Opinion

This case arises under the 1978 death penalty initiative, codified as Penal Code sections 190 through 190.5 (further statutory references, unless otherwise indicated, are to this code). Defendant was convicted of first degree murder, rape, and sodomy, accompanied by two special circumstance findings, namely, murder committed while engaged in the commission of rape or the attempted commission of rape (§ 190.2, subd. (a)(17)(iii)), and murder committed while engaged in sodomy or attempted commission of sodomy (id., subd. (a)(17)(iv)). He was also found to have used a deadly weapon. The jury fixed the punishment at death. The appeal from the death penalty is automatic (§ 1239, subd. (b)).

We conclude that the judgment should be affirmed in its entirety.

I

Facts

A. Guilt Phase Evidence

Shirley Hill left her home in San Francisco on September 5, 1979, to attend a real estate course in neighboring Daly City. She was driven to school by her former husband, with whom she lived although they were divorced. They discussed how Hill would get home by bus, and the bus route. Hill was seen in classes from noon to 3 or 3:15 p.m., and was last seen at 3:30 at the Westlake Shopping Center in Daly City.

Hill’s body was found the afternoon of September 6, 1979, in a bungalow adjoining the Mission High School football field. The field was several miles from the Westlake Shopping Center, but only a few blocks from a bus transfer stop which Hill would have used on her way home from Daly City. The floor of the bungalow was dusty everywhere except around the body, and the desk chairs in the bungalow were in disarray. Hill’s personal belongings, including her underpants, a real estate binder and a coin purse containing some money, were scattered around her. She had died of ligature strangulation, and her slacks were knotted tightly around her neck.

*757The medical examination revealed numerous abrasions recently inflicted and consistent with the victim thrashing about. Death had occurred between 5 p.m. and 9 p.m. on September 5. Tests found sperm four to five inches inside the anal cavity and a similar amount of sperm five inches inside the vagina. No sperm was found on the pubic hair, perineum or thighs, and there were no semen stains on the victim’s underpants, as would be likely if she had dressed after intercourse. The cavity tests and the lack of leakage or staining on the body’s surfaces led the examining doctor in the coroner’s office to conclude that two separate ejaculations had occurred, each shortly before death.

Other evidence recovered in the bungalow included a palm print found on a windowsill and a thumbprint and smudged fingerprints found on the back of a chair. Hill’s fingerprints were found on the windowsill near the palm print and on either side of it. The position of her fingerprints indicated that when they were made Hill was on the floor probably facing the window. The position of the palm print also indicated that the other person was facing the window. The thumbprint and smudged fingerprints on the chair were visible as brown stains, as though the fingers had been sticky with some substance. It was impossible to date the fingerprints, thumbprint and palm print, which could have been placed at different times.

In January 1980, police officers matched defendant Russell Coleman’s thumbprint to the one found on the chair in the bungalow. The palm print from the windowsill also matched defendant’s. The fingerprints on the chair could not be identified, because they were smudged.

After police made the print identification, they questioned defendant who had been arrested on unrelated charges. After being given Miranda admonitions (Miranda v. Arizona (1966) 384 U.S. 436 [16 L.Ed.2d 694, 86 S.Ct. 1602, 10 A.L.R.3d 974]), defendant agreed to talk to a police officer. Defendant said he did not know Shirley Hill. He also stated that he knew where Mission High School was but had never been on the football field there or in the bungalows adjacent to the field. He denied that the thumbprint found in the bungalow was his. When told that he was thought to be responsible for Shirley Hill’s death, and would be booked, he terminated the interview.

Defendant presented an alibi defense. He testified that on September 5, 1979, he was in class at City College of San Francisco until 3:30 p.m., had left school in the company of a friend, Carlton McAllister, and had gone home. Defendant’s wife and children were home when he arrived. Defendant, his wife and McAllister visited until McAllister left at approximately 5 or 6 p.m. Defendant stayed home all night and did not leave until the next day. He had a clear memory of September 5, because it was the first day of *758classes at City College. Contrary to his statement to the officers, defendant explained the presence of his thumb and palm prints by stating that he had been in the bungalow on Labor Day, September 4, 1979. He had jogged on the football field, as was his habit, and had ducked into the bungalow for awhile to smoke a “joint.” He claimed he had lied in the initial interview with the police because he was “scared, paranoid, and nervous.” Testimony also revealed that defendant felt his memory for detail of the period was impaired because, after his arrest in January 1980, he had suffered a mental breakdown, was confined in San Francisco General Hospital, and given Haldol, a powerful drug which adversely affected his mental state.

McAllister, a pastor and bible college student, was also taking courses at City College in September 1979. McAllister testified that on September 5, he had attended a math course taught by Mr. Lindsey, and about 2 or 3 p.m., had encountered defendant in Cloud Hall. The men went by bus to defendant’s house, a trip which took between 30 minutes and one hour. McAllister stated that he had stayed for a couple of hours, and that defendant was there the whole time and was still there when McAllister left. On cross-examination, McAllister stated that to the best of his recollection, he had met defendant on September 5, but it could have been during the first week of classes.

Defendant’s former wife began her testimony by stating that she was divorced from him, did not have good feelings about him, and did not want to see him again. Nevertheless, she corroborated defendant’s and McAllister’s testimony as to their whereabouts the afternoon and evening of September 5. She also said that defendant stayed in the house all that evening and did not leave until 7 a.m. on September 6. She had not told the police or the district attorney’s investigator of the alibi earlier, even though it might have absolved her husband, because she did not think it “her place to volunteer information.”

The dean of admissions and records at City College brought copies of defendant’s and McAllister’s class schedules for fall 1979. On September 5, both McAllister and defendant were registered for a class taught by Lindsey, in Batmale Hall. McAllister and defendant both transferred to a different class, taught in Cloud Hall, on September 10 and 11, respectively. Lindsey produced his class attendance records for September 1979, which showed that defendant had been absent from class on September 5, 6 and 7. The absence on September 5 was marked excused, but not the others. Lindsey did not know when he had marked the September 5 absence excused, and conceded that it was possible that defendant came to him with an excuse after class on September 5. McAllister was marked present in Lindsey’s class on the 5th. Defendant retook the stand after Lindsey’s testimony, and explained that he had skipped class because of a class *759conflict. He remembered seeing Lindsey on the 5th and telling him he was going to drop the class. Defendant could not remember the conflicting class.

B. The Medical Experts’ Testimony

Dr. Edward Blake, a forensic serologist, conducted tests on semen fouñd in Hill’s vagina, as well as on her blood. Prior to any testimony by Dr. Blake before the jury, defense counsel requested a foundational hearing pursuant to Evidence Code section 402. The record does not show the request or whether any objection or offer was made in connection with the request.1 After the section 402 hearing, the only objection defendant made to Dr. Blake’s testimony was that he should have performed a test in addition to those tests that were performed.

At the Evidence Code section 402 hearing, the prosecutor did not question Dr. Blake or otherwise attempt to lay a foundation for his testimony. Dr. Blake’s testimony at the hearing was largely repeated at trial, and the following summary of Dr. Blake’s testimony includes that given at the hearing and before the jury.

Blake has a doctorate in criminology from the University of California at Berkeley. He wrote his doctoral dissertation on the. determination of genetic markers in human semen. He has also published numerous articles on the subject.

Blake received a vaginal wash of Shirley Hill from Dr. Stephens of the coroner’s office. Blake testified that a vaginal wash is generally obtained by inserting a saline solution into the cavity and then withdrawing it into a syringe.2 He also obtained a sample of Hill’s blood, and blood and saliva samples from her former husband and from defendant. Tests performed showed that the semen found in the vaginal wash could not have been her former husband’s, and that defendant was a member of a class consisting of 8 percent of the population that could have deposited the semen.3

Blake explained the two tests he used to classify the bodily substances he had received. The first test used can identify, in some instances, the blood *760type of a person by the antigens found in bodily fluids other than blood. Blake explained that every person has one of four possible blood antigen types—A, B, AB or O. Most people (80 percent of the general population) secrete their antigens into other bodily fluids such as saliva, semen, urine and vaginal secretions. Thus the blood type of secretors can be determined by examination of one of these fluids.

The vaginal wash obtained from Hill’s body, although it contained a high level of semen, did not contain any blood antigens, indicating that the semen donor was one of the 20 percent of the population that does not secrete antigens—a “nonsecretor.”4

Blake also did an analysis of PGM in the semen sample. PGM analysis categorizes an enzyme found in every person’s bodily fluids and secretions; the various types have been found to occur with established frequencies in the population. There are three PGM groups: 1/1, 2/1 and 2/2. PGM types 2/1 and 2/2, together, occur in 40 percent of the population. Decedent’s blood was type 1/1; the vaginal wash contained type 2, which, then, must have come from the semen donor. Defendant proved to be a type O nonsecretor (so that his blood antigens would not have been found in the wash), PGM type 2/1.

According to Blake, because only 20 percent of the general population are nonsecretors and 40 percent of the general population are PGM type 2/1 or 2/2, statistically only 8 percent of the general population could have donated the semen found in Hill’s body. Blake testified that these are established statistical frequencies, not projected possibilities. A frequency computation is valid only where the variables (secretor-nonsecretor and PGM type 1/1, 2/1 or 2/2) are independent, as these are. In other words, there is (and must be) no preexisting relationship between antigen secretors and any particular PGM type. Studies have shown that the frequency of PGM 2 is the same among secretors and nonsecretors. Further, for the frequency *761calculation to be valid, the characteristic must be stable at all times in the subject. Here, both tests looked for genetic traits or markers which do not change, but remain constant through life.

Defendant presented testimony directed to the medical evidence other than the tests performed by Dr. Blake. A defense investigator stated that he had talked with Sisson, a doctor in the coroner’s office, who had stated that the medical evidence was not inconsistent with consensual sexual intercourse, and that the presence of sperm in the victim’s rectal cavity could be present for reasons other than sodomy. Sisson was also reported as having been surprised that a Peptidase A test was not done on the vaginal wash.

