned that Longoria would be unable to reach a verdict independently or to consider seriously and vote for the death penalty. He also stated that he had excused her because she was female and younger than he would have preferred. In criticizing these explanations, the majority attacks only the *706last two—dismissal because of age or sex. The opinion omits any consideration of the merits of the other reasons Faller gave for dismissing Longoria. Certainly Faller could legitimately excuse a juror he felt was equivocal during the Witherspoon examination. Similarly, Faller’s concern that Longoria would have been unable to form an opinion independently and reach a verdict was a valid ground for dismissal.

Conclusion

So that there will be no misunderstanding, I should make it clear that this dissent is not intended as a “jury argument” in favor of the prosecutor’s professed reasons for challenging the six Hispanics. If the trial court had concluded that the prosecutor’s explanations were not genuine and had found that the peremptory challenges to the six Hispanic jurors had been exercised because of supposed group bias, I would have no reason to quarrel with that determination. The trial court came to the opposite conclusion, however, and given its ability to view the prosecutor and the potential jurors in person and to evaluate the credibility of the prosecutor’s explanations from that perspective, I submit that it is inappropriate—and contrary to the teaching of Wheeler and Hall—for us to brush aside the trial court’s finding and substitute our own view of the matter.

*676Opinion

Jeremiah Trevino appeals from a conviction of murder (Pen. Code, § 187).1 The People appeal the dismissal of murder charges against his codefendant, Leonard L. Rivas. The parties raise two separate and distinct claims.

Trevino raises the question whether the term “Spanish surnamed” is sufficiently descriptive of a cognizable group to satisfy the Wheeler2 test for identifying prosecutorial abuse of the peremptory challenge in violation of a defendant’s constitutional right to be tried by a jury drawn from a representative cross-section of the community. We conclude that the phrase does describe a cognizable' class, namely Hispanics, and that the prosecution failed to sustain its burden of proving that Hispanics were not impermissibly excluded from jury service on the basis of group bias. Trevino’s conviction, therefore, must be reversed.

The People assert that the trial court’s order granting Rivas a new trial was not tantamount to an acquittal, regardless of the language used in the order. This argument overlooks the threshold inquiry as to whether the trial court erred earlier in the trial in denying Rivas’ motions for judgment of acquittal based on insufficiency of the evidence. (§ 1118.1.) Because the evidence was “insufficient as a matter of law,” as the trial court belatedly acknowledged in ruling on Rivas’ motion for a new trial, and Rivas was therefore entitled to acquittal, we conclude that the trial court responsible for conducting a new trial correctly discharged Rivas on the basis of his once in jeopardy plea.

The facts as to the two defendants are interwoven. On June 11, 1979, Rollo “Ted” Hinton’s body was found in the bedroom of his ransacked apartment. Cause of death was determined to be a strong blow to the back of the head. The possible murder weapon, a steam iron, was found next to the body.

On January 26, 1982, an information was filed charging Jeremiah Trevino and Leonard L. Rivas, friends of the deceased, each with one count of murder (§ 187) with a special circumstances allegation that the murder was committed during the commission or attempted commission of a robbery. (§§ 211, 190.2, subd. (a)(17).) Both men pleaded not guilty.

During the jury selection phase of the trial, the district attorney used peremptory challenges to exclude six Spanish surnamed individuals called *677to sit on the panel of twelve: Gloria Longoria, Justina Gonzalez, Mary Silvas, Catalina Martinez, Alex Fació and Robert Guerrero.3 No Spanish surnamed jurors remained on the panel.

Following the prosecutor’s exclusion of the sixth Spanish surnamed juror, the defense advised the court that it wished to make a Wheeler motion regarding the district attorney’s seemingly automatic removal of this distinct group of jurors. The court apparently told the defense to wait until the three alternate jurors had been selected before pursuing the motion. During the selection of alternates, the district attorney peremptorily challenged one Spanish surnamed juror, Arnold Rocha, without questioning him on voir dire, but allowed another Spanish surnamed juror, Marcene Rocha, to remain as an alternate.

The defense then made a Wheeler motion, chronicling the district attorney’s behavior as a violation of the defendants’ constitutional right to a jury drawn from a representative cross-section of the community. The trial court called upon the prosecutor to explain his exclusion of the Spanish surnamed jurors and accepted the proffered justifications. The motion was denied.

The trial commenced on June 29, 1982. Wilma Nyberg, a neighbor of the deceased, testified for the prosecution that, at approximately 3 p.m. on the day in question, she saw a man she identified as Trevino exiting Hinton’s apartment. At the same time, she heard a groan, and what sounded like Hinton saying “oh no,” coming from inside the apartment. Immediately thereafter, a tall, slender, clean-shaven man emerged from the apartment. This man carried himself in the same manner as Rivas did in the courtroom. Both he and the first man appeared intoxicated. The two unsuccessfully attempted to start a light colored car4 parked in front of Hinton’s apartment. As Nyberg entered her apartment, she saw that the two men were rocking the car to try to start it. Later that afternoon she noticed that the car and the men were gone.

Nyberg testified that she had attended a physical lineup in November 1981 where she had been able to identify Trevino as one of the men she saw, but she was unable to positively identify Rivas. Two days prior to the physical lineup she had been shown a photographic lineup, which included Rivas’ picture. Nyberg identified a different person as the second man she had seen at the Hinton apartment. Nyberg testified that she had been shown other photographic displays and had identified Rivas, but the parties later stipulated to the fact that Nyberg was mistaken, as she had never been shown any other such lineup.

*678Sergeant Wittman, the investigating officer who had shown Nyberg the photo display in 1981, testified that the picture of Rivas included in that display was taken the day after Hinton’s death. Contrary to Nyberg’s description of the second man she saw, the picture revealed that Rivas did not appear to be slender at the time and did have a moustache.

Nyberg admitted on the stand that “even today I can’t positively identify Mr. Rivas other than he is the type of man that came through that corridor. ” The only other evidence linking Rivas to the scene of the crime was one fingerprint found on a dresser drawer. The expert who dusted Hinton’s apartment for prints testified that he also found one of Trevino’s prints on the dresser, and two of Hinton’s prints and two of another individual’s on a whiskey bottle in the living room. The expert was unable to estimate the age of the prints, admitting that they could have been made months earlier. Another prosecution witness confirmed that Trevino, Rivas and members of Trevino’s family visited Hinton on occasion and referred to him affectionately as “Primo.”5

The prosecution also introduced evidence that Trevino and Rivas were seen together on the day of the crime. The two had visited Trevino’s girlfriend in the hospital sometime after 6 a.m. when she gave birth to a child. The prosecution failed in its efforts to pinpoint the time with any greater accuracy.

At the close of the prosecution’s case, the defendants moved for judgment of acquittal based on insufficient evidence. (§ 1118.1.) The court denied the motions.

Defense witnesses testified that they had seen Hinton alive on the day in question at 2:30 p.m. and at 4 p.m. The defendants did not testify. At this point the defense renewed the motions for acquittal, which the court again denied.

The jury found the defendants guilty of first degree murder with a finding that the special circumstances allegation was true. The court set aside the special circumstances finding as to both Trevino and Rivas. The defendants then moved for a new trial based on insufficiency of the evidence to support the murder convictions. (§ 1181.) The court denied the motion as to Trevino, but granted as to Rivas. Trevino received a 25 years to life sentence.

Rivas later pled not guilty to the charges of murder and robbery contained in an amended information. The second trial court then allowed him to *679change his plea to once in jeopardy (§ 1016), and discharged him on that basis.

I

We deal first with the issue whether the prosecutor’s use of peremptory challenges to exclude the jurors in question violated Trevino’s constitutional right to trial by a jury drawn from a representative cross-section of the community. (Cal. Const., art. I, § 16; Wheeler, supra, 22 Cal.3d 258.) Before reaching the merits of this claim, a review of the fundamental nature of the right in question, and the role of the prosecutor in safeguarding that right, is appropriate.

“The purpose of [the] jury is to guard against the exercise of arbitrary power—to make available the commonsense judgment of the community as a hedge against the overzealous or mistaken prosecutor and in preference to the professional or perhaps overconditioned or biased response of a judge [citations].” (Taylor v. Louisiana (1975) 419 U.S. 522, 530 [42 L.Ed.2d 690, 698, 95 S.Ct. 692].) The representative cross-section requirement is designed to serve several related functions, each of which complements this central role of the jury in our system of justice. First, and foremost in the eyes of the criminal defendant whose fate lies in the jurors’ hands, the representative cross-section requirement protects, to the greatest extent possible, the right to be tried by an impartial jury. (Id., at pp. 527-528 [42 L.Ed.2d at pp. 696-697]; Wheeler, supra, 22 Cal.3d at pp. 266, 271-272; People v. White (1954) 43 Cal.2d 740, 754 [278 P.2d 9].) Of equal import to our overall concept of justice, the fair cross-section rule serves to “legitim[ize] the judgments of the courts, promot[e] citizen participation in government, and prevent[] further stigmatizing of minority groups. ” (Italics added.) (Wheeler, supra, 22 Cal.3d at p. 267, fn. 6.) Certainly these interests are not advanced, but thwarted, by systematic exclusion from the jury of an identifiable group within the community. (See Note, Limiting the Peremptory Challenge: Representation of Groups on Petit Juries (1977) 86 Yale L.J. 1715, 1735-1738.)

These interests, we have noted, are essential to the integrity and continued vitality of our judicial system. For this reason, the representative cross-section rule is not limited in application to traditional minority groups alone. Thus, in Theil v. Southern Pacific Co. (1946) 328 U.S. 217 [90 L.Ed. 1181, 66 S.Ct. 984, 166 A.L.R. 1412], the high court held that exclusion of daily wage earners from jury service violated the cross-sectional requirement, threatening “the subtle undermining of the jury system.” (Id., at p. 225 [90 L.Ed. at p. 1187].) The court cautioned that the rule does not mean “that every jury must contain representatives of all the economic, social, reli*680gious, racial, political and geographic groups in the community; . . . [b]ut it does mean that prospective jurors shall be selected by court officials without systematic and intentional exclusion of any of these groups.” (Id., at p. 220 [90 L.Ed. at p. 1185].) To hold otherwise, the high court stressed, “would breathe life into any latent tendencies to establish the jury as the instrument of the economically and socially privileged. That we refuse to do.” (Id., at p. 224 [90 L.Ed. at p. 1186].)

These concerns, nonetheless, are amplified in the context of discriminatory exclusion of a racial or ethnic minority group, as alleged in the present case.6 As this court emphatically stated nearly 50 years ago, “ ‘one who is on trial for an alleged crime is entitled to a jury from which individuals of his own race who are otherwise qualified as jurors in the particular case, have not been arbitrarily excluded merely because of their nationality, race or color.’” (Italics added.) (People v. Hines (1939) 12 Cal.2d 535, 539 [86 P.2d 92].) Such exclusion is suggestive of “decision-making by racial stereotype, a technique that should be anathema in our courts.” (People v. Johnson (1978) 22 Cal.3d 296, 299 [148 Cal.Rptr. 915, 583 P.2d 774].)

Over 40 years ago, Justice Black admonished the courts that “[f]or racial discrimination to result in the exclusion from jury service of otherwise qualified groups not only violates our Constitution and the laws enacted under it but is at war with our basic concepts of a democratic society and a representative government.” (Fn. omitted.) (Smith v. Texas (1940) 311 U.S. 128, 130 [85 L.Ed. 84, 86, 61 S.Ct. 164].) In Wheeler, we added that “in such a war the courts cannot be pacifists.” (Wheeler, supra, 22 Cal.3d at p. 267.) Vigilant enforcement of a criminal defendant’s right to a jury drawn from a representative cross-section of the community is the single most effective weapon available in our constitutional arsenal to combat such exclusion.

Responsibility for the preservation of a defendant’s cross-sectional rights does not fall exclusively on the shoulders of the court overseeing the selection of the petit jury.7 The prosecutor also shares in that responsibility.

The prosecutor’s role is a unique one within the criminal justice system. Though the district attorney must diligently discharge the duty of *681prosecuting individuals accused of criminal conduct, the prosecutor may not seek victory at the expense of the defendant’s constitutional rights. (People v. Superior Court (Greer) (1977) 19 Cal.3d 255, 266 [137 Cal.Rptr. 476, 561 P.2d 1164]; People v. Fuller (1982) 136 Cal.App.3d 403 , 424 [186 Cal.Rptr. 283].) Thus, the prosecution is obligated to respect the defendant’s right to a fair and impartial trial in compliance with due process of law. (Berger v. United States (1935) 295 U.S. 78, 88 [79 L.Ed. 1314, 1321, 55 S.Ct. 629]; People v. Lyons (1956) 47 Cal.2d 311, 318, 319 [303 P.2d 329]; Fuller, supra, 136 Cal.App.3d at p. 424, fn. 26.) Moreover, the prosecutor may not bring criminal charges against an individual unless supported by probable cause, and, once charges are instituted, must reveal to the court any information which negates the existence of probable cause. (Rules Prof. Conduct, rule 7-102.)

The prosecutor’s dual role, as the defendant’s adversary and as a guardian of the defendant’s constitutional rights, extends to the prosecutor’s use of peremptory challenges to exclude individuals from the jury. The prosecutor is free to use the statutorily provided peremptory challenges in any manner he or she chooses, so long as that use does not obstruct the defendant’s constitutional rights. (See Wheeler, supra, 22 Cal.3d 258; People v. Hall (1983) 35 Cal.3d 161, 170 [197 Cal.Rptr. 71, 672 P.2d 854],)8 As we held in Wheeler, use of peremptory challenges against jurors on the ground of group bias alone violates the right to a jury drawn from a representative cross-section of the community. (Id., at pp. 276-277.) The prosecutor must be ever watchful to insure that in his or her conscientious efforts to efficiently discharge official duties the statutory right to peremptorily challenge jurors is not exalted over the defendant’s constitutional right to a jury drawn from a representative cross-section of the community. (See Wheeler, supra, 22 Cal.3d at p. 281, fn. 28; Fuller, supra, 136 Cal.App.3d at p. 416.)

The prosecutor further shares responsibility for protecting the accused’s cross-sectional rights because he or she acts on behalf of the state in conducting criminal prosecutions. (See Imbler v. Pactman (1976) 424 U.S. 409 *682[47 L.Ed.2d 128, 96 S.Ct. 984].) As we noted in Greer, supra, 19 Cal.3d 255, “the role of the prosecutor ... [is not] simply a specialized version of the duty of any attorney not to overstep the bounds of permissible advocacy .... In all his activities, his duties are conditioned by the fact that he ‘is the representative not of any ordinary party to a controversy, but of a sovereignty whose obligation to govern impartially is as compelling as its obligation to govern at all . . . (Id., at p. 266.) Thus, the prosecutor must execute the duties of this representative office diligently and fairly, avoiding even the appearance of impropriety that might reflect poorly on the state.

In the context of jury selection, the prosecutor must take care to exercise peremptory challenges for legitimate purposes only, ever mindful of the prosecutor’s role as a representative of the state. The use of group stereotypes as the sole criterion for the exercise of these challenges not only violates the defendant’s constitutional rights, but attributes discriminatory intent to the state as well, detracting from the legitimacy of the state-imposed process. “[P]art of the function of the jury is to ensure that the public will feel its conscience is reflected in the verdict .... To allow the prosecutor to exercise peremptory challenges based on group affiliation alone casts a shadow on the appearance of integrity which must of necessity underlie public identification with the verdict.” (Note, The Defendant’s Right to Object to Prosecutorial Misuse of the Peremptory Challenge (1979) 92 Harv. L. Rev. 1770, 1782.) The prosecutor, as an officer of the state, must take extra pains to insure that use of peremptory challenges does not result in a violation of an accused’s right to be tried by a jury drawn from a representative cross-section of the community. The courts, in turn, must be vigilant in seeing to it that prosecutors do not overstep the constitutional bounds of the cross-section requirement in selecting members of the jury.9

We turn to the merits of the case at bench. In Wheeler we formulated a three-part seriatim test by which a defendant can challenge the prosecutor’s misuse of peremptory challenges to strike jurors on the basis of group bias alone.10 To make out a prima facie case of discriminatory exclusion, the defendant must establish: (1) “[A]s complete a record of the circumstances as is feasible”; (2) “that the persons excluded are members *683of a cognizable group . . and (3) “a strong likelihood that such persons are being challenged because of their group association rather than because of any specific bias.” (Wheeler, supra, 22 Cal.3d at p. 280.) Once the defendant satisfies these requirements to the court’s satisfaction, the burden shifts to the prosecutor to show that the jurors in question were legitimately excused.

In the case at bench, the trial court did call upon the prosecutor to justify his use of peremptories to eliminate the Spanish surnamed jurors after the defendants, in a Wheeler motion, brought this systematic exclusion to the court’s attention. In ruling on the motion, the court explicitly referred to the prosecutor’s “burden of justification.” Nonetheless, the People assert that Trevino failed to satisfy the second prong of the Wheeler test in that he did not establish that the excluded jurors were members of a cognizable class.

