OPINION OF THE COURT

The single question presented by the consolidated instant appeals is whether retroactive effect should be accorded to our en banc holding in United States ex rel. Matthews v. Johnson,1 that due process was violated in a Pennsylvania murder trial when the defendant’s request for a voluntary manslaughter instruction was denied for lack of evidentiary basis.2

The question arises by reason of the denial below of the appellants’ habeas corpus petitions on the ground that “full retroactivity should not be accorded to Matthews.” United States ex rel. Cannon v. Johnson, 396 F.Supp. 1362, 1365 (E.D.Pa. 1975). The petitions disclosed that prior to Matthews, the Supreme Court of Pennsylvania affirmed the first degree murder convictions of the appellants after rejecting their contention that the trial court had erred in denying their request for a voluntary manslaughter instruction for lack of evidentiary basis.3

It must be noted parenthetically that the court below chose to limit its holding as to Matthews’ retroactivity to habeas corpus cases on the ground that “this case does not squarely raise the question of the applicability of Matthews to direct appeals before the Pennsylvania Supreme Court.”4

The conclusions underlying the holding below, expounded in Judge Becker’s exhaustive opinion in Cannon, supra5 may be epitomized as follows:

*1015(1) Matthews was not designed “to enhance the reliability of the truth-finding function;”6 “the purpose of the Matthews rule is the elimination of unseemly arbitrariness from the judicial process . [and] not to assure that guilt or innocence is reliably determined . or to prevent convicting the innocent;”7 (2) Pennsylvania “relied heavily on the old rule” until May 2, 1974, when its Supreme Court, exercising its “supervisory power,” ruled in Commonwealth v. Jones, 457 Pa. 563, 573-74, 319 A.2d 142, that thereafter “a defendant . . . would be entitled, upon request, to have the jury advised of its power to return a verdict of voluntary manslaughter;” 8 and (3) “any type of retroactive application of Matthews would have a substantially adverse effect upon the administration of justice in Pennsylvania,"9 and “retroactivity would threaten a far greater number of cases than the 294 presently pending on direct appeal [to the Supreme Court of Pennsylvania].”10 (emphasis supplied).

The foregoing three-pronged determination was made by Judge Becker in consonance with the three-fold guideline criteria spelled out for resolution of the issue of retroactivity in Stovall v. Denno, 388 U.S. 293, 87 S.Ct. 1967, 18 L.Ed.2d 1199 (1967) at page 297, 87 S.Ct. at page 1969:

“Our recent discussion of the retroactivity of other constitutional rules of criminal procedure make unnecessary any detailed treatment of that question here. Linkletter v. Walker, supra [381 U.S. 618, 85 S.Ct. 1731, 14 L.Ed.2d 601]; Tehan v. Shott, supra [382 U.S. 406, 86 S.Ct. 459, 15 L.Ed.2d 453]; Johnson v. New Jersey, supra [384 U.S. 719, 86 S.Ct. 1772, 16 L.Ed.2d 882]. ‘These cases establish the principle that in criminal litigation concerning constitutional claims, “the Court may in the interest of justice make the rule prospective . . . where the exigencies of the situation require such an application” . . Johnson, supra, 384 U.S., at 726-727 [86 S.Ct. [1772] at 1777]. The criteria guiding resolution of the question implicate (a) the purpose to be served by the new standards, (b) the extent of the reliance by law enforcement authorities on the old standards, and (c) the effect on the administration of justice of a retroactive application of the new standards." (emphasis supplied). Judge Becker concluded that Matthews,

in sum, announced a new criminal procedural rule, which, in light of his criteria determinations, should not be accorded retroactive effect, pursuant to “the wholly prospective approach” spelled out in Johnson v. New Jersey, 384 U.S. 719, 86 S.Ct. 1772, 16 L.Ed.2d 882 (1966), and “frequently followed since that time.” 396 F.Supp. at 1372.

We agree with Judge Becker’s view that Matthews should not be accorded retroactive effect. Inasmuch as he confined application of his holding to cases on habeas corpus review, for the reason earlier stated, we make this further holding:

Matthews is inapplicable to pending, and/or future appeals from pre-Matthews murder verdicts.

Discussion of our holdings must be prefaced by the observation that Matthews did not decide the issue of its retroactive or prospective application, albeit, three of the judges participating in the majority opinion,11 citing our holding in United States v. Zirpolo, 3 Cir., 450 F.2d 424, 431-433 (1971), expressed the view that while “full retroactivity should not be accorded this decision,” *1016it should apply “to those cases ... on direct appeal [in the Pennsylvania courts] on this date.”12

Coming now to our holding that Matthews is not to be accorded retroactive effect in any respect:

Judge Becker, in resume, held that “the Matthews rule does not implicate the integrity of the fact-finding process, that the Commonwealth [of Pennsylvania] relied heavily upon the prior rule, and that retroactive application of the Matthews standard would have an adverse effect upon the administration of justice,” and that Linkletter v. Walker, 381 U.S. 618, 85 S.Ct. 1731, 14 L.Ed.2d 601 (1965), and its cited progeny, “require[s] us to hold that the decision in Matthews should not be retroactively applied to relator.”13

