Jimmy Lee Gray appeals from the district court's order denying his petition for habeas corpus relief. We affirm.

On June 25, 1976, Derissa Jean Scales, a three-year old girl, disappeared from her parents' apartment in Pascagoula, Mississippi. Because Gray, who lived nearby, was the last person seen with Derissa, he was questioned by the police about her disappearance. Gray admitted to the police that he had taken Derissa for a ride in the country, stopped the car on a back road, talked with Derissa and touched her vaginal area. He claimed that Derissa then wandered away from the car and fell into a shallow ditch filled with water. Gray stated that he pulled her out of the ditch while she was still breathing, put her into the trunk of his car half alive and began to drive back to Pascagoula. On the way back, Gray crossed a bridge over Black Creek. He stopped the car, opened the trunk, and threw Derissa Scales into the water.1

Gray was convicted in state court of capital murder and sentenced to death. The conviction, however, was reversed on appeal and the case was remanded for a new trial. See Gray v. State, 351 So. 2d 1342 (Miss.1977). About three weeks before the second trial, Fielding Wright and James Heidelberg were appointed to represent Gray. At the second trial, Gray was again convicted and sentenced to death. The second conviction was upheld on appeal. See Gray v. State, 375 So. 2d 994 (Miss.1979), cert. denied, 446 U.S. 988, 100 S. Ct. 2975, 64 L. Ed. 2d 847 (1980).

Gray filed a petition for habeas corpus alleging twenty-two separate constitutional violations. Chief among Gray's claims were that he had been denied effective assistance of counsel and that the record developed by the trial court was insufficient to determine the merits of his claim under Witherspoon v. Illinois, 391 U.S. 510, 88 S. Ct. 1770, 20 L. Ed. 2d 776 (1968). Although the district court only discussed these two claims in detail, it dismissed all of Gray's claims as lacking merit. On appeal, Gray reasserts his ineffective assistance and Witherspoon claims. He also contends that the evidentiary hearing held by the district court was inadequate, its findings of fact were inadequate and its disposition of the claims raised in his habeas petition was improper.

Because Gray's ineffective assistance claims are directed solely at his counsels' efforts during the sentencing phase of his second trial, we will discuss only that portion of the trial. Hindsight shows us that Wright and Heidelberg properly perceived that unless they could get Gray's statements to the police suppressed, they had little chance of avoiding a conviction at the guilt phase of his trial. Their primary aim became to avoid the death penalty at the sentencing phase. To this end, Heidelberg spoke with Gray frequently, explained to him the nature of the bifurcated trial, and asked him if there were anyone whom he wanted to testify on his behalf. Even though Gray steadfastly maintained that he did not want character witnesses, Heidelberg explored the possibility of having family members testify. This proved unsuccessful, however, since Gray's mother believed that her son should be executed and Gray's brother was unable to come to Mississippi for the trial. Heidelberg also spoke to a police officer who had befriended Gray and contacted Gray's local parole officer. He tried to get in touch with Gray's former employer and girlfriend, but both had left Pascagoula. Heidelberg talked with Louis Fondren, the attorney for Gray's first trial, and discussed possible witnesses with him.

Heidelberg also considered investigating two other sources of witnesses. Before Gray moved to Pascagoula, Gray had been imprisoned in Arizona for second degree murder. Because Gray had been a model prisoner, it was possible that people who had worked with Gray in prison could testify on his behalf. Heidelberg, however, chose not to pursue this testimony for two reasons. First, Gray did not provide Heidelberg with names of any of these witnesses and Arizona refused to release Gray's prison record, which might have contained such information. Although a suit could have been instituted in Arizona to compel the release of Gray's records, Heidelberg testified that the defense lacked the funds to file suit in Arizona. Moreover, even if the information had been available, Heidelberg was hesitant to rely on these witnesses. He felt the weight of their testimony on Gray's good character while imprisoned would be overcome by the emphasis they would place on Gray's previous conviction in Arizona for the murder of another young girl.

A second source of witnesses which Heidelberg considered using was a group of people who had written to Gray after the murder of Derissa Scales. Most of their correspondence concerned religious matters. Heidelberg decided, however, that these people would not make good witnesses because they had never known Gray personally and had only corresponded for a short time. Moreover, Heidelberg was concerned that some of these correspondents had only been interested in exploiting Gray.

Instead of trying to establish that Gray had some worth in his life, Wright and Heidelberg decided Gray's best chance to avoid the death penalty lay in trying to convince the jury he was mentally disturbed.2 Gray had told Wright and Heidelberg that at times he became emotionally disturbed, that various things set him off and that while he knew what he was doing he could not stop himself from doing it. Being subject to such an irresistible impulse does not render a person legally insane in Mississippi. However, this emotional impairment can be offered as a mitigating factor at the sentencing phase.

To support this line of proof, Wright had Gray analyzed by a local psychiatrist, Dr. Bridges. Dr. Bridges, however, advised Wright that his testimony would not support Wright's theory. Wright therefore decided to rely on a standardized personality test, the Minnesota Multiphasic Personality Inventory, which Gray had taken approximately nine years before trial. This test concluded that there was a 95% probability that Gray was mentally disturbed. Wright had discussed the validity of the test with Dr. Bridges and was aware that it was dated. He felt, however, that its conclusion would be clear to the lay jury and that it was the most persuasive piece of evidence available to the defense.

To get this test admitted into evidence Wright made a deal with the prosecutor. The prosecutor would agree to let the test come into evidence if Wright would agree to let the state's psychologist, Dr. Stanley, testify in rebuttal. Stanley had previously examined Gray to determine if he were competent to stand trial. Wright stated that he had wanted Stanley to testify since Stanley's testimony could be used to establish that Gray was mentally disturbed. Although Wright did not interview Stanley before trial, he did review the reports which Stanley had compiled on Gray.

Gray's present counsel now faults the course pursued by Wright and Heidelberg for two reasons.3 First, he claims that their failure to interview certain witnesses prevented them from making an informed decision on what defenses to offer at the sentencing phase. Second, he claims that having chosen to argue only that Gray was mentally disturbed, Wright and Heidelberg failed to prepare adequately for Gray's defense. He claims that these omissions deprived Gray of his right to the effective assistance of counsel.

In this circuit, the standard for constitutionally effective assistance of counsel is "not errorless counsel, and not counsel judged ineffective by hindsight, but counsel reasonably likely to render and rendering reasonably effective assistance." Herring v. Estelle, 491 F.2d 125, 127 (5th Cir. 1974). The determination of whether a counsel rendered reasonably effective assistance turns in each case on the totality of facts in the entire record. See Washington v. Estelle, 648 F.2d 276 (5th Cir.), cert. denied, --- U.S. ----, 102 S. Ct. 402, 70 L. Ed. 2d 216 (1981); United States v. Gray, 565 F.2d 881 (5th Cir.), cert. denied, 435 U.S. 955, 98 S. Ct. 1587, 55 L. Ed. 2d 807 (1978). Thus, we must consider a counsel's performance in light of "the number, nature, and seriousness of the charges ... the strength of the prosecution's case and the strength and complexity of the defendant's possible defenses." Washington v. Watkins, 655 F.2d 1346, 1357 (5th Cir. 1981), cert. denied, --- U.S. ----, 102 S. Ct. 2021, 71 L.Ed.2d ---- (1982). In this context, we recently recognized that while attorneys are not held to a higher standard in capital cases, the severity of the charge is part of the " 'totality of circumstances in the entire record' that must be considered in the effective assistance calculus." Id.

Gray does not contend that Wright and Heidelberg were not reasonably likely to render effective assistance. He claims instead that they failed to render effective assistance by not fully investigating the use of character witnesses at the sentencing phase of his trial. Gray argues that Wright and Heidelberg should have investigated the people who had worked with him in the Arizona prison and people who had corresponded with him after he was imprisoned in Mississippi for Derissa Scales' murder. Gray established that the people who had worked with him in the Arizona prison would have testified that he had a satisfactory prison record and could function without problems in a prison environment. The people who had corresponded with Gray about religious matters would have testified that Gray had become very spiritual and exhibited a "caring, sharing type of interest in other people." Gray contends that the failure to investigate these witnesses prevented Wright and Heidelberg from making an informed decision about what defenses to present at the sentencing hearing.

We disagree.