In People v. Harris (1984) 36 Cal.3d 36 [201 Cal.Rptr. 782, 679 P.2d 433], this court expressly recognized Hispanics as a cognizable class for purposes of determining whether a defendant has made out a prima facie case of constitutionally invalid selection of the jury pool.11 Responding to the defendant’s contention that “Blacks and Hispanics” were discriminatorily excluded from jury service, we concluded: “Both Blacks and Hispanics share with other members of their groups a common perspective arising from their respective experiences as a group, and no other members of the community are capable of adequately representing their perspective. Thus, Blacks and Hispanics are cognizable groups for purposes of fair cross-section analysis. (See, e.g., Castaneda v. Partida, supra, 430 U.S. 482, 495 []; Hovey v. Superior Court (1980) 28 Cal.3d 1, 20, fn. 45 [].)” (Id., at p. 51.)

Many ties bind Hispanics together as a cognizable group within the community. Hispanics often share an ethnic and cultural “community of interest,” including language, history, music and religion. In addition, Hispanics have made notable achievements in the professions, the arts, industry and public life. On a more somber note, Hispanics, in relation to other Americans, share a host of harsh realities, such as relatively high unemployment, poverty, relative lack of educational opportunity and, of import to the present case, discrimination directed at them precisely because they are Hispanic.

In the case at bar, the defense presented its Wheeler motion by referring to the district attorney’s systematic elimination of “Spanish surnamed” ju*684rors.12 The issue, therefore, is whether the term “Spanish surnamed” is sufficiently descriptive of Hispanics, a cognizable class within the community. We conclude that it is.

In Hernandez v. Texas (1954) 347 U.S. 475 [98 L.Ed. 866, 74 S.Ct. 667], the high court expressly rejected the argument that Spanish surnames do not identify a definable segment of the population. In holding that the systematic exclusion of persons of Mexican descent from jury service violated the defendant’s equal protection rights under the Fourteenth Amendment, the court stated: “The State challenges any reliance on names as showing the descent of persons in the County. However, just as persons of a different race are distinguished by color, these Spanish names provide ready identification of the members of this class.” (Id., at p. 480, fn. 12 [98 L.Ed. at p. 871].)

Twenty-three years after Hernandez, the high court reaffirmed the propriety of using Spanish surnames to define the relevant class of persons discriminatorily excluded from jury service in a given case. In Castaneda v. Partida (1977) 430 U.S. 482 [51 L.Ed.2d 498, 97 S.Ct. 1272], the court noted: “For our purposes, the terms ‘Spanish-surnamed’ and ‘Mexican-American’ are used as synonyms for the census designation ‘Persons of Spanish Language or Spanish Surname.’” (Id., at p. 486, fn. 5 [51 L.Ed.2d at p. 505].) The court then went on to determine that automatic elimination of persons from grand jury service on the basis of the “readily identifiable” characteristic of having a Spanish surname (id., at p. 497 [51 L.Ed.2d at p. 512]) unconstitutionally discriminated against “Mexican-Americans”13 in the county in question. (Id., at pp. 496-499 [51 L.Ed.2d at pp. 511-513].)

*685We judicially notice14 that governmental agencies have made extensive use of the category “Spanish surnamed” in compiling statistics on various aspects of American life; 2,222,185 Californians chose the category “Spanish surname” in responding to the 1970 census. (Californians of Spanish Surname (Fair Employment Practices Com., Div. of Fair Employment Practices (June 1976) p. 57).)15 Although the 1980 census did not include a “Spanish surname” response,16 the Bureau of the Census continues to maintain and update a census list of Spanish surnames.

The Equal Employment Opportunity Commission also compiles and publishes employment statistics on the basis of Spanish surnames. The commission defines the designation as follows: “The term Spanish surnamed American has been adopted ... to describe all persons of Mexican, Puerto Rican, Cuban, or Spanish origin. The term is meant to coincide with other terms employed by data collectors and analysists to denote the same group— for instance, Hispanic, persons of Spanish origin, Spanish Americans, persons of Spanish language.” (The Employment Status of Spanish Surnamed Americans in the New York SMS A (U.S. Equal Employment Opportunity Com., Research Rep. No. 42 (June 1974) p. iv).)17

Governmental categorization by Spanish surname signifies more than simple utility in statistical compilation. It demonstrates a broader understanding that individuals bearing those names represent a discrete segment of the population based on ethnic commonality.

In common parlance, the Spanish surname designation refers to Hispanics. The term Hispanic is formally defined as “American[s] of Spanish or *686Latin-American descent.” (American Heritage Dict. (2d college ed. 1982) pp. 613-614.) The Hispanic Policy Development Project18 defines “Hispanic” in the following manner: “[T]he term ‘Hispanic’ refers to all individuals who were classified as ‘Persons of Spanish Origin or Descent’ by the 1980 census. ‘Persons of Spanish Origin or Descent,’ and therefore ‘Hispanics,’ are those who classified themselves in one of the specific Spanish origin categories listed on the census questionnaire—Mexican, Puerto Rican, or Cuban—as well as those who indicated that they were of other Spanish/ Hispanic origin.” (Hispanic Policy Development Project, The Hispanic Almanac (1984) p. 161.)

The bulk of the case law analyzing representative cross-section claims involves discrimination against a cognizable class defined by race or sex. (See, e.g., Duren v. Missouri (1979) 439 U.S. 357 [58 L.Ed.2d 579, 99 S.Ct. 664] (women); Taylor, supra, 419 U.S. 522 (women); Wheeler, supra, 22 Cal.3d 258 (blacks); Johnson, supra, 22 Cal.3d 296 (blacks); Hall, supra, 35 Cal.3d 161 (blacks).) This type of discrimination, based on physical attributes, often is more easily isolated than ethnic discrimination, which may turn on some nonphysical characteristic such as a name, an accent or a style of dress. Nonetheless, these more subtle forms of ethnic discrimination are no less reprehensible than the most overt race or sex discrimination. In Wheeler we condemned group bias on the basis of “racial, religious, ethnic, or similar grounds.” (Italics added.) (Wheeler, supra, 22 Cal.3d at p. 276.) We must not close our eyes to such bias simply because it may be cloaked in the guise of neutrality.