It would not serve the judicial economy to engage in extended discussion of Linkletter, and its progeny, in view of their exhaustive analysis in Judge Becker’s opinion. We must, however, stress the specific holding in Desist v. United States, 394 U.S. 244, 249, 89 S.Ct. 1030, 1033, 22 L.Ed.2d 248 (1969), that “foremost ” consideration should be given “to the purpose to be served by the new constitutional rule,” and the later holding in Williams v. United States, 401 U.S. 646, 653, 91 S.Ct. 1148, 1152, 28 L.Ed.2d 388 (1971) that retroactivity should be accorded “[wjhere the major purpose of new constitutional doctrine is to overcome an aspect of the criminal trial that substantially impairs its truth-finding function and so raises serious questions about the accuracy of guilty verdicts in past trials . . ..” (emphasis supplied).

The vitality of the principles stated in Desist and Williams is attested to by their re-affirmation in two recent Supreme Court cases, decided after the filing of Judge Becker’s opinion. United States v. Peltier, 422 U.S. 531, 535, 95 S.Ct. 2313, 45 L.Ed.2d 374 (1975), and Bowen v. United States, 422 U.S. 916, 918, 95 S.Ct. 2569, 45 L.Ed.2d 641 (1975).

What has been said is relevant to the consideration given by Judge Becker to the Linkletter-Stovall “purpose” criteria, and his conclusion that “the Matthews rule does not implicate the integrity of the fact-finding process.”14 The stated conclusion, it may be said, is supported by the statement in Matthews that “[n]othing we decide today affects the fact-finding process.”15

No useful purpose would be served by discussion of Judge Becker’s conclusions with respect to the Linkletter-Stovall criteria of reliance on the “old standards,” and the effect on the administration of justice of the “new standards,” in light of his thorough analysis of these factors. The record fully sustains these criteria determinations.

We come now to our earlier stated holding that Matthews is inapplicable to pending or future direct appeals from pre-Matthews murder verdicts.

The Supreme Court, since Johnson v. New Jersey, supra, has for the most part made prospective from the date of its announcement, a new criminal procedural rule which does not implicate the truth-finding process of a trial.16 It did so in the recently decided Daniel v. Louisiana, 420 U.S. 31, 95 S.Ct. 751, 42 L.Ed.2d 766 (1975) and Peltier, supra, and Bowen, supra. It must be noted at this juncture that Daniel held prospective from the date of its decision the ruling in Taylor v. Louisiana, 419 U.S. 522, 95 S.Ct. 692, 42 L.Ed.2d 690 (1975) that Louisiana’s method of selection of petit juries violated *1017the Sixth and Fourteenth Amendments. The stated holding in Daniel is in sharp contrast with our action in United States v. Zirpolo, supra, in making applicable to cases “on direct appeal”17 our holding therein that the procedure utilized in the selection of federal grand juries in Newark, New Jersey, violated federal law.

Our “direct appeal” ruling in Zirpolo is also inconsistent with Michigan v. Payne, 412 U.S. 47, 93 S.Ct. 1966, 36 L.Ed.2d 766 (1973). It was there held that retroactive effect was not to be accorded to North Carolina v. Pearce, 395 U.S. 711, 89 S.Ct. 2072, 23 L.Ed.2d 656 (1969), and thus the Michigan Supreme Court erred in applying Pearce on a “direct appeal” from a prePearce sentence.18

This, too, must be said:

Although our holding with respect to direct appeals is broader than is essential for the disposition of the instant appeals, it is consistent with the position taken by the Supreme Court in Stovall v. Denno, supra,19 and DeStefano v. Woods, 392 U.S. 631, 88 S.Ct. 2093, 20 L.Ed.2d 1308 (1968).

It is consistent, too, with the rule announced by the Pennsylvania Supreme Court on May 2, 1974, in Commonwealth v. Jones, supra, which provides that henceforth a voluntary manslaughter instruction shall be given, upon request, in all murder trials.

For the reasons stated, the orders below denying the habeas corpus petition of the appellants will be affirmed.

(concurring):

I favored meeting the retroactivity issue during our adjudication of United States ex rel. Matthews v. Johnson, 503 F .2d 339 (3d Cir. 1974) (in banc), cert. denied, 420 U.S. 952, 95 S.Ct. 1336, 43 L.Ed.2d 430 (1975), and, at that point, would have followed the limitations set forth in United States v. Zirpolo, 450 F.2d 424, 431-33 (3d Cir. 1971). In the interim since the Matthews opinion was filed, the Supreme Court decided Daniel v. Louisiana, 420 U.S. 31, 95 S.Ct. 751, 42 L.Ed.2d 766 (1975), which was concerned with essentially the same issue as Zirpolo. Daniel held that retroactivity was not required as to trial procedures which occurred before the date of the decision. The same standard should apply in this case, and I, therefore, join in the majority opinion.