Gilday v. Callahan

866 F. Supp. 611, 1994 U.S. Dist. LEXIS 6666, 1994 WL 643224

• Court: District Court, D. Massachusetts
• Decided: March 23, 1994
• Docket: Civ. A. 81-2886-REK
• Status: Published

Opinion

Opinion

KEETON, District Judge.

I.

After full consideration of the contentions of the parties in this habeas corpus petition, I conclude that the petition should be denied.

The petition was filed in this court more than a decade after the state court conviction it challenges. Now, more than another decade after it was filed, it is before the court for decision on voluminous submissions (the most recent of which was filed in 1994).

II. Factual and Procedural Background

On September 22, 1970, Officer Walter A. Schroeder of the Boston Police was killed in the course of an armed robbery of the Brighton Branch of the State Street Bank and Trust Company. On March 10, 1972 a jury convicted petitioner of first degree murder and two counts of armed robbery for his involvement in the events of September 22, 1970.

The others charged in the robbery and murder were Stanley R. Bond, Robert J. Valeri, Susan E. Saxe, Michael Fleischer and Katherine A. Power. Bond, who testified as a defense witness at petitioner’s trial, died in prison. Valeri, who testified as a prosecution witness against petitioner, pleaded guilty to manslaughter and is now free. Fleischer testified as a prosecution witness against petitioner; his indictments were ultimately dismissed. After a period as a fugitive, Saxe was tried in 1976. That trial ended in a hung jury; she then pled guilty to manslaughter and is now free. The court takes judicial notice of the widely publicized fact that Katherine Power recently ended a twenty-three year period as a fugitive, pled guilty to charges against her in the Superior Court of Massachusetts and is currently serving a term of imprisonment in Massachusetts.

On October 4, 1972, petitioner filed his first motion for a new trial in light of Furman v. Georgia, 408 U.S. 238, 92 S.Ct. 2726, 33 L.Ed.2d 346 (1972). His motion was denied, but his death sentence was changed to a life sentence, which he is currently serving.

On August 20, 1973, petitioner filed a second motion for a new trial. He alleged that a) the prosecution had suppressed exculpatory statements by a witness, Michael Finn, b) the prosecution had suppressed an arrangement of leniency Valeri had allegedly been granted in return for his testimony, c) he had been denied a trial before an impartial jury due to prejudicial pretrial publicity, d) the judge’s instructions on the meaning of “reasonable doubt” violated his due process rights, and e) certain other constitutional violations had occurred at his trial. Evidentiary hearings were held on this motion, after which the Superior Court denied the motion; the Supreme Judicial Court affirmed in Commonwealth v. Gilday, 367 Mass. 474, 327 N.E.2d 851 (1975) (“Gilday 7”).

On January 11, 1979, petitioner filed a third motion for new trial, alleging that the prosecution had suppressed exculpatory evidence of an arrangement of leniency Fleischer had been granted in exchange for his testimony. The motion was denied, and petitioner appealed. The Supreme Judicial Court remanded the case for an evidentiary hearing to determine if in fact such an arrangement of leniency had been made. The Superior Court, though finding that an arrangement of leniency had been made, denied relief. Petitioner appealed again. The Supreme Judicial Court affirmed the trial court’s finding that an arrangement of leniency had been made. The Court ruled, however, that any constitutional error thereby committed was harmless beyond a reasonable doubt. Commonwealth v. Gilday, 382 Mass. 166, 415 N.E.2d 797 (1980) (“Gilday II").

On October 13, 1981, Gilday filed his petition for writ of habeas corpus with this court. The Commonwealth filed a motion to dismiss the petition on the grounds that some of the issues had not been exhausted in state court. Judge McNaught of this court issued a memorandum and order dismissing petitioner’s original petition and treating it as resubmitted with the then-unexhausted state claims deleted (Docket No. 22, December 17, 1982).

On May 7, 1983, petitioner filed a motion to expand the record before the court to include a recently obtained affidavit of the witness Valeri, which petitioner asserted disclosed that an arrangement of leniency had been made in return for Valeri’s testimony. Petitioner moved for an evidentiary hearing on the matter. That motion was denied by Magistrate Judge Joyce L. Alexander on September 16, 1983, Gilday v. Callahan, 99 F.R.D. 308 (no docket number). The denial was affirmed by Judge McNaught (marginal notation on Docket No. 53, made on January 18, 1983).

On December 8, 1990, Judge McNaught ordered the case dismissed for want of prosecution (Docket No. 93). On February 1, 1991, Judge McNaught retired. On June 25, 1991, Judge Skinner granted a motion vacating the order of dismissal, and the case was reassigned to the judge before whom the matter is now pending (Docket No. 95).

Meanwhile, on June 1, 1987, petitioner had filed his fourth motion for a new trial with the state courts, in order to litigate the thenunexhausted claims. He alleged in that motion that the charge to the jury had impermissibly removed specific intent as an element of the crime, substituting mandatory presumptions of intent and malice. The Superior Court denied this motion and the Supreme Judicial Court affirmed. Common wealth v. Gilday, 409 Mass. 45, 564 N.E.2d 577 (1991) (“Gilday III”).

On January 24, 1992, this court allowed petitioner to amend his petition to include the claims that were exhausted by Gilday III (Docket No. 101, marginal notation).

III. Reasonable doubt instructions

A. Constitutional Error

Petitioner challenges the charge to the jury on the government’s burden of proof.

In 1970, before the trial of petitioner, the Supreme Court held that due process requires in any criminal proceeding that the factfinder be persuaded “beyond a reasonable doubt” of facts necessary to show the essential elements of the offense. In re Win-ship, 397 U.S. 358, 90 S.Ct. 1068, 25 L.Ed.2d 368 (1970).

A trial court need not define reasonable doubt, but any definition the court states for the jury must not lessen the government’s burden of proof. United States v. Olmstead 832 F.2d 642, 644-46 (1st Cir.1987). To show on collateral review that a trial court erred in defining “beyond a reasonable doubt,” a petitioner bears a heavy burden. The Supreme Court has declared that “a single instruction ... may not be judged in artificial isolation, but must be viewed in the context of the overall charge.” Cupp v. Naughten, 414 U.S. 141, 146-47, 94 S.Ct. 396, 397, 38 L.Ed.2d 368 (1973). To show prejudice, a petitioner must show more than merely that the instruction was “undesirable, erroneous, or even ‘universally condemned.’” The burden is to show that the error “so infected the entire trial that the resulting conviction violates due process.” Id. Once constitutional error in a reasonable-doubt charge has been established, however, the error can never be regarded as harmless. Sullivan v. Louisiana, — U.S. --, 113 S.Ct. 2078, 124 L.Ed.2d 182 (1993); see discussion below at 628.

Courts have determined to be unconstitutional instructions that, among other things, defined the reasonable doubt standard as one of “moral certainty.” For example, in Cage v. Louisiana, 498 U.S. 39, 41, 111 S.Ct. 328, 330, 112 L.Ed.2d 339 (1990), the trial court had instructed the jury that if they were not convinced of guilt to a “moral certainty,” and the evidence does not give rise to “grave uncertainty,” or to “actual substantial doubt,” then guilt beyond a reasonable doubt had been established. The Supreme Court found this instruction to be constitutionally deficient. In Lanigan v. Maloney, 853 F.2d 40, 47-48 (1988) the Court of Appeals for the First Circuit held that a charge that, among other things, had required jurors to be convinced of guilt to “a degree of moral certainty” (emphasis added) was constitutionally deficient.

Other First Circuit precedent makes clear, however, that whether a reasonable doubt instruction that contains “moral certainty” language is unconstitutional depends on context, including the entire charge in which the challenged language appears.

Id. at 48 (citing cases in which charge as a whole was not unconstitutional but finding that in the case before it the entire thrust of charge deemphasized the government’s burden).

See also Dunn v. Perrin, 570 F.2d 21, 25 (1978) (cumulative effect of three incorrect definitions was to obfuscate an essential of due process).

Thus, for example, in United States v. Drake 673 F.2d 15, 20 (1982) the court determined that language referring to “moral certainty,” without more, did not render constitutionally deficient a charge that, as a whole, properly instructed the jury on the government’s burden and on the presumption of innocence.

Objecting to “moral certainty” language in the instructions in this ease, petitioner points to five occasions on which the court used the phrase “moral certainty” in charging the jury. Respondent counters by noting that on five occasions, the court emphasized to the jury the high standard of proof. Respondent also notes that unlike the charge in Dunn, the charge here did not contain the words “to a degree of.” In the opinion of the court in Dunn, this latter phrase had been particularly harmful in conjunction with the phrase “moral certainty.” Respondent notes also that, unlike the charge in Cage, the charge in this case did not use the phrases “grave uncertainty” or “actual substantial doubt,” which were there determined to lower the government’s burden of proof.

Petitioner points to other language as having a tendency to trivialize the government’s burden. For example, the trial court charged the jurors that before convicting they must have the same level of certainty as they would have “when they take action in the major affairs of their lives” and “when they take vital action in [their] everyday fives.” In Dunn, the court found that where jurors are advised that the appropriate level of certainty is one that would cause the jurors to refrain from acting in the significant affairs of their lives, the instruction was not infirm. 570 F.2d at 24-25. Here, however, the instruction spoke of the certainty jurors required before acting. The Dunn court noted that such instructions had been disapproved by the Court of Appeals for the District of Columbia Circuit and by the Supreme Judicial Court of Massachusetts for their tendency to trivialize the government’s burden. Id. at 24-25 (citing Scurry v. United States, 347 F.2d 468, 470 (1965), Commonwealth v. Ferreira, 373 Mass. 116, 364 N.E.2d 1264 (1977).

An instruction that referred to “the degree of certainty upon which you would act in the important affairs of your life” was, however, later upheld in Rogers v. Carver, 833 F.2d 379, 382 (1st Cir.1987). In that case, the Court of Appeals distinguished Ferreira on the grounds that in Ferreira the court had given specific examples of “major affairs,” such as deciding whether to continue one’s education or leave one’s job, which tended to “trivialize the awesome duty of the jury” to find defendant guilty beyond a reasonable doubt. In Rogers, by contrast, the Court of Appeals determined the instructions to be far less harmful, because they had not given any examples of “important affairs,” and thus had not trivialized the government’s burden. Id.

Petitioner urges that, in the present case, the use of the term “everyday fives” further trivialized the charge, as had the reference to specific instances in Ferreira. In isolation, the term “everyday” might be taken to trivialize the government’s burden more than would language referring to major events; here, however, the term “everyday” was joined with the phrase “vital action” — “when they take vital action in [their] everyday fives.” In this context, “everyday” does not have the trivializing effect it otherwise might.

Although the phrasing of the instruction could have been improved, I conclude that it did not sufficiently impair the charge to amount to a denial of due process.

Petitioner objects to two other passages:

All we can do is weigh the pros and cons against any contemplated course of action; and then with the wisdom and the intellect that we possess, make a decision. We may be right, we may be wrong (emphasis added).

This language, petitioner argues, brings to mind the scales of justice, and strongly suggests the preponderance of the evidence standard. Even worse, petitioner argues, is the following language:

When you get all through analyzing this evidence, it has to be a doubt nagging your mind, leaving you with an uncertainty of conviction to that moral certainty which you can stand up and argue in the jury room with principle and integrity and honesty to your fellow jurors. And if you don’t believe in it yourself, you haven’t got a reasonable doubt.

This latter passage, petitioner argues, did not merely convey an impression of balancing; rather, it conveyed the impression that the law tilted the balance against the petitioner. Petitioner characterizes the above passage as similar to language found infirm in Dunn:

[Reasonable doubt] does not mean a trivial or frivolous or fanciful doubt nor one which can be readily or easily explained away, but rather such a strong and abiding conviction as still remains after careful consideration of all the facts and arguments.

570 F.2d at 23-24. In Dunn, the court called this an “inescapable violation” of Winship, noting that it

was the exact inverse of what it should have been ... [i]nstead of requiring the government to prove guilt, it called upon the defendants to establish doubt in the jurors’ minds.

Id.

Petitioner’s contention that the present charge involved a similar inversion of the burden of proof is based, however, on a reading of one sentence, in isolation. It is true that the trial judge did misspeak in stating that “if you don’t believe in it yourself, you haven’t got a reasonable doubt.” Aside from this sentence, however, nothing in the charge can reasonably be understood as even suggesting, much less instructing the jury to apply, an inversion of the burden of proof. And it is well settled that a single sentence is not to be interpreted “in artificial isolation.” The preceding sentence plainly referred to a “conviction” that the defendant was guilty as charged, not to a conviction of a “doubt” or belief in a “doubt.” The two sentences together cannot reasonably be interpreted as inverting the burden of proof. The conclusion that the context does not support an interpretation of inversion of the burden is reinforced when one takes account of the repeated instructions elsewhere in the charge of the point that it is the government, not the defendant, that has a burden of proof beyond a reasonable doubt.