We conclude, therefore, that the use of Spanish surnames to exclude jurors is sufficiently indicative of a systematic effort to excuse Hispanics to satisfy the requirements of the Wheeler test. Spanish surnames identify Hispanics, a cognizable class. Although the correlation between surname and group membership is not exact, such precision is unnecessary. Where, as here, no one knows at the time of challenge whether a particular individual who has a Spanish surname is Hispanic, a showing that jurors are being excluded on the basis of surname alone constitutes a prima facie case of exclusion of a cognizable class.

The argument that Trevino failed to establish a “cognizable class” for fair cross-section purposes because he did not prove up the national origin of each excluded Spanish surnamed juror, mischaracterizes Trevino’s motion and the Wheeler test itself. National origin is a relevant inquiry only where the motion is premised on that basis. Where, as here, the motion is *687based on ethnic identification, voir dire regarding national origin is unnecessarily intrusive upon the defendant’s legitimate interests. In Wheeler, we imposed no requirement that the defendant establish that systematically excluded black jurors were of Afro-American, Caribbean, African or Latin American descent. To graft such a requirement on to the Wheeler test in the context of exclusion of an ethnic group easily identified on the basis of surname is to place an unwarranted and impractical burden on the defendant’s ability to preserve his representative cross-section rights.

In Wheeler we stressed that “veniremen are not required to announce their race, religion, or ethnic origin when they enter the box, and these matters are not ordinarily explored on voir dire. The reason, of course, is that the courts of California are—or should be—blind to all such distinctions among our citizens.” (Wheeler, supra, 22 Cal.3d at p. 263. See also People v. Jones (1972) 25 Cal.App.3d 776, 782, fn. 5 [102 Cal.Rptr. 277] (“our officials are very properly colorblind”).) In the present case, defense counsels’ decision to make the Wheeler motion on the basis of easily identifiable surnames, rather than risk juror animosity in quizzing selected individuals as to whether or not they are Mexican-American, was proper. The prosecution cannot avoid the fact that the defendant identifies a cognizable class by suggesting an additional requirement of statistical perfection based on national origin. “A party does not sustain his burden of justification by attempting to cast a different burden on his opponent.” (1Wheeler, supra, 22 Cal.3d at p. 283, fn. 30.)

Although the defense need not establish the nationality of excluded Spanish surnamed jurors, the composition of the relevant Hispanic community may be telling. This is particularly true in the instant case. For purposes of fair cross-section analysis, the relevant demographic community is the county in which the trial takes place. We judicially notice the 1980 census results, which indicate that 73,298 persons of Spanish origin live in Tulare County, approximately 30 percent of the total population of the county (total=245,738). This Spanish origin population is overwhelmingly of Mexican descent. A full 94 percent of the Spanish origin population, 68,985 individuals, chose the Mexican descent response in completing the census questionnaire. (U.S. Dept. of Commerce, 1980 Census of Population, vol. 1, Characteristics of Population, ch. B, General Population Characteristics, pt. 6, Cal. table 51 (July 1984).)19 These statistics further indicated that the defense did identify a cognizable class in making the Wheeler motion.

In addition to identifying a cognizable group, Trevino satisfied the final prong of the Wheeler test, that of showing “a strong likelihood” that *688peremptories are being used to exclude the group in question. (Wheeler, supra, 22 Cal.3d at p. 280.)20 Trevino and Rivas are Hispanic. The victim of the crime was Caucasian, as were most of the members of the jury. The district attorney exercised peremptory challenges to remove six Hispanics from the jury after asking them few if any questions on voir dire. In making the Wheeler motion, the defense described these excluded jurors as “male and female, young, middle-aged and older, [of] differing backgrounds, from different areas of the county.” (Fn. 10, ante.) The prosecution left a solitary Spanish surnamed juror on the panel as an alternate, but only after the defense advised the court that it intended to make the Wheeler motion. These factors, taken together, create the requisite “strong likelihood” that the prosecution was abusing the peremptory challenge by eliminating Hispanics, as a group, from jury service.

Once the defense made out a prima facie case, the burden shifted to the prosecution to rebut the inference of impermissible group bias. In determining whether the prosecution has satisfied this burden, the court must not consider the proffered justifications in terms of the district attorney’s “ ‘judgment’ or ‘sincerity,’ ” but must determine whether the “ground of challenge was a specific bias on the part of the individual juror.” (Wheeler, supra, 22 Cal.3d at p. 284, fn. 32.)21 It appears that the trial court failed to adhere to this principle in ruling on the motion.

After calling upon the prosecutor to justify his use of peremptories against the six Spanish surnamed jurors and listening to the reasons offered, the trial court denied the Wheeler motion. In so ruling, the court stated: “Mr. Faller has sustained his burden of justification that the use of the peremptory *689challenges against these particular named jurors was for specific reasons which he explained on the record; [n]ot that they were necessarily members of a certain ethnic group, but because of their particular age or sex, he excused them for those reasons and other specific reasons such as with one juror he mentioned the fact that he had a relative convicted of [a] similar offense or a crime. And these explanations seem to fall within the reasonable use of the peremptory challenges for bias or implied bias that counsel might want to use. For that reason the motion would be denied.”

It appears that the trial court was confusing “specific reason” with “specific bias.” In Wheeler, we defined “specific bias” as “a bias relating to the particular case on trial or the parties or witnesses thereto.” (Id., 22 Cal.3d at p. 276.) A review of the record demonstrates that the prosecutor did not, in fact, satisfy his burden of showing that he excluded the Spanish surnamed jurors on the grounds of specific, not group, bias.22 *690Our conclusion in Hall, supra, 35 Cal.3d at p. 168, that disparate treatment of the members of the excluded group and the unchallenged jurors is indicative of group bias, guides our analysis.