We have previously recognized that adequate investigation is a requisite of effective assistance. See Rummel v. Estelle, 590 F.2d 103, 104 (5th Cir. 1979) (per curiam); Gaines v. Hopper, 575 F.2d 1147 (5th Cir. 1978) (per curiam). To establish a constitutional violation, a defendant must show both a failure to investigate adequately and prejudice arising from that failure. See Washington v. Watkins, 655 F.2d 1346, 1362 & n.32 (5th Cir. 1981). The determination of whether a defendant has been prejudiced varies of course with each breach alleged. When, as in this case, a defendant alleges that his counsel's failure to investigate prevented his counsel from making an informed tactical choice, he must show that knowledge of the uninvestigated evidence would have altered his counsel's decision. The fact that an investigation would have turned up admissible evidence is in itself insufficient to show prejudice. Cf. Washington v. Strickland, 673 F.2d 879 (5th Cir. 1982). A defendant must demonstrate that the bases underlying his counsel's tactical choice to pursue or forego a particular course would have been invalidated.

Even if we assume that Wright and Heidelberg failed to make an adequate investigation, there is no indication that knowledge of the proffered testimony would have changed their decision. Their decision not to attempt to develop proof that Gray had some good in his makeup was not affected by what counsel has later discovered could have been shown. The decision was based on the fact that (1) such proof was inconsistent with a mentally disturbed defense, (2) the testimony of the Arizona prison officials would have reinforced the fact that Gray had murdered another young girl and (3) the weight of the testimony of the people who corresponded with Gray was undercut by their limited opportunity to know him. Even though Wright and Heidelberg did not know the exact nature of these witnesses' testimony, they were aware of both the potential and the weaknesses in the testimony.4 Because the testimony proffered by Gray at the evidentiary hearing did not invalidate the reasons behind Wright and Heidelberg's tactical choice not to rely on character evidence, Gray was not prejudiced by their decision not to investigate.

Moreover, the investigation undertaken was adequate. Heidelberg sought out those people, Gray's relatives, friends, and employer, whom a counsel would have reasonably been expected to contact. Heidelberg did not fail to investigate other witnesses whom Gray had identified as crucial. Had he failed to do so, we might reach a different result. See Kemp v. Leggett, 635 F.2d 453 (5th Cir. 1981) (per curiam) (failure to interview the single eyewitness); Gaines v. Hopper, 575 F.2d 1147 (5th Cir. 1978) (per curiam) (failure to interview known eyewitnesses); cf. Baty v. Balkcom, 661 F.2d 391 (5th Cir. 1981) (failure to interview any witnesses). Gray, however, steadfastly maintained that he did not want anyone to testify on his behalf and refused to identify any witnesses.5 While Gray's refusal did not negate Heidelberg's duty to investigate, the scope of that duty was limited by Gray's refusal. See Akridge v. Hopper, 545 F.2d 457 (5th Cir.), cert. denied, 431 U.S. 941, 97 S. Ct. 2657, 53 L. Ed. 2d 260 (1977). Moreover, the witnesses which Gray now claims should have been interviewed were not eyewitnesses to the crime and therefore crucial to the defense of the case. They were character witnesses, the value of whose testimony was assessed on a basis which assumed that they would testify favorably to Gray's character. Given Gray's refusal to identify anyone and given the marginal effectiveness of these character witnesses, the investigation Heidelberg undertook provided Gray with the assistance the constitution requires. See Lovett v. Florida, 627 F.2d 706, 708 (5th Cir. 1980).

Gray purports not to challenge Wright and Heidelberg's choice of trial strategy. However, some of his arguments are in fact directed solely at their decision to rely on a mentally disturbed defense rather than testimony on Gray's good character. We have rejected them. We have consistently recognized that a counsel's decision to pursue one course rather than another is not to be judged by hindsight. See Winfrey v. Maggio, 664 F.2d 550 (5th Cir. 1981). Moreover, the fact that a particular strategy may prove to be unsuccessful does not by itself establish ineffective assistance. See id. at 553. In the case at bar, Wright and Heidelberg could have reasonably decided that testimony on Gray's good character would not be persuasive to a jury aware of the nature of his crime. Moreover, their decision that good character testimony would be inconsistent with a mentally disturbed defense is perfectly reasonable. Although these two types of testimony need not be inconsistent, Wright and Heidelberg could have reasonably determined that in this case a local jury would perceive them to be inconsistent. In short, counsel's tactical decision did not constitute ineffective assistance.

Gray argues alternatively that Wright and Heidelberg failed to prepare adequately for the defense they did present-that he was mentally disturbed. He claims that they should have investigated more thoroughly the Minnesota Multiphasic Personality Inventory which was admitted into evidence at the sentencing hearing and that they should have interviewed Dr. Stanley, the state's witness, before he testified.

The personality test which Wright and Heidelberg relied on was flawed by the fact that Gray had taken the test nine years before trial. Because of its age, the test's methodology was slightly outmoded and its relevance to Gray's mental state at the time of the murder was lessened. Wright, however, was completely aware of the test's weaknesses. He had discussed both the validity of the test and its interpretation with Dr. Bridges, the private psychiatrist who had interviewed Gray. Although Gray claims that Wright should have investigated the test more thoroughly, he has not indicated what further investigation his counsel should have taken or that the investigation undertaken did not allow Wright to anticipate the test's weaknesses.

Gray's main objections appear to be primarily that Wright should not have relied on such stale evidence and that he should have introduced the test through an expert. However, because Gray has not shown that more favorable evidence was available on his mental state, we have no reason to fault Wright's use of the test. With respect to Gray's claim that the test should have been explained by an expert, we agree that the use of an expert might have been preferable. However, the test set out its results clearly and explicitly. The test stated that Gray showed a 95% probability of marked disturbance. Immediately below that finding, the test stated:

INTERPRETATION ...

THE MI, MULTIPHASIC INDEX, REFLECTS AN EMOTIONAL DISORDER OF MARKED SEVERITY. RESPONSES ARE INAPPROPRIATE, UNREALISTIC AND SELF-DEFEATING.

TEST EVIDENCE STRONGLY SUPPORTS THE PICTURE OF A SEVERELY DISTURBED PERSON.

These statements clearly supported the defense theory and were readily comprehensible by a layman.6 Although it might have been better practice for the defense to have called an expert witness to explain the test to the jury, the right to effective assistance does not entitle a defendant to the best defense possible, but only a reasonably effective defense. Because the test's results were clear enough to have been comprehended by the jury, presentation of the report to the jury was sufficient.

Gray, finally, faults his counsel for not interviewing Stanley, the state's rebuttal witness, prior to his testimony. Gray contends that Wright's familiarity with Stanley's reports was insufficient preparation for trial. We disagree. We have previously recognized the "critical interrelation between expert psychiatric assistance and minimally effective representation of counsel." United States v. Fessel, 531 F.2d 1275, 1279 (5th Cir. 1976). Thus, we have held that the failure to seek psychiatric testimony may constitute ineffective assistance. See id. (failure to move for a court appointed psychiatrist); United States v. Edwards, 488 F.2d 1154, 1164 (5th Cir. 1974) (failure to pursue a successful motion for a court appointed psychiatrist); Greer v. Beto, 379 F.2d 923, 925 (5th Cir. 1967) (failure to offer any evidence on defendant's mental condition). The case at bar does not concern a failure to seek any psychiatric testimony; instead, it concerns the extent of preparation required prior to presenting a psychiatrically related defense. We need not, however, explore the perimeters of this issue since we find that Gray was not prejudiced by his counsel's failure to interview Stanley.

Stanley's testimony on direct examination falls into two parts. In the first part, Stanley presented the jury with an analysis of Gray's personality based on his earlier observations of Gray. This part of Stanley's testimony repeats almost verbatim a report which Stanley had prepared on February 1, 1978, and which Wright had previously studied. Because Wright could not have gained a clearer understanding of this part of Stanley's testimony if he had interviewed Stanley extensively, Gray was not prejudiced by the fact that Wright had only read Stanley's report.

The second part of Stanley's testimony concerns the Minnesota Multiphasic Personality Inventory and Stanley's interpretation of it. Stanley had not seen the results of the test as administered to Gray. This had been done by the Arizona prison officials.7 Thus, Wright could have obtained Stanley's views on Gray's test only by allowing Stanley, the state's rebuttal witness, to examine the test before trial. We refuse to fault Wright for not giving the state an extended chance to study the defense's evidence.