The defense also called Dr. Jindrich, the Marin County Coroner and a forensic pathologist, who testified that sperm in the rectal area was not necessarily indicative of sodomy, in that it could have been caused by leakage from the vagina after intercourse or as the result of the movement of the body by the coroner’s office. Also, because there was no damage to the vagina, perineum or rectum, it was not clear that the intercourse was not consensual; the other abrasions on the body could have been caused by Hill’s struggle with her killer, or by rough handling of the body by the coroner’s employees. Jindrich did agree that because the semen deposits were both four or five inches inside two different cavities, and there was no staining of the clothing or leakage onto other areas of the body, it was likely that there had been two separate deposits of semen, rather than leakage from one area to another.

As previously indicated, the jury found defendant guilty of first degree murder with special circumstances, and other lesser offenses.

C. Penalty Phase Evidence

The prosecution introduced as aggravating circumstances evidence of other violent behavior by defendant. The most serious offenses of which defendant was convicted were the false imprisonment of a female college student, carrying a concealed weapon, the rape and oral copulation of a 13-year-old girl, and lewd conduct involving an 11-year-old girl. Defendant had been convicted of each of these crimes prior to the murder trial. Evidence was also presented of the seizure of a stabbing weapon from defendant’s cell and of a series of assaults committed by defendant while in prison, on other inmates, guards and medical personnel.5 A deputy testified that *762defendant was transferred to the Adjustment Center at San Quentin because his violent behavior made him a hazard to himself, other inmates and staff. The deputy was not aware of any other inmate who had been transferred on his own request.

The defense presented evidence in mitigation consisting of psychiatric testimony, and the testimony of defendant’s mother and of a clergyman.

Dr. Eugene Studenski, a parole outpatient clinic psychiatrist, interviewed defendant for one hour in December 1979. Studenski related that defendant had reported a number of emotional problems, and was sad and depressed as he related his family history, which was quite violent. At one point, defendant noted that while in prison he had successfully requested a transfer to the San Quentin Adjustment Center to avoid another inmate whose friend had killed defendant’s sister. The center is a restricted, single-cell unit for the most difficult or dangerous prisoners. In Studenski’s opinion, most prisoners at the center undergo some emotional damage over time because of the isolation, hostility among prisoners, and lack of social stimulation. Studenski said defendant might have been paranoid before he entered the center, or his stay there may have made him so. On cross-examination, Studenski further testified that his written psychiatric report described defendant as having conscious control of himself and not psychotic. Studenski’s diagnosis of defendant was that he suffered from depression with possible schizophrenia.

Defendant’s mother, a nurse, testified to the difficult family environment in which defendant was raised. The father of the family was a violent, jealous man who regularly beat and abused defendant’s mother, often in the presence of the children. The police had to be called frequently. At the age of eight, defendant saw his father, whom he loved very much, stab and kill a neighbor. Defendant was forced to testify at the murder trial, and he blamed his mother for not protecting him from the ordeal. Defendant never saw his father after the trial. The father was an alcoholic and defendant’s mother would not tell the children where he was living. Defendant was in and out of juvenile hall from age 10 to 14, and from age 14 to 18 was beyond his mother’s control. At age 18 or 19, he moved in with his sister, with whom he was very close. She was later killed. Defendant was arrested for selling drugs, and from age 20 to 27 was never out of jail for more than 3 or 4 months. In May 1979, he was again released, and he joined a church. He practically lived at the church, and went to services every evening. In August 1979, he was married.

The defense also offered the testimony of Charles J. Jones, a Pentecostal minister and prison outreach coordinator connected with Teen Challenge *763Ministry, who in 1979 and 1980 was with Christ Center, a halfway house. Defendant attended the halfway house program voluntarily from May to July 1979. Jones testified defendant believed he had an uncontrollable, violent nature and was obsessed with the homicide committed by his father. Defendant told Jones he had been afraid of his father, who had whipped him and beat his mother. He appeared highly jealous and mistrustful of women, had no real self-control and had to be given a private room at the center. His marriage was unsuccessful because of his jealousy and mistrust of his wife. Jones felt that defendant was making an effort, and always looked for work to do. He painted the entire building and worked in the garden. He was friendly, loving and respectful with the minister’s family.

II

Jury Selection Error

Defendant contends the court erred in failing to exclude for cause four members of the venire, each of whom, he asserts, made it unmistakably clear that he or she would automatically vote for the death penalty.6 In each instance, although defense counsel’s challenge for cause was denied, defendant later used peremptory challenges to excuse all four prospective jurors. Defendant ultimately used 24 of his 26 available peremptory challenges. After the defense had accepted the jury, the prosecutor, who had 11 challenges left, also accepted the jury. The trial court announced that four alternate jurors would be chosen, and pursuant to section 1089,7 each side received an additional four peremptory challenges to the prospective alternate jurors. Defendant used all four of his allotted peremptory challenges to alternate prospective jurors. One of the alternates was seated at the trial.

Defendant asserts that the failure to exclude for cause jurors with an unmistakable bias for the death penalty is reversible Witherspoon error (Witherspoon v. Illinois (1968) 391 U.S. 510 [20 L.Ed.2d 776, 88 S.Ct. 1770]). Witherspoon and its progeny developed the standard for determining when a prospective juror must be excused because he opposes the death penalty. As explained below, we conclude that the ruling on a challenge for cause when a prospective juror appears biased in favor of the death penalty should be examined in light of the same Witherspoon standard. Applying that standard, we find that the court here improperly denied one challenge for cause in this case. However, we find the erroneous denial of a challenge *764for cause to a death-biased juror is not analogous to the wrongful exclusion of a prospective juror. In the former case, the defendant is forced to use one peremptory challenge; in the Witherspoon situation, the defendant is deprived of the jury to which he was entitled. We also find that the erroneous denial of a challenge for cause, with the result that a defendant uses one of his peremptory challenges, is not the type of constitutional error which should automatically result in a per se reversal. As a result, we analyze the error to see if it results in prejudice to the defendant; finding none here, we reject defendant’s Witherspoon argument.

A. The Witherspoon Standard

In assessing defendant’s argument, we must first determine whether the Witherspoon standard, supra, 391 U.S. 510, for assessing bias as to the death penalty is applicable to the facts of this case. In Witherspoon, the Supreme Court considered under what circumstances a prospective juror may be excluded for cause based on his or her views of the death penalty. The court held that procedural due process forbids exclusion for cause of prospective jurors who have mere conscientious objections to the imposition of the death penalty. The court further stated, however, that: “nothing we say today bears upon the power of a State to execute a defendant sentenced to death by a jury from which the only veniremen who were in fact excluded for cause were those who made unmistakably clear (1) that they would automatically vote against the imposition of capital punishment without regard to any evidence that might be developed at the trial of the case before them, or (2) that their attitude toward the death penalty would prevent them from making an impartial decision as to the defendant’s guilt. ” (Witherspoon v. Illinois, supra, 391 U.S. at pp. 522-523, fn. 21 [20 L.Ed.2d at p. 785], italics in original.)

The proper standard for exclusion was restated in Wainwright v. Witt (1985) 469 U.S. 412 [83 L.Ed.2d 841, 105 S.Ct. 844] as follows: “We therefore take this opportunity to clarify our decision in Witherspoon, . . . as [to] the proper standard for determining when a prospective juror may be excluded for cause because of his or her views on capital punishment. That standard is whether the juror’s views would ‘prevent or substantially impair the performance of his duties as a juror in accordance with his instructions and his oath, . . . [T]his standard . . . does not require that a juror’s bias be proved with ‘unmistakable clarity.’ ” (Id., at p. 424 [83 L.Ed.2d at pp. 851-852].)

We adopted the Witt standard in People v. Ghent (1987) 43 Cal.3d 739, 767 [239 Cal.Rptr. 82, 739 P.2d 1250], noting that “California courts have generally followed the teachings of the high court in determining when a prospective juror properly may be excused for cause because of his views *765regarding capital punishment. . . .” (See also People v. Guzman (1988) 45 Cal.3d 915 [248 Cal.Rptr. 467, 755 P.2d 917], applying the Witt standard in reviewing a claim of Witherspoon-Witt error.) However, although our opinions in Ghent and Guzman discussed the Witt standard where the issue was whether a juror may be properly excluded for his views on capital punishment, neither they nor any other opinion of this court has applied them to a situation where the purported error is not wrongful exclusion, but wrongful inclusion of a prospective juror in the face of a “for cause” challenge.

Although neither Witherspoon, supra, 391 U.S. 510, nor Witt, supra, 469 U.S. 412, on its face concerns exclusion of a prospective juror for cause due to his or her view favoring the death penalty, we think Witt makes clear that a challenge on the basis of bias meeting the Witherspoon standard is no different from any other challenge for cause, where the trial court is asked to determine whether the juror lacks impartiality on an issue relevant to the case. (See Witt, supra, 469 U.S. at p. 423 [83 L.Ed.2d at p. 851].) When the adversary seeking exclusion is the People, and the basis for exclusion is an inability to conscientiously consider all of the sentencing alternatives, Witt offers particular guidance as to when the People have shown that partiality. We conclude that the same standard of partiality would also have to be shown when the defendant asks the state to exclude a prospective juror for cause, based on a view of the death penalty. A defendant seeking to exclude a prospective juror for cause, based on the person’s views of the death penalty, must therefore demonstrate that those views “would ‘prevent or substantially impair the performance of his duties as a juror in accordance with his instructions and his oath’ . . . .” (Witt, supra, 469 U.S. at p.424 [83 L.Ed.2d at pp. 851-852]; see also Ross v. Oklahoma (1988) 487 U.S. _, _ [101 L.Ed.2d 80, 88, 108 S.Ct. 2273] [applying Witt in such a case].)