*691The district attorney relied on age as the most compelling factor underlying his use of peremptories against this group. He voiced his concern that these particular jurors lacked maturity and that they might be close to the same age as the defendants. However, the district attorney made no effort to determine the jurors’ ages on voir dire nor did he inquire into any age-based biases the jurors might harbor. Justina Gonzalez, to whom the prosecutor objected as “quite young” had two children, ages eight and six. Gloria Longoria who was “somewhat younger than ... I would have ideally preferred” had children ages 14, 10 and 7. Mary Silvas’ oldest child was 18. It is doubtful that these jurors were the same age as Trevino who was 22 at the time of trial.23 Moreover, the prosecutor did not challenge Richard Mifko who was about the same age as the defendants. His asserted explanation for this disparity, that Mifko was a professional, is questionable. Mifko was an unemployed accountant. Taken as a whole, the district attorney’s age justifications do not rise to the level of specific bias required by Wheeler.

The prosecution also found it troublesome that Silvas’ eighteen-year-old child was close in age to at least one defendant. Although the district attorney feared that this might influence Silvas to identify with the defendants, he apparently had no similar concerns regarding juror Bernor who had a. 16-year-old son, juror Hobbs who had a 19-year-old son or juror Miller whose son was either 20 or 22 years old.24

The district attorney’s inconsistent approach is further exemplified by his admission that the fact that jurors Longoria and Silvas were female influenced his decision to excuse them. Not only is it difficult to imagine how these jurors’ sex in any way translates into specific bias in the case, it is even harder to do so given the fact that five women sat on the final jury panel.

Likewise, the district attorney indicated he was bothered by the fact that several of the Spanish surnamed jurors had no strong opinion on the death *692penalty, but he did not seem to mind that most of the members of the group ultimately empaneled were similarly noncommittal. During the Witherspoon examination, juror Burford expressed ambivalence about capital punishment and juror Foster was “perplexed.” Although juror Hicks had done some reading on the subject, he had yet to develop an opinion pro or con, and juror Bernor not only had not read anything on the death penalty, but also “really hadn’t given that much thought in it” and had no opinion. In all, seven of the empaneled jurors had no particular opinion about the death penalty.

Finally, the district attorney’s reason for excluding Robert Guerrero based on his body language and mode of answering certain questions, is particularly untenable in light of Wheeler’’ s requirement of a showing of specific bias.25 In a similar vein, the district attorney condemns Longoria for expressing herself “rather haltingly,” then takes the opposite approach and criticizes Gonzalez’ perceived verbosity. The internal inconsistencies in the district attorney’s reasoning, coupled with the obvious lack of uniformity in application of this reasoning, lead to the conclusion that, with the exception of Alex Fació, and possibly Catalina Martinez, the Spanish surnamed jurors were excluded for reasons other than those presented to the court.

In Wheeler, this court noted that the prosecution might sustain its burden by a showing that it peremptorily challenged similarly situated members of the majority group on identical or comparable grounds. (Id., 22 Cal.3d at p. 282.) Such a comparison must be presented to the trial court at the time the prosecution attempts to justify the challenges made to members of the cognizable class.26 This the district attorney failed to do. Moreover, the reasons given for excluding the cognizable class are scarcely more believable than those we rejected in Hall, where we emphasized the trial court’s substantial responsibility to seriously consider the prosecutor’s justifications in light of the defendant’s constitutional right to a jury drawn from a representative cross-section of the community. (Id., 35 Cal.3d at pp. 167-169.)

As the Court of Appeal correctly noted, “The reasons given for excluding Gloria Longoria, Justina Gonzalez, Mary Silvas, Catalina Martinez and Robert Guerrero are suspect to say the least. If such reasons suffice to show that the prosecutor did not predicate his peremptory challenges to the Mex*693ican-American jurors on the basis of group bias, then the cross-representative rule could be evaded easily in every case.”

Our Constitution and system of justice cannot and should not tolerate such a result. “It demeans the Constitution to declare a fundamental personal right under that charter and at the same time make it virtually impossible for an aggrieved citizen to exercise that right.” (Wheeler, supra, 22 Cal.3d at p. 287.)

Trevino stated a prima facie case of systematic exclusion of a cognizable group from the petit jury, and the prosecution failed to sustain its burden of showing that the jurors were not peremptorily excluded as a result of group bias. The conviction must, therefore, be set aside because of a denial of Trevino’s right to an impartial jury under article I, section 16 of the California Constitution. As a result, we need not reach Trevino’s remaining contentions.27

*694II

We now consider whether requiring Rivas to stand trial again would constitute double jeopardy after the trial court, in granting his motion for a new trial, stated that the evidence was “insufficient as a matter of law.” We conclude that it would, but not because of error committed in granting that motion. Instead, the trial court erred earlier in not reaching the identical result when it considered Rivas’ motions for judgment of acquittal (§ 1118.1). As Rivas was entitled to a judgment of acquittal based on insufficiency of the evidence to support the conviction, the double jeopardy clause bars his retrial.

In granting Rivas’ motion for a new trial pursuant to section 1181, subdivision 628 the trial court prefaced its statement by defining what it considered to be the appropriate standard of review: “A motion for a new trial made by a defendant on the basis that substantial evidence does not support his conviction requires a review of the entire record in the light most favorable to the judgment to determine whether it discloses substantial evidence such that a trier of fact could find the defendant guilty beyond a reasonable doubt. This is the test applied by the reviewing courts, and by this trial court, in considering this motion.” (Italics added.) The court then reviewed the evidence and concluded: “. . . [t]his court cannot view Ms. Nyberg’s identification testimony as other than insubstantial and not sufficient to identify Rivas as being at the scene of the crime. [H] Other than the testimony of Ms. Nyberg the evidence linking Rivas to the slaying is negligible .... [|] Viewing the evidence in a light most favorable to the prosecution, leaves this court with the conclusion that the evidence as to the defendant Rivas is insufficient as a matter of law.”

jn fact, trial courts do not employ the same standard as reviewing courts do in weighing the sufficiency of the evidence on a section 1181 motion for a new trial. (People v. Serrato (1973) 9 Cal.3d 753 [109 Cal.Rptr. 65, 512 P.2d 289].)29 This is not the case, how*695ever, with a motion for judgment of acquittal under section 1118.1.30 In considering a section 1118.1 motion, the trial court is duty bound to apply the same test as does an appellate court in reviewing the sufficiency of the evidence to support a judgment of conviction. (People v. Blair (1979) 25 Cal.3d 640, 666 [159 Cal.Rptr. 818, 602 P.2d 738].)

The standard of review applicable to a section 1118.1 motion is essentially the one the trial court erroneously used in ruling on Rivas’ motion for a new trial. The court must determine “ ‘whether from the evidence, including reasonable inferences to be drawn therefrom, there is any substantial evidence of the existence of each element of the offense charged. [Citations.]”’ (People v. Lines (1975) 13 Cal.3d 500, 505 [119 Cal.Rptr. 225, 531 P.2d 793], quoting People v. Valerio (1970) 13 Cal.App.3d 912, 919 [92 Cal.Rptr. 82].) “[T]he court must review the whole record in the light most favorable to the judgment below to determine whether it discloses substantial evidence—that is, evidence which is reasonable, credible and of solid value—such that a reasonable trier of fact could find the defendant guilty beyond a reasonable doubt.” (People v. Johnson (1980) 26 Cal.3d 557, 578 [162 Cal.Rptr. 431, 606 P.2d 738, 16 A.L.R.4th 1255].)

“[T]his inquiry does not require a [reviewing] court to ‘ask itself whether it believes that the evidence at the trial established guilt beyond a reasonable doubt. ’ [Citation.] Instead, the relevant question 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.” (Original italics.) (Jackson v. Virginia (1979) 443 U.S. 307, 318-319 [61 L.Ed.2d 560, 573, 99 S.Ct. 2781].) Where the section 1118.1 motion is made at the close of the prosecution’s case-in-chief, the sufficiency of the evidence is tested as it stood at that point. (People v. Belton (1979) 23 Cal.3d 516, 521 [153 Cal.Rptr. 195, 591 P.2d 485]; People v. Camp (1980) 104 Cal.App.3d 244, 247 [163 Cal.Rptr. 510].)

In the present case, Rivas moved for acquittal under section 1118.1 at the close of the prosecution’s case-in-chief, and renewed the motion after the defense rested. In each instance, the trial court expressed serious reservations as to the adequacy of the evidence, but chose not to grant the motion. *696The inherent contradiction between these rulings and the court’s later ruling on the motion for a new trial is manifest. It strains logic to rule during trial that the evidence is sufficient to preclude acquittal, and rule after trial that the evidence is insufficient as a matter of law to support the conviction, all the while applying the same standard of review.

These conflicting rulings require us to make an initial determination as to the sufficiency of the evidence presented on the murder charge. If the evidence was insufficient as a matter of law, then the trial court erred in not granting the motion for acquittal and Rivas’ subsequent motion to dismiss. If, on the other hand, substantial evidence existed to support the conviction on appeal, then all of the trial court’s rulings would be correct, and the reasons given for granting the new trial could be safely ignored.

A review of the evidence suggests that it is insufficient as a matter of law. Because the trial court, in ruling on the motion for a new trial, accurately summarized the state of the evidence at the time the prosecution rested its case, we excerpt that discussion:31 “[T]he People rely heavily on the testimony of Mrs. Nyberg to show the presence of Rivas at the apartment at or about the time of Ted Hinton’s death. However, Mrs. Nyberg was never able to positively identify Rivas as the second man at the apartment. She described the second man as tall and slender; that he walked very erect with his head up. She was not positive that it was Rivas but she saw a similarity in the way Rivas carries himself and that the second man had no moustache. In contrast to this identification testimony, Rivas was not, at that time, slender and he did have a moustache [elicited on cross-examination of Sergeant Wittman regarding a photograph of Rivas taken by the police on the day following the alleged murder]. Additionally, Mrs. Nyberg, despite her recollection of viewing several photo lineups of Rivas, [actually] only looked at one set of photos which included Rivas’ picture, and she identified a Pat Juarez as the second person she saw at the apartment. Moreover, evidence was adduced indicating that Mrs. Nyberg attended a live lineup in which Rivas was present, but no further testimony was offered as to any identification she may have made at the live lineup. Further contradictions appear in Mrs. Nyberg’s testimony as to the posture and manner of walking of the person alleged to be Rivas [Wittman testified that Nyberg told him that the man ‘walked like he was intoxicated’]. . . . [t] The fingerprint evidence cannot be considered substantially incriminating. The fact that Rivas’ thumbprint was found on one of the unopened dresser drawers [the fourth drawer down of a five drawer chest-high dresser] in the victim’s bedroom is susceptible of various interpretations .... The expert witness could not determine the age of the print, hence there is no reason *697to presume the print was made on the day of the murder as opposed to some previous occasion. [Rivas had been a guest in Hinton’s home.] The evidence as to how or when the print came to be placed on the dresser is fraught with uncertainty, leaving the triers of fact to speculate as to how and when the print was made. This kind of guesswork as to the facts does not elevate speculation to the level of reasonable inference. [Citation.] [f] The testimony that Rivas was seen with Trevino before the crime [sometime after 6 a.m. Rivas, Trevino and Trevino’s sister visited Trevino’s girlfriend in the hospital] . . . does ‘no more than generate suspicion, which will not support a conviction.’ [Citations.]”

The only prosecution evidence that the trial court failed to chronicle in the above discussion is the testimony of the doctor who performed the autopsy estimating the time of death as a full day before both the defense and the prosecution agree the crime was committed. This, then, is the sum total of the evidence presented by the prosecution in its case in chief.