Although Gray alleges that he was prejudiced in two respects, neither withstands scutiny. First, Gray claims that if Wright had interviewed Stanley he would have been aware that Stanley would discount the validity of the personality test. Gray notes Stanley's testimony that, because of the test's age, its results should be taken "with a grain of salt." Wright, however, was fully aware, because of his discussions with another psychiatrist, that the test results were dated. Gray has not shown that Wright's preparation prevented him in any way from anticipating Stanley's objections.

Gray also claims that if Wright had interviewed Stanley, he would have been aware of Stanley's opinion that Gray was likely to repeat his crime. An examination of the testimony at the sentencing phase of the trial, however, reveals that it was the prosecutor, and not Stanley, who put this gloss on Stanley's analysis of Gray.

The prospect that Gray presented a future danger was raised three times during Stanley's direct testimony. Each time, it was the prosecutor who interjected this idea into Stanley's testimony. Stanley had testified extensively about Gray without concluding that Gray would be likely to repeat his crime. The prosecutor then raised the issue for the first time:

Q. Now, in regard to this hostile personality, Doctor, if in the future, he gets mad at somebody, will he repeat the act of murder?

A. The best statistics psychologically, where psychological research shows, is that the best prediction of future behavior is to look at past behavior. And, I would suggest that you know, looking at past behavior worth the personality profile he shows, I would have to answer yes.-That the chances are better than average, very good, as a matter of fact.

Later, after questioning Stanley about the validity of the personality test, the prosecutor asked:

Q. All right, Sir. One question. I ask you this. One phrase here, probability-.

A. Of marked disturbance?

Q. Marked disturbance, or ability to repeat the crime. I believe they had it marked, what?

A. 95%

Q. And does that match your findings you just said about the opportunity, or the chances of the defendant repeating his crime at this time?

A. Well, as a scientist, I hate to put behavioral things in percentages, but I can't argue with it.

The personality test did not indicate, however, that Gray exhibited a 95% probability of "(m)arked disturbance, or ability to repeat the crime." It stated only that he showed a probability of marked disturbance. While Stanley quoted the test accurately, it was the prosecutor who interjected the idea of future dangerousness into the test results. Finally, after Stanley had discussed the contents of the personality test without any reference to future dangerousness, the prosecutor concluded:

Q. All right. I wasn't asking you in terms of percentages, but that report does agree that the-with your statement a few minutes ago, that the defendant's hostility would lead him to (be) a repeater in this type of crime?

A. Oh, yes, I think so. It would all fit together.

Although both the personality test and Stanley's observations of Gray may have been compatible with the prosecutor's suggestion that Gray was likely to repeat his crime, neither the personality test nor Stanley had advanced that conclusion independently. The idea of future dangerousness was interjected solely by the prosecutor. There is no indication that a pretrial interview of Stanley would have revealed that this issue would arise at trial.8 While it would have been better practice for Wright to have interviewed Stanley, we cannot say that, on these facts, Gray was prejudiced by the failure to do so.9

During voir dire, seven prospective jurors stated that they had scruples against capital punishment. These jurors were then questioned to determine the extent of their opposition to the death penalty. Although the attorneys' questions and the jurors' responses were transcribed, the subsequent challenges for cause and preemptory challenges were not recorded.

Gray did not contend on direct appeal to the state supreme court, nor does he contend now,10 that any of the jurors were improperly excluded under Witherspoon v. Illinois, 391 U.S. 510, 88 S. Ct. 1770, 20 L. Ed. 2d 776 (1968). Instead, he argues that Mississippi's death penalty statute requires, as a constitutionally mandated procedure, that the state supreme court review each death sentence to determine if it were arbitrarily imposed. See Miss.Code Ann. § 99-19-105(3). Gray contends that because the inadequate state records prevented the state supreme court from determining whether any Witherspoon violations occurred, the court was precluded from engaging in its constitutionally required review.

The premise of Gray's claim is that the Mississippi Supreme Court was constitutionally required to scan the record of his trial to determine if any jurors were improperly excluded under Witherspoon. Gray contends that this is so because a sentence imposed by a jury organized to return the death penalty poses the same risk of arbitrariness that the Court identified as unconstitutional in Furman v. Georgia, 408 U.S. 238, 92 S. Ct. 2726, 33 L. Ed. 2d 346 (1972).

In Stephens v. Zant, 631 F.2d 397, 402-03 (5th Cir. 1980),11 we recognized that when the record presented to a state appellate court is so deficient that it would be impossible for the court to perform its constitutionally required review, the sentence cannot stand. However, Stephens is only applicable if the omitted portion of the record concerns a matter which the state court is required to review. Because we find that the constitutionally mandated review of death sentences does not require a state supreme court to search for possible Witherspoon violations, the trial court's failure to develop a record on this point did not affect the validity of Gray's sentence.

In Furman v. Georgia, supra, three of the Justices reasoned that because juries had previously been given complete discretion to return the death penalty, the penalty had been arbitrarily imposed in violation of the eighth amendment. See 408 U.S. at 240, 92 S.Ct. at 2727 (Douglas, J. concurring); 408 U.S. at 306, 92 S.Ct. at 2760 (Stewart, J. concurring); 408 U.S. at 310, 92 S.Ct. at 2763 (White, J. concurring). Justice Stewart found:

These death sentences are cruel and unusual in the same way that being struck by lightning is cruel and unusual. For, of all the people convicted of rapes and murders in 1967 and 1968, many just as reprehensible as these, the petitioners are among a capriciously selected random handful upon whom the sentence of death has in fact been imposed. My concurring Brothers have demonstrated that, if any basis can be discerned for the selection of these few to be sentenced to die, it is the constitutionally impermissible basis of race.

498 U.S. at 310-11, 92 S. Ct. at 2762 (footnotes omitted).

Four years later in Gregg v. Georgia, 428 U.S. 153, 96 S. Ct. 2909, 49 L. Ed. 2d 859 (1976), the Supreme Court upheld the imposition of a death penalty because the procedures adopted by the state of Georgia were sufficient to remedy the randomness which had been condemned in Furman. One of the procedures which the Court noted with approval was the fact that the state supreme court automatically reviewed each conviction to determine whether it had been imposed under the influence of passion or prejudice and whether the sentence of death were excessive or disproportionate to the penalty imposed in similar cases. This review helped ensure that the death sentence "would be imposed in a more consistent and rational manner and ... that there would be a 'meaningful basis for distinguishing the ... cases in which it is imposed ... from the many cases in which it is not.' " Lockett v. Ohio, 438 U.S. 586, 601, 98 S. Ct. 2954, 2963, 57 L. Ed. 2d 973 (1978). A state supreme court is thus required by Furman and Gregg to review the record for any Witherspoon violations only if such a search would promote the consistent imposition of the death penalty. We find that it would not.

Witherspoon v. Illinois found that the exclusion of jurors who merely had scruples against the death penalty violated a defendant's sixth amendment right to an impartial jury. The decision was premised on the grounds that while a state has no valid interest in excluding a juror who states that he can follow the law, a defendant is seriously prejudiced by the exclusion of all jurors opposed to capital punishment. See Adams v. Texas, 448 U.S. 38, 100 S. Ct. 2521, 2525, 65 L. Ed. 2d 581 (1980). Witherspoon was thus concerned with the question of who decides whether the death penalty should be imposed. It sought to ensure that the jury that determines the death penalty be drawn from a fair cross section of the community.

Inclusion of a fair cross section of the community, however, does not advance the interest identified in Furman and Gregg, the consistent and reasoned imposition of the death penalty. If anything, expanding the ideological base from which jurors are drawn, as Witherspoon does, leads to greater diversity within and among juries and less uniform results. Thus, we find that a state appellate court may satisfy the constitutional objections noted in Furman and Gregg, without making an independent assessment of the jury composition under Witherspoon. Cf. Spinkellink v. Wainwright, 578 F.2d 582, 598-99 (5th Cir. 1978), cert. denied, 440 U.S. 976, 99 S. Ct. 1548, 59 L. Ed. 2d 796 (1979).