The charge as a whole instructed the jurors to acquit unless they had a conviction beyond reasonable doubt that the defendant was guilty as charged. There was no inversion or shifting of burdens. The charge did not so diminish or trivialize the burden imposed upon the government as to deny petitioner due process.

IY. Mandatory Presumptions

A. Sandstrom Claims

The jury was charged that they might find petitioner guilty of first degree murder on either a theory that he had participated in a killing with premeditated malice aforethought or on a felony murder theory (that the killing had been committed as part of an armed robbery). Thus it is possible that the jury, in finding defendant guilty of first degree murder, did so on a finding that he had participated in an armed robbery, in the course of which a homicide occurred.

Petitioner asserts that the charge to the jury allowed the jury to reach such a felony-murder verdict without finding one of the elements of armed robbery, namely, the specific intent to commit an armed robbery. Petitioner alleges that the charge to the jury, instead of instructing the jurors that in order to convict of felony murder they must find specific intent to commit armed robbery, instructed them to rely on a presumption that petitioner intended the natural and probable consequences of his actions. Interpreted in this way, petitioner contends, the charge is in violation of Sandstrom v. Montana, 442 U.S. 510, 520-24, 99 S.Ct. 2450, 2457-59, 61 L.Ed.2d 39 (1970).

B. Retroactivity

1. The Teague Doctrine

Winship, decided in 1970, announced a constitutional rule that in any criminal proceeding the factfinder must be persuaded “beyond a reasonable doubt” of facts necessary to establish each of the elements of the crime charged.

Petitioner was tried in 1972. By April 25, 1975, the Supreme Judicial Court of Massachusetts had rejected petitioner’s efforts to set aside his murder conviction.

On June 9, 1975, the Supreme Court in Mullaney v. Wilbur, 421 U.S. 684, 95 S.Ct. 1881, 44 L.Ed.2d 508, invalidated a jury charge that established a rebuttable presumption that where a homicide was intentional and unlawful, it was committed with malice aforethought. The charge had stated that the presumption could be rebutted by defendant on a showing by a preponderance of the evidence that he had acted in the heat of passion on sudden provocation.

In 1979, the Supreme Court decided in Sandstrom that a charge instructing the jury to presume that the defendant intended the natural and probable consequences of his actions unconstitutionally shifted the government’s burden of proof as to an essential element of an offense.

In Teague v. Lane, 489 U.S. 288, 109 S.Ct. 1060, 103 L.Ed.2d 334 (1989) the Supreme Court decided that a new rule of constitutional interpretation will not be applied retroac tively on collateral review unless the rule is within one of two exceptions.

Under this array of precedents, to decide whether petitioner presents a meritorious constitutional claim under Sandstrom, a court hearing petitioner’s Sandstrom-based claim must consider whether Teague and its progeny stand as an obstacle to petitioner’s asserting a Sandstrom-based claim on collateral review.

The Court has explained the Teague doctrine as “validat[ing] reasonable, good-faith interpretations of existing precedents,” Butler v. McKellar 494 U.S. 407, 414, 110 S.Ct. 1212, 1217, 108 L.Ed.2d 347 (1990). The Court has stated that a constitutional rule will not be considered new if “reasonable jurists” at the time petitioner’s conviction became final would have considered it to be “dictated” by precedent.

Teague, 489 U.S. at 301, 109 S.Ct. at 1070;

see also Sawyer v. Smith, 497 U.S. 227, 234-41, 110 S.Ct. 2822, 2827-31, 111 L.Ed.2d 193 (1990);

Saffle v. Parks, 494 U.S. 484, 488-495, 110 S.Ct. 1257, 1260-64, 108 L.Ed.2d 415 (1990);

Caspari v. Bohlen, — U.S. -, ----, 114 S.Ct. 948, 954-957, 127 L.Ed.2d 236 (1994).

At the time petitioner’s conviction became final, did existing precedents provide clearly that jury instructions are unconstitutional if they impose mandatory presumptions on issues of fact that are essential to the offense charged? Supreme Court precedent existing at the time petitioner’s conviction became final had foreshadowed Sandstrom to some extent. Did it go far enough that Sandstrom announced no “new rule?”

In Morissette v. United States, 342 U.S. 246, 273-76, 72 S.Ct. 240, 255-56, 96 L.Ed. 288 (1952), the Supreme Court ruled that where an intent is an element of an offense, the existence of that intent is a question of fact that cannot be removed from the jury by instructing them that the law raises a mandatory presumption of intent. The court stated that “this presumption would conflict with the overriding presumption of innocence with which the law endows the accused and which extends to every element of the crime.” Id. at 275, 72 S.Ct. at 256. Morissette involved a federal offense and the decision did not explicitly rest on any constitutional principle.

In its later decision in Winship, however, the Court ruled that in any criminal proceeding the factfinder must be persuaded “beyond a reasonable doubt” of facts necessary to establish each of the elements of the crime charged. Thus the requirement of proof “beyond a reasonable doubt” in criminal cases was raised to constitutional status, applicable to cases tried in state courts and not alone to federal prosecutions.

After Winship, then, it was possible to frame an argument that mandatory presumptions violated the constitution; this hardly indicates, however, that “reasonable jurists” would have agreed that the argument must be sustained. It was not until the landmark Mullaney and Sandstrom decisions — both decided subsequent to petitioner’s conviction having become final — that such instructions were definitively invalidated.

In applying the “reasonable jurist” standard, a court is to apply a foresight rather than a hindsight standard. With the passage of time and experience under the rule of Mullaney and Sandstrom, it is tempting to look back and conclude that those decisions were foreordained by precedents such as Morissette and Winship. A close examination, however, of contemporaneous views and practices of the bench and bar before Mullaney and Sandstrom were handed down is at least instructive, if not controlling.

Petitioner cites DeJoinville v. Commonwealth, 381 Mass. 246, 248, 251, 408 N.E.2d 1353 (1980) for the proposition that the Sandstrom rule is retroactively applicable on collateral review in Massachusetts. First, however, this was a decision under pre-Teague standards. Thus it is not on point for purposes of considering retroactivity on federal habeas corpus review; it is questionable whether it is even good law for purposes of Massachusetts collateral review, insofar as it seems to rely upon federal case law.

See, e.g., DeJoinville, 408 N.E.2d at 1356 n. 7 (citing Hankerson v. North Carolina, 432 U.S. 233, 97 S.Ct. 2339, 53 L.Ed.2d 306 (1977) for proposition that Mullaney rule is retroactive.)

Moreover, the reasoning of the court in DeJoinville suggests that under the now-applicable Teague standard, the Sandstrom rule should not be viewed as retroactive, because the Sandstrom decision radically altered the practices of the bench and bar. In DeJoinville the court determined that the United States Supreme Court cases decided before De Joinville’s appeal did not provide De Joinville with “sufficient guidance” to give him a genuine opportunity to raise an objection to a mandatory presumption. Id., 408 N.E.2d at 1356. Significantly, the Supreme Judicial Court came to this conclusion even though, by the time De Joinville made his appeal, Mullaney had been decided. If it was not reasonable to expect someone to predict the Sandstrom claim after Mullaney, it would have been even less reasonable to expect someone to predict the claim before Mullaney — that is, at the time petitioner’s conviction became final. And if it was not a claim that could reasonably have been predicted, it was a fortiori not a claim that reasonable jurists would have agreed was “dictated” by then-existing precedents.

Cf. Commonwealth v. Callahan, 380 Mass. 821, 406 N.E.2d 385, 388 (1980) (noting that the Supreme Judicial Court began to scrutinize jury instructions more carefully after the Supreme Court’s Mullaney decision, and applied a stricter standard of scrutiny to post-Mullaney instructions);

Commonwealth v. Stokes, 374 Mass. 583, 374 N.E.2d 87, 91 (1978) (failure of a petitioner to object to assertedly invalid burden-shifting charge, at a trial conducted before the Supreme Court decided Mullaney, would not bar review of that assertedly invalid charge; to do so would be to require “clairvoyance” on the part of defense counsel);

Caspari, — U.S. at ---, 114 S.Ct. at 954-57 (then-existing decisions of state courts are relevant to a determination as to whether reasonable jurists, at the time, would have regarded the precedents as dictating a particular constitutional rule).

I conclude that for purposes of applying the Teague doctrine, the Sandstrom prohibition against mandatory presumptions as to the essential elements of criminal offenses was a new rule.

See Cain v. Redman, 947 F.2d 817, 820-22 (6th Cir.1991) cert. denied — U.S. -, 112 S.Ct. 1299, 117 L.Ed.2d 521 (1992) (Sandstrom “broke new ground,” and was a “new” rule for Teague purposes);

Goodwin v. McQuen, 812 F.Supp. 181 (D.Kan.1993).

Cf. Prihoda v. McCaughtry, 910 F.2d 1379, 1382 (7th Cir.1990).

Since Sandstrom announced a “new” rule, petitioner may not avail himself of the benefit of that decision unless he can show that the rule falls into one of two narrow exceptions to the Teague doctrine.

The first exception relates to “primary, private individual conduct beyond the power of the criminal law-making authority to proscribe,” 489 U.S. at 307, 109 S.Ct. at 1073. It is not implicated in this case.

The second exception is for watershed rules of criminal procedure that are necessary to the fundamental fairness of the criminal proceeding. Id. at 311, 109 S.Ct. at 1075 (plurality opinion), Saffle, 494 U.S. at 486, 110 S.Ct. at 1259. Petitioner asserts that the Sandstrom rule implicated a bedrock procedural element essential to the fundamental fairness of the trial, and therefore that Sandstrom should be considered to fall into the second Teague exception. For the reasons stated below, I conclude that this contention cannot be sustained.

The Teague plurality, in carving out the second exception, made clear that it was narrow and was not intended to implicate every constitutional rule that might have a bearing on fairness, or a bearing on accuracy. Rather, the exception encompasses only such rules as both implicate the accuracy of the trial and are “absolute prerequisite^] to fundamental fairness.” Teague, 489 U.S. at 314, 109 S.Ct. at 1077. The Teague plurality described the second exception without applying it. In that case, petitioner urged the adoption of a new rule, that petit juries — as opposed to jury venires — be subject to a fair-cross-section rule; without deciding whether the proposed rule had merit, the plurality expressed the view that such a new constitutional rule would not fall within the second exception. Id. at 313-14, 109 S.Ct. at 1076-77.

The Supreme Court has stated in dicta that the rule of Gideon v. Wainwright, 372 U.S. 335, 83 S.Ct. 792, 9 L.Ed.2d 799 (1963) is a rule the “primacy and centrality” of which would justify applying the second exception. Saffle, 494 U.S. at 494, 110 S.Ct. at 1263. The Supreme Court, however, has found no occasion to hold that a new rule fell within the second exception.

See Saffle 494 U.S. at 494, 110 S.Ct. at 1263 (proposed rule that jury in capital sentencing be allowed to mitigate sentence based upon the sympathy they feel for the defendant after hearing mitigating evidence would not fall within second Teague exception);

Sawyer, 497 U.S. at 243-44, 110 S.Ct. at 2832 (rule forbidding inaccurate prosecutorial argument diminishing juror responsibility in capital sentencing is not within exception);

Butler, 494 U.S. at 415-17, 110 S.Ct. at 1218 (rule forbidding interrogation following invocation of right to counsel is not within exception).

In Gilmore v. Taylor, —r- U.S.-,-, 113 S.Ct. 2112, 2119,124 L.Ed.2d 306 (1993), the Supreme Court held that the rule of Falconer v. Lane, 905 F.2d 1129 (7th Cir. 1990), did not fall ■within the second Teague exception. The Falconer decision invalidated jury instructions that failed to state expressly that the jury could not return a murder conviction if it found that the defendant had a mental state that would reduce the offense to voluntary manslaughter. 905 F.2d at 1136. Although the Gilmore decision is not precisely on point, it indicates that a new rule that it is constitutional error to instruct the jury in a manner that has a tendency to confuse the jury on the government’s burden of proof is not of the sort of “bedrock” or “watershed” importance as would implicate the second exception. Id. This point applies with special force to a Sandstrom error, because the impact of the Sandstrom rule is an exceedingly narrow one. The constitutional error of stating that the jurors “must presume” a fact is avoided if the trial court instead states that the jurors “may infer” that fact. Error will probably also be avoided if the jurors are instructed that they “may presume” that fact.

See, e.g. Hardy v. United States 691 F.2d 39, 40 (1st Cir.1982).

