This case presents important questions touching the federal habeas corpus jurisdiction, 28 U.S.C. § 2241 et seq., in its relation to state criminal justice. The narrow question is whether the respondent Noia may be granted federal habeas corpus relief from imprisonment under a New York conviction now admitted by the State to rest upon a confession obtained from him in violation of the Fourteenth Amendment, after he was denied state post-conviction relief because the coerced confession claim had been decided against him at the trial and Noia had allowed the time for a direct appeal to lapse without seeking review by a state appellate court.

Noia was convicted in 1942 with Santo Caminito and Frank Bonino in the County Court of Kings County, New York, of a felony murder in the shooting and killing of one Hammeroff during the commission of a robbery. The sole evidence against each defendant was his signed confession. Caminito and Bonino, but not Noia, appealed their convictions to the Appellate Division of the New York Supreme Court. These appeals were unsuccessful, but subsequent legal proceedings resulted in the releases of Caminito and Bonino on findings that their confessions had been coerced and their convictions therefore procured in violation of the Fourteenth Amendment.1 Although it has been stipulated that the coercive nature of Noia's confession was also established,2 the United States District Court for the Southern District of New York held in Noia's federal habeas corpus proceeding that because of his failure to appeal he must be denied relief under the provision of 28 U.S.C. § 2254 whereby 'An application for a writ of habeas corpus in behalf of a person in custody pursuant to the judgment of a State court shall not be granted unless it appears that the applicant has exhausted the remedies available in the courts of the State * * *.' 183 F.Supp. 222 (1960).3 The Court of Appeals for the Second Circuit reversed, one judge dissenting, and ordered that Noia's conviction be set aside and that he be discharged from custody unless given a new trial forthwith. 300 F.2d 345 (1962). The Court of Appeals questioned whether § 2254 barred relief on federal habeas corpus where the applicant had failed to exhaust state remedies no longer available to him at the time the habeas proceeding was commenced (here a direct appeal from the conviction), but held that in any event exceptional circumstances were present which excused compliance with the section. The court also rejected other arguments advanced in support of the proposition that the federal remedy was unavailable to Noia. The first was that the denial of state post-conviction coram nobis relief on the ground of Noia's failure to appeal barred habeas relief because such failure constituted an adequate and independent state ground of decision, such that this Court on direct review of the state coram nobis proceedings would have declined to adjudicate the federal questions presented. In rejecting this argument, the court—while expressing the view that '(j)ust as it would be an encroachment on the prerogatives of the state for the Supreme Court upon direct review to disregard the state ground, equally—if not more so—would it be a trespass against the state for a lower federal court, upon a petition for habeas corpus, to disregard the state ground in granting relief to the prisoner,' 300 F.2d at 359—held that the exceptional circumstances excusing compliance with § 2254 also established that Noia's failure to appeal was not a state procedural ground adequate to bar the federal habeas remedy: 'The coincidence of these factors: the undisputed violation of a significant constitutional right, the knowledge of this violation brought home to the federal court at the incipiency of the habeas corpus proceeding so forcibly that the state made no effort to contradict it, and the freedom the relator's codefendants now have by virtue of their vindications of the identical constitutional right leads us to conclude that the state procedural ground, that of a simple failure to appeal, reasonable enough to prevent federal judicial intervention in most cases, is in this particular case unreasonable and inadequate.' 300 F.2d, at 362. The second argument was that Noia's failure to appeal was to be deemed a waiver of his claim that he had been unconstitutionally convicted. The Court of Appeals rejected this argument on the ground that no waiver could be inferred in the circumstances. Id., at 351—352.