Moreover, we note that when this issue was raised in the district court, the court was able to reconstruct which jurors had been struck for cause and which had been struck preemptorily. The court based its determination on testimony by the counsel for both parties and on a jury list which Wright had marked during voir dire. Although Gray attacks the validity of this reconstruction, he does not claim that any of the jurors were improperly excluded under Witherspoon. Thus, even if the state court were required to search out any Witherspoon violations, Gray was not prejudiced by its failure to do so.12 See Stephens v. Zant, 631 F.2d 397, 404 (5th Cir. 1980).

Gray objects to the district court's reconstruction of the strikes made at voir dire because he claims that the marked jury list was improperly admitted and because the district court's findings were clearly erroneous. Gray objected to the admission of the marked jury list on the ground that it was unreliable. However, because the list was marked as the objections were made, there is no indication that its accuracy was marred by a lapse of time. Moreover, because Wright marked the list to keep track of preemptory strikes and strikes for cause for his own use at trial, there is every indication that the list was carefully and accurately marked. We cannot say that the district court abused its discretion in overruling Gray's objection to the list's admission. See Page v. Barko Hydraulics, 673 F.2d 134 (5th Cir. 1982), Richardson v. McClung, 559 F.2d 395, 396 (5th Cir. 1977) (per curiam).13

Gray also argues that the district court's reconstruction of the strikes was clearly erroneous. The basis of his claim is that the district court judge mistakenly attributed the testimony given by the state prosecutor to the state judge. Although the district court did refer to testimony given by the state prosecutor as the state judge's testimony, the reference is somewhat unclear since the state prosecutor had been appointed a state judge prior to the district court's evidentiary hearing. The district court continually referred to both men as "judge" throughout the hearing. However, even if we disregard the allegedly misidentified testimony, there is still sufficient evidence to convince us that the district court's reconstruction was not clearly erroneous. See Van Ooteghem v. Gray, 654 F.2d 304 (5th Cir. 1981) (en banc).

Gray argues that the district court failed to apply the proper legal standards, failed to make adequate factual findings and improperly ruled on the admission of evidence.14

Gray contends that, in evaluating his ineffective assistance claims, the district court applied the wrong law. It considered the course of Wright and Heidelberg's conduct over the whole trial instead of considering whether they failed to conduct an adequate investigation in preparation for the sentencing phase. Gray argues that the district court should be reversed for applying the wrong legal standards. This point in Gray's argument is unclear. In any event, on this appeal Gray has enumerated the alleged legal errors of the district court, we have considered them fully and affirm the result reached by the district court.

Gray also argues that because the district court misperceived the applicable law, it failed to make the factual findings necessary for appellate review. This claim raises a slightly greater problem since the district court's resolution of factual issues facilitates appellate review. See Hart v. United States, 565 F.2d 360, 362 (5th Cir. 1978) (per curiam). However, an appellate court may supplement the district court's opinion when the record permits us to make a complete and fair resolution of the issues presented. See Pullman-Standard v. Swint, --- U.S. ----, ----, 102 S. Ct. 1781, 1792, 71 L.Ed.2d ---- (1982); Matter of Legel, Braswell Government Securities Corp., 648 F.2d 321, 326 n.8 (5th Cir. 1981); Perry v. Texas, 456 F.2d 879, 881 (5th Cir.), cert. denied, 409 U.S. 916, 93 S. Ct. 248, 34 L. Ed. 2d 178 (1972).

Although Gray claims that the testimony in the record is too conflicting to allow a fair resolution of the issues, he only notes one instance of that conflict. He claims that Wright's and Heidelberg's testimony differed on a critical issue, the choice of defense tactics. Gray notes that Heidelberg investigated the use of character witnesses before rejecting that line of defense while Wright appears to have rejected this defense from the start of the case. Gray does not argue that there was any dispute about the course which each counsel pursued. Instead, he argues that their inconsistent courses cast doubt on the basis of their decision not to rely on character witnesses. There is no difficulty, however, in resolving the testimony of these two attorneys. Wright adopted a general defense strategy after reviewing the facts of the case. Heidelberg did more of the investigative or background work. This pattern is perfectly consistent with the allocation of responsibility between co-counsel. Heidelberg's investigation simply did not convince him that Wright's initial strategy should be altered and both counsel ultimately adhered to that course. There is no indication that their testimony was inconsistent or that Gray did not receive the benefit of Heidelberg's investigation.

Gray finally faults the district court for twice preventing his counsel from questioning Wright and for refusing his offers of proof. Because we find that the questions asked were irrelevant to the claims Gray presented, we uphold the court's rulings. We note, however, that the court clearly erred in refusing to allow Gray to make an offer of proof. See Fed.R.Evid. 103(a)(2).

An offer of proof is required to perfect an objection to the exclusion of evidence. See Fed.R.Evid. 103(a)(2). It allows an appellate court to reevaluate the trial court's decision in light of the actual evidence to be offered and to determine whether any prejudice resulted from the exclusion of evidence. See Fortunato v. Ford Motor Co., 464 F.2d 962 (2d Cir. 1972). An offer of proof is not necessary, however, when the substance of the evidence is "apparent from the context within which the questions were asked." Fed.R.Evid. 103(a)(2). Although this later exception is normally relied on by litigants who have failed to make an offer of proof, we rely on it now as providing an adequate basis for reviewing the district court's evidentiary rulings. Because the substance of the evidence Gray sought to elicit was apparent from the context of his questions, we are able to discern whether the questions were in fact relevant to his claims.

Gray's counsel at the evidentiary hearing questioned Wright about a computer "hit." It was developed that when Gray had been picked up by the Pascagoula police, they ran a computer check on him. The computer erroneously showed a "hit"; it indicated that Gray was wanted by the Wisconsin police. The Pascagoula police held Gray overnight, during which time he confessed to the police.

After questioning Wright about the circumstances of Gray's arrest, Gray's counsel began to ask whether Wright had checked to see if the police had acted in good faith in holding Gray overnight. Wright replied that he had not inquired into the police's good faith since their treatment of Gray accorded with what he understood to be normal police practice. When Gray's counsel continued to probe into why Wright had not investigated further, the district court sustained the state's objection to this line of questioning.

Gray argues that because these questions were relevant to his claims of ineffective assistance, the district court's ruling was incorrect. Gray, however, did not allege in his petition for habeas corpus that Wright had been ineffective for failing to investigate his detention. The extent of Wright's investigation into Gray's detention was thus irrelevant to the effectiveness of counsel claims raised in the habeas petition.

Alternatively, Gray claims that a complete development of the facts surrounding his detention was necessary to determine whether his subsequent confessions had been properly admitted at trial. Although information on this point would have been relevant to the voluntariness claims raised in Gray's habeas petition, the questions asked by Gray's counsel could not lead to a development of such information. The questions asked Wright went to the extent of his investigation and why Wright had chosen not to inquire into the officers' good faith in holding Gray. Since Wright stated he made no such investigation in view of his past experience, further questions addressed to Wright could not have illuminated the factual issues surrounding Gray's detention. Indeed, Gray had already raised these claims in state court and the testimony developed there provided an adequate record of the events that transpired. If Gray's counsel needed to develop any further information about the officers' motives, he should have questioned the officers themselves since Wright had already provided the court with all the information he had on this point. The district court did not abuse its discretion in sustaining the state's objection.

Gray also faults the district court for preventing his counsel from inquiring into Wright's knowledge of Witherspoon v. Illinois. On direct examination, he had questioned, without any objection, whether Wright was familiar with Witherspoon. On redirect, he sought to question Wright about his interpretation of Witherspoon. It was this later line of questioning to which the district court objected.

Wright's understanding of Witherspoon, however, was irrelevant since Gray had not claimed in his habeas petition that Wright was ineffective because he lacked knowledge of Witherspoon or failed to object to the exclusion of jurors. Gray argues that Wright's knowledge of Witherspoon was relevant because it bore on his ability to recall whether the voir dire challenges were valid. This misstates the issue. The issue was not whether certain challenges were valid15 or whether Wright was aware of their validity. Because the district court sought to reconstruct the prosecutor's strikes, the only relevant issue was which jurors were struck for cause under Witherspoon and which were struck preemptorily. Wright's precise understanding of Witherspoon would not have affected his ability to recall the prosecutor's stated reasons for striking a particular juror.

Gray's petition for habeas corpus contained twenty-two separate claims. At the evidentiary hearing held by the district court, Gray presented evidence on two of the issues, ineffective assistance and the state trial court's failure to record Witherspoon objections. In his Post-Hearing Memorandum, Gray presented arguments on the two claims which were pursued at the evidentiary hearing and on six other claims which he had raised in his habeas petition. Thus, Gray neither presented evidence nor any further legal arguments on fourteen of the twenty-two claims raised in his habeas petition.

In its Memorandum Opinion, the district court only discussed the two claims Gray pressed at the evidentiary hearing. Although it alluded generally to the constitutionality of Mississippi's capital punishment statute, the district court never discussed the particular arguments raised by Gray on that point. In the concluding paragraph of its opinion, the district court stated, "(I)t is the considered judgment of this court ... that every one of (Gray's) constitutional rights (has) been respected in every phase of this entire case (and) that the petition of Jimmy Lee Gray is without merit in its entirety ...."

Gray does not argue that the district court's decision on the merits of his claims is incorrect. He argues that the summary manner in which it dismissed his claims was improper. The claims raised in Gray's habeas petition may be divided into three categories. First, there are the two claims which Gray admits were considered and discussed by the district court. Second, there are six additional claims which Gray raised in his habeas petition and reasserted in his Post-Hearing Memorandum which the district court may have denied in the sentence quoted above. Third, there are fourteen remaining claims which Gray raised in his habeas petition but neither pursued nor presented again to the district court.

Because Gray does not contend that the district court failed to offer a sufficient explanation of its ruling on the first category of his claims, the validity of the manner in which it treated those claims is not at issue. With respect to the second category of claims, those raised in the habeas petition and argued in the Post-Hearing Memorandum, we agree with Gray that the district court's summary dismissal was inadequate. In Washington v. Strickland, 673 F.2d 879 (5th Cir. 1982), we recently recognized that a district court is required to offer some explanation of its ruling on each ground of relief raised in a petition for habeas corpus. Although Strickland recognized that in many instances only a brief explanation is required, the district court failed to offer any explanation in this case. We thus agree that the court erred in the manner in which it treated this second category of claims.

With respect to the third category of claims, those which were raised in Gray's habeas petition but neither pursued nor pressed before the district court, we find that the district court did not need to address those claims. See Harper v. Merckle, 638 F.2d 848, 855 (5th Cir. 1981); Corenswet, Inc. v. Amana Refrigeration, 594 F.2d 129, 139 (5th Cir. 1979); accord Bast v. Department of Justice, 665 F.2d 1251 (D.C.Cir.1981); Stanspec Corp. v. Jelco, Inc., 464 F.2d 1184, 1187 (10th Cir. 1972); King v. Stevenson, 445 F.2d 565, 570-71 (7th Cir. 1971); Ark-Tenn Distributing Corp. v. Breidt, 209 F.2d 359, 361 (3d Cir. 1954).

Sound reasons exist for the application of this rule. Our adversary system of justice depends in large part on "that concrete adverseness which sharpens the presentation of issues" and illuminates the disposition of difficult claims. Baker v. Carr, 369 U.S. 186, 204, 82 S. Ct. 691, 703, 7 L. Ed. 2d 663 (1962). By raising a claim in the pleadings but failing to pursue it any further, a litigant fails to apprise the court fully of the legal and factual underpinnings of his claims. Instead of having issues clarified during the course of a trial by hearings, arguments or memoranda, the district court is left to divine the legal and factual bases of a claim on its own. This deprives the court of the benefits the adversary system was designed to confer. Moreover, when a party fails to press his claims, a district court is left in doubt as to whether any live controversy remains on that point. While a party must of course raise his claims in the pleadings, the pleadings only mark the beginning of a lawsuit. Having plead his claims correctly, a litigant may not abandon any further pursuit of his claims and then complain because the district court failed to address his contentions.

The requirement that a litigant must pursue as well as raise his claims is particularly appropriate in light of our holding in Strickland. The burden which Strickland places on the district court to address each claim raised in a habeas petition puts a concomitant duty on counsel to give the court a full development of every claim he does not waive. We note also that raising multiple undeveloped claims gives rise to the possibility of using pleadings for delay. Without attributing any such motive to Gray's counsel, we note that such a tactic may be appealing in death penalty cases where delay may be a defendant's main hope.

By requiring a party to pursue his claims actively, the issues are clarified and can be resolved by the district court with greater skill and speed. Adequate development of the issues also allows the district court to identify frivolous claims more easily and remove them from consideration. To the extent that attorneys are discouraged from filing frivolous claims, the habeas petitioner himself is aided since both the court and the petitioner's attorney are able to focus on those claims which merit the most attention. We thus find that those claims which Gray raised in his petition for habeas corpus but never pressed or presented again to the district court were abandoned.

The district court's summary dismissal of Gray's claims was therefore improper only with respect to the second category of claims, those claims which were raised in the petition for habeas corpus and presented to the district court in Gray's Post-Hearing Memorandum. In Strickland, supra, we remanded the petitioner's claims which had been summarily dismissed since an explanation by the district court of its rulings "will enable this court to give meaningful review to the district court's factual and legal conclusions." 673 F.2d at 907. However, the issues posed by the remaining six claims raise questions only of law which do not require fact development to resolve. Because we are in as good a position as the district court to answer these inquiries, the interests of justice indicate we should pass on them here and now rather than remand the cause and delay their ultimate disposition. 28 U.S.C. § 2106.

1. The Constitutionality of the Mississippi State Statute

Gray's attack on Mississippi's capital punishment statute is itself composed of nine separate arguments. First, Gray claims that Mississippi allows a jury to impose the death sentence when there has been no finding of intent. See Miss.Code Ann. § 97-3-19(2). Gray argues that because there was no finding either at the guilt or sentencing phase of his trial that he purposefully or intentionally acted to kill Derissa Scales, death is an excessive penalty in violation of the eighth amendment. Although this claim might pose a difficult question of constitutional law, we need not resolve it in this case since the assumption underlying the argument is missing in Gray's case.

The jury found beyond a reasonable doubt that Gray had purposefully taken Derissa Scales' life to avoid capture. In reviewing Gray's conviction, the Mississippi Supreme Court stated:

It is now contended in Gray's behalf that a "very serious" dispute exists as to whether Gray had the deliberate design to effect the death of the child. However, the jury had before it all of the evidence given during the trial of the guilt phase.... Upon the entire record, the jury was justified in finding beyond a reasonable doubt that the child had been killed by Gray intentionally and maliciously for the purpose of silencing her outcries or preventing the report by her of acts of molestation.

Gray v. State, 375 So. 2d at 1004. The state court's determination of what the jury was permitted to find and obviously did find is presumptively correct unless rebutted, which Gray has not attempted to do. See Sumner v. Mata, 449 U.S. 539, 101 S. Ct. 764, 66 L. Ed. 2d 722 (1981).

Second, Gray argues that the Mississippi statute is unconstitutional because it is underinclusive. He points out that the Mississippi statute allows a jury to impose the death penalty solely because a murder is committed during the course of certain specified felonies, but it does not authorize the death penalty for a simple murder, no matter how atrocious. Gray claims that this distinction violates the eighth amendment, due process and equal protection. We disagree.

With respect to Gray's eighth amendment claim, it is the imposition of a penalty which triggers eighth amendment scrutiny. See Gregg v. Georgia, 428 U.S. at 199, 96 S.Ct. at 2937. A particular punishment may violate the eighth amendment if the punishment itself is particularly "barbaric" or if it is excessive in relation to the crime committed. See Coker v. Georgia, 433 U.S. 584, 592, 97 S. Ct. 2861, 2866, 53 L. Ed. 2d 982 (1977). In the context of capital punishment, the Court has recognized that the eighth amendment requires regularity in the imposition of the death penalty. See Gregg, supra. Thus, Mississippi's death penalty statute violates the eighth amendment if it relies on barbaric or inhumane methods, is excessive in relation to the crime committed or fails to ensure that the death penalty is imposed in a reasoned manner within a particular class of cases. The fact that Mississippi chooses not to impose the death penalty for the crime of simple murder committed in an atrocious manner simply does not implicate the eighth amendment.

An allegedly improper classification scheme may, however, violate either the due process or equal protection clauses. In reviewing equal protection and due process claims, our initial inquiry is to identify the proper level of scrutiny. See Arceneaux v. Treen, 671 F.2d 128 (5th Cir. 1982). Traditionally, courts have accorded legislative classifications of criminal activity wide latitude. See Patterson v. New York, 432 U.S. 197, 201, 97 S. Ct. 2319, 2322, 53 L. Ed. 2d 281 (1977); Lambert v. California, 355 U.S. 225, 228, 78 S. Ct. 240, 242, 2 L. Ed. 2d 228 (1957). Legislatures may recognize degrees of evil and may seek to deal with undesirable conduct one step at a time. See Skinner v. Oklahoma, 316 U.S. 535, 540, 62 S. Ct. 1110, 1113, 86 L. Ed. 1655 (1942). This deference has been accorded the legislature's judgment because it is particularly within a legislature's competence both to identify conduct which offends the community and to determine appropriate punishment. Because Gray does not claim that the Mississippi statute is impermissibly directed at a suspect class, the only factor which could arguably justify a departure from the wide latitude traditionally accorded a legislature's judgment is the presence of the death penalty. See Skinner v. Oklahoma, 316 U.S. at 541-42, 62 S.Ct. at 1113-14. Gregg v. Georgia, however, teaches that this factor does not require a higher level of scrutiny.

In Gregg, the Supreme Court stressed, in the context of the death penalty, the deference which a court should give a legislature's judgment. See 428 U.S. at 174-76, 98 S.Ct. at 2925-26. The Court reaffirmed the proposition that it is particularly within the province of the legislature to respond to the moral values of the community and to determine the appropriate degree of punishment, including the death penalty, for certain classes of offenses. Because of the complexity of determining the need for the death penalty, the Court found that the decision to authorize capital punishment for some classes of crimes was one that was best left to the legislature unless "clearly wrong." See id. 428 U.S. at 187-88, 96 S. Ct. at 2931. Although the issue in Gregg was whether the legislature's decision to impose the death penalty violated the eighth amendment because of the severity of the punishment, the degree of deference which the Court's "clearly wrong" test accorded the legislative judgment convinces us that the due process and equal protection clauses do not require a higher level of scrutiny for legislative classifications that may result in the death penalty. Thus, Gray's claims are to be assessed under a rational basis test.

The basis of Gray's claim under both equal protection and due process is that there is no rational basis for imposing the death penalty on people who commit murder during the course of a felony but not imposing it on people who commit especially atrocious simple murders. However, Mississippi could have rationally decided that felony murders pose a problem different from atrocious simple murders and could have sought to cure the felony murder problem first. Alternatively, the legislature could have decided that the death penalty would be more effective in deterring felony murders since an experienced felon is more likely to assess the consequences of his acts. Conversely, it could have rationally determined that the death penalty might not effectively deter atrocious simple murders since such people are likely as a group to act on passion or impulse and thus be unmindful of the consequences of their crime. In short, the legislature could have rationally decided that the one class of murders either presented a different problem from the other or that the death penalty would be a more effective deterent to felony murders than atrocious simple murders.

Third, Gray argues that under the Mississippi statute a defendant convicted of a felony murder will "automatically" be eligible for the death sentence since one of the aggravating factors, commission of a murder in the course of a felony, will have already been proved. He argues that if a defendant is convicted of a nonfelony murder, the state is still required to prove aggravating circumstances at the sentencing phase. He claims that because this disparity leads to the possibility that defendants convicted of felony murder will receive a disproportionately higher number of death sentences, it violates the eighth amendment. See State v. Cherry, 298 N.C. 86, 257 S.E.2d 551, 568 (1979).

This argument assumes that to meet constitutional muster the death penalty must be imposed proportionately among the aggravating factors established by the state. This argument, however, misperceives the eighth amendment's requirements. The constitutional defect identified in Furman v. Georgia was the arbitrary imposition of the death penalty among similarly situated defendants. Because each aggravating factor normally identifies a particular class of defendants, Furman and Gregg only require that the death penalty be consistently imposed within that class. They do not require that the death penalty be equally distributed among discrete classes of defendants.

Fourth, Gray complains that he was sentenced to death on the basis of an unconstitutionally vague aggravating factor, that the offense was especially heinous, atrocious or cruel. See Miss.Code Ann. § 99-19-101(5)(h). In Gregg v. Georgia, however, the Supreme Court upheld a similar aggravating factor as facially valid. See 428 U.S. at 201, 96 S.Ct. at 2938. The Court noted that while the factor could be construed broadly to encompass almost any murder, there was no reason to assume that the state supreme court would adopt such a construction. In Godfrey v. Georgia, 446 U.S. 420, 100 S. Ct. 1759, 64 L. Ed. 2d 398 (1980), the Court later reversed the imposition of a death sentence based solely on this factor. The Court reasoned that this factor had been applied to include an impermissibly large number of cases. It noted that because Godfrey's "crime cannot be said to have reflected a consciousness materially more 'depraved' than that of any person guilty of murder," the presence on this factor failed to provide any "principled way to distinguish (Godfrey's) case in which the death penalty was imposed, from the many cases in which it was not." Id. 100 S.Ct. at 1767. While Gregg indicates that Mississippi's aggravating factor is facially valid, Godfrey shows that its application may render it unconstitutionally vague.

In the case at bar, Gray does not argue that Mississippi has adopted an open-ended construction of this factor. Indeed, Mississippi has consistently applied a limiting construction to this factor. See Coleman v. State, 378 So. 2d 640, 648 (Miss.1979). In Coleman, the Mississippi Supreme Court reiterated that this factor applied only to "the conscienceless or pitiless crime which is unnecessarily torturous to the victim." Moreover, Gray does not contend that the application of this factor to his case was improper. It would indeed be difficult to argue that abducting a three-year-old child, sexually molesting her, entombing her half-alive in a car trunk and then tossing her out to drown in a river is not especially heinous, atrocious, or cruel. Because Gray has not shown that Mississippi has either adopted an open-ended construction of this factor or applied it in an open-ended manner and because the Court has applied it properly to date, we reject his claim.

Fifth, Gray argues that once an aggravating factor is established by the state, the burden of proof is impermissibly shifted to the defendant. Gray contends that to avoid the death penalty, the defendant must show that the mitigating factors outweigh the aggravating factors. Gray argues that this allocation of proof creates an unconstitutional presumption of death.

Gray's argument, however, is based on a misapprehension of the applicable state law. Section 99-19-103 of the Mississippi Code provides that the jury may not impose the death penalty unless it finds that at least one statutory aggravating circumstance exists and that the aggravating circumstances are not outweighed by the mitigating circumstances. See Miss.Code Ann. § 99-19-103. The burden of proving an aggravating circumstance exists as part of the prosecution's general bundle. Gray v. State, 351 So. 2d 1342, 1345 (Miss.1977) (appeal from Gray's first trial). The defendant is permitted but not required to offer any relevant mitigating circumstances not encompassed in the prosecutor's presentation. No burden of proof or persuasion is assigned, just as the defendant's option to develop proof of alibi, good character, or, in this case, mental instability does not operate to shift the burden of proof, so the choice to present any number or kind of mitigating circumstances does not. Moreover, even if the jury finds that the aggravating circumstances outweigh the mitigating circumstances, it is not required to impose the death penalty. It may still sentence a defendant to life imprisonment. See Coleman v. State, 378 So. 2d 640, 647 (Miss.1979). Thus, the statutory procedures act only to require the prosecution to build the aggravating circumstance bridge the jury must cross to be entitled to impose the death penalty. Cf. Stephens v. Zant, --- U.S. ----, 102 S. Ct. 1856, 71 L.Ed.2d ---- (1982). They do not limit its ability to impose a life sentence. Every mandatory element of proof is assigned to the prosecution. Neither the burden of production nor the burden of proof ever shifts to the defendant.16

Sixth, Gray argues that by authorizing the introduction of any evidence that is relevant to the sentencing process, the state introduces a substantial possibility that the death sentence will be imposed for reasons other than those enumerated in the eight aggravating circumstances. See Miss.Code Ann. § 99-19-101(1). Gray does not argue that any evidence was improperly admitted at his trial because of this provision. Instead, he claims that the sentencing procedure is invalid on its face since it fails to channel the jury's discretion. Gray's claim is frivolous. The Mississippi Supreme Court has construed this provision to allow the admission of only that evidence which is relevant to the enumerated aggravating factors. See Coleman v. State, 378 So. 2d 640, 648 (Miss.1978). Because Gray has not shown that the statute is unconstitutional on its face or as applied, we reject his claim.

Seventh, Gray contends that Mississippi violates the equal protection and due process clauses by denying capital defendants the right to a presentence report which is afforded to all other felons in the state. Gray does not claim that he was denied access to a presentence report which the jury relied on in imposing the death sentence. See Gardner v. Florida, 430 U.S. 349, 97 S. Ct. 1197, 51 L. Ed. 2d 393 (1977). Instead, his claim turns on the fact that the state does not prepare a presentence report for capital defendants.

Because capital defendants are not a suspect class and because the right to a presentence report is not fundamental, Mississippi's practice is invalid only if it fails the minimum rationality standard. See Arceneaux v. Treen, supra. Mississippi might rationally conclude that a capital defendant's ability to present all relevant mitigating evidence at the sentencing phase of his trial would remove any need for a presentence report. We reject his claim.

Eighth, Gray argues that Mississippi fails to provide any standard for the jury in weighing the aggravating factors against the mitigating factors and fails to provide that the prosecution bears the burden of production and proof on these issues. The first part of Gray's claim is frivolous. Mississippi prevents a jury from imposing the death penalty if the mitigating factors outweigh the aggravating factors. Mississippi also allows every jury that unique discretion only juries possess in our system of law to decline to impose the death penalty though a hundred other panels might properly find it mandated. Moreover, its statutes comport completely with those found facially valid by the Supreme Court. See Gregg v. Georgia, 428 U.S. at 193, 203, 96 S. Ct. at 2934-35, 2939.

The second part of Gray's claim, that the Mississippi statute does not provide that the prosecution bears the burden of proof, also lacks merit. As noted, the Mississippi Supreme Court has construed its capital punishment statute to place the burden of proving any aggravating circumstances on the prosecution. See Gray v. State, 351 So. 2d 1342, 1345 supra. This is the critical burden since the jury's failure to find an aggravating circumstance precludes it from imposing the death penalty. See Miss.Code Ann. § 99-19-103.

Ninth, Gray argues that the jury should be required to find beyond a reasonable doubt that the aggravating circumstances outweigh the mitigating circumstances before it can impose the death penalty. This states the rule too strongly. While Mississippi requires that jurors find the existence of each aggravating circumstance beyond a reasonable doubt, the jury may return the death penalty if the aggravating circumstances are not outweighed by the mitigating circumstances. This allocation of proof accords with the capital sentencing procedures which the Court has upheld as facially valid. See Gregg v. Georgia, 428 U.S. 153, 96 S. Ct. 2909, 49 L. Ed. 2d 859 (1976); Proffitt v. Florida, 428 U.S. 242, 96 S. Ct. 2960, 49 L. Ed. 2d 913 (1976); Jurek v. Texas, 428 U.S. 262, 96 S. Ct. 2950, 49 L. Ed. 2d 929 (1976). We therefore reject Gray's claim.

Gray claims that the inculpatory statements which were introduced at trial were obtained by threatening him and depriving him of sleep. The determination of whether Gray's confession was voluntary presents a mixed question of law and fact and may be assessed independently by this court.17 See Sumner v. Mata, --- U.S. ----, ----, 102 S. Ct. 1303, 1305, 71 L. Ed. 2d 480 (1982) (per curiam). However, in assessing the voluntariness of his confession we presume that the underlying facts found by the state court are correct unless shown otherwise. See id.; Sumner v. Mata, 449 U.S. 539, 101 S. Ct. 764, 66 L. Ed. 2d 722 (1981).

Gray's claim that his confession resulted from threats obviously turns in the first instance on the factual issue of whether any threats occurred. At the hearing on the motion to suppress his confession, Gray testified that he had been threatened three times by one of the officers, McIlrath. Gray alleged that, as they were leaving the police station to go to the jail, McIlrath threatened to beat him up if he did not take the police to Derissa Scales. When they got to the jail, they entered an elevator to go to the fourth floor. Gray claimed that McIlrath stopped the elevator between floors and threatened to beat in his face with the jail keys unless he led the police to Scales. After Gray agreed to take the police to Scales, Gray claims that McIlrath told him that if they failed to find the little girl, Gray wouldn't come back.

Both McIlrath and another officer who had been present testified that they had neither threatened Gray nor offered him any help to induce him to tell them where to find the little girl. They testified that Gray was frightened that once he was put in jail he would not get out again. They claimed that in the elevator on the way up to the jail, Gray spontaneously offered to lead them to the child to avoid being put in jail.

Although the trial court judge did not articulate the factual basis of his denial of Gray's motion to suppress, the state supreme court found:

A further statement of the Miranda warnings was given to Gray, but a brief preliminary interrogation (proved) to be unfruitful. However, as the officers and Gray were ascending in the elevator, Gray spontaneously said, "If I take you to her will you help me?" No offer of help was made to Gray but he offered to take the officers to where he had left Derissa. Needless to say, this offer was accepted and Gray and the officers entered an automobile to be directed by Gray to the place where he had left the child.

Gray v. State, 375 So. 2d at 996. The state court obviously accepted the officer's version of the events and rejected Gray's testimony.18 Gray has not articulated any reason to reject the state court's finding that Gray spontaneously offered to lead the officers to Derissa Scales, nor are we aware of any. See Sumner v. Mata, 449 U.S. 539, 101 S. Ct. 764, 66 L. Ed. 2d 722 (1981); Branch v. Estelle, 631 F.2d 1229, 1232-33 (5th Cir. 1980). Because we are bound by the state court's resolution of this pure credibility fact issue, Gray's renewed claim that his confession resulted from the officer's threats is frivolous.

Gray's claim that the statement was the result of a lack of sleep, also lacks merit.19 On the day Derissa Scales disappeared, the police found Gray around midnight at a fast food restaurant where his girlfriend worked. When Gray offered to take the police to find Derissa Scales, the police set out immediately and discovered her body about 4:30 in the morning. After the body was recovered, Gray was returned to the police station and given Miranda warnings. Gray made an oral statement at that time and signed a copy of the statement the next afternoon. Gray testified at the motion to suppress that, except for the three alleged threats made by McIlrath, his subsequent statements were made freely and voluntarily. See Gray v. State, 375 So. 2d at 1001. Given Gray's testimony, we cannot say that any lack of sleep vitiated the voluntariness of his confessions.

Gray claims that the method of jury selection used by the State of Mississippi violated the sixth and fourteenth amendments because his counsel were not able to inquire adequately into the attitudes of prospective jurors. Because Gray notes that this inquiry is indispensable when a defendant's life is at stake and cites Witherspoon v. Illinois, supra, it would appear that the basis of his claim is that the method of jury selection prevented his counsel from adequately determining prospective jurors' views on the death penalty.

Under the jury selection process used by the trial court, the prospective jurors were addressed as a group. The trial judge asked whether any jurors had scruples against the death penalty. Those that responded affirmatively were then questioned further to determine the extent of their objection. Because both counsel were allowed an unlimited opportunity to question the jurors who indicated that they had some scruples against imposing the death penalty, Gray's objection cannot be leveled against this stage of the proceeding. Gray's claim must therefore be that the initial inquiry was insufficient to reveal all jurors who had scruples against the death penalty. This claim is frivolous.

To the extent that this method of jury selection fails to reveal jurors who oppose the death penalty, a defendant is aided. Because these prospective jurors remain in the jury pool, there is a greater chance that jurors opposed to the death penalty will serve on the petit jury. There is thus no indication that the use of this jury selection method prejudiced Gray's interests.4. Failure to Instruct on a Lesser-Included Offense

Gray argues that the fourteenth amendment was violated because the jury was not permitted to consider a lesser included offense. See Beck v. Alabama, 447 U.S. 625, 100 S. Ct. 2382, 65 L. Ed. 2d 392 (1980). In the case at bar, Gray's counsel requested an instruction on the lesser included offense of kidnapping, which was given. Gray's counsel did not, however, request that the jury be instructed on any other lesser offense.

Even if it were possible to have given an instruction on another lesser included offense, Gray is barred from raising this issue on habeas. Under state law, Gray's failure to request an instruction on another lesser included offense precluded him from later raising that point. See Newell v. State, 308 So. 2d 71, 78 (Miss.1975). Because an independent and adequate state procedural ground exists for denying Gray's claim, he cannot raise it on habeas absent a showing of cause and prejudice. See Wainwright v. Sykes, 433 U.S. 72, 97 S. Ct. 2497, 53 L. Ed. 2d 594 (1977); Washington v. Watkins, 655 F.2d 1346 (5th Cir. 1981). Because Gray has shown neither, the state's independent procedural ground bars our review of his claim.

5. The Construction of the Aggravating Circumstances

Gray argues that the four aggravating circumstances found in his case have been interpreted in such a broad manner that there is no way to distinguish those cases where the death penalty has been imposed from those where it has not. The four aggravating circumstances found to exist are: i) that the murder was committed under a sentence of imprisonment; ii) that the murder was committed during the course of a kidnapping; iii) that the murder was committed for the purpose of avoiding or preventing the detection or lawful arrest of the defendant; and iv) that the murder was especially heinous, atrocious or cruel.

Gray claims that because the first factor, that the murder was committed under a sentence of imprisonment, is susceptible of widely divergent interpretations, it is vague in violation of Grayned v. Rockford, 408 U.S. 104, 92 S. Ct. 2294, 33 L. Ed. 2d 222 (1972). In evaluating a vagueness challenge to a state law, we are required to consider the law in light of the state court's interpretation of it. See Village of Hoffman Estates v. Flipside, --- U.S. ----, 102 S. Ct. 1186, 1191 n.5, 71 L. Ed. 2d 362 (1982).

Gray concedes that this first factor clearly covers murders committed while a defendant is physically imprisoned and that the Mississippi Supreme Court has "sanctioned (its) application to a person on parole at the time of the murder." See Gray v. State, 375 So. 2d at 1005. We fail to understand how the state could have more clearly defined this factor nor do we see any danger of arbitrary application. Gray's claim is completely without merit.

In attacking the second aggravating circumstance, that the murder was committed during the course of a kidnapping, Gray admits that he restates an earlier argument, and that this factor was proven when he was convicted of felony murder at the trial stage. We have already considered this argument and have rejected it.

Gray's attack is directed primarily at the third aggravating factor. He claims that because the trial judge's instructions differed from the statutory definition, the jury was permitted to base its imposition of the death sentence on a nonstatutory aggravating circumstance. Section 99-19-101(5)(e) of the Mississippi Code provides that murder committed "for the purpose of avoiding or preventing a lawful arrest or effecting an escape from custody" constitutes an aggravating circumstance. See Miss. Code Ann. § 99-19-101(5)(e). In this case, the trial judge instructed the jury that this factor could be found if the murder "was committed for the purpose of avoiding or preventing the detection and lawful arrest of the Defendant ...."

Although the first part of this aggravating circumstance might have been interpreted to apply only to a murder committed while a defendant was being pursued by a policeman, the trial judge interpreted this circumstance to apply to the murder of a witness to a crime to avoid detection. The state supreme court accepted the trial court's construction of this factor. The court stated "Upon the entire record, the jury was justified in finding beyond a reasonable doubt that the child had been killed by Gray intentionally and maliciously for the purpose of silencing her outcries or preventing the report by her of acts of molestation." Gray v. State, 375 So. 2d at 1004.20 The Mississippi courts' interpretation of that state's statute is definitive and controlling. The jury was not allowed to impose the death sentence on the basis of a nonstatutory factor.

The question which remains, however, is whether this aggravating circumstance has been interpreted so broadly that it fails to offer a meaningful distinction between those cases where the death penalty may be imposed and those where it may not. See Godfrey v. Georgia, 446 U.S. 420, 100 S. Ct. 1759, 64 L. Ed. 2d 398 (1980). Gray argues that the state court's interpretation of this factor impermissibly renders it applicable to all murders, since every murderer silences his victim and thus is aided in evading or preventing arrest. We do not think, however, that Mississippi has adopted such a broad construction. The court has construed this aggravating circumstance to refer to purposefully killing the victim of an underlying felony to avoid or prevent arrest for that felony. Like the murder of a clerk who has delivered the cash to the gunman who robs her, the state supreme court noted that Gray purposefully killed Derissa Scales to prevent her from reporting that he had indulged his repulsive sexual aggressions on her little body. As such, this factor merely achieves the state's interest in protecting victims of felonies from being killed to prevent the felon's detection. This construction is not so broad that it comprehends an impermissibly large group of murderers. See Godfrey v. Georgia, 446 U.S. 420, 100 S. Ct. 1759, 64 L. Ed. 2d 398 (1980); Riley v. State, 366 So. 2d 19, 22 (1979) (per curiam).

Gray claims that the fourth aggravating factor, that the murder is especially heinous, atrocious or cruel has never been construed by the state supreme court. This claim, however, is incorrect. See Coleman v. State, 378 So. 2d 640, 648 (1978).

Gray also claims that this factor has been inconsistently applied and contrasts Irving v. State, 361 So. 2d 1360 (Miss.1978), cert. denied, 441 U.S. 913, 99 S. Ct. 2014, 60 L. Ed. 2d 386 (1979), with Voyles v. State, 362 So. 2d 1236 (Miss.1978), cert. denied, 441 U.S. at 956, 99 S. Ct. 2184, 60 L. Ed. 2d 1059 (1979). On its face, Gray's argument appears meritorious since Voyles involved a particularly gruesome murder and Irving involved the killing of a victim by a single shotgun blast. The difficulty with Gray's argument, however, is that Irving, as it is presented by the state supreme court, did not rely on the fact that the murder was especially heinous, atrocious or cruel. The two aggravating circumstances which the court noted were that Irving had committed the murder during the course of a robbery and that he had a prior criminal record. See 361 So. 2d at 1371, 1372. The factual basis of Gray's argument is thus incorrect.

Moreover, we note that the state supreme court has consistently applied this factor to pitiless crimes which are unnecessarily torturous to the victim. See Reddix v. State, 381 So. 2d 999 (Miss.) (elderly man beaten to death with iron wrench), cert. denied, 449 U.S. 986, 101 S. Ct. 408, 66 L. Ed. 2d 251 (1980); Jones v. State, 381 So. 2d 983 (Miss.) (same), cert. denied, 449 U.S. 1003, 101 S. Ct. 543, 66 L. Ed. 2d 300 (1980); Culbertson v. State, 379 So. 2d 499 (Miss.1979) (man knocked to ground with stick and shot while begging for mercy), cert. denied, 449 U.S. 986, 101 S. Ct. 406, 66 L. Ed. 2d 250 (1980); Voyles v. State, 362 So. 2d 1236 (Miss.1978) (woman beaten, run over twice with car and thrown in creek). The only arguable exception to the consistent application of this factor has been Washington v. State, 361 So. 2d 61 (Miss.1978), cert. denied, 441 U.S. 916, 99 S. Ct. 2016, 60 L. Ed. 2d 388 (1979). In Washington the defendant robbed a convenience store and shot the clerk while leaving the store. Although the murder was cold blooded, it was arguably not a "pitiless crime which is unnecessarily torturous to the victim." See Coleman v. State, 378 So. 2d at 648. We note that Washington was the first case to apply this factor and was decided before Godfrey v. Georgia, supra. Even if Washington construed this term too broadly, there is every indication that the state supreme court has refined its views, applied this term consistently since then, and will continue to do so. See Coleman v. State, 378 So. 2d at 650 (invalidating death penalty because penalty was disproportionate to that imposed in similar cases).

Gray claims that the Mississippi Supreme Court's comparative review of death sentences is flawed since the court only compared Gray's case with those cases where the death sentence had been imposed and not with all the cases where it could have been imposed. Because the Supreme Court has rejected a similar argument in Proffitt v. Florida, 428 U.S. 242, 258-59 & n.16, 96 S. Ct. 2960, 2969-70 & n.16, 49 L. Ed. 2d 913 (1976), we reject this claim as well.

Gray received constitutional assistance of counsel in his trial before a lawfully chosen jury. He was sentenced to death for his crime under statutory and decisional procedures which contain no constitutional deficit. He is not entitled to federal habeas corpus relief from the due execution of the sentence imposed and affirmed on review by the courts of Mississippi.

AFFIRMED.