Cf. Sandstrom, 442 U.S. at 527-28, 99 S.Ct. at 2461 (concurring opinion of Rehnquist, J.) (expressing doubts whether jury was so attentively attuned to instructions to divine the difference between “infer” and “presume.”)

Furthermore, an otherwise ambiguous instruction may be cured of any constitutional infirmity if the surrounding language of the charge is such that jurors would not be “reasonably likely” to interpret the charge as creating a mandatory presumption. Boyde v. California 494 U.S. 370, 110 S.Ct. 1190, 108 L.Ed.2d 316 (1990).

For all these reasons, I conclude that the Sandstrom rule does not come within the second Teague exception.

See Cain, 947 F.2d at 822 (Sandstrom rule not encompassed within second Teague exception;

Goodwin, 812 F.Supp. at 182 (same).

C. Procedural Default of Non-Intoxication Sandstrom Claims

There is a separate reason why petitioner may not assert certain aspects of his Sandstrom claims; insofar as the Sandstrom claims relate to portions of the charge other than the charge on intoxication, petitioner’s claims are barred by an independent and adequate state procedural ground.

The Supreme Judicial Court, in considering petitioner’s fourth motion for new trial, chose to consider on the merits only those portions of the instructions that related to intoxication. Petitioner had asserted that certain other portions of the charge also contained Sandstrom errors, but the Supreme Judicial Court refused to consider these claims (“non-intoxication claims,” here inafter), ruling that 1) the issue of error in non-intoxication instructions was not “properly before the Court, not having been addressed by defendant in his motion [for a new trial], accompanying memorandum, and supplemental letter to the [trial] court,” and 2) that “since the issues raised in the motion for reconsideration could have been raised ... but were not, they must be deemed waived[,] See Mass.R.Crim.P. 30(c)(2).” Gilday III, 564 N.E.2d at 578 & n. 3. Respondent in the present action repeats the assertion that petitioner’s non-intoxication Sandstrom claims were barred by Mass.R.Crim.P. 30(c)(2).

Petitioner takes issue with the first determination. Petitioner argues, first, that although he asserted state law grounds for finding the intoxication instructions invalid, he explicitly argued, as part of his motion for post-conviction relief, that the intoxication instructions were also violative of the federal constitution. He points to specific language in the motion for new trial and also in a supplemental letter supplying an inadvertently-dropped footnote, which referred to the presumption issue and cited Winship, Mullaney, and Sandstrom.

Petitioner apparently understands the Supreme Judicial Court’s determination in Gil-day III that he had waived his non-intoxication claims as a determination based on the ground that no Sandstrom claims at all were before the motion judge. If this had in fact been the basis of the waiver determination in Gilday III, then petitioner would be correct in asserting that he had not waived the non-intoxication claims; he had clearly grounded his objection to the intoxication instructions, at least in part, on Sandstrom grounds. Indeed, respondent does not argue that the assertion of Sandstrom error in the intoxication instructions is proeedurally barred.

Petitioner’s reading of Gilday III is strained, however. A more sensible reading is that the Supreme Judicial Court refused to consider non-intoxication instructions because petitioner had only pointed to Sandstrom error in the intoxication portion of the instructions in his motion for a new trial. Thus he had “waived” any objection to other Sandstrom errors in the instructions. Petitioner does not now point to any part of that motion in which he pointed to the non-intoxication instructions that he now asserts were Sandstrom errors.

Petitioner points to the motion judge’s statement that he read the “charge as a whole.” This statement does not establish, however, that the motion judge chose to consider separate Sandstrom errors in the non-intoxication portion of the instructions. No such errors had been briefed. The statement only shows that the motion judge conscientiously considered the “charge as a whole” in evaluating the asserted Sandstrom error in the intoxication instructions, as the motion judge was required to do.

See Francis v. Franklin 471 U.S. 307, 315, 105 S.Ct. 1965, 1971-72, 85 L.Ed.2d 344 (1985). (if a specific portion of the jury charge, considered in isolation, could reasonably have been understood as creating a presumption, the potentially offending words must be considered in the context of the charge as a whole, in case other instructions explain the particular infirm language).

Petitioner argues that the Supreme Judicial Court’s ruling that his claim is procedurally barred is contrary to “undisputed” facts, is therefore a ruling on a question of law, and as such is not entitled to a presumption of correctness due to questions of fact under 28 U.S.C. 2254(d). Petitioner cites as authority for this contention a case involving an insurance dispute in which it was held that, where facts are undisputed, the question whether waiver occurred is an issue of law. Making an assumption arguendo that this is so does not help petitioner. Ouimette v. Moran 942 F.2d 1, cited by petitioner, stands for the proposition that certain issues that are “mixed questions of law and fact” are not entitled to the 2254(d) presumption of correctness. This refers, however, to “mixed” questions of fact and federal constitutional law. See id. at 4-5. It does not alter the general principle that the highest state court is the final arbiter of state law. Thus if, as petitioner asserts, the issue of “waiver” here were a pure question of law, the state court ruling that petitioner’s claim was “waived” would constitute an independent and ade quate state ground for refusing to hear his petition. Absent a showing of cause and prejudice, the Supreme Judicial Court’s ruling on the “law” would be dispositive. Wainwright v. Sykes, 433 U.S. 72, 97 S.Ct. 2497, 53 L.Ed.2d 594 (1977).

In any event, I do not accept petitioner’s contention that this is a pure question of law. There is a factual issue as to whether petitioner failed to point to any Sandstrom error in the non-intoxication portions of the instructions. This factual issue is covered by the presumption of correctness of section 2254(d). I find that petitioner did not point to any Sandstrom error in the non-intoxieation portion of the instructions. I would so find even if the state court determination on this matter were not to be accorded a presumption of correctness; petitioner has not indicated any portion of his motion for a new trial in which he pointed to Sandstrom error in the non-intoxication portion of the instructions.

The fact that petitioner noted a Sand strom-type error in the intoxication instructions does not show that he preserved the right to object to a different Sandstrom-type error in other parts of the instruction. The trial judge in the course of those instructions stated that the presumption that

a man who of his own accord, puts himself in a condition which causes him to commit a crime ... ought to be treated as if he had voluntarily committed a crime,” applied “[a]t least, insofar as the element of liquor is concerned.

This intoxication instruction, taken alone, was not reasonably subject to the interpretation that it told the jury that they were to apply a general presumption that a person intends the natural consequences of his actions. Petitioner failed to raise a timely claim, in his motion for new trial, that certain non-intoxication instructions might have been taken by the jury to create such a presumption, and petitioner has not asserted cause for his failure to assert this claim. Wainwright v. Sykes, supra.

I conclude that petitioner has failed to show any basis for overturning the Supreme Judicial Court’s ruling that the claims of Sandstrom error in the non-intoxication instructions are procedurally barred.

Y. Jury Bias

A. Standard of Review

Petitioner asserts that extensive news coverage — about the robbery, about the state’s efforts to bring petitioner to trial, about petitioner’s alleged attempt to escape from the Charles Street Jail before his trial, and about petitioner’s prior criminal record — made it impossible to have an impartial jury. Petitioner contends that the trial judge either should have granted continuances until prejudicial community sentiment abated or else should have granted a change of venue to a location where the prejudicial community sentiment did not prevail. Petitioner also asserts that at least jurors actually seated displayed bias against petitioner.

Petitioner need not show actual juror bias to show that he was prejudiced by negative publicity. It is sufficient to prove inflammatory, sensational publicity “so saturate[d] a community from which [the jury was] drawn as to render it virtually impossible to obtain an impartial jury.” United States v. Angiulo 897 F.2d 1169, 1181 (1st Cir.1990). Alternatively, one may make an adequate showing of prejudice by proving that a high percentage of the venire admits to a disqualifying prejudice. Id. Of course, petitioner may also show actual prejudice among the jurors seated at trial. Id. at 1182.

To show jury partiality, however— whether circumstantially or by directly showing bias — petitioner must meet a heavy burden. Since the challenge here is on collateral review in a habeas corpus petition, the finding of the trial judge that the assertedly biased jurors were not biased in fact is entitled to a presumption of correctness under 28 U.S.C. § 2254. Patton v. Yount 467 U.S. 1025, 1036, 104 S.Ct. 2885, 2891, 81 L.Ed.2d 847 (1984). This is so because the determination of bias of individual jurors is a question of historical fact. The trial court’s determination, made only after a voir dire proceeding “designed specifically” to identify bias, is one which is “essentially one of credibility, and therefore largely one of demean- or.” Id. at 1038, 104 S.Ct. at 2892. To the extent that the trial court made a more general determination that pervasive publicity had not biased the jury, that determination will be reversed only if it is shown to be “manifest error.”

Compare Irvin v. Dowd 366 U.S. 717, 723, 81 S.Ct. 1639, 1642-43, 6 L.Ed.2d 751 (1961) (applying “manifest error” standard to claim of jury partiality) with Patton, 467 U.S. at 1032 n. 7, 104 S.Ct. at 2889 n. 7 (doubting whether there is any practical difference between the manifest error standard and the presumption of correctness, which presumption was added to the statute subsequent to Irvin)-,

see also Amirault v. Fair 968 F.2d 1404 (1st Cir.1992) (trial court findings on issues of juror credibility and honesty are peculiarly within a trial judge’s province and are accorded great deference).

Similarly, to show that the trial judge erred in failing to grant a continuance or change of venue, petitioner must also meet a heavy burden. The Court of Appeals for the First Circuit, in reviewing direct appeals of defendants in federal criminal proceedings, has applied an abuse of discretion standard to cases in which the trial judge failed to grant a continuance or change of venue on the grounds of prejudicial publicity.

See United States v. Moreno Morales 815 F.2d 725, 739 (1st Cir.1987) (failure to grant continuance).

United States v. Drougas 748 F.2d 8, 29 (1st Cir.1984) (failure to grant change of venue).

These eases involved the exercise of the supervisory power of the Court of Appeals. Thus it is possible that the standard of review that would apply on collateral review would be even less favorable to the claim of error than an abuse of discretion standard. In any event, however, there is no reason to think that the standard on collateral review would not be any more favorable to the petitioner than the abuse of discretion standard; the eases just cited, like petitioner’s case, involved asserted denials of a constitutional right to a fair trial. Applying an abuse of discretion standard, I determine that the trial court did not err in failing to grant a second continuance.

Great deference to the findings of the trial court is particularly appropriate here in light of the efforts made by the trial court to prevent prejudice to the petitioner. The trial was originally set to begin in April, 1971, already at least five months after the commission of the crime in question. The trial court granted a continuance, and the trial was delayed until the beginning of February 1972; thus the trial took place sixteen months after the robbery. Subsequent requests for a continuance, and requests for a change of venue, were denied.

[The judge] addressed all potential jurors in the strongest language as to publicity and its possible effect on them. The judge’s interrogation of individual jurors was resourceful and thorough; he invited suggestions from counsel for questions to be put by the court to jurors as to publicity. His interrogation of the jurors as to matters of prejudice was painstaking and resourceful, and he resolved all doubts in favor of the defendant. He sequestered the jury after empaneling____ We think it was within the judge’s discretion, in light of the extensive and careful measures taken by him, whether to accept, without more, the declaration of the jurors as to their disinterest and freedom from emotional or intellectual commitment.

Gilday I, 327 N.E.2d at 862.

I cannot conclude that both the trial judge and the Supreme Judicial Court erred in reaching these essentially fact-based determinations.

B. Actual Bias

Petitioner asserts that “many of the jurors who were eventually seated had shown considerable bias.” Pet.Memo Supp.Pet. at 57.

In considering an allegation of jury bias, one must keep in mind that the law does not require

that the jurors be totally ignorant of the facts and issues involved. In these days of swift, widespread and diverse methods of communication, an important case can be expected to arouse the interest of the public in the vicinity, and scarcely any of those best qualified to serve as jurors will not have formed some impression or opinion as to the merits of the case. This is particularly true in criminal cases. To hold that the mere existence of any preconceived notion as to the guilt or innocence of an accused, without more, is sufficient to rebut the presumption of a prospective juror’s impartiality would be to establish an impossible standard. It is sufficient if the juror can lay aside his impression or opinion and render a verdict based on the evidence presented in court.

Irvin, 366 U.S. at 722-23, 81 S.Ct. at 1642-43. With this substantive standard in mind, and according great deference to the trial court’s findings that the jurors actually seated met this standard, I conclude that petitioner has failed to show that the jury actually seated was biased.

According to petitioner, Juror Faretra “agreed that he knew that Gilday was involved with the bank robbery or murder.” Pet.Mot.Supp.Pet. at 57. A review of the record shows that Faretra, under questioning by the trial court, stated that although he had “heard or read some papers” about the events in question, he had not formed any opinion about the guilt or innocence of the accused, and that he would be able to decide that question based solely on the evidence he heard in the courtroom. Trial Record at 1280-82. Under questioning by defense counsel, Faretra testified that he had “no details in [his] mind” about the case. Id. at 1286. When asked, however, whether the name “Gilday” meant anything to him, however, he said that it did, and when asked by defense counsel whether it was “that he was involved in a bank robbery and the murder of a police officer,” Faretra answered affirmatively. Id. at 1286-87. Defense counsel then asked the following:

So that, at least, even though you don’t remember the details as you sit there now, your memory of what you have read and what you have learned was that Gilday in some way was involved in a bank robbery or murder?

Id. Faretra answered:

He is involved with some act like that, yes. I haven’t recollected anything. I mean, I didn’t anticipate that. That is the reason.

Id. Defense counsel then stated:

Right. But at least, as you sit there now, you are willing to be unconvinced of that, depending on the evidence.

Id. The government objected to this question, and the Court asked Faretra whether he felt

actively or subconsciously so that if you sat as a juror, Gilday would have the burden of putting that out of your mind? Has he got to unconvince you?

I haven’t any thought or feeling on it at this moment, absolutely nothing. I can’t say anything one way or another about the matter.

Id. Defense counsel then interjected:

Except that you did tell me that Gilday was involved in a bank robbery and murder, at least so far as what you read in the paper?

Id. at 1288. Faretra answered:

I haven’t followed this thing, to be perfectly honest about it. I didn’t follow this case at all.

Id. Defense counsel then challenged Faretra for cause, having exhausted his peremptory challenges. The court denied the motion, and Faretra was sworn. Id.

Reviewing this exchange, I am unable to determine that the trial judge erred in finding that Faretra had no bias. As noted, it was not required that Faretra’s mind be a blank slate, devoid of any knowledge of media coverage, or even devoid of any opinion on the merits of the case. Faretra stated that he had formed no opinion, and would base his judgment solely on the evidence in the court; he maintained this position despite the attempt of defense counsel, by means of leading questions, to elicit from him a statement to the effect that he had formed such an opinion with respect to Gilday and that he would regard it as Gilday’s burden to “unconvince him.” This court must give deference to the determination of the trial court that Faretra had no bias. This is particularly so because the trial court had the opportunity to observe Faretra’s demeanor in responding to the leading questions asked by defense counsel.

Petitioner asserts that three other members of the jury actually seated demonstrated bias. It appears, however, that petitioner has misread the record, for although Faretra was seated, the other three members of the venire identified by petitioner as having been seated (Chamberlain, Comerford, and Stone) were not.

See Trial Record at 4375 (jury returns with its verdict, and Jurors Luck, Tartaglini, Pallozzi, McCann, Mulkerron, Marnell, Goodwin, Smith, Todisco, Faretra, McDonald, and Choo answer present, as well as four alternates, Kobialka, Egdall, Holden, and DeAngelis).

Petitioner may have been mislead by the practice of the trial court and court reporter of referring to members of the venire as jurors. In this regard it should also be noted that questions posed to venire members as to whether other “jurors” had been speaking about media coverage of the alleged offenses were referring to discussions among members of the venire, not among members of the jury finally seated. Since these venire members were not actually seated, petitioner’s contentions with regard to them are irrelevant to a finding of actual bias of the deliberating jury.

Furthermore, review of the voir dire of these three venire members does not reveal anything more indicative of bias than did the voir dire of Faretra. Thus, if I take these venire members as a sample of community sentiment as a whole, keeping in mind that petitioner has drawn my attention to these venire members as examples of bias, review of their testimony buttresses my finding below that media coverage did not lead to pervasive community sentiment against petitioner such as would render the trial constitutionally invalid.

C. Community Sentiment

As explained above, jury prejudice may be shown by (or, as is often stated, will be presumed from) proof to a high level of probability that community sentiment was pervasively against the defendant. To show such conditions, however, is extremely difficult. Petitioner asserts that publicity was extensive and inflammatory. Respondent concedes that there was widespread pretrial publicity. Answer at 1.

As evidence of pervasive community sentiment, petitioner asserts that 51 out of 132 (38.6%) members of the venire who were not previously excused for other reasons were dismissed by the judge on the grounds that they were biased. Although this assertion, if true, would tend to indicate that pretrial publicity was pervasive, it does not tend to show that sentiment against petitioner was so overwhelming that the jury actually selected by the trial court should be presumed to be biased. The Court of Appeals for the First Circuit has stated that even if a greater percentage of the venire had a fixed opinion of petitioner’s guilt, this would not be enough to impeach the impartiality of those jurors actually seated.

Moreno Morales, 815 F.2d at 735 n. 12 (stating that 42.85% proportion of committed venire members would be insufficient to allow presumption of prejudice and citing Patton, supra for the proposition that even a proportion of 77% venire members would be insufficient).

I conclude that the percentage of biased venire members was not high enough to give rise to a presumption that the jurors actually seated were biased. This conclusion is particularly warranted in light of the extensive precautions taken by the trial court here to ensure that those actually seated had not been impermissibly influenced by outside opinion.

More generally, petitioner’s contentions as to the nature of the publicity in this case prove too much. Petitioner notes that the trial judge and prosecutor expressed the view that community knowledge of the robbery were never going to entirely abate. Petitioner’s contention that pretrial publicity was pervasive is premised, at least in part, on his assertion that “every pretrial motion and argument of the defense or prosecution caught the media’s attention.” Petitioner cites as an example of prejudicial publicity a story in the Herald Traveler on petitioner’s request for a second continuance.

The proffered evidence tends to show that if petitioner had succeeded in obtaining further continuances, media interest would have continued unabated. Indeed, recent events have shown that murder of Officer Schroeder continues to excite considerable interest even after more than twenty years have elapsed. Petitioner’s assertion that he was entitled to avoid trial so long as his case remained a matter of intense public interest, under the circumstances, amounts to an assertion that he was entitled to delay trial indefinitely.

A reviewing court, in considering the totality of the circumstances to determine whether the decisions of the trial court to deny continuances, to deny venue changes, and to seat particular jurors were “manifest error,” or an abuse of discretion, must consider the options available to the judge. Where circumstances strongly suggest that the legal maneuvering surrounding the trial was generating publicity on its own, the trial judge can hardly be faulted for deciding that granting a further continuance, in addition to the initial continuance of ten months, would not significantly lessen (and might worsen) the publicity surrounding the case. To the extent that petitioner’s argument suggests that he is entitled to continuances ad infinitum, the contention must be rejected.

In Moreno Morales, the Court of Appeals for the First Circuit noted:

The denial of a motion for an indefinite or substantial continuance predicated upon widespread adverse pretrial publicity about a defendant is all the more warranted when, as here, there is sound reason to believe that the defendant will continue to be a controversial, pubhcity-invoking figure and, hence, that there is little assurance that the passage of time will result in an abatement or subsidence of critical publicity in the foreseeable future.

815 F.2d at 737 (internal citation omitted). The Court of Appeals noted that the Sixth Amendment guarantee of a speedy trial is one to which the government and society, as well as the accused, has a claim, and stated that society “cannot be utterly deprived of the right to prosecute and try someone for a crime within a reasonable time merely because of widespread community knowledge.” Id. at 739.

For similar reasons, petitioner’s assignment of error to the failure to grant a venue change must also fail. In Moreno Morales, defendant had declined to seek a venue change, and because of this failure the Court of Appeals placed a heavier burden on the defendant to show error in the failure to grant a further continuance. Id. at 739. Nevertheless, the reasoning of Moreno Morales does not suggest that a defendant will ever have an absolute right to be granted either a venue change or indefinite continuance, should he request both and show the present existence of prejudicial community sentiment. The reasoning in Moreno Morales suggests that where publicity is so pervasive that neither a change of venue nor a continuance would be likely to diminish community prejudice, the trial court is within its discretion in denying both.

Here petitioner cannot reasonably assert that prejudice to the defendant would have been significantly reduced had the trial been moved to another part of the Commonwealth. Petitioner has cited United States v. Abrahams, 453 F.Supp. 749 (D.Mass.1978) in support of the proposition that a change of venue is warranted. In fact, the case supports the opposite conclusion. That was a federal case, and there the trial court was able to order that the trial take place in a United States district court in Texas, where, it was reasonable to believe, community sentiment was less inflamed. Petitioner here was tried in state court. No showing has been made that the trial judge could have ordered that the trial occur outside the Commonwealth. The court in Abrahams expressly found that a change of venue from Boston to Springfield could not have significantly reduced prejudice in a well-publicized case, noting that although Springfield had its own newspapers and radio and television stations, it was also a major market for the two Boston daily papers that carried prejudicial information in that case. Id. at 754. Since petitioner in his briefs argues that “the very nature of [petitioner’s] case served to inflame the public in a way that Abrahams’ case never could,” Abrahams does not support his argument that a change of venue to another site within Massachusetts would have reduced prejudice to him.

For these reasons, petitioner’s motion for an evidentiary hearing on pretrial publicity (Docket no. 125) and his request for funds to secure a news service are denied. Petitioner has not shown the existence of a factual dispute that, if resolved in his favor, would support his claim for relief.

See, e.g., United States v. Hadfield, 918 F.2d 987, 992 (1st Cir.1990) (evidentiary hearing not required unless defendant makes threshold showing that material facts are in doubt or dispute), cert. denied, 500 U.S. 936, 111 S.Ct. 2062, 114 L.Ed.2d 466 (1991);

see also Copp v. United States, 968 F.2d 1435, 1438 n. 1 (1st Cir.1992) (threshold showing required);

Weinberger v. Great Northern Nekoosa Corp., 925 F.2d 518, 527-8 (1st Cir.1991) (same);

VI. Trial Errors

Until recently, federal courts, including the Supreme Court in “a handful of eases,” Brecht v. Abrahamson, — U.S. -, -, 113 S.Ct. 1710, 1712, 123 L.Ed.2d 353 (1993), have applied to collateral review of constitutional errors in state court proceedings the harmless error standard of Chapman v. California, 386 U.S. 18, 87 S.Ct. 824, 17 L.Ed.2d 705 (1967), namely, whether the error was “harmless beyond a reasonable doubt.” In Brecht, the Supreme Court held that, in a collateral proceeding, a different harmless error standard applies to some alleged constitutional errors. The precise issue in Brecht concerned the prosecution’s use, for impeachment purposes, of petitioner’s silence after a Miranda warning. The Court held that this matter is appropriately analyzed not under Chapman but rather under the harmless-error standard of Kotteakos v. United States, 328 U.S. 750, 776, 66 S.Ct. 1239, 1253, 90 L.Ed. 1557 (1946) — namely, whether the error “had substantial and injurious effect or influence in determining the jury’s verdict.” Brecht, — U.S. at -, 113 S.Ct. at 1712. Until Brecht, the Kotteakos standard had been reserved for review of noneonstitutional errors in federal criminal trials.

In Brecht the Court reasoned that the Kotteakos standard was more appropriate on habeas review than the Chapman standard because the Kotteakos standard makes it harder to overturn a judgment. This standard, more demanding of the petitioner, is more accommodating of society’s interest in the finality of convictions. Id. at---, 113 S.Ct. at 1711-12.

Nevertheless, the Court noted that there is a continuum of constitutional errors. At one end are errors — called “trial errors” — that are to be reviewed under the Kotteakos standard. At the other end are errors so fundamental that they are not subject to harmless-error review at all. Id. at---, 113 S.Ct. at 1710-11.

The prosecutorial misconduct at issue in Brecht was a “trial error.” It occurred during the presentation of the case to the jury and might be assessed in the context of other evidence presented at trial. Id. (citing Arizona v. Fulminante, 499 U.S. 279, 111 S.Ct. 1246, 113 L.Ed.2d 302 (1991)).

The Court cited, as an example of an error that was not subject to harmless-error review, the deprivation of counsel — an error that “de[fles] harmless error analysis” because it concerns a structural defect of the trial mechanism that “infect[s] the entire trial process.” Id. (also citing Fulminante).

A second example of an error that would not be subject to the Kotteakos harmless-error analysis because it concerns a “structural defect” is a charge to the jury that fails to instruct them correctly on the government’s burden of proving guilt beyond a reasonable doubt.

See Sullivan v. Louisiana, — U.S. -, 113 S.Ct. 2078, 124 L.Ed.2d 182 (1993) (constitutionally deficient instruction on reasonable doubt can never be harmless error; even if the evidence against defendant were overwhelming, defendant was denied his constitutional right to have the determination of guilty-beyond-a-reasonable-doubt made by a jury).

Respondent concedes that, in light of Sullivan, the reasonable doubt instructions are not subject to harmless-error analysis, but he urges that the Kotteakos standard be applied to the other errors asserted in this case.

The reasoning of Sullivan is compatible with a contention that the Kotteakos standard does not apply to claims of Sandstrom error as well, since these claims similarly involve the trial court’s instructions to the jury, and not the presentation of evidence to the jury. I need not reach that question, however, since I have rejected the Sandstrom claims on other grounds.

Similarly, I need not decide whether constitutionally impermissible jury bias is subject to Chapman or to Kotteakos harmless error-analysis, or no harmless error analysis at all, because I have determined that no such error has been shown in this case.

Petitioner’s remaining claims, however, are subject to the Kotteakos standard, because they do concern errors of the “trial type” in that they affected the presentation of evidence to the jury. These are contentions that the Commonwealth failed to disclose material exculpatory information in its possession, in violation of Brady v. Maryland, 373 U.S. 83, 83 S.Ct. 1194, 10 L.Ed.2d 215 (1963), and that the Commonwealth purposely used false testimony in violation of Giglio v. United States, 405 U.S. 150, 92 S.Ct. 763, 31 L.Ed.2d 104 (1972).

Petitioner urges, to the contrary, that the Chapman standard be applied to all the trial-type errors in this case, in light of the following reservation expressed by the Brecht Court in a footnote:

Our holding does not foreclose the possibility that in an unusual case, a deliberate and especially egregious error of the trial type, or one that is combined with a pattern of prosecutorial misconduct, might so infect the integrity of the proceeding as to warrant the grant of habeas relief, even if it did not substantially influence the jury’s verdict.

Brecht, — U.S. at - n. 9, 113 S.Ct. at 1722 n. 9. Reviewing the errors that were alleged in this case, however, as discussed in more detail below, I do not conclude that this is the sort of “unusual” case referred to by the Court. Thus I do not reach the question reserved by the Supreme Court, namely whether in such an “unusual case” an exception should be carved out of the general rule that “errors of the trial type” are to be reviewed for harmless error according to the Kotteakos standard.

As the Court noted in Brecht, application of the Kotteakos standard means that habeas petitioners will no longer be “entitled to habeas relief based on trial error unless they can establish that it resulted in ‘actual prejudice.’ ” Id. at 4340 (citing United States v. Lane, 474 U.S. 438, 449, 106 S.Ct. 725, 732, 88 L.Ed.2d 814 (1986)). This is a substantial shift.

See Brecht, — U.S. at-, 113 S.Ct. at 1723 (Stevens, J., concurring) (Kotteakos harmless error analysis is “appropriately demanding” for collateral review, where review need not be as demanding as that required on direct review);

Lane, 474 U.S. at 472 n. 11, 106 S.Ct. at 744 n. 11 (Stevens, J. concurring in part and dissenting in part) (noting “considerable difference” between Kotteakos and Chapman standards).

Nevertheless it should be kept in mind that the Kotteakos standard, like the Chapman standard, is still an inquiry into what would or would not have influenced the jury verdict, rather than a direct inquiry by the reviewing court as to whether the evidence introduced at trial showed defendant to be guilty. See Brecht, — U.S. at---, 113 S.Ct. at 1722-23 (concurring opinion of Stevens, J.). “Overwhelming evidence of guilt” was previously relevant under the Chapman standard not because it showed petitioner to be guilty but because it showed that petitioner would have been convicted despite the error. Similarly, “actual prejudice” may be shown despite the fact that there was sufficient evidence to support the conviction.

In order for the suppression of evidence by the prosecution to be a constitutional violation, the evidence suppressed must be both exculpatory and material. Brady, 373 U.S. at 87, 83 S.Ct. at 1196-97. Whether exculpatory information is “material” goes to the impact the excluded information would likely have had on the jury, had it been disclosed.

See United States v. Agurs, 427 U.S. 97, 96 S.Ct. 2392, 49 L.Ed.2d 342 (1976) (stating a three-tiered test of materiality);

United States v. Bagley, 473 U.S. 667, 682, 105 S.Ct. 3375, 3383-84, 87 L.Ed.2d 481 (1985) (stating unitary test of materiality).

In this respect the materiality inquiry is similar to the harmless-error analysis. Since this case is a collateral review of alleged errors in a state court proceeding, the Brecht decision requires that^ the reviewing court deny habeas relief unless the errors had substantial or injurious effect in determining the jury verdict. Thus, whether or not an error is “material,” petitioner will not prevail unless the error actually prejudiced him by having a substantial and injurious effect or influence in determining the jury’s verdict.

Reviewing the alleged errors in the context of the totality of the evidence presented at trial, I determine that petitioner was not actually prejudiced at trial. The discussion that follows refers to those aspects of the evidence presented at trial that particularly support this determination. This is not to say, however, that the evidence I have described in this opinion is the only evidence in the trial record that tends to support this determination. A summary of the evidence that was presented at trial appears in an appendix at the end of this opinion.

B. Petitioner’s Conviction

For two reasons it may be useful to state briefly the inferences to be drawn from the charge and the verdict about what findings the jury made in order to arrive at their verdict. First, engaging in this analysis helps one think about what evidence would have been “exculpatory.” Second, it helps one think about whether the failure to pror vide the “exculpatory” information was “material,” and whether it “had a substantial and injurious effect or influence” in determining the verdict.

Here, petitioner was convicted of first degree murder. The judge charged the jury that they could find petitioner guilty of first degree murder if they found either that petitioner had committed homicide with deliberately premeditated malice aforethought, or (under the felony murder instruction) that petitioner had committed (or attempted to commit) an armed robbery, in the course of which the homicide occurred. The judge also instructed the jury that criminal liability could attach based upon a finding that petitioner had participated in a joint venture to commit the armed robbery. Since the jury also convicted defendant of two counts of armed robbery, and since a killing undeniably occurred in that robbery, it is a compelling inference that the jury found defendant guilty on the felony murder theory, although they may have found him guilty on a premeditation theory as well.

In Gilday I, the Supreme Judicial Court made the following observation (declaring, however, that it had not served as a basis for the court’s determination that the errors there charged were harmless beyond a reasonable doubt):

The admissions of the defendant on the witness stand would have been even more extraordinary if the Commonwealth had chosen to proceed on proof that the defendant was an accessory before the fact____ Among other things the defendant testified that he procured the guns and an automobile that were used in the bank robbery; that he stole a license plate and affixed it to another of the vehicles used in the robbery; that he disposed of one of the vehicles after the robbery; that he took a substantial share of the loot; and that he knew Bond, Valeri and the two girls were planning something but did not know exactly what.

327 N.E.2d at 866 n. 4.

Petitioner argues that

The Supreme Judicial Court’s dictum in Gilday II that Gilday could have been found guilty as an accessory before the fact (where presence is not an element) has no place in our system of justice. Gil-day’s case was never premised on such a theory of guilt, which requires different evidence and a separate indictment. Commonwealth v. Di Stasio, 297 Mass. 347, 357, 8 N.E.2d 923, 930 (1937) (“Since the offenses are distinct, proof of guilt as an accessory before the fact would constitute a variance upon an indictment as princi pal”); Nolan & Henry, Criminal Law, 32 Mass. Practice Series §§ 632-633. Gilday was indicted, tried, and found guilty solely as a principal.

Petitioner’s Memorandum at 84 n. 39. Petitioner’s statement of Massachusetts law is not quite accurate. It is true that at one time there were two distinct offenses of “principal” and “accessory before the fact.” But that distinction was abolished in 1968— before petitioner’s trial. St.1968 e. 206, (amending M.G.L. ch. 274, §§ 2 & 3 so that accessories before the fact shall be indicted, tried, and punished as principals). In Commonwealth v. Benjamin, 358 Mass. 672, 266 N.E.2d 662 (1971), the Supreme Judicial Court explained that

[wjhether the defendant be indicted as a principal under the 1968 statute or under the preexisting provisions ... the substance of the crime is the same____ The new statute merely calls the offender a principal instead of calling him an accessory____ That formerly, for some purposes, the crimes were separate was an artificial matter going chiefly to the form of the indictment.

Id. at 680, 266 N.E.2d at 667. The authorities cited by petitioner for the proposition that liability as an accessory before the fact is distinct from liability as principal are not on point, for they rely on a differently-worded statute than the one in effect at the time of petitioner’s offense and trial.

Thus, as the Supreme Judicial Court noted, petitioner’s own admissions went a long way toward establishing that he was guilty of the offense charged by way of accessorial liability. Accessorial liability is established where the defendant assists the principal in the commission of the crime, sharing with the principal the mental state required for that crime. Commonwealth v. Richards, 363 Mass. 299, 293 N.E.2d 854, 860 (1973). Petitioner’s admissions showed at the very least that he assisted others (who were “principals” even in the pre-1968 sense) in their commission of an armed robbery, in the course of which a killing occurred. Petitioner did not go so far as to admit that he had the specific intent to commit or to assist the commission of an armed robbery. (As we shall see, in Part VI.D below, however, this shortfall does not render petitioner’s admissions insignificant).

At various points in the briefs, petitioner appears to assert that to find petitioner guilty, the jury would have had to find that he was the person who actually shot Officer Sehroeder. This is not true. First, as explained above, under the statutory provision in effect in 1970, he might have been indicted, tried, and charged as a principal even if he was only liable as an accessory before the fact. Second, even under the pre-1968 law, the distinction between accessories before the fact and principals did not place all accomplices in the category of accessory before the fact; someone present at, or near, the scene of the crime, who aided and abetted the commission of the crime, could be found guilty as a principal.

See, e.g. Commonwealth v. Locke, 335 Mass. 106, 111, 138 N.E.2d 359, 363 (1956) (defendant seen running shortly after robbery with another person, who held a ladies handbag, and who escaped with defendant in defendant’s automobile, could be found guilty as a principal);

Commonwealth v. Knapp, 27 Mass. (9 Pick.) 495, 518 (1830) (defendant is a principal in the second degree if he is present at the scene of the crime, is in a situation where he might render aid to the perpetrator, and by agreement or previous knowledge, has consented to do so, with a purpose of rendering aid and encouragement to the perpetrator).

See also Commonwealth v. Mannos, 311 Mass. 94, 108-110, 40 N.E.2d 291 (1942) (“It is a general rule in criminal law that one who aids and assists another person to commit an offense which only such other person can commit may be equally guilty with the latter as a principal or as an accessory before the fact, depending upon the evidence showing his relation to the actual commission of the crime; or as principal to such other person in jurisdictions where the distinction between principals and accessories has been abolished by statute”).

As explained below, the record shows that the jury, in returning a finding of guilty, must have at the very least found that petitioner was present at the scene of the crime, whether or not they found that he physically did the shooting.

The court notes that the although the statute was amended in 1968 to abolish the distinction between principals and accessories before the fact, the legislature in 1973 amended the same statute yet again, so that it read exactly as it had before 1968. See St.1973, c. 529 § 1 (amending M.G.L. e. 275 § 2). This has no bearing on the present case, however, for the offense was committed in 1970 and petitioner was tried in 1972. During that time the relevant law was the 1968 amendment, and the law was as stated in Benjamin: the distinction between principal and accessory before the fact had been abolished.

At trial, petitioner primarily relied upon an alibi defense. Petitioner asserts that as a matter of law one cannot be found liable on a joint venture theory unless one is present at the scene of the crime, citing Commonwealth v. Giang, 402 Mass. 604, 608, 524 N.E.2d 383, 386 (1988), and contends that the trial court failed to instruct the jury that to find petitioner liable on a joint venture theory they must find that he was present at the scene of the crime. The statute and ease law do not indicate that presence at the scene of the crime is a prerequisite to accessorial liability, but this court need not reach that question, because despite petitioner’s assertion to the contrary, the jury was instructed that they must find that petitioner was present at the scene of the crime.

The trial court instructed the jury that the Commonwealth had the burden of disproving petitioner’s alibi, and that a failure to do so would negate an essential element of the offense. Trial Record at 4325. Whether or not the law required the court so to instruct the jury, it is appropriate in this collateral proceeding to presume that the jury followed these instructions. Thus in returning a verdict of first degree murder, the jury must have found that petitioner was present at the scene of the crime, and that he intentionally assisted the commission of an armed robbery.

In determining whether evidence was “exculpatory,” “material,” or “had substantial or injurious effect,” it is appropriate to look to the law as the jury was charged. It is not appropriate for the court instead to evaluate the evidence independently according to its own understanding of the law. Thus, for purposes of this decision, evidence is exculpatory to the extent that it would have tended to assist petitioner in his effort to show that he was not present at the scene of the crime. Evidence that merely would have thrown doubt on whether petitioner fired the fatal shots will not be viewed as exculpatory, however, unless it also would have thrown doubt on the petitioner’s presence at the scene of the crime, and then only as to its effect in this latter way.

C. The Alleged Errors

1. Use of Allegedly False Testimony of Witness Fleischer

Petitioner asserts that the witness Fleischer gave “false” testimony at the trial as to whether an arrangement of leniency existed with the prosecution. This asserted error was the basis for petitioner’s second motion for a new trial. This court will adopt the factual findings rendered by the motion judge in that proceeding. Gilday II, 415 N.E.2d 797, 802-04. See 28 U.S.C. 2254(d) (presumption of correctness of state court findings). The prosecutor, without reaching a formal agreement with Fleischer’s attorney, told that attorney before trial that he would do everything he could to bring it to the court’s attention that Fleischer had cooperated in giving a statement to the police and that if Fleischer did testify, his cooperation would also be made known to the court so that a disposition could be made of his case that would not leave him with a criminal record. Evidence was offered at the hearing on the motion for new trial that Fleischer’s attorney had told Fleischer that it would be “in his best interests to testify in the case.”

At trial, Fleischer denied on cross-examination that he had made any deal with the prosecution, and instead told the jury that the only promise that had been made was that the prosecution would “ask for high bail but not demand it.” He said that the only reason he was testifying was because a man had been killed and he wanted to “see justice done.” The motion judge found that this was not perjury, there being no evidence that Fleischer knew of the informal arrangement that had been made.

Petitioner argued that even without proof of intentional falsehood, the prosecutor had a duty not to remain silent when Fleischer stated facts the prosecutor knew to be untrue or misleading. Id. at 802-03. The Supreme Judicial Court agreed that, in the circumstances, the prosecutor had “a duty to make a seasonable disclosure to the defense and the trial judge.” The Court reasoned that the prosecutor was chargeable with knowledge that Fleischer testified with the expectation of leniency. Id.

Petitioner asserts that the Supreme Judicial Court found a violation of Giglio, although he concedes that the Court found the error to be harmless. Pet.Memo.Supp.Pet. at 5, 20. This is not strictly true, although the decision is not explicit on this point. The court stated that the prosecutor had a duty to make a seasonable disclosure, and that he was not excused from that duty by the lack of a formal agreement with the witness. 415 N.E.2d at 803. The Court stated that “[i]f we failed to reach these conclusions on the facts of this case, we would in effect approve the evasion of the Giglio rule by means of artful device.” Id.

Although this last statement might initially appear to be a finding of a Giglio violation, the remaining discussion makes it clear that the court had not reached that ultimate determination yet. The court next turned to the question whether the prosecutor’s failure “did not influence the jury so as to require a new trial.” This was a materiality inquiry. The court recited the three-tiered Agurs standard for determining materiality, and then stated that “even applying the most stringent standard, however, we like the motion judge conclude that Gilday is entitled to no relief’ because the error was harmless beyond a reasonable doubt. At a very minimum, the court declined to reach the materiality question, predicating its ruling on the harmlessness of the error. A fair reading of the decision, however, is that the court held that an error that is “harmless beyond a reasonable doubt” cannot meet “even ... the most stringent” materiality test, that is, the test most favorable to the claim of error. In other words, the court’s comment that the prosecutor had a Giglio duty under the circumstances meant only that the prosecution’s structuring of the arrangement of leniency would not, by itself, immunize the prosecution against a potential Giglio error. The statement did not mean, however, that, given all the evidence that was presented in that ease, the information so suppressed was constitutionally material.

In any event, even if petitioner were correct that the Supreme Judicial Court found a Giglio violation, and even if this court were to find such a violation, the finding would be of no importance if the Giglio error were harmless. As explained above, this court must apply the Brecht/Kotteakos standard, which is more lenient toward the government than the Chapman standard that the Supreme Judicial Court applied.

Thus, I do not reach the question as to what standard of materiality to apply, and whether that standard is met, since I do not find that the prosecutor’s conduct “actually prejudiced” petitioner under the Kotteakos standard. I find that there would have been no substantial and injurious effect or influence in determining the jury’s verdict if the information — material or not — that was suppressed had instead been disclosed to the defense. I so find, even assuming arguendo that the disclosure of this evidence would have enabled the defense to completely discredit Fleischer’s testimony. This finding is based in part on my determination that Fleischer’s testimony was merely cumulative of other information put before the jury by the witness Valeri and to some extent by the witness McGrory. As stated below, the testimony of Valeri and McGrory is not tainted by constitutional error and should not be disregarded when conducting a harmless-error analysis. My determination on this matter is further buttressed by other inculpatory evidence that was before the jury, some of which is discussed below.

2. Use of Allegedly False Testimony of Witness Valeri

Petitioner asserts that a similar deal was arranged with the witness Valeri, and that a similar Giglio violation took place. Two jailhouse inmates, and defense counsel, claimed to have spoken with Valeri and elicited an admission from him to this effect. This issue was raised in petitioner’s first motion for new trial. At the hearing on the motion, Valeri testified that he had testified truthfully at trial. The motion judge found that Valeri had testified truthfully and that no promises or threats had been made him. Gilday I, 327 N.E.2d at 860-61. Respondent urges that this finding is entitled to a presumption of correctness.

Petitioner asserts, however, that the factual question as to whether Valeri did in fact perjure himself at trial should be reopened. Petitioner has brought to the attention of this court an affidavit signed by Valeri, dated April 1983, subsequent to Gilday I. Petitioner asserts that this affidavit contains “new information,” namely that Valeri lied at the state hearing, thus rebutting the presumption of correctness that would otherwise be accorded the state court finding. Petitioner has moved for an evidentiary hearing on this question.

Petitioner had previously moved to depose Valeri. (Docket No. 34, filed July 18, 1983). Magistrate Judge Alexander denied the motion on the ground that petitioner had not shown that the deposition of Valeri would lead to relevant evidence, and had only reiterated the findings of fact made by the motion judge and Supreme Judicial Court. Gilday v. Callahan, 99 F.R.D. 308 (1983) Judge McNaught of this court affirmed.

I adopt Magistrate Judge Alexander’s finding that Valeri did not commit perjury at trial, and her conclusion that this affidavit does not indicate that further evidentiary exploration of this issue would yield relevant evidence. In relevant part, the affidavit states that

prior to the trial of William M. Gilday, both [the prosecutor] and my attorney advised me that they could not make additional promises to me personally other than that my cooperation would be brought to the attention of the Court at the disposition of the charges pending against me. Thus, I would be able to truthfully testify that no specific promises had been made to me. [My lawyer] advised me that my cooperation at Mr. Gilday’s trial would “be in my best interest,” and [the prosecutor] repeated this advice to me.

Petitioner argues that this affidavit is a strong indication that an arrangement (a “deal,” petitioner asserts) parallel to that worked out with Fleischer took place with Valeri. He also argues that even if it were true that no “specific” promises had been made to him, it is not correct to assert that no deal was made, in that a more general promise to seek leniency was made. Petitioner asserts that a vague promise is actually more likely to induce favorable testimony, because the witness is kept in anxiety as to the leniency he is to receive.

There are aspects of this affidavit that indicate that petitioner may well be correct that an arrangement parallel to that with Fleischer existed with Valeri. Although the affidavit does not directly say that such a parallel arrangement was made, the statement, “thus I would be able to truthfully state that no specific promises had been made,” combined with the advice that it would be in his best interests to testify, support an inference that the same arrangement was made as was made with Fleischer. Since this affidavit was not brought to the attention of the motion judge, and because evidence of the arrangement with Fleischer was not known to the defense or to the motion judge at the time of that ruling, there may be reason not to accord the presumption of correctness to the state court finding that “no promises were made.”

Nevertheless, petitioner’s ultimate claim fails. The relevant factual question that must be answered is not whether any promises were made, but whether any promises were made that were not disclosed at trial. Assuming, arguendo, that a parallel arrangement was made with Valeri, the situation is distinguishable as to impact on the jury verdict. The key difference is that Valeri and the government were considerably more forthcoming in their testimony and arguments at trial regarding the existence of any such arrangement with Valeri, than were Fleischer and the government regarding the arrangement with Fleisher.

At trial, Valeri testified that he had been indicted for armed robbery and for first degree murder. Trial Record at 2370. He also testified that he had pled guilty to a charge of larceny of government property, for which he had not been sentenced yet. Id. at 2395.

The following testimony was given under cross-examination by defense counsel:

Q: Have any promises or considerations of any nature been made to you by anyone concerning your testimony in this ease?

A: Promises? You mean so far as bringing it to the Court’s attention of my cooperation, or—

Q: Of any nature.

A: No out-and-out promises.

Q: Pardon?

A: No out-and-out promises, no sir.

Q: Has anything at all been said to you that any consideration would be given to you for your testimony in this case?

A: No, sir. It was just mentioned that it would be brought to the Court’s attention that I have cooperated in this case.

sj: * :{: ‡ ‡ sfc

Q: And who said that?

A: Mr. Gaffney [the prosecutor] told me that, sir.

Id. at 2370-71.

It is true that later in the proceedings, under direct examination by the prosecutor, Gaffney, Valeri was asked whether he had a “deal with anybody concerning your testimony in the case,” and answered that he did not. Id. at 3948. It should have been apparent at that point, given his previous testimony, that Valeri did not consider the promise to bring his cooperation to the attention of the court to be a “deal.” Nothing in his response to Gaffney’s question, or in the surrounding testimony, appears to indicate that he was recanting his earlier statement that such an arrangement had been made. Rather, this testimony was a denial of an allegation that Valeri had made a “deal” to testify against petitioner in exchange for being allowed to plead to manslaughter.

Furthermore, Gaffney, in his closing argument, drew the jury’s attention to the fact that Valeri had testified that an arrangement existed between Gaffney and Valeri, and confirmed that such an arrangement existed:

[Valeri] said there were no promises made, but it is true. It will be called to the attention of the Court, whatever consideration he may be entitled to as a result of telling the truth, and cooperating. But there were no deals.

Id. at 4233-34.

The relevant question this court must address is not whether an agreement to bring a witness’s cooperation to the attention of the court should be characterized as a “deal.” The relevant consideration is whether the arrangement — whatever label is applied to that arrangement — was disclosed to the jury. Even if we assume that Valeri, like Fleischer, was testifying with a generalized expectation of leniency, there was no prejudice to the petitioner, because this expectation was disclosed to the jury. Even assuming that the arrangement was purposely designed to conceal Valeri’s expectation of leniency, the arrangement failed to accomplish that purpose as to Valeri, although it apparently succeeded as to Fleischer. Petitioner had a sufficient opportunity at trial to exploit Valeri’s admission of a generalized expectation of leniency to undermine Valeri’s credibility. As further support for this finding, I note that the trial judge pointedly instructed the jurors to consider carefully the reasons witnesses may have for testifying and to weigh those reasons in their assessment of the credibility of the witnesses. Trial Record at 4292-94.

The law does not condone prosecutorial efforts to conceal whether a witness is testifying with an expectation of leniency. Brady, however, “is not a rule of pretrial discovery,” United States v. Valencia-Lucena, 925 F.2d 506, 514 (1st Cir.1991). Since the prosecutorial conduct with regard to the alleged arrangement with Valeri does not indicate that any arrangement of leniency existed that was not revealed at trial, petitioner has failed to show that any material evidence was suppressed, and thus has failed to show that his constitutional rights were violated. Furthermore, even if the conduct complained of did rise to the level of a constitutional error, the error did not actually prejudice him.

See United, States v. Sepulveda, 15 F.3d 1216, 1220-21 & n. 5 (1st Cir.1993) (determining that evidence that was allegedly negligently withheld by government was immaterial, and noting that “[njeither our decisions nor those of other circuits have been sympathetic to new trial claims based solely on the discovery of additional information useful for impeaching a government witness.”)

Petitioner’s motion for an evidentiary hearing on the alleged Giglio violation regarding Valeri’s testimony is denied. Petitioner has not shown the existence of a factual dispute that, if resolved in his favor, would support his claim for relief. See Hadfield, supra, Copp, supra, Weinberger, supra.

3. Alleged Failure to Disclose Exculpatory Evidence of Witness Finn

Michael Finn claimed to have been at the scene of the crime on the day of the robbery and witnessed the shooting. Neither the prosecution nor the defense called him as a witness. Petitioner asserts that the Commonwealth had in its possession, either actually or constructively, information from the witness that was exculpatory. Petitioner’s second motion for new trial was premised in part on this assertion.

At the hearing on the motion, Finn gave a description of the shooter that was consistent with the appearance of the petitioner. However, the motion judge found that Finn had, before trial, given the FBI a description that was inconsistent with the description he gave at the hearing. Finn had, apparently, given such an inconsistent description to the FBI shortly after the shooting, and when shown a photo array, had failed to identify petitioner. The judge also found that the FBI did not give that prior statement, or any information concerning it, to any state authorities, and that neither the police nor prosecution ever told Finn not to talk to defense counsel.

Petitioner argues that the motion judge’s findings that the prosecution had no possession of the FBI information should not be credited, given the close cooperation between the agencies in this case. He further argues that the question of “possession” of this information is a mixed question of law and fact, since even if the prosecution did not have actual possession of the information, they should be charged with constructive possession of the information. Petitioner further argues that the evidence is strong that whatever information was given by Finn to the FBI was also given to the Boston Police, and thus that it is all the more appropriate to charge the prosecution with constructive possession. Petitioner claims that trial counsel did not call the witness because he expected the prosecution to call him, and that had he known of the exculpatory information, he would have called Finn himself.

I do not undertake to determine whether, in these circumstances, the prosecution should be held vicariously accountable for (or, as the concept is more commonly expressed, “charged with constructive possession of’) the exculpatory information, because I do not find the information to be material. I find it to be information not sufficiently likely to have swayed the jury to be material. Several circumstances combine to support this finding.

First, the motion judge found that Finn would have been an unreliable witness and that the decision of the defense not to call him was a tactical judgment. I credit these findings. Indeed, on the submissions before me I would make the same findings even if not according the trial judge’s findings a presumption of correctness. At trial, the court conducted a colloquy with both counsel, at which they expressed an awareness of Finn’s claim to have witnessed the shooting, and of his reputation for mental instability and alcoholism. Gilday I, 327 N.E.2d at 859.

I find also that the failure to provide this exculpatory information was harmless. Another way of putting the point is to say that, assuming as true the claim of petitioner that, had he known of the exculpatory information he would have called Finn as a witness, I find that the testimony so elicited would not have substantially assisted petitioner. For one thing, there are the credibility problems noted above, which would have detracted from the force of Finn’s testimony even if Finn had testified that petitioner was not the shooter. I find it far more likely, however, given Finn’s testimony at the motion hearing, and given his pattern of cooperation with the authorities, that had either the prosecution or the defense called Finn as a witness at trial, his testimony would have been unfavorable to the defense. Of course, petitioner would then have been able to impeach Finn with his prior statement, but that does not show that the overall effect of Finn’s testimony would have been to petitioner’s benefit.

Finn did not testify, and so the jury that actually convicted petitioner did not rely on any purported identification by Finn. Petitioner now asserts, however, that had he called Finn, or had he even been able to draw the jury’s attention to Finn’s failure to identify petitioner in the photo array, “there may very well have been an outright acquittal.” This assertion cannot be credited. The jury had before it a prosecution witness who failed to identify petitioner in a far more dramatic fashion, pointing mistakenly to a member of the audience other than Gilday. Trial Record at 2589-91 (witness Gaudette). Finn’s failure to identify petitioner in the photo array, had it been before the jury, would only have been cumulative of the exculpatory evidence actually admitted at trial. I cannot ascribe to it the impact petitioner suggests it would have had, particularly in light of the credibility problems of the witness. See Sepulveda, supra, 15 F.3d at 1220-21 & n. 5.

4. Alleged Failure to Disclose Exculpatory Evidence of Witnesses Becker and Fleischer

Petitioner makes two other allegations of failure to disclose exculpatory information in FBI reports. The information in these reports was obtained later through Freedom of Information Act requests. These reports were apparently not the subject of a motion hearing as was the FBI report concerning Finn; thus this court does not have as full a picture of the circumstances regarding the FBI’s possession, or the prosecution’s lack of physical possession. Nevertheless, I will assume arguendo that the prosecution may be held accountable as if it had possession of these reports.

An FBI report taken three weeks after the shooting states that the eyewitness Bernard Becker could not provide a description of the shooter. This information is exculpatory in that, had it been provided to defense counsel, it could have provided a basis for impeaching the witness’s later testimony that petitioner was the gunman. I do not find this information material, however. It was not sufficiently likely to have swayed the jury. For the same reason, I find that any error in failing to disclose the information was harmless. In reaching these findings I have taken account of the fact that another witness also failed to identify petitioner. Also relevant, of course, is the other evidence discussed in this opinion that strongly implicated petitioner at trial.

Petitioner also alleges that another FBI report, not disclosed to petitioner, contained the information that when first asked by the FBI whether he knew Gilday, Fleischer denied it. This information was exculpatory, in that it could have been used to show Fleischer’s willingness to tell a lie.

In the assessment of evidence about an arrangement of leniency with Fleischer, Part VLC, supra, I determined that the suppression of information regarding that disclosure was harmless. For the same reasons, I find harmless the failure of the prosecution to disclose the fact that Fleischer attempted to conceal his involvement with the defendant when first approached by the FBI.

5. Miscellaneous Allegations of Prosecutorial Misconduct

Petitioner raises a series of claims of prosecutorial misconduct in the course of the trial, which petitioner concedes are not of the same significance as the “major” claims discussed above, but which he asserts nevertheless should be considered because of their cumulative effect on the trial. Reviewing the allegedly improper conduct, I do not find anything to change the determinations I have made in the foregoing discussion. Also, I do not find in the cumulative effect of all the alleged errors any reason for allowing the present petition.

The prosecutor characterized petitioner as an “old pro” at least six times during the trial, and at one point referred to his “professional criminal mind.” Although these comments may have been inappropriate exaggeration, it is significant that defense counsel himself referred to defendant as an “ex-con,” and “old man,” and a “drunk.” Gil-day I, 327 N.E.2d at 851. In view of the tactical choices defense counsel made in characterizing petitioner, petitioner cannot now reasonably complain that the prosecutorial comments were likely to have had any noteworthy effect on the jury.

Petitioner complains that the prosecutor tried to introduce a lengthy letter by the defendant that was allegedly inculpatory, without disclosing the letter to defense. The defendant denied writing the document, however, and the letter was not received in evidence. Id. at 863. Since the evidence was never before the jury, it did not affect the jury verdict.

Petitioner complains that the prosecutor never disclosed that their dusting of the automobile driven by the gunman revealed no fingerprints of petitioner, and that this constituted a violation of Brady. I do not find that the undisclosed information was material, or that the failure to disclose it had a substantial and injurious effect or influence in determining the jury’s verdict.

Petitioner complains that the prosecution never disclosed that money and other assistance were provided to the witness McGrory in exchange for his cooperation. The trial record indicates that such an arrangement was disclosed to the jury. See Trial Record at 2207 and 2235.. Petitioner’s sole claim would thus appear to be that he should have been notified of the arrangement by the prosecution. Assuming arguendo that the prosecution deliberately failed to bring this arrangement to defense counsel’s attention, I find the error to be harmless in light of the fact that the arrangement was disclosed at trial. I find that the prosecution’s action did not have a substantial and injuri-

ous effect or influence in determining the jury’s verdict.

Petitioner suggests that the photo arrays shown to the eyewitnesses Becker and Gaudette were unconstitutionally suggestive, in that the arrays were shown only after extensive publicity depicting petitioner as the shooter. The trial court conducted an extensive voir dire regarding the procedure used to display these photographs. I determine that the procedure used was not unconstitutionally suggestive. I further determine that the procedure used did not have a substantial and injurious effect or influence in determining the jury’s verdict, particularly in light of the fact that petitioner had an opportunity to suggest through cross-examination and argument to the jury that the pre-trial identifications were of questionable value because of the extensive publicity. This is particularly true with respect to the witness Gaudette, who failed to identify petitioner in the courtroom. Trial Record at 2589-91.

D. Overview of the Inculpatory Evidence Properly Before the Jury

The foregoing discussion has focussed on individual allegations of error, and in each case I have determined that the asserted errors were either immaterial or harmless or both. I further determine that the cumulative weight of all these asserted errors is not such as would have had a substantial and injurious effect or influence in determining the jury’s verdict. A summary of key parts of the inculpatory evidence that was before the jury strongly supports this determination. Petitioner asserts throughout his briefs that the inculpatory evidence, excluding such evidence as he asserts was tainted by constitutional error, was such that it was unlikely to have been given much weight by the jury. This contention must be rejected because of the evidence noted here even though this is not an exhaustive account.

When petitioner took the stand, he made numerous admissions that, although not exactly amounting to a confession, certainly would have supported strong inferences of guilt. Moreover, the inference of guilt that these admissions suggested was strength ened by the other evidence before the jury, thus making it more strongly likely that they would find guilt beyond a reasonable doubt.

For example, petitioner admitted extensive conduct that aided and assisted the other participants in the robbery. He further admitted knowing that they were planning something for the day of the robbery, even though he claimed not to have known exactly what. Obvious from his admissions, however, was that the plans involved the use of weapons and cars appropriate for use in getting away. His own admissions indicated, at the very least, a strong likelihood that he was aware that he was assisting criminal conduct of a kind using such devices. Furthermore, Bond, a witness friendly to petitioner, testified that he was sure that petitioner knew that Bond was involved in robbing banks. Thus, even if no evidence had been introduced indicating that petitioner was present at the scene of the crime, sufficient evidence was introduced to strongly support, if not virtually compel, the inference that petitioner had the specific intent to assist the commission of an armed robbery. In fact, of course, support for felony murder was even more compelling because of the evidence, discussed below, tending to show that petitioner was at the scene of the crime.

As explained above, the jury was charged that presence at the scene of the crime was an essential element of the offense. Petitioner did not admit to being present, and defense counsel presented evidence in support of an alibi. There was, however, very substantial evidence before the jury that, despite his denial, petitioner had in fact been a participant at the scene of the crime.

Petitioner admitted that he took a substantial part of the loot after the crime and that he had asked to join Bond in flight after the commission of the crime. Furthermore, Valeri specifically placed petitioner at the scene of the crime. Valeri’s testimony to that effect would have had more weight than that of eyewitnesses, because Valeri knew petitioner and was unquestionably a knowledgeable participant in the crime. Petitioner urges this court to recognize that Fleischer and Valeri were “star witnesses” for this reason. (Memorandum at 31). That contention has some force.

Petitioner argues that since the “star” testimony is tainted by error, the conviction cannot stand. This contention fails.

Even assuming arguendo that Fleischer’s testimony must be discounted, this court has rejected the argument that Valeri’s is similarly vulnerable. Valeri’s testimony, taken in conjunction with the inculpatory evidence that was admitted by petitioner, strongly supported the inference that petitioner was in fact a participant at the scene of the crime.

Furthermore, another witness, McGrory, testified that petitioner had made a statement to McGrory that tended to implicate petitioner directly in participation in the crime. Petitioner’s arguments that this testimony would have had little effect on the jury cannot be credited. Petitioner notes that the testimony that petitioner said “I did it,” does not necessarily implicate petitioner as the shooter, only as a participant in the robbery. This distinction is without legal significance,

' since a finding of knowing participation in the armed robbery would have sustained a conviction for felony murder. Petitioner argues that the statement by petitioner that he would kill McGrory even if he were on death row, “even if believed,” would at most “allow a juror to infer that Gilday was somehow involved with the crime.” It is true that the information would not compel a juror to reach that finding; nevertheless, this was very strong evidence of petitioner’s guilt.

Petitioner makes a further argument for discounting the effect that McGrory’s testimony would have had on the jury — an argument that since McGrory “was implicated in the crime and had reasons to want to please the prosecutor,” his testimony had inherent credibility problems. Thus, petitioner asks the court to presume that the jury knew that potential co-defendants have an interest in pleasing the prosecutor, and thus that the jury places little weight on their testimony. At the same time, however, petitioner asks the court to accept his assertion that Fleischer and Valeri, also involved with the participants, were “star witnesses,” precisely because of their involvement in the crime. Finally, petitioner asks this court to determine that, had the jury been told that Fleischer’s lawyer had told him that “it [was] in his best interest” to testify for the prosecution, this information would have entirely changed the jury’s estimate of Fleischer’s credibility. These contentions have some tendency to cancel each other out. In any event, they are not persuasive.

In summary, the conduct complained of by the petitioner did not have a “substantial and injurious effect or influence in determining the jury’s verdict.” Brecht, — U.S. at-, 113 S.Ct. at 1722.

VII.

In April of 1993, petitioner moved for oral argument. During the intervening months the court has allowed supplemental submissions in addition to extensive memoranda filed earlier. The issues have been thoroughly examined in the written submissions and the request for oral argument will be denied.

ORDER

The clerk is directed to enter on a separate document a Final Judgment as follows: “The petition for writ of habeas corpus is denied.”

APPENDIX

In Gilday I, 327 N.E.2d at 854-59, the Supreme Judicial Court summarized the evidence presented at trial as follows:

“We summarize only the most significant evidence, and observe at the same time that there was much other testimony which tended to support an inference of the defendant’s guilt on all indictments. The witness Francis Goddard testified that he had seen a man firing at the bank, and his testimony as to the man’s description was consistent with the defendant’s appearance. The witness Bernard Becker identified Gilday before the jury as the man who fired the rifle at the bank on the day in question. The witness Andrew Gaudette testified that he saw a man in a white sedan firing a weapon toward the bank, and that he had identified a photograph of Gilday, among a group of photographs shown to the witness, as the man who fired the gun. Gaudette failed to identify Gilday before the jury, at a time when Gilday was seated among the spectators in the court room.

“The witness James A. Fox, a licensed firearms dealer in New Hampshire, testified that he had sold a .45 caliber semiautomatic rifle, and other weapons, to the defendant on September 5, 1970, and that the defendant was accompanied by Bond. The defendant and Bond fired test rounds from the weapon into a sandbank. He identified a semiautomatic rifle, which had been found in Bond’s luggage at the time of Bond’s arrest in Colorado after the murder and robbery, as the one he sold to the defendant. A ballistics expert testified that bullets and spent casings recovered from the New Hampshire sandbank, from the police ear of Officer Schroeder, from in and around the Brighton bank, and from the area where the white Ambassador had been, all were fired from the semiautomatic rifle in evidence that had been sold to Gilday and later found in Bond’s luggage.

“The blue Chevrolet, which had been used by Bond, Saxe, and Valeri in the robbery, was recovered by the police. It was shown that an Ontario license plate on the vehicle had been stolen from a vehicle in the parking lot of the Huntington Avenue Y.M.C.A. in Boston where Gilday at the time lived. Gil-day’s thumb print was found on the license plate.

“Alan McGrory testified that he knew the defendant and that he also knew Stanley Bond and Robert Valeri; that in September, 1970, he was living in a Northeastern University dormitory, and the defendant Gilday was residing at the Huntington Avenue Y.M.C.A., a few blocks away. On September 20, 1970, he was with Gilday in a bar; that Gilday told him he, Gilday, was ‘in on something good,’ a revolutionary cause, with all kinds of professors and businessmen involved who had no police records. Gilday showed him a .38 caliber pistol and bullets. Gilday told him there were girls involved, who would do away with anybody who would hurt the group.

“McGrory stated that on September 23, 1970, at 12:30 a.m. Gilday came to see McGrory at the latter’s apartment, and McGrory saw Bond and Valeri outside on the sidewalk. Gilday then told McGrory that he had told Bond and Valeri about him. Gilday said that he had told Bond and Valeri that McGrory wanted to see change and did not have much time to effect it, and that McGrory would have no qualms about killing people. He implied that McGrory should confirm these statements to Bond and Valeri. The witness McGrory testified that he and Gilday went out to meet Bond and Valeri and entered the white Ambassador; that Bond asked him why he wanted in, and the witness stated he wanted to see change; that he did not have long to achieve it; and that he was prepared to go to war with the establishment, depending on whom he was with, and the means used. Gilday then stated again the kinds of people involved, who had no criminal records, and who would ‘blow the wall[s] of Walpole down if need be.’ Gilday repeated that there were girls involved who were willing to kill if necessary, and that the group had police and army radios, submachine guns, and bazookas.

“About 2:30 p.m. that same day, the day of the robbery and murder, the defendant came to McGrory’s room. Bond was outside on the street, casually talking to a driver in a taxicab. At that time the witness McGrory said to Gilday, T was sick to hear that two guys and a girl and an old reprobate had robbed a bank and critically wounded a police officer.’ Gilday answered that he had not beard about it. Then the two tried to get the news on the radio without success. Gil-day then said that he did not think the officer was going to die, and that, even if the officer did die, he, Gilday, had nothing to lose. Gil-day warned McGrory that even if Gilday were imprisoned on death row, he would take care of McGrory if he said anything. At this point Gilday gave the witness some money from a large roll of bills, and asked the witness whether he would be ready to leave that afternoon for the west coast. Gilday offered to get the witness a gun.

“Robert J. Valeri stated that he had met Gilday and Bond in 1969, and Saxe and Power in 1970 at Brandéis University, and that in September he was a student at Northeastern University, and the group met in two apartments, one at 337 Beacon Street, Boston, rented by Bond, and the other at 163 Beacon Street, Boston, rented by Power. The witness testified that he knew the witness McGrory, and corroborated McGrory’s account of the meeting in the early morning hours of September 23, 1970; that after McGrory had been left at his residence, Gil-day, Bond, and he, Valeri, returned to 337 Beacon Street, where they met with Saxe, Power, and one Michael Fleischer; that, earlier that evening, sometime after 7 p.m., the same group had discussed the robbing of a bank; and that Bond had gone to locate the bank, and on his return, told them the bank was to be the branch of the State Street Bank and Trust Company in Brighton. The witness testified that plans were made that Saxe, Bond, and Gilday were to enter the bank, and Valeri was to stay outside, but the following morning Gilday asked Valeri to go in while Gilday remained outside. Bond was to carry a 9 mm. Browning; Gilday, a shotgun and a .38 caliber pistol; Saxe, a .30 caliber carbine and pistol; and Valeri, a semiautomatic weapon and a .357 Magnum, but the plans were changed, and Gilday took the semiautomatic weapon, and he, Valeri, took the shotgun. According to Valeri’s testimony plans were made for Bond, Saxe, and Valeri to use the blue Chevrolet; Gilday was to have the white Ambassador; and Power was to use a 1961 or 1962 Ford station wagon, which Gilday had purchased. The witness testified that he stole the blue Chevrolet in Springfield, Massachusetts, and that just before leaving for the robbery, Gilday changed the plates and put on the Ontario plate. The witness testified that he had seen the semiautomatic weapon prior to September 23, and Gilday told him that he and Bond had bought the weapon in New Hampshire.

“Valeri further stated that, when they left the Beacon Street area, Bond was driving the blue Chevrolet, Gilday the white Ambassador, and he, Valeri, was driving the station wagon; that, on a side street, Power took the station wagon as the switch car, and he, Valeri, joined Bond and Saxe in the Chevrolet; that the two vehicles proceeded to the bank, and Gilday took his position in the white Ambassador with the New Jersey plates, on Western Avenue near the corner of Everett Street; that he, Valeri, Bond, and Saxe entered the bank, and when the tellers were slow in giving up the money, Bond fired two quick shots in the bank; that Bond took the money in bags, and they left the bank; that they drove from the bank, met Power, abandoned the Chevrolet, and drove away in the Ford station wagon, with Bond and himself lying in the back under a rug. The witness said he was dropped at Roberts station in Waltham from where he took a train to North Station, Boston, and proceeded to 163 Beacon Street. When he arrived, Bond and Fleischer were present, and the girls later returned. Gilday came in after the girls, and he had with him a suitcase containing the semiautomatic weapon. Bond asked Gilday what had happened, and Gilday said he stayed there and waited, and when the police officer came, he opened fire on him, and Bond and Valeri of the others asked why he, Gilday, had waited, and Gilday said that ‘he had always wanted to shoot a police officer.’

“Stanley R. Bond was called as a witness by the defendant. He testified that in 1970 he was paroled from the Massachusetts Correctional Institution at Walpole, where he had known the defendant Gilday and the witnesses Valeri and McGrory; that after his release from prison, he became acquainted with Susan Saxe, Katherine Power, and Michael Fleischer; that, in the summer of 1970 and thereafter, the four persons, with Valeri after his release, formed a revolutionary group. In the summer of 1970, he, with Valeri, Saxe, and Power, had robbed banks in Chicago, Philadelphia, and California.

“Bond also testified that he saw Gilday in the summer; that Gilday was not politically oriented; that he was present when Gilday purchased the semiautomatic weapon, with other guns; and that he and Gilday bought the Ford station wagon. The witness admitted that he saw McGrory on the night before the robbery; that Valeri was the one who wanted McGrory in with the group; and that the meeting with McGrory did take place substantially as testified to by McGrory.

“Bond stated that Gilday had stolen license plates for the group, one of which was used in the State Street bank holdup, but the witness was not sure on which car the plate had been used. However, Bond testified that Gilday was not directly involved in a prior robbery planned in New York but not executed.

“Bond also testified that he thought he drove the blue Chevrolet, that Michael Fleischer drove the white Nash, and that Power drove the station wagon. At one point, the witness said, ‘Mike had pulled up across the street.’ They drove around the bank. Saxe then went into the bank to see if the vault was open. After she returned to them, they parked the car and Bond, Saxe, and Valeri entered the bank together. Bond took the guard’s gun. When the tellers hesitated, he fired one shot from his weapon. Bond and the others took the money in bags and left. They drove to the switch car site, joined Power, then dropped Power off on Route 9, and Valeri in Waltham. The witness Bond left the station wagon in Waltham, and took a cab back to the apartment on Beacon Street. When he arrived, Power was there and told him a ‘cop’ had been shot. Saxe returned, then Valeri, and finally the defendant Gilday and Fleischer arrived at the same time. Later that evening, Bond drove Gilday to Brandéis University and told him to drive the station wagon somewhere, and he gave Gilday $1,000. Bond testified that he and Power left Boston by car; that Saxe and Fleischer went to the Logan airport; and that the four were together again in Philadelphia.

“Bond further testified that he was sure that Gilday knew that he, Bond, was involved in robbing banks and in other activities, and that Gilday had participated in a breaking and entering of the Newburyport armory together with himself, Saxe, Power, and Valeri. The witness testified that he did not see Gilday on September 23, 1970, until after noon, and that Gilday must have put the Ontario plate on the blue Chevrolet the night before and that, when he saw Gilday at noon, Gilday was drunk.

“The defendant testified and admitted that he knew Bond and that he had bought the semiautomatic weapon, but he claimed that he was sick from drinking at the time. Gil-day acknowledged that he was a frequent visitor with Bond, Valeri, Saxe, Power and Fleischer at the Beacon Street apartments, and admitted he had been carrying a gun on his person from about September 16, 1970. He further stated that he had stolen the Ontario plate and that he had purchased the Ford station wagon, the sale being made in the name of one Sheldon Gelman, as Bond had identification papers for Sheldon Gel-man. The defendant testified that on September 20, 1970, he had driven Kathy Power to the armory at Newburyport and waited near by, but he asserted that he was drunk at the time and had fallen asleep in the back seat and only later learned that Bond, Valeri, and Saxe had obtained radios and other material from the armory. The defendant admitted, however, that he had shown them a circuitous route from Newburyport to Boston.

“The defendant also testified that he knew the witness MeGrory; that on the evening of September 22, 1970, he was at 337 Beacon Street, and knew that Bond, Valeri, and two girls were planning something, but did not know exactly what; that he saw MeGrory with Bond and Valeri, and further verified substantially what MeGrory had stated; that, after the meeting with MeGrory, he returned to his room at the Y.M.C.A. at which time he slept, awakening about 7 a.m. on September 23, the day of the robbery. Gilday stated that on that morning he had taken a few drinks, went to downtown Boston for some shopping, visited a political headquarters, and registered to vote at a booth on Boston Common shortly after 11 a.m. According to Gilday’s testimony he then went to Northeastern University. Unable to operate the elevator, at the university, he went to Dean Garland at the school about 12:30 p.m. to get a key. He visited another bar and heard news reports of the holdup and first went to 337 Beacon Street. He then went to the other apartment where Bond, Saxe, Power, and Fleischer were ‘yelling and hollering’ and charging each other with things, and he took some of the holdup money and left. The defendant further testified that he went to pick up some clothes that had previously been purchased, then went for more drinks. He testified, although not entirely certain of the sequence of events, that he then revisited Bond, finally returning to the Y.M.C.A. where Bond picked him up and drove him to Waltham to get the station wagon. At this point Bond gave him $1,000, but, although he asked Bond to take him with him, Bond refused.

“In rebuttal, Michael Fleischer testified that he knew Bond, Saxe, Power, and Valeri, and met Gilday in September, 1970; that he was in sympathy with aims to change society, but did not agree with Bond’s plan to rob banks to finance a general uprising; that on September 22, 1970, he heard a discussion between the five concerning the robbery of a bank on September 23, 1970; that he saw many guns in the apartment at 337 Beacon Street; and that he helped the five load things into the three ears, and he then went back to the apartment at 163 Beacon Street. The witness testified Power returned first, then Bond and Valeri. Gilday arrived approximately fifteen minutes later, and finally Saxe returned. There was a discussion about shooting the ‘cop,’ and Saxe and Power accused Gilday of being ‘trigger-happy’ and said that it was ‘a real stupid thing to do to shoot the cop,’ and Gilday said, *What did you want me to do, the cop was right there, he was only thirty seconds behind you.’”