B. Application of the Witherspoon-Witt Standard

The voir dire relating to the four jurors in question can be summarized as follows. The first prospective juror, Ms. Ng, initially indicated that she had a conscientious objection to the death penalty and would therefore vote for a verdict of less than murder in the first degree, to stop the trial from going further. Her answers to further questions from the court indicated no bias as to penalty. In examination by defense counsel, Ng stated that she would vote for the death penalty if it were shown that a defendant “deliberately” as opposed to “emotionally” acted in raping and murdering someone. In response to questions from the prosecutor, Ng then said she would listen to all evidence, and would be guided by the court in determining the appropriate penalty. Under questioning by the defense, she again seemed to take the position that she would automatically vote for the death penalty where someone intentionally murdered another. Defense counsel challenged Ng *766for cause, “based on the fact that she is an automatic-death-penalty person in this situation.” The court denied the challenge, stating that it was far from clear that Ng would automatically vote for the death penalty.

Ms. Shorter was another prospective juror giving widely conflicting answers to questions exploring her ability to vote for the death penalty. Shorter first indicated she would not automatically vote for either penalty. She then indicated that for certain crimes, including premeditated murder, she felt that the appropriate penalty was death. Her answers changed again, however, and she stated that she had some hesitancy about the death penalty even where a murder was involved. Finally, she stated that she would vote for the death penalty for someone convicted of child molestation.8 The court denied a challenge for cause.

Mr. Low’s answers to the court’s initial voir dire indicated no prejudice as to the appropriate penalty. His answers to further questioning indicated that he would vote for death if premeditated murder was involved; then his responses indicated that even in that circumstance it would depend on the case; his responses to further questioning made it clear, however, that so long as the circumstances indicated that the murder was premeditated, he would automatically vote for the death penalty. A challenge for cause was denied.

The last of the four jurors challenged was Ms. Hastings, who was called as a prospective alternate juror.9 Her first responses indicated she would not favor one penalty over another. She then stated she would “most likely” vote for death if a defendant was guilty of rape, sodomy and murder, but that her decision would depend on the evidence. Further questioning elicited the response that in such a case, she would probably vote for death. She could not think of any mitigating factors which would convince her not to vote for the death penalty in such a case.

In People v. Ghent, supra, 43 Cal.3d at page 768, we held that: “[W]here equivocal or conflicting responses are elicited regarding a prospective juror’s ability to impose the death penalty, the trial court’s determination as to his true state of mind is binding on an appellate court. [Citations.]” We went on in Ghent to find no error where four prospective jurors were *767excused even though they “gave equivocal answers which, if taken in isolation from the remainder of their voir dire examination, might be held insufficient to justify their exclusion under then applicable Witherspoon principles. [Citations.]” (Ibid.) We further noted: “But at some point during the examination of these venirepersons, each of them demonstrated an inflexible inability to impose death.” (Ibid.)

The issue is therefore whether there are equivocal or conflicting responses to the voir dire questioning such that the determination as to the prospective jurors’ state of mind should be left to the trial court. In Ghent, supra, 43 Cal.3d 739, the questioning of the prospective jurors led to equivocal answers; there was also clearly support for the court’s ruling, because at some point the prospective juror had demonstrated an inflexible inability to impose death.10 In the case of prospective jurors Ng, Shorter and Hastings, it is clear that their varying answers depended on how questions were phrased, and in some cases, the same question was answered differently depending on whether it was asked by defense counsel or the prosecutor, The court’s determination as to these prospective jurors’ state of mind is therefore binding on this court and we must conclude the court’s failure to exclude these prospective jurors for cause was not error.

The voir dire as to Low was different, however. Low’s initial responses to the court’s standard questions were innocuous, but those responses were clarified on examination by defense counsel and resulted in Low’s unequivocal statement that he would always vote for the death penalty in a case of premeditated murder. 11 Neither the court nor the prosecutor asked any question of Low at the conclusion of defense counsel’s inquiry. The uncer*768tainty and inconsistency present with the other three prospective jurors was not present with respect to Low. He firmly stated a point of view, he made no conflicting statements, and the prosecutor did not in any way “rehabilitate” his answers. Nor did the court, in denying the challenge for cause, indicate what led to its decision.12 In these circumstances, we cannot accept the court’s denial of the challenge for cause, and conclude the court erroneously denied the challenge for cause.

C. The Erroneous Ruling on the Challenge for Cause

In the instant case the foregoing error resulted in the erroneous, temporary inclusion of a juror. The sole result of the failure to exclude venireperson Low for cause was that defendant had to use one of his remaining peremptory challenges to excuse Low. Even after using this peremptory challenge, however, defendant had two challenges left. After the exercise of 24 peremptory challenges, the defense indicated that it was satisfied with the jury.

Thus, the present case is distinguishable from the situation presented in Gray v. Mississippi, supra, 481 U.S. 648 [95 L.Ed.2d 622, 107 S.Ct. 2045], involving the wrongful exclusion of a prospective juror, where the error was deemed prejudicial per se. In Gray, the prosecutor could only assert in hindsight that he would have used his challenges to exclude a venireman. Here, on the other hand, we know exactly how defendant exercised his peremptory challenges, and we also know that, as the result of that exercise, defendant was not tried by a jury which included a juror to whom he properly had objected. In addition, we know that the erroneous inclusion of a prospective juror was an isolated incident. Most importantly, the error here did not result in a jury particularly apt to impose the death penalty, and there is no indication that the jury before which defendant was tried was anything other than fair and impartial.

We conclude that these facts distinguish the present case from the situation considered in Gray, and that an erroneous ruling on a challenge for cause which results in the inclusion of a prospective juror is subject to a harmless-error analysis. (Chapman v. California (1967) 386 U.S. 18, 23 [17 L.Ed.2d 705, 710, 87 S.Ct. 824, 24 A.L.R.3d 1065].) Here, the erroneous ruling on the challenge for cause resulted merely in the *769temporary inclusion of a veniremen. He was not a regular or alternate member of the jury eventually impaneled, and there is no indication that the bias for which he should have been excluded infected any other member of the venire who later sat on the jury.13

It might be suggested, however, that the erroneous denial of the challenge for cause was not harmless because the effect of that ruling was that defendant was forced to use one of his peremptory challenges to remove the prospective juror. Such an argument was recently rejected by the United States Supreme Court in Ross v. Oklahoma, supra, 487 U.S. __ [101 L.Ed.2d 80, 108 S.Ct. 2273], involving similar facts. In Ross, as in the present case, the defendant was forced to use one of his peremptory challenges to remove a juror who should have been removed for cause. (As here, the juror had indicated he would automatically vote for death if the defendant were found guilty.) The defense ultimately used all of its peremptory challenges. Ross distinguished Gray, supra, on the basis that unlike Gray, “In the instant case there is no need to speculate whether [the excluded juror] would have been removed absent the erroneous ruling by the trial court; [the juror] was in fact removed and did not sit.” (487 U.S. at p. _ [101 L.Ed.2d at p. 90].) The court rejected the argument that the loss of a peremptory challenge constituted a violation of the constitutional right to an impartial jury, stating “So long as the jury that sits is impartial, the fact that the defendant had to use a peremptory challenge to achieve that result does not mean the Sixth Amendment was violated.” (Ibid.)

As for the defendant’s due process rights under the Fourteenth Amendment, the Ross court ruled that although the right to exercise peremptory challenges is an important one, “it is for the State to determine the number of peremptory challenges allowed and to define their purpose and the manner of their exercise. [Citations.] As such, the ‘right’ to peremptory challenges is ‘denied or impaired’ only if the defendant does not receive that which state law provides.” (487 U.S. at p. __ [101 L.Ed.2d at pp. 90-91].) According to Ross, under Oklahoma law a defendant who disagrees with a trial court’s ruling on a challenge for cause must exercise a peremptory challenge to remove the juror, and the error is ground for reversal only if the defendant exhausts his peremptory challenges and an incompetent juror is forced on him. (Ibid.) “As required by Oklahoma law, petitioner exercised one of his peremptory challenges to rectify the trial court’s error, and consequently he retained only eight peremptory challenges to use in his unfettered discretion. But he received all that Oklahoma law allowed him, and therefore his due process challenge fails.” (487 U.S. at p. _ [101 *770L.Ed.2d at pp. 91-92], fn. omitted.) In a footnote, the high court left open the question whether the same result would follow “in the absence of Oklahoma’s limitation on the ‘right’ to exercise peremptory challenges . . . .” (487 U.S. at p. _, fn. 4 [101 L.Ed.2d at p. 92].)

The California courts hold that the defendant must exercise his peremptory challenges to remove prospective jurors who should have been excluded for cause, and that to complain on appeal of the composition of the jury, the defendant must have exhausted those challenges. (Kimbley v. Kaiser Foundation Hospitals (1985) 164 Cal.App.3d 1166, 1169 [211 Cal.Rptr. 148].) As stated in Kimbley, “It has long been the rule in California that exhaustion of peremptory challenges is a ‘condition precedent’ to an appeal based on the composition of the jury. [Citation].” Although our courts have never considered whether an erroneous ruling on a challenge for cause is, in effect, the denial of a peremptory challenge, in other cases involving the denial of peremptory challenges, we have found a potential effect on the selection of the jury only where the defense used all of its available challenges, but remained dissatisfied with the venire.

As we explained in People v. Armendariz (1984) 37 Cal.3d 573, 584 [209 Cal.Rptr. 664, 693 P.2d 243], “California courts have consistently held that ‘the failure to grant a defendant the prescribed number of peremptory challenges when the record reflects his desire to excuse a juror before whom he was tried is reversible error.’ [Citations.]” (Italics added.) (See also People v. Yates (1983) 34 Cal.3d 644, 654 [194 Cal.Rptr. 765, 669 P.2d 1] [reversible error where defendant entitled to 26 challenges and court only permitted 10; after exercising all permitted challenges, counsel stated dissatisfaction with jury and requested additional challenges]; People v. Shaw (1965) 237 Cal.App.2d 606, 611 [47 Cal.Rptr. 96] [reversible error where defendant entitled to 20 peremptory challenges but court only permitted 10; after exercising 10 permitted challenges, defense sought to excuse additional venirepersons and objected to swearing of jury]; People v. Diaz (1951) 105 Cal.App.2d 690 [234 P.2d 300] [reversible error where court erroneously denied defendant two peremptory challenges, defendant used all available challenges, sought to exercise another, and moved for mistrial on the basis he was deprived of his rights with respect to the selection and impaneling of the jury]; cf. People v. Crowe (1973) 8 Cal.3d 815, 831-832 [106 Cal.Rptr. 369, 506 P.2d 193] [no prejudicial error where, after defendant exercised six peremptory challenges, court mistakenly announced defendant had completed his peremptory challenges and swore jury; after recess, court announced its error but counsel declined the court’s invitation to exercise additional challenges].)14

*771Here, the defense indicated its satisfaction with the jury after having exercised only 24 of its 26 peremptory challenges. No juror to whom defendant objected was seated on the jury that heard his case, and he did not object to the jury as constituted, even though he had peremptory challenges remaining.15 In this situation, we cannot find defendant was denied a peremptory challenge in a way which affected his right to a fair and impartial jury. Because there was no possible prejudice to defendant in such circumstances, no reversible error occurred.

III

Guilt Phase Contentions

Defendant argues that errors at the guilt phase of his trial require reversal of the verdict. We examine each of his arguments in turn and conclude that none requires reversal.

A. Motion to Set Aside Information

Defendant’s family retained an attorney who represented him at a preliminary examination in early March 1980. The magistrate found probable cause to hold defendant on all charges and two weeks later, an information was filed. One month after the preliminary examination, retained counsel withdrew, and the public defender was appointed. Subsequently defendant moved to set aside the information on the ground that his retained attorney rendered ineffective assistance at the preliminary examination.

The motion was supported by an affidavit by the retained attorney attesting that he had never tried a homicide case or represented a client in a death *772penalty case, that although defendant exhibited bizarre behavior and was hospitalized by the authorities, counsel did not seek a psychiatric examination, and that he failed to render effective assistance.

At the hearing on the motion to set aside the information, however, retained counsel painted a different picture. Counsel testified that his practice consisted of 25 percent criminal cases, that during 16 years of practice he handled numerous felony cases, including homicides, rapes, and child molestations, and that he had appeared at several hundred preliminary examinations and at about 100 felony trials.16 When hired to represent defendant, counsel considered whether to move for a change in venue, inspected the physical evidence, and discussed the case with police investigators, defendant, and various members of defendant’s family. Counsel did not visit the scene of the crime nor did he employ investigators to interview potential witnesses or consult experts to test the physical evidence. At the preliminary examination, the attorney stipulated to the admissibility and content of a necropsy report prepared by a doctor who was available for testimony at the hearing. Counsel cross-examined the fingerprint expert.

Defendant also presented the testimony of three other criminal attorneys in support of his motion to set aside the information. After reviewing documents related to the case and the transcript of the preliminary examination, each of these witnesses was of the opinion that the retained counsel had rendered ineffective assistance. All agreed that defendant’s case was thereby damaged, but only one witness insisted that a potentially meritorious defense was withdrawn within the meaning of People v. Pope (1979) 23 Cal. 3d 412 [152 Cal.Rptr. 732, 590 P.2d 859, 2 A.L.R.4th 1], and its progeny.17

The court denied the motion to set aside the information. Defendant’s subsequent petition for writ of mandate to compel the requested relief was summarily denied by the Court of Appeal, and this court and the United States Supreme Court each denied petitions to review the matter.

When a defendant is denied a substantial right at a preliminary examination, the commitment is deemed unlawful under section 995, and the information must be set aside. (People v. Pompa-Ortiz (1980) 27 Cal.3d 519, 523 [165 Cal.Rptr. 851, 612 P.2d 941]; Jennings v. Superior Court (1967) 66 Cal.2d 867, 874 [59 Cal.Rptr. 440, 428 P.2d 304].) If a defendant is denied the assistance of counsel at the preliminary hearing, a substantial right has been denied. (Jennings v. Superior Court, supra, 66 Cal.2d at *773p. 874.) The right to counsel includes the right to effective assistance of a reasonably competent attorney during the pretrial stage of a proceeding. (People v. Pope, supra, 23 Cal.3d at p. 423.)

Assuming that retained counsel’s performance at the preliminary examination was inadequate, we find no reversible error. The standard of review for determining whether a substantial right was denied at a preliminary examination was set forth in Pompa-Ortiz, supra, 27 Cal.3d at page 529, in which we explained that if a defendant makes a timely motion in the trial court to set aside the information and shows that a substantial right was denied, the information must be set aside without a showing of prejudice. In a postconviction context, however, when an appellate court reviews the denial of a motion to dismiss under section 995, irregularities in the preliminary examination require reversal only if the defendant can show that the error resulted in an unfair trial or that he was otherwise prejudiced. (Ibid.)

Defendant claims that counsel’s inadequate assistance forced him to trial for his life in an inadequate state of preparation and thus denied him a fair trial. We have previously expressed reluctance to overturn a conviction on the ground that counsel was inadequately prepared for trial in the absence of a specific showing of what favorable evidence might have been obtained. (People v. Jackson (1980) 28 Cal.3d 264, 291-292 [168 Cal.Rptr. 603, 618 P.2d 149].) Defendant attempts to show specific prejudice by pointing to the fact that the prosecution’s case here was based largely on scientific evidence, and urges that it should have been fully explored at or prior to the preliminary examination. The record shows, however, that this evidence received extensive scrutiny at trial, and it does not appear that the stipulation to admit the necropsy report or counsel’s failure to perform a further investigation affected the ability of trial counsel to advocate defendant’s case at trial. Defendant has not therefore met the burden of showing prejudice as a result of his attorney’s pretrial performance.

B. Admission of Testimony Regarding the Hemostick

Defendant claims that the court erred in admitting, over objection,18 testimony by Inspector Ihle, a police expert *774from the San Francisco Crime Laboratory, that as part of his investigation of the murder scene he had tested the stained thumbprint found on the back of the chair with a “hemostick.”19 The hemostick, according to Ihle, was used to perform a presumptive test for blood, changing color in the presence of blood. The test of the thumbprint on the chair resulted in a color change. Ihle’s testimony made clear that although the hemostick indicated the presence of blood, the test could not show whether or not the substance was human blood, or determine blood type. Ihle stated that he had used the test in crime laboratory work for three years and found it reliable after performing the test on substances with an appearance similar to blood, and obtaining negative results. He also indicated that he had previously testified at criminal trials as to the results of hemostick tests, but did not know why blood caused the stick to change color, and had not read any scientific literature on the subject.

Following Ihle’s testimony, Dr. Blake, the forensic serologist who was called to testify on other matters at trial, stated that he was aware of the hemostick test, and explained that the hemoglobin in blood caused the stick to change color. He also stated that the test was a presumptive test for blood.20

Defendant argues that the admission of the testimony regarding the hemostick was error because the prosecutor failed to show that the test was reliable.21 As a general rule, evidence of a scientific test should not be *775admitted unless the scientific basis for the test and its reliability are generally recognized by competent authorities. (People v. Kelly (1976) 17 Cal. 3d 24, 30 [130 Cal.Rptr. 144, 549 P.2d 1240].) Ihle’s testimony was insufficient to permit the introduction of the hemostick evidence in this case. Ihle was not an expert on blood or blood tests, nor did he otherwise have a scientific basis on which to testify as to the reliability of the test. No alternative or supplemental basis for admission of the test was presented, such as prior appellate court decisions allowing use of such a test or extensive scientific literature supporting the test. (People v. Kelly, supra, 17 Cal.3d at pp. 32, 35.) It should be noted that presumptive tests for blood are generally accepted by California courts, and have been used by criminologists and admitted into evidence in this state for over 40 years. (See, e.g., People v. Alcalde (1944) 24 Cal.2d 177, 182 [148 P.2d 627]; see also People v. Burgener (1986) 41 Cal.3d 505, 526-528 [224 Cal.Rptr. 112, 714 P.2d 1251]; People v. Talbot (1966) 64 Cal.2d 691, 708 [51 Cal.Rptr. 417, 414 P.2d 633].) However, because we have no information as to the type of presumptive test the “hemostick” was, we cannot simply assume it is one of the generally accepted presumptive tests for blood. The hemostick test should not, therefore, have been admitted.

Despite the erroneous admission of the hemostick testimony, however, we conclude the error was harmless on the facts of this case. The wrongly admitted evidence did not point to defendant as the perpetrator of the murder, nor was it the most incriminating evidence in the case. The record shows that Ihle, as well as both the prosecution and defense counsel, noted the limitations of Ihle’s testimony. Ihle testified that the test showed the presence of blood, but that it could not and did not indicate whether it was human blood. This point was reemphasized by defense counsel on cross-examination. In addition, the other evidence against defendant was quite substantial. Defendant’s prints were found in the bungalow, he lied about his presence there when he first discussed the matter with police, and his genetic characteristics placed him within a class consisting of only 8 percent of the population that could have committed the rape. His alibi defense was significantly undermined by the other evidence presented by the prosecutor. We conclude that it is not reasonably probable that a result more favorable to defendant would have occurred but for the error. (People v. Watson (1956) 46 Cal.2d 818, 836 [299 P.2d 243].)

C. Admission of Photographs

Defendant next contends the court erred in admitting into evidence, over his objections, photographs of the victim taken at the scene of the murder and at the autopsy. Defendant argues that the court failed to *776weigh the probative value of the photographs against the danger of undue prejudice to him (see Evid. Code, § 352), and that if the court had weighed these factors, it would have excluded the photographs. Neither of these points is persuasive on this record.

We have repeatedly stated that the court has wide discretion in determining the admissibility of photographs of a murder victim. (People v. Green, supra, 27 Cal.3d 1, 19; People v. Frierson (1979) 25 Cal.3d 142, 171 [158 Cal.Rptr. 281, 599 P.2d 587].) We have previously held that a court may admit even “gruesome” photographs if the evidence is highly relevant to the issues raised by the facts, or if the photograph would clarify the testimony of a medical examiner. (People v. Murphy (1972) 8 Cal.3d 349, 365 [105 Cal.Rptr. 138, 503 P.2d 594].) However, when a defendant objects that the proffered evidence is more prejudicial than probative, the record must affirmatively show that the court weighed these factors, in order to allow proper appellate review of abuse of discretion claims. (People v. Green, supra, 27 Cal.3d 1, at p. 25.)

Here, the record reflects that the court did consider the probative value of the photographs and their potential prejudicial effect before admitting them. The photographs were not unusually gruesome, and were highly relevant to the issues raised by the facts. As noted by the court, significant issues in the case were whether and how a forcible sexual attack occurred and whether the location of semen could be explained by leakage. The photographs admitted showed abrasions on the abdomen and hip bones of the victim, the position of the body and the condition of the room in which the body was found. The photos were therefore clearly relevant to the issues raised by the defense. No abuse of discretion therefore appears. (People v. Pierce (1979) 24 Cal.3d 199, 211 [155 Cal.Rptr. 657, 595 P.2d 91]; People v. Murphy, supra, 8 Cal.3d 349, 365.)

D. Admission of Dr. Blake’s Testimony

Relying on People v. Collins (1968) 68 Cal.2d 319, 328-329 [66 Cal.Rptr. 497, 438 P.2d 33, 36 A.L.R.3d 1176], defendant now asserts that the court erred in admitting Dr. Blake’s testimony that results of the antigen test (ABO blood typing) showed that only 20 percent of the male population could have deposited the semen found in Hill’s body, that PGM blood-type analysis showed only 40 percent of the male population could have deposited it, and that because the frequencies are independent of each other, only 8 percent of the population could have been the donor.

Collins, supra, 68 Cal.2d 319, deals with the issue of the proper use of statistical probability figures in a criminal trial. Defense counsel did not, however, object at trial to the testimony on Collins grounds, nor did he *777argue that the evidence was otherwise inadmissible or that the tests performed by Blake were unreliable. Counsel did obtain a foundational hearing under Evidence Code section 402, but after the hearing, the only objection raised to Blake’s testimony as to the semen tests was that he should have performed a different test.22 The failure to object to the admission of the evidence on any other ground is fatal to defendant’s current contentions.

Evidence Code section 353 states: “A verdict or finding shall not be set aside, nor shall the judgment or decision based thereon be reversed, by reason of the erroneous admission of evidence unless: [(¡] (a) There appears of record an objection to or a motion to exclude or to strike the evidence that was timely made and so stated as to make clear the specific ground of the objection or motion; and [([] (b) The court which passes upon the effect of the error or errors is of the opinion that the admitted evidence should have been excluded on the ground stated and that the error or errors complained of resulted in a miscarriage of justice.” (See also Cal. Const., art. VI, § 13.) The rationale for this rule is clear; a contrary rule would deprive the party offering the evidence of any opportunity to cure the defect at trial and would permit the nonobjecting party to gamble that the error will provide grounds for reversal of the matter. (People v. Rogers (1978) 21 Cal.3d 542, 547-548 [146 Cal.Rptr. 732, 579 P.2d 1048].)

We have previously held that in a death penalty case, we may review an admissibility issue, even though there was a technical flaw in the form of the objection made at trial. (People v. Frank (1985) 38 Cal.3d 711 [214 Cal.Rptr. 801, 700 P.2d 415]; People v. Bob (1946) 29 Cal.2d 321 [175 P.2d 12].) In Bob, for example, the prosecutor had read a statement by the defendant’s companion that defendant had struck the blows which resulted in death of the victim. The defendant’s attorney objected at length to the evidence, calling it “secondary evidence” and requesting that, as the companion was available, he be produced as a witness. (29 Cal.2d at p. 324.) We held that counsel’s failure to use the word “hearsay” in his objection was not fatal to our review of the question of admissibility, because the tenor of his remarks was plainly directed to that ground. (Id., at p. 325.) In Frank, defendant’s attorney similarly objected to evidence on the ground of over-breadth of the search warrant which produced the challenged evidence, although he did not state which specific clauses in the warrant were over-broad or clarify the precise theory of his claim. Although we recognized *778that the overbreadth objection could have been more specific, we nevertheless held that a technical insufficiency in the form of an objection may be disregarded on an appeal from a judgment imposing the death penalty. (38 Cal.3d at p. 730, fn. 3.)

Unlike Frank, supra, 38 Cal.3d 711, and Bob, supra, 29 Cal.3d 321, in the present case no objection was raised on the same grounds as are now asserted on appeal. The purpose of the Evidence Code section 402 hearing does not appear on the record, but at the conclusion of that hearing, defendant’s only stated basis for excluding Blake’s evidence was that one of the possible tests on the sample had not been performed before the sample was consumed by other tests. The prosecutor therefore had no incentive to provide additional testimony to lay a foundation for Blake’s testimony. Moreover, defendant’s failure to object could have been part of a tactical strategy on his part to avoid a parade of experts who would attest to the validity of the scientific tests that were conducted. This is not a situation in which we can conclude that the substance of an objection was made, and undertake review even though the form of the objection was technically incorrect.

Nor is this a case where a clear miscarriage of justice will result unless the admissibility of the semen evidence is reviewed. There is no evidence that in fact the tests conducted by Blake were unreliable. We have recognized the evidentiary value of semen analysis in People v. Nation (1980) 26 Cal.3d 169 [161 Cal.Rptr. 299, 604 P.2d 1051], in which we held that if a semen sample is recovered following an attempted or actual rape, the authorities must take reasonable measures to adequately preserve this evidence. We also noted in Nation that: “While there are many possible analyses that may be performed on semen to identify the donor [class], and by corollary, to eliminate others from the class of possible donors, the two analyses deemed most commonly feasible are ABO blood typing and identification of the genetic marker phosphoglucomutase (PGM) [Citation.]” (Id., at p. 176.)23

*779The ABO-typing test and PGM test were the two tests which were performed on the sample recovered from Hill’s body. In addition, as noted, defendant’s palm print and thumbprint, found at the scene of the crime, combined with his denial that he had ever been at the scene constituted persuasive evidence of guilt, especially when viewed in conjunction with the People’s evidence undermining defendant’s alibi defense. On these facts, we cannot find that a miscarriage of justice would occur unless we were to review defendant’s evidentiary objections, made for the first time on appeal.

E. The Felony-murder Rule

Defendant contends the felony-murder rule is unsound as a matter of policy and that this court should abolish it, and that the rule is unconstitutional as a violation of due process. We rejected both of these arguments in People v. Dillon (1983) 34 Cal.3d 441 [194 Cal.Rptr. 390, 668 P.2d 697]. Defendant presents us with no compelling reason to reexamine that holding.

IV

Special Circumstances Contentions

Defendant contends the court failed to instruct in accord with Carlos v. Superior Court (1983) 35 Cal.3d 131 [197 Cal.Rptr. 79, 672 P.2d 862], that intent to kill is an element of the felony-murder special circumstance, and that the special circumstances—murder during the commission of rape and murder during the commission of sodomy—must therefore be vacated. We rejected this claim in People v. Anderson (1987) 43 Cal.3d 1104, 1138-1148 [240 Cal.Rptr. 585, 742 P.2d 1306]. Defendant has never argued that more than one person was involved in the crime, and the evidence showed that defendant either killed Hill or was not involved in the crime at all. We find that the record establishes beyond doubt that the defendant was the actual killer in this case, and that the finding required by Enmund v. Florida (1982) 458 U.S. 782 [73 L.Ed.2d 1140, 102 S.Ct. 3368], is therefore satisfied. (Cabana v. Bullock (1986) 474 U.S. 376, 389-391 [88 L.Ed.2d 704, 718-720, 106 S.Ct. 689, 699-700].)

Defendant next argues that the sweep of death eligibility for unintentional felony murder is unjustifiably and irrationally broader than that for deliberate, premeditated murder. He contends that insofar as it makes persons guilty of felony murder eligible for the death penalty, the 1978 death penalty law is unconstitutional for failure to “provide a rational *780basis for distinguishing between those murderers who deserve to be considered for the death penalty and those who do not.” (People v. Green (1980) 27 Cal.3d 1, 61 [164 Cal.Rptr. 1, 609 P.2d 468].)

This claim was also rejected in Anderson, supra. There we found no constitutional impediment to a statutory scheme which would render persons who commit felony murder without intent to kill eligible for the death penalty. {People v. Anderson, supra, 43 Cal.3d at p. 1146.) As we noted, “[wjhether or not we approve of the wisdom of the statutory classification, it appears to be generally accepted that by making the felony murderer but not the simple murderer death-eligible, a death penalty law furnishes the ‘meaningful basis [required by the Eighth Amendment] for distinguishing the few cases in which [the death penalty] is imposed from the many cases in which it is not.’ (Furman v. Georgia (1972) 408 U.S. 238, 313 [33 L.Ed.2d 346, 392, 92 S.Ct. 2726] (conc. opn. of White, J.); accord, Godfrey v. Georgia (1980) 446 U.S. 420, 427 [64 L.Ed.2d 398, 405-406, 100 S.Ct. 1759] (plur. opn.).)” (People v. Anderson, supra, 43 Cal.3d at p. 1147 (brackets in original; fn. omitted).)

Defendant next argues that use of the underlying felony murder as an element of the crime, a special circumstance, and an aggravating factor is an indiscriminate and grossly unfair multiple use of facts. We have previously rejected this contention (People v. Gates (1987) 43 Cal.3d 1168, 1188-1190 [240 Cal.Rptr. 666, 743 P.2d 301]; People v. Balderas (1985) 41 Cal.3d 144, 200-201 [222 Cal.Rptr. 184, 711 P.2d 480]) and defendant advances no convincing reason why we should reconsider our previous conclusion.

V

Penalty Phase Contentions

Defendant contends that a number of errors were committed during the penalty phase of his trial. We find that none of the errors requires reversal on this record.

A. Instruction as to Governor's Power to Commute Life Sentence

The jury was given the so-called “Briggs instruction” at the conclusion of the penalty phase of defendant’s trial.24 We have held the giving of this instruction to be error. (People v. Ramos (1984) 37 Cal.3d 136 [207 *781Cal.Rptr. 800, 689 P.2d 430] (Ramos IT).) We find, however, that because the court here also gave an instruction directing the jury to disregard the former instruction, the error was not prejudicial.25 (Accord, People v. Hamilton (1988) 45 Cal.3d 351, 375-376 [247 Cal.Rptr. 31, 753 P.2d 1109].)

After giving the Briggs instruction, the court added: “So that you will have no misunderstandings relating to a sentence of life without possibility of parole, you have been informed generally as to the Governor’s commutation modification power. You are now instructed, however, that the matter of a Governor’s commutation power is not to be considered by you in determining the punishment for this defendant, [if] You may not speculate as to if or when a Governor would commute the sentence to a lesser one which includes the possibility of parole, [if] I instruct you again that you are to consider only those aggravating and mitigating factors which I have already read to you in determining which punishment shall be imposed on this defendant.”26 About an hour after retiring, the jury requested that the instructions be reread. They were, including the above portion.

We held in Ramos II that it is a violation of the state Constitution to give the Briggs instruction. We noted that the instruction is misleading, particu*782larly because it does not advise the jury that a penalty of death as well as life without possibility of parole may be commuted, and because it invites the jury to consider matters which are speculative and irrelevant to its decision. In addition, we held that such an instruction is improper because it may divert the jury from its proper function, either by reducing the jury’s understanding of its personal responsibility for the decision or by inviting the jury to second guess the future actions of a governor. (People v. Ramos, supra, 37 Cal.3d at pp. 155-158.)

As we explained in People v. Hamilton, supra, 45 Cal.3d 351, 375-376, the concerns expressed in Ramos II are, however, largely undercut by the additional instruction given by the court in this case. The additional instruction in essence negated the Briggs instruction by advising the jury that it was to disregard the Governor’s commutation power completely in determining punishment, and that it was not to “speculate as to if or when a governor would commute the sentence to a lesser one . . . .” The fear that a jury might consider speculative and irrelevant matters is thus expressly addressed by this instruction, as is the concern that a jury might be tempted, by virtue of the unadorned Briggs instruction, to speculate regarding the actions of a future governor.

Here, the jury was told it should not consider the possibility of commutation in determining the sentence.27 The jury, having been so advised, could not have been misled as to the nature and scope of the commutation power. Viewing the two instructions together, we conclude that under any standard, the Briggs instruction error was not prejudicial.

B. Failure to Give Proper Instructions on “Other-crimes Evidence”

Defendant contends, and the Attorney General concedes, that the court erred in failing to instruct the jury sua sponte that evidence of defendant’s prior criminal acts would be considered in aggravation only if proved beyond a reasonable doubt. (People v. Robertson (1983) 33 Cal.3d 21, 53-55 [188 Cal.Rptr. 77, 655 P.2d 279] (plur. opn. by Kaus, J.), pp. 60-63 (conc. opn. by Broussard, J.).)

On this record, we do not find that the court’s failure to give the Robertson instruction would have affected the outcome of the penalty phase of defendant’s trial. Most of the other-crimes evidence introduced related to *783felony crimes for which defendant had already been convicted, including the crimes involving the 13 and 11 year olds, and the false imprisonment of the college student. The remainder of the crimes28 and their violent nature were clearly proved by testimony of either the victim of defendant’s assault or an eyewitness to the event, and defendant introduced no evidence denying or contradicting the evidence presented by the prosecution related to the crimes.29 Because the most serious of the prior criminal acts were acts for which the defendant was convicted, and the remainder of the crimes were situations in which the evidence is compelling, we find that, under any standard, the failure to give the Robertson, supra, 33 Cal.3d 21, instruction was not prejudicial to defendant.

C. Instructions on Mitigating Evidence and Jury Discretion

The jury instructions given at the conclusion of defendant’s penalty trial simply listed the statutory aggravating and mitigating factors and did not otherwise advise the jury that it could consider in mitigation all of defendant’s proffered evidence, even though it may not literally “extenuate the gravity of the crime.” (Former CALJIC No. 8.84.1.) The jury was also informed that it “shall” impose a sentence of death if it found aggravating circumstances outweighed mitigating circumstances. (Former CALJIC No. 8.84.2.) We have previously held that in future cases these instructions should be amplified to avoid any potential that the jury might thereby be misled about the scope of mitigating evidence, or its sentencing discretion.

Thus, in People v. Easley, supra, 34 Cal.3d 858, we imposed the requirement that in connection with statutory factor (k) (see § 190.3, factor (k)), courts should inform the jury that it may consider as a mitigating factor any other “ ‘aspect of [the] defendant’s character or record . . . that the defendant prolfers as a basis for a sentence less than death.’ ” (People v. Easley, supra, 34 Cal.3d 858, 878, fn. 10.)

Similarly, in People v. Brown, supra, 40 Cal.3d 512, we observed that possible confusion might be engendered by use of the unadorned words of the statute that the jury “shall impose a sentence of death if [it] concludes *784that the aggravating circumstances outweigh the mitigating circumstances.” As we explained in People v. Allen (1986) 42 Cal.3d 1222 [232 Cal.Rptr. 849, 729 P.2d 115], our concern as to this instruction was that the instruction might, in “two interrelated ways,” lead the jury to misunderstand its weighing and decisionmaking responsibility. First, we were concerned that the jury might believe that its weighing responsibility required merely that it count the aggravating and mitigating factors. Second, we were concerned that the instruction might be understood to allow the jury to return a death sentence without making a personal moral judgment that death was the appropriate punishment in the case. (People v. Allen, supra, 42 Cal.3d at pp. 1276-1277.)

Having reviewed the record as a whole (People v. Brown, supra, 40 Cal.3d at p. 544, fn. 17), we conclude that a reasonable jury would not have been misled here as to the scope of the evidence it could consider under factor (k). Nor do we believe a reasonable jury would have been misled about its sentencing discretion under Brown.

1. Factor (k) Instruction

Defendant presented significant “sympathy” evidence at the penalty phase of his trial. His mother described his character and difficult family background; Dr. Studenski, the psychiatrist, related his interview with defendant, who evidenced mental anguish as to various events in his life; and Reverend Jones testified about defendant’s attempts to bring his life under control.

The court in delivering the instructions told the jury, “In determining which penalty is to be imposed on defendant, you shall consider all of the evidence which has been received during any part of this trial.”30 The court also stated, however, “I instruct you again that you are to consider only those aggravating and mitigating factors which I have already read to you in determining which punishment shall be imposed on this defendant.” After retiring, the jury asked that the instructions be read to them again, which they were.

In his closing argument discussion of factor (k), the prosecutor told the jury: “And the last catchall, any other circumstance which extenuates the gravity of the crime, even though it is not a legal excuse for the crime. And perhaps the defense team thought that that includes bringing a mother into court and saying ‘Please don’t kill my son,’ except if you listen to that instruction carefully, it is this: Any other circumstance which extenuates *785the crime, [^j] So, we are talking about the crime.” In addition, the prosecutor argued, “I ask you to listen to the law, don’t be swayed by cries of mercy, because this is not a mercy hearing. If it were, the law would so say.”

The prosecutor’s argument is troublesome. Initially, the argument might be viewed as misinforming the jury about the scope of its consideration of defendant’s mitigating evidence under factor (k). Review of the argument as a whole, however, reveals that the prosecutor refrained from expressly telling the jury that the evidence of defendant’s poor childhood and adult-life problems as testified to by Studenski, defendant’s mother, and Jones (ante, pp. 762-763) was irrelevant to its penalty determination, under factor (k) or otherwise. Instead, the prosecutor’s statements were reasonably understood as simply telling the jury not to be swayed by “mere” pleas for mercy that were unrelated to actual evidence about defendant as a person. (See, e.g., California v. Brown, supra, 479 U.S. at pp. 541-542 [93 L.Ed.2d at pp. 939-940, 107 S.Ct. at p. 840] (plur. opn. by Rehnquist, C. J.) & 544-548 [93 L.Ed.2d at pp. 941-944, 107 S.Ct. at pp. 841-842] (conc. opn. by O’Connor, J.).)

Even assuming the prosecutor left the jury with ambiguous directions about its consideration of mitigating evidence under factor (k), we are confident that defense counsel’s argument left the jury with a proper understanding. He first discussed factors (a) through (j), and concluded that the evidence demonstrated four factors in mitigation. With respect to factor (k), he told the jury: “And then, K, which is the last one and which leaves you a great catchall, for within your conscience, the words weigh any other circumstance which extenuates the gravity of the crime, even though it is not a legal excuse for the crime.” We believe that a reasonable jury would interpret this to mean that factor (k) gave it the flexibility to consider all relevant evidence relating to defendant as a person.

Viewing the record as a whole—i.e., counsel’s arguments, the three defense witnesses, and the court’s repeated instruction to consider “all evidence,” we do not believe the jury was misled to defendant’s prejudice about the scope of the evidence it could consider.

2. Sentencing Discretion

At no point did either counsel attempt to tell the jury to “count” or merely “add up” the various aggravating and mitigating factors in deciding the penalty issue. In fact, both counsel in their closing arguments told the jury to weigh the factors.

The prosecutor stated near the opening of his closing argument, “the law is clear that you will balance and weigh the circumstances in aggravation *786against the circumstances in mitigation . . . .” Defense counsel noted: “There are a set of things called aggravating factors. What you have to do is weigh them. There is no burden of proof. There is no burden of persuasion. I am going to talk to you in a little while about what I think the word ‘weigh’ means, because the law doesn’t define it for you . . . .” Later in his argument, defense counsel stated: “What you are supposed to do with these things, these mitigating circumstances and these aggravating circumstances, the Judge will tell you is you are going to weigh them. What does that mean? And you are going to be looking at the Judge and looking at me, and we can’t tell you what it means. [jf] And there is good reason we can’t tell you what it means, because it means what you twelve people in your conscience, in your humanity, in your compassion, in your want [szc] of justice, in your hearts, in your religion, in your lack of religion, in your wisdom, in what you want this society to be, that is how you will weigh them. That is what weigh means. [j[] He cannot tell you that you add up five against three and five wins. It doesn’t work that way. Even though I have told you, most of the factors here are mitigating factors, so if we were to add the mitigating factors up, just add them up, Coleman would win.” The arguments clearly advised the jury of the nature of the weighing process.

Nor was the jury misled as to its responsibility for deciding for itself whether death was appropriate for defendant. Both counsel referred to the jury’s role, and correctly told the jury that it was responsible for the penalty decision. The prosecutor told the jury more than once that it had the difficult decision of deciding a man’s fate, and at no point attempted to derail the jury from that understanding.31 The prosecutor noted: “[T]he job that each and every one of you are going to embark on, . . . is much more difficult than my job, because you are going to have to make that decision.” He also stated: “It is a decision you have to make, and I want to impress upon you, you make that decision,” and near the end of his argument said: “As hard as it may sound, it is a decision that you must make, based upon the factors which have been presented to you, in accordance with one another, a decision which you must not make lightly, which you should take seriously as it is, as gravely as it is, but a decision which you nonetheless must make.”

The jury’s responsibility for its decision was a point also repeatedly made by defense counsel in his closing argument. Counsel noted that “It is an *787individual thing, an individual verdict on your part.” He argued that each juror would have to take responsibility for the verdict, and that the choice of verdict was up to each juror.

Based on a review of the arguments of both counsel, we believe the jury properly understood its role in the decisionmaking process, and that it was to decide, in the context of the weighing process, the appropriate penalty in this case. (See People v. Allen, supra, 42 Cal.3d at pp. 1276-1280.)

D. Instruction on Inapplicable Statutory Factors

Defendant contends that the trial court erred in reading the entire list of statutory aggravating and mitigating factors, although several factors were assertedly inapplicable to the case. We have recently rejected this argument in People v. Ghent, supra, 43 Cal.3d 739, 776-777, and defendant presents no compelling argument for us to reconsider that holding.

E. Constitutionality of the 1978 Sentencing Statute

Defendant contends that section 190.3 is invalid on the ground that it lacks the necessary procedural safeguards essential to a constitutional capital sentencing scheme, namely, that the statute: (1) does not specifically enumerate aggravating and mitigating factors, (2) permits consideration of aggravating evidence other than that specified in the statute, (3) does not require written findings of aggravating factors, (4) does not require that aggravating factors be found beyond a reasonable doubt, (5) does not require jury unanimity regarding aggravating factors supporting a sentence of death, and (6) does not require comparative appellate review to prevent arbitrary and disproportionate imposition of the death penalty. With the exception of defendant’s second argument above, we recently rejected each of these arguments. (People v. Rodriguez (1986) 42 Cal.3d 730, 777-779 [230 Cal.Rptr. 667, 726 P.2d 113].) Defendant presents no arguments which would lead us to reconsider the holding in Rodriguez or in the cases on which it relies.

The second argument was addressed in People v. Boyd (1985) 38 Cal.3d 762, 773 [215 Cal.Rptr. 1, 700 P.2d 782]. We held in Boyd that the court should admit in aggravation only that penalty phase evidence which is relevant to the statutory listed factors (exclusive of factor (k)). Defendant’s second argument is therefore based on an incorrect premise, because the statute, properly interpreted, does require exclusion of aggravating evidence not relevant to the specified aggravating factors.

Defendant made no objection to the admission of any of the other-crimes evidence on the ground that the evidence did not relate to one of the statutory factors. In addition, our review of the record finds only one item *788of evidence introduced at the penalty trial which was not admissible under section 190.3, factors (b) or (c). Deputy Sheriff Clich testified to defendant’s stabbing of another inmate, his possession while in prison of a “shank” (a stabbing weapon); and to his assault on deputies trying to obtain a hair sample pursuant to a court order. During that testimony, however, Clich also stated that defendant had threatened him after the last incident involving the hair sample. It is not clear that the threat constituted criminal activity or the express or implied threat of force or violence.32 (People v. Boyd, supra, 38 Cal.3d at pp. 777-778.) No objection was made to this testimony, however, and the point may be deemed waived on appeal. Moreover, even if admission of the evidence was erroneous, we find it nonprejudicial under any standard.

F. Prosecutorial Misconduct

Defendant contends the prosecutor committed misconduct in his comments regarding the testimony of a defense psychiatrist.33 Defendant did not object to-the prosecutor’s comments at trial. A claim of prosecutorial misconduct made for the first time on appeal will not be considered if a timely objection and admonition would have cured the harm. {People v. Green, supra, 27 Cal.3d at pp. 27-34.) We are satisfied that a timely admonition would have cured any possible harm here, and therefore reject defendant’s claim of misconduct. In any event, any error was harmless on this record under any standard.

The judgment of guilt, the findings of two special circumstances and the judgment of death are affirmed.

Panelli, J., Arguelles, J., Eagleson, J., and Kaufman, J., concurred.

I agree with the majority that the error in admitting expert testimony in violation of the Kelly-Frye rule (People v. Kelly (1976) 17 Cal.3d 24 [130 Cal.Rptr. 144, 549 P.2d 1240]; Frye v. United States (D.C. Cir. 1923) 293 Fed. 1013 [54 App.D.C. 46, 34 *789A.L.R. 145]) was not reversible under the Watson test. (People v. Watson (1956) 46 Cal.2d 818, 836 [299 P.2d 243].) Therefore I would affirm the judgment of guilt.

On the other hand, I agree with Justice Broussard that the inculpatory evidence in this case is demonstrably thin. Even as catalogued by the majority, it cannot possibly be deemed “quite substantial.” (Maj. opn., p. 775.)

In the interest of justice and in good conscience I cannot send this defendant to his death in the gas chamber on a record of such frail evidence. Therefore, I would exercise our authority under Penal Code sections 1260 and 1181, subdivision 7 (see, e.g., People v. Lucero (1988) 44 Cal.3d 1006, 1034-1036 [245 Cal.Rptr. 185, 750 P.2d 1342] (conc. and dis. opn. of Mosk, J.)) and vacate the judgment as to penalty and remand the cause to the trial court with directions to impose a sentence of life imprisonment.

I dissent. The trial court erred in admitting the evidence that defendant’s thumbprint was bloody. The court should have excluded the evidence for lack of reliability. The majority recognize the error. (Ante, at p. 775.)

The evidence in this case to identify defendant as the perpetrator of the crime is so weak that the error must be held prejudicial. The evidence to connect defendant to the crime was the presence of his thumbprint and palm print in the bungalow, his false statement to the homicide inspector denying that he had been in the bungalow, and the sperm tests. In finding the error nonprejudicial, the majority omit a crucial item of the evidence, place undue weight on the sperm tests, and erroneously claim that the prosecution significantly undermined the alibi evidence.

At the outset, it should be pointed out that the erroneously admitted evidence was not cumulative of other evidence. The inspector did not testify that the discoloration on the thumbprint appeared to be caused by blood. The record does not show whether his reason for testing for blood was merely that his testing equipment was handy or whether he had tentatively concluded the discoloration was blood before testing. In addition, the fact that the thumbprint was bloody was given substantial attention at the trial. The inspector had written the words “bloody print” on the back of the chair, and the back of the chair was received in evidence.

In some cases, fingerprints may provide devastating evidence of guilt. But this is not one of them. The police officers found the victim’s prints, a thumbprint and a palm print of defendant, several smudged prints, and an unidentified fingerprint. The unidentified print was found on the same chair as the thumbprint. The unidentified print is not mentioned in the majority opinion. When we exclude from consideration the evidence that the thumb*790print was bloody, the print evidence does not suggest that there is any greater likelihood that defendant was the killer than the donor of the unidentified print. Further, the crime involved substantial movement over a substantial period of time, so that one would expect that the killer would have left more prints unless he wore gloves or unless there were very few surfaces in the bungalow which would retain prints. The fact that only three prints for two people were found indicates a substantial possibility that others entered the bungalow as well as the victim, defendant, and the donor of the unidentified print.

Defendant’s lie, although a circumstance to be considered, is not entitled to much weight. At the time defendant’s thumbprint was matched to the thumbprint on the chair, about four months after the homicide, defendant was in jail on other charges. A police inspector brought defendant from his cell to the offices of the homicide department. After being given Miranda admonitions (Miranda v. Arizona (1966) 384 U.S. 436 [16 L.Ed.2d 694, 86 S.Ct. 1602, 10 A.L.R.3d 974]), he agreed to talk to the inspector. Defendant said he did not know Shirley Hill, the victim. He knew where Mission High School and Dolores Park were but had never been on the football field or in the bungalows. He categorically denied the thumbprint was his. When told he was thought to be responsible for Shirley Hill’s death and would be booked, he terminated the interview.

Obviously, his statement that he had never been in the bungalow was false; defendant lied. However, the problem is not merely whether he lied. The question is whether his lie provides the basis for a strong or substantial inference that he committed the murder. While no doubt defendant would lie if he were guilty, we must also consider whether defendant would tell the truth if he were innocent or whether he would lie to avoid involvement in a homicide matter. On the record before us, I have little confidence in a conclusion that defendant, if innocent, would not lie to avoid involvement in a homicide matter and am unable to give much weight to his lie.

The proper print evidence and the lie are not strong evidence of guilt; indeed there is some question whether they would be sufficient to sustain a conviction in view of the unidentified fingerprint and the possibility that someone who did not leave prints was the murderer. It seems clear that any conclusion that the error was nonprejudicial must be based on a conclusion that the sperm evidence was highly persuasive of guilt. However, it was not.

Even viewing that evidence most favorably to the prosecution I have difficulty concluding that it is entitled to great weight. And viewing it on the basis of the whole record, I doubt whether it is entitled to any substantial weight.

*791Dr. Blake testified that he received a vaginal wash (a solution of loose cellular matter and spermatozoa in saline solution) obtained from decedent’s body, and he subjected it to phosphoglucomutase analysis (PGM) and a test for blood antigens. On the PGM test, he concluded that the vaginal wash contained type 2. Since the decedent was type 1/1, the donor of the sperm had to be either a type 2/1 or 2/2. Dr. Blake failed to find any antigens, so he concluded that decedent and the donor were nonsecretors, people who do not secrete their blood antigens into other body fluids such as saliva, urine, semen and vaginal secretions. Tests have showed that only 40 percent of the population are PGM type 2, and only 20 percent are nonsecretors. There is no preexisting relationship between PGM types and nonsecretors, and based on this fact the frequencies may be multiplied so that only 8 percent of the male population could have donated the sperm. Tests of defendant showed he was within the 8 percent.

Dr. Blake emphasized that these tests can never demonstrate that a particular sample came from a particular person, although they may show that it absolutely did not come from a particular person. The absence of antigens in the instant case excluded the victim’s former husband as a suspect.

California and a majority of other states have long admitted blood-typing evidence. (See Evid. Code, §§ 891-897; People v. Poggi (1988) 45 Cal.3d 306, 324 [246 Cal.Rptr. 886, 753 P.2d 1082]; People v. Brown (1985) 40 Cal.3d 512, 536, fn. 6 [220 Cal.Rptr. 637, 709 P.2d 440], revd. on other grounds sub nom. California v. Brown (1987) 479 U.S. 538 [93 L.Ed.2d 934, 107 S.Ct. 837]; People v. Lindsey (1978) 84 Cal.App.3d 851, 863-866 [149 Cal.Rptr. 47, 2 A.L.R.4th 485]; People v. Vallez (1978) 80 Cal.App.3d 46, 56 [143 Cal.Rptr. 914].) While such testing in general has been approved, controversy has arisen where the sample is not secured in the laboratory but obtained from a dried stain exposed to drying, aging, temperature, and contamination (particularly with bacteria or other organic substances), and where the composition of the test sample is unknown. (People v. Brown, supra, 40 Cal.3d 512, 530.) Recently, the Court of Appeal after an eight-day trial court hearing upheld the admissibility of electrophoretic typing of dried bloodstain evidence. (People v. Reilly (1987) 196 Cal.App.3d 1127, 1135 [242 Cal.Rptr. 496]; but cf. People v. Young (1986) 425 Mich. 470 [391 N.W.2d 270, 284].)

In determining prejudice, the issue in the present case is not admissibility but the weight to give to the evidence. Nevertheless, cases dealing with the issue of admissibility point out the difficulty in determining the proper weight to be given to the evidence. As pointed out in People v. Kelly (1976) 17 Cal.3d 24, 30-31 [130 Cal.Rptr. 144, 549 P.2d 1240], scientific evidence may in some cases assume a quality of infallibility, and care must be espec*792ially exerdsed when the sdentific technique is used to identify the perpetrator of crime.

Under the Kelly/Frye rule, the proponent of the scientific evidence must establish (1) the generally accepted reliability of the method usually by expert testimony, (2) that the witness furnishing such testimony is properly qualified as an expert to give an opinion on the subject, and (3) that correct scientific procedures were used in the particular case. (People v. Kelly, supra, 17 Cal.3d 24, 30-32; Frye v. United States (D.C. Cir. 1923) 293 Fed. 1013, 1014 [54 App.D.C. 46, 34 A.L.R. 145]; People v. Brown, supra, 40 Cal.3d 512, 530.)

Although the defense requested a hearing pursuant to Evidence Code section 402, the prosecution did not attempt to provide a foundation for the evidence or meet the Kelly/Frye rule.1 Thus there was no showing of the generally accepted reliability of the method used.

Neither was there a showing that correct scientific procedures were used. Dr. Blake testified that he received a vaginal wash from Dr. Stephens of the coroner’s office, and that a vaginal wash is obtained by inserting a saline solution into the cavity and then withdrawing it into a syringe. However, Dr. Sisson testified that he obtained the sample by using a swab. There is nothing to indicate that the latter method was correct.

The primary objection to giving substantial weight to the test is the unreliability of the results secured by the test. In People v. Reilly, supra, 196 Cal.App.3d 1127, 1141-1146, which involved in part electrophoretic testing for PGM, the witnesses appear to have recognized that a small error rate affects the reliability of results.2 It does not appear whether Dr. Blake used electrophoretic testing for PGM, whether he used some other method, or whether such other method would also involve a recognized rate of error. If there was a rate of error, it becomes very difficult to evaluate the evidence that the donor of the semen was donated by a person in the 40 percent of the population having PGM 2. In other words, a slight margin of error which would fail to exclude the other 60 percent means that the evidence that the donor was limited to the 40 percent group cannot be given much weight.

While the possibilities that the swab method of obtaining the sample is improper and that there is a recognized rate of error in the PGM test may *793be matters of speculation, Dr. Blake’s testimony itself limits the value of the testimony as to the ABO antigens. His result, it will be recalled, was that there were no antigens found, a negative result. The absence of antigens, he concluded, established that the victim and defendant were nonsecretors, who comprise only 20 percent of the population.

Dr. Blake recognized that the absence of antigens could also be caused by the fact that the donor was a low-level secretor or by the destruction of the antigens. He stated that the possibility that the donor was a low-level secretor was not a “consideration” in this case because the sample was sufficiently concentrated for even a low-level secretor to be detected.

In contrast, however, to his total rejection of the possibility that the donor was a low-level secretor, he never ruled out the possibility that the antigens were destroyed. Indeed, the plain implication of his testimony was that there was a possibility that there were antigens present but they were destroyed, although he doubted this occurred.

He started out by stating that the possibility that antigens were destroyed was difficult to assess. He acknowledged that correct procedures in securing the sample, including its freezing, could destroy them. He testified that antigens in general were very stable, that if destruction had occurred in this case 99.9 percent of the antigen material would have had to be destroyed, and that if destruction had taken place, he would “expect” to see acid phosphotase and PGM destroyed or damaged. The latter results were clear-cut. He concluded that his “assessment” was that it was “likely” and “reasonable” that destruction did not occur and “unlikely” that destruction occurred.

His lack of complete certainty as to whether destruction occurred and his reasoning in reaching his assessment preclude any substantial weight being given to the antigen evidence. The test results obtained by him were apparently relatively rare results under his own testimony. Only 20 percent of the population are nonsecretors so it seems to follow that tests on couples, if accurate, would result in a finding of no antigens in only 4 percent of the cases. (.20 X .20 = .04.) A direct correlation between clear-cut acid phosphotase and PGM results and lack of destruction of antigens should result in finding antigens 96 percent of the time. If his tests or scientific literature indicated that antigens were present in as many as 90 percent of the cases where acid phosphotase and PGM were clear—a high correlation—his assessment still would not be entitled to any weight but would indicate destruction was more likely than accuracy. His statement that he would “expect” a correlation seems far short of a direct correlation. Further, his *794use of “likely,” “reasonable,” and “unlikely” concede that his prior tests and prior information did not show a direct correlation.3

Absent a direct correlation, or at least a precise correlation, any attempt to give weight to the failure to find antigens involves pure speculation. Thus any use of the 20 percent figure to reduce the 40 percent figure to 8 percent would be improper. The 40 percent figure, which may also be suspect, does not by itself furnish substantial evidence of guilt, and when combined with the print evidence and the lie does not provide a strong case of guilt.

Not only was the prosecution’s guilt case weak, but there was substantial alibi evidence. The victim attended classes at a school some distance from City College of San Francisco from noon to 3 p.m., and around 3:30 p.m. a friend saw her at a shopping center. The scene of the crime is a few blocks from the bus route she was expected to take to return home after school.

Defendant testified that he was at City College of San Francisco until 3:30 p.m. on the date of the murder, that he met Pastor McAllister and that they went to the home he shared with his former wife. His wife and her two children were home and Pastor McAllister stayed until early evening, five or six o’clock. Pastor McAllister, who was pastor at defendant’s wife’s church, testified that he met defendant around two or three o’clock in Cloud Hall after he attended Mr. Lindsey’s math class. They went by bus to defendant’s house which required between 30 minutes and an hour. He stayed a couple of hours, and defendant was there when he left. On cross-examination Pastor McAllister said that to the best of his recollection the meeting occurred on September 5, which was the first day of classes, but it could have occurred during the first week of classes. Defendant’s former wife confirmed Pastor McAllister’s visit and said defendant did not leave the house until the following morning.

The prosecution in rebuttal showed that Pastor McAllister attended Mr. Lindsey’s class on the first day of class but defendant did not attend on that day or the rest of the week. Mr. Lindsey had marked “excused” on the first day but not the others. The “excused” entry meant that at some time after the class either the same day or later defendant had attempted to excuse his absence. Both Pastor McAllister and defendant dropped the class the following week in favor of another class.

Defendant was not charged with the murder until four months after it occurred, and it does not appear when Pastor McAllister learned of the date of the crime or whether he recognized the coincidence of dates and volun*795teered his testimony or the defense contacted him. Given the passage of at least four months, his failure to be absolutely certain of the date is understandable. So far as appears, the most likely time that decedent met her murderer, assuming Pastor McAllister’s best recollection was correct, was during the period when defendant was with him. Since defendant missed the math class for the whole week, the prosecution evidence on rebuttal does not impeach the pastor’s testimony.

Viewing the whole record of the guilt trial, this has to be one of the weakest cases to identify defendant as the perpetrator of the crime that I have seen. Further, there is substantial unimpeached alibi evidence. The evidence that defendant’s thumbprint was bloody was erroneously admitted, as the majority recognize. That error gave weight to defendant’s thumbprint as establishing defendant as the perpetrator of the crime, and it is reasonably probable that a result more favorable to him would have been reached in the absence of the error. (People v. Watson (1956) 46 Cal.2d 818, 836 [299 P.2d 243].) Accordingly the judgment should be reversed.

Appellant’s petition for a rehearing was denied November 3, 1988, and on November 2, 1988, the opinion was modified to read as printed above. Mosk, J., and Broussard, J., were of the opinion that the petition should be granted.