In denying Rivas’ motion for acquittal at this point in the proceedings, the trial court opined: “I think it’s enough [evidence] . . . certainly to put Trevino there. As far as substantial evidence, it’s more difficult to make that assessment as far as Rivas is concerned because the testimony regarding identification of him is based primarily on the feeling as far as concepts he has about the way he walks and holds himself. I suppose that’s sufficient identification testimony to put before the jury, albeit, pretty weak. So far as putting the defendants at the scene, I think the direct evidence plus some circumstantial evidence gives rise to circumstantial evidence from which inferences can be derived that I could say could be reasonable—not totally unreasonable or totally speculative or guesswork about the crime itself of Murder . . . .”

The evidence, or lack thereof, indicates that this ruling was in error with regard to Rivas. The highly speculative and equivocal identification testimony and the solitary fingerprint of some unknown vintage do not constitute evidence which is “reasonable, credible and of solid value—such that a reasonable trier of fact could find the defendant guilty beyond a reasonable doubt.” (Johnson, supra, 26 Cal.3d at p. 578.) Because the evidence was insufficient as a matter of law, Rivas was entitled to a judgment of acquittal when he made his initial section 1118.1 motion.

Likewise, the trial court’s denial of the renewed motion for acquittal after the defense rested was in error. This motion generated a good deal of discussion between the bench and counsel, centered around the credibility of the witnesses with regard to the actual time of death. Countering defense counsel’s repeated assertion that the final defense witness’ testimony con*698clusively undermined Nyberg’s testimony that the victim was killed at 3 p.m., the court simply emphasized and re-emphasized that the argument boiled down to a question of which witness the jury would believe. The court concluded that the credibility of the witnesses was a question for the trier of fact and denied the motion.

What the court overlooked in ruling on this motion was that the dispute as to the time of death was of little consequence to Rivas’ right to acquittal. Viewing the evidence in the light most favorable to the prosecution, as the court must do on a motion for acquittal, there simply was no substantial evidence putting Rivas at the scene at the time Nyberg asserted the crime occurred.

The court actually indicated as much when it observed, during this discussion that, “[t]he evidence is . . . inconsistent. In my opinion it raises . . . reasonable doubt. There is some question in my mind whether the evidence is substantial. I think that term ‘substantial’ bothers me.”

This concern over the question of substantiality evaporated when the court, in evaluating the motion for a new trial, emphasized: “Even if the court assumes that the jury discounted entirely the testimony of [the defense witnesses] as to the time they last observed the victim alive, this court cannot view Mrs. Nyberg’s identification testimony as other than insubstantial and not sufficient to identify Rivas as being at the scene of the crime. Other than the testimony of Mrs. Nyberg, the evidence linking Rivas to the slaying is negligible.” The court was compelled to conclude that “the evidence as to the defendant Rivas is insufficient as a matter of law.”

Considering the evidence, the trial court’s rulings on the motions for acquittal and the reasoning it offered to justify the grant of a new trial, it is difficult to avoid the conclusion that the trial court may have been holding off on dismissing the charges against Rivas to allow the prosecution time to somehow bolster its case. Indeed, the trial court’s ruling on Rivas’ motion to dismiss (§ 1385),32 issued the same day as its ruling on the motion for a new trial, supports this conclusion. The court denied the motion to dismiss the murder charge as to Rivas for three reasons: “the evidence indicative of guilt, the nature of the crime involved, and the likelihood of additional evidence being presented upon a retrial.” As we have seen, the trial court clearly did not believe the first *699reason, nor is it supported by the evidence. As the second reason is necessarily tied to the first, it too must fall. This leaves the third reason which, in this case, would represent a direct violation of the double jeopardy clause. (See People v. Pierce (1979) 24 Cal.3d 199, 209-210 [155 Cal.Rptr. 657, 595 P.2d 91].) As the United States Supreme Court observed in Burks v. United States (1978) 437 U.S. 1 [57 L.Ed.2d 1, 98 S.Ct. 2141], “the purposes of the Clause would be negated were we to afford the government the opportunity for the proverbial ‘second bite at the apple.’ ” (Id., at p. 17 [57 L.Ed.2d at p. 13].)

Rivas’ situation, in this respect, is indistinguishable from that of the defendant in Hudson v. Louisiana (1981) 450 U.S. 40 [67 L.Ed.2d 30, 101 S.Ct. 970], In Hudson, the high court held that Louisiana violated the double jeopardy clause by retrying a defendant for first degree murder after the trial court granted a new trial because the evidence was insufficient as a matter of law. The court also noted however, that Louisiana law does not allow trial courts to enter judgments of acquittal in jury trials, making the motion for a new trial a defendant’s sole means of challenging the sufficiency of the evidence against him. (Id., at p. 41, fn. 1 [61 L.Ed.2d at p. 32].)

Although California law ordinarily provides both methods of challenge, it is virtually indistinguishable from the Louisiana system when the trial court precludes relief under the first of these methods by failing to seriously consider the evidence in light of the proper standard. This court cannot condone the evasion of a defendant’s former jeopardy rights where, as here, the defendant is entitled to, but denied a judgment of acquittal. Given the facts of this case, the Hudson rule bars Rivas’ retrial on the murder charge.

The conclusion of the second trial court that Rivas has been once in jeopardy on these charges (§ 1016, subd. 5), and is therefore entitled to be discharged, should be affirmed.

The judgment as to defendant Trevino is reversed and his cause is remanded for further proceedings consistent with the views expressed herein. The judgment as to defendant Rivas is affirmed; retrial is barred by the double jeopardy clause.

Bird, C. J., Mosk, J., Broussard, J., and Grodin, J., concurred.

I concur in the judgment dismissing the charges against defendant Rivas, based on the insubstantiality of the evidence against him.

I dissent, however, to the reversal of the judgment as to defendant Trevino. For the reasons aptly stated by Justice Kaus in his concurring and dissenting opinion, the trial court properly ruled that any inference of group bias was rebutted by the prosecutor. In addition, and unlike Justice Kaus, I believe that People v. Wheeler (1978) 22 Cal.3d 258 [148 Cal.Rptr. 890, 583 P.2d 748], was incorrectly decided and should be reexamined. (See id., at p. 288 et seq. [dis. opn. by Richardson, J.].)