We granted certiorari. 369 U.S. 869, 82 S.Ct. 1140, 8 L.Ed.2d 274. We affirm the judgment of the Court of Appeals but reach that court's result by a different course of reasoning. We hold: (1) Federal courts have power under the federal habeas statute to grant relief despite the applicant's failure to have pursued a state remedy not available to him at the time he applies; the doctrine under which state procedural defaults are held to constitute an adequate and independent state law ground barring direct Supreme Court review is not to be extended to limit the power granted the federal courts under the federal habeas statute. (2) Noia's failure to appeal was not a failure to exhaust 'the remedies available in the courts of the State' as required by § 2254; that requirement refers only to a failure to exhaust state remedies still open to the applicant at the time he files his application for habeas corpus in the federal court. (3) Noia's failure to appeal cannot under the circumstances be deemed an intelligent and understanding waiver of his right to appeal such as to justify the withholding of federal habeas corpus relief.

The question has been much mooted under what circumstances, if any, the failure of a state prisoner to comply with a state procedural requirement, as a result of which the state courts decline to pass on the merits of his federal defense, bars subsequent resort to the federal courts for relief on habeas corpus.4 Plainly it is a question that has important implications for federal-state relations in the area of the administration of criminal justice. It cannot be answered without a preliminary inquiry into the historical development of the writ of habeas corpus.

We do well to bear in mind the extraordinary prestige of the Great Writ, habeas corpus ad subjiciendum,5 in Anglo-American jurisprudence: 'the most celebrated writ in the English law.' 3 Blackstone Commentaries 129. It is 'a writ antecedent to statute, and throwing its root deep into the genius of our common law * * *. It is perhaps the most important writ known to the constitutional law of England, affording as it does a swift and imperative remedy in all cases of illegal restraint or confinement. It is of immemorial antiquity, an instance of its use occurring in the thirty-third year of Edward I.' Secretary of State for Home Affairs v. O'Brien, (1923) A.C. 603, 609 (H.L.). Received into our own law in the colonial period,6 given explicit recognition in the Federal Constitution, Art. I, § 9, cl. 2,7 incorporated in the first grant of federal court jurisdiction, Act of September 24, 1789, c. 20, § 14, 1 Stat. 81—82, habeas corpus was early confirmed by Chief Justice John Marshall to be a 'great constitutional privilege.' Ex parte Bollman and Swartwout, 4 Cranch 75, 95, 2 L.Ed. 554. Only two Terms ago this Court had occasion to reaffirm the high place of the writ in our jurisprudence: 'We repeat what has been so truly said of the federal writ: 'there is no higher duty than to maintain it unimpaired,' Bowen v. Johnston, 1939, 306 U.S. 19, 26, 59 S.Ct. 442, 446, 83 L.Ed. 455, and unsuspended, save only in the cases specified in our Constitution.' Smith v. Bennett, 365 U.S. 708, 713, 81 S.Ct. 895, 898, 6 L.Ed.2d 39.

These are not extravagant expressions. Behind them may be discerned the unceasing contest between personal liberty and government oppression. It is no accident that habeas corpus has time and again played a central role in national crises, wherein the claims of order and of liberty clash most acutely, not only in England in the seventeenth century,8 but also in America from our very beginnings, and today.9 Although in form the Great Writ is simply a mode of procedure, its history is inextricably intertwined with the growth of fundamental rights of personal liberty. For its function has been to provide a prompt and efficacious remedy for whatever society deems to be intolerable restraints. Its root principle is that in a civilized society, government must always be accountable to the judiciary for a man's imprisonment: if the imprisonment cannot be shown to conform with the fundamental requirements of law, the individual is entitled to his immediate release. Thus there is nothing novel in the fact that today habeas corpus in the federal courts provides a mode for the redress of denials of due process of law. Vindication of due process is precisely its historic office. In 1593, for example, a bill was introduced in the House of Commons, which, after deploring the frequency of violations of 'the great Charter and auncient good Lawes and statutes of this realme,' provided:

'Fore remedy whereof be it enacted: That the provisions and prohibicions of the said great Charter and other Lawes in that behalfe made be dulie and inviolatelie observed. And that no person or persons be hereafter committed to prison but yt be by sufficient warrant and Authorities and by due course and proceedings in Lawe * * *.

'And that the Justice of anie the Queenes Majesties Courts of Recorde at the common Lawe maie awarde a writt of habeas Corpus for the deliverye of anye person so imprisoned * * *.'10

Although it was not enacted, this bill accurately prefigured the union of the right to due process drawn from Magna Charta and the remedy of habeas corpus accomplished in the next century.

Of course standards of due process have evolved over the centuries. But the nature and purpose of habeas corpus have remained remarkably constant. History refutes the notion that until recently the writ was available only in a very narrow class of lawless imprisonments. For example, it is not true that at common law habeas corpus was exclusively designed as a remedy for executive detentions; it was early used by the great common-law courts to effect the release of persons detained by order of inferior courts.11 The principle that judicial as well as executive restraints may be intolerable received dramatic expression in Bushell's case, Vaughan, 135, 124 Eng.Rep. 1006, 6 Howell's State Trials 999 (1670). Bushell was one of the jurors in the trial, held before the Court of Oyer and Terminer at the Old Bailey, of William Penn and William Mead on charges of tumultuous assembly and other crimes. When the jury brought in a verdict of not guilty, the court ordered the jurors committed for contempt. Bushell sought habeas corpus, and the Court of Common Pleas, in a memorable opinion by Chief Justice Vaughan, ordered him discharged from custody. The case is by no means isolated,12 and when hebeas corpus practice was codified in the Habeas Corpus Act of 1679, 31 Car. II, c. 2, no distinction was made between executive and judicial detentions.13

Nor is it true that at common law habeas corpus was available only to inquire into the jurisdiction, in a narrow sense, of the committing court. Bushell's case is again in point. Chief Justice Vaughan did not base his decision on the theory that the Court of Oyer and Terminer had no jurisdiction to commit persons for contempt, but on the plain denial of due process, violative of Magna Charta, of a court's imprisoning the jury because it disagreed with the verdict:

'* * * (W)hen a man is brought by Habeas Corpus to the Court, and upon retorn of it, it appears to the Court, That he was against Law imprison'd and detain'd, * * * he shall never be by the Act of the Court remanded to his unlawful imprisonment, for then the Court should do an act of Injustice in imprisoning him, de novo, against Law, whereas the great Charter is, Quod nullus libet homo imprisonetur nisi per legem terrae; This is the present case, and this was the case upon all the Presidents (precedents) produc'd and many more that might be produc'd, where upon Habeas Corpus, many have been discharg'd * * *.

'This appears plainly by many old Books, if the Reason of them be rightly taken, For insufficient causes are as no causes retorn'd; and to send a man back to Prison for no cause retorn'd, seems unworthy of a Court.' Vaughan, at 156, 124 Eng.Rep., at 1016, 9 Howell's State Trials, at 1023.

To the same effect, we read in Bacon's Abridgment:

'(I)f the commitment be against law, as being made by one who had no jurisdiction of the cause, or for a matter for which by law no man ought to be punished, the court are to discharge him * * *; and the commitment is liable to the same objection where the cause is so loosely set forth, that the court cannot adjudge whether it were a reasonable ground of imprisonment or not.'14

Thus, at the time that the Suspension Clause was written into our Federal Constitution and the first Judiciary Act was passed conferring habeas corpus jurisdiction upon the federal judiciary, there was respectable common-law authority for the proposition that habeas was available to remedy any kind of governmental restraint contrary to fundamental law. In this connection it is significant that neither the Constitution nor the Judiciary Act anywhere defines the writ, although the Act does intimate, 1 Stat. 82, that its issuance is to be 'agreeable to the principles and usages of law;—the common law, presumably. We need not pause to consider whether it was the Framers' understanding that congressional refusal to permit the federal courts to accord the writ its full common-law scope as we have described it might constitute an unconstitutional suspension of the privilege of the writ. There have been some intimations of support for such a proposition in decisions of this Court. Thus Mr. Justice (later Chief Justice) Stone wrote for the Court that '(t)he use of the writ * * * as an incident of the federal judicial power is implicitly recognized by Article I, § 9, Cl. 2, of the Constitution.' McNally v. Hill, 293 U.S. 131, 135, 55 S.Ct. 24, 26, 79 L.Ed. 238. (Italics supplied.) To the same effect are the words of Chief Justice Chase in Ex parte Yerger, 8 Wall. 85, 95, 19 L.Ed. 332: 'The terms of this provision (the Suspension Clause) necessarily imply judicial action.' And see United States ex rel. Turner v. Williams, 194 U.S. 279, 295, 24 S.Ct. 719, 724, 48 L.Ed. 979 (concurring opinion).15 But at all events it would appear that the Constitution invites, if it does not compel, cf. Byrd v. Blue Ridge Rural Elec. Cooperative, 356 U.S. 525, 537, 78 S.Ct. 893, 900, 2 L.Ed.2d 953, a generous construction of the power of the federal courts to dispense the writ conformably with common-law practice.

The early decision of this Court in Ex parte Watkins, 3 Pet. 193, 7 L.Ed. 650, which held that the judgment of a federal court of competent jurisdiction could not be impeached on habeas, seems to have viewed the power more narrowly; see also Ex parte Kearney, 7 Wheat. 38, 5 L.Ed. 391. But Watkins may have been compelled by factors, affecting peculiarly the jurisdiction of this Court, which are not generally applicable to federal habeas corpus powers. It was plain from the decision in Marbury v. Madison, 1 Cranch 137, 174—175, 2 L.Ed. 60, which had narrowly construed the grant of original jurisdiction to the Supreme Court in Article III, that the Court would have the power to issue writs of habeas corpus only if such issuance could be deemed an exercise of appellate jurisdiction. Confronted with the question in Ex parte Bollman and Swartwout, 4 Cranch 75, 2 L.Ed. 554—like Watkins, a case of direct application to the Court for the writ—the Court held that the jurisdiction 'which the court is now asked to exercise is clearly appellate. It is the revision of a decision of an inferior court, by which a citizen has been committed to gaol.' 4 Cranch, at 100. This answer sufficed to enable the discharge of the petitioners, who had been committed (but not tried or convicted) for treason; but at the same time it virtually dictated the result in Watkins. The Court had no general jurisdiction of appeals from federal criminal judgments, see pp. 834—835, infra; if, therefore, the writ of habeas corpus was appellate in nature, its issuance to vacate such a judgment would have the effect of accomplishing indirectly what the Court had no power to do directly. This reasoning is prominent in Chief Justice Marshall's opinion for the Court in Watkins. See 3 Pet. at 203.

Strictly, then, Watkins is authority only as to this Court's power to issue the writ; the habeas jurisdiction of the other federal courts and judges, including the individual Justices of the Supreme Court, has generally been deemed original. In re Kaine, 14 How. 103, 14 L.Ed. 345; Ex parte Yerger, 8 Wall. 85, 101, 19 L.Ed. 332. But cf. Ex parte Clarke, 100 U.S. 399, 25 L.Ed. 715. But even as to this Court's power, the life of the principles advanced in Watkins was relatively brief.16 In Ex parte Lange, 18 Wall. 163, 21 L.Ed. 872, again a case of direct application to this Court for the writ, the Court ordered the release of one duly convicted in a Federal Circuit Court. The trial judge, after initially imposing upon the defendant a sentence in excess of the legal maximum, had attempted to correct the error by resentencing him. The Court held this double-sentencing procedure unconstitutional, on the ground of double jeopardy, and while conceding that the Circuit Court had a general competence in criminal cases, reasoned that it had no jurisdiction to render a patently lawless judgment.

This marked a return to the common-law principle that restraints contrary to fundamental law, by whatever authority imposed, could be redressed by writ of habeas corpus. See also Ex parte Wells, 18 How. 307, 15 L.Ed. 421; Ex parte Parks, 93 U.S. 18, 21, 23 L.Ed. 787. The principle was clearly stated a few years after the Lange decision by Mr. Justice Bradley, writing for the Court in Ex parte Siebold, 100 U.S. 371, 376—377, 25 L.Ed. 717:

'* * * The validity of the judgments is assailed on the ground that the acts of Congress under which the indictments were found are unconstitutional. If this position is well taken, it affects the foundation of the whole proceedings. An unconstitutional law is void, and is as no law. An offence created by it is not a crime. A conviction under it is not merely erroneous, but is illegal and void, and cannot be a legal cause of imprisonment. It is true, if no writ of error lies, the judgment may be final, in the sense that there may be no means of reversing it. But personal liberty is of so great moment in the eye of the law that the judgment of an inferior court affecting it is not deemed so conclusive but that * * * the question of the court's authority to try and imprison the party may be reviewed on habeas corpus * * *.'

The course of decisions of this Court from Lange and Siebold to the present makes plain that restraints contrary to our fundamental law, the Constitution, may be challenged on federal habeas corpus even though imposed pursuant to the conviction of a federal court of competent jurisdiction.17

The same principles have consistently been applied in cases of state prisoners seeking habeas corpus in the federal courts, although the development of the law in this area was at first delayed for several reasons. The first Judiciary Act did not extend federal habeas to prisoners in state custody, Ex parte Dorr, 3 How. 103, 11 L.Ed. 514; and shortly after Congress removed this limitation in 1867, it withdrew from this Court jurisdiction of appeals from habeas decisions by the lower federal courts and did not restore it for almost 20 years.18 Moreover, it was not until this century that the Fourteenth Amendment was deemed to apply some of the safeguards of criminal procedure contained in the Bill of Rights to the States. Yet during the period of the withdrawal of the Supreme Court's jurisdiction of habeas appeals, the lower federal courts did not hesitate to discharge state prisoners whose convictions rested on unconstitutional statutes or had otherwise been obtained in derogation of constitutional rights.19 After its jurisdiction had been restored, this Court adhered to the pattern set by the lower federal courts and to the principles enunciated in Ex parte Siebold and the other federal-prisoner cases.20 More recently, further applications of the Fourteenth Amendment in state criminal proceedings have led the Court to find correspondingly more numerous occasions upon which federal habeas would lie.21

Mr. Justice Holmes expressed the rationale behind such decisions in language that sums up virtually the whole history of the Great Writ:

'* * * (H)abeas corpus cuts through all forms and goes to the very tissue of the structure. It comes in from the outside, not in subordination to the proceedings, and although every form may have been preserved opens the inquiry whether they have been more than an empty shell.

'The argument for the appellee in substance is that the trial was in a court of competent jurisdiction * * *. But * * * (w)hatever disagreement there may be as to the scope of the phrase 'due process of law,' there can be no doubt that it embraces the fundamental conception of a fair trial * * *. We are not speaking of mere disorder, or mere irregularities in procedure, but of a case where the processes of justice are actually subverted. In such a case, the Federal court has jurisdiction to issue the writ. The fact that the state court still has its general jurisdiction and is otherwise a competent court does not make it impossible to find that a jury has been subjected to intimidation in a particular case. The loss of jurisdiction is not general, but particular, and proceeds from the control of a hostile influence.'22

We do not suggest that this Court has always followed an unwavering line in its conclusions as to the availability of the Great Writ. Our development of the law of federal habeas corpus has been attended, seemingly, with some backing and filling. E.g., Ex parte Parks, 93 U.S. 18, 23 L.Ed. 787; Ex parte Bigelow, 113 U.S. 328, 5 S.Ct. 542, 28 L.Ed. 1005; In re Belt, 159 U.S. 95, 15 S.Ct. 987, 40 L.Ed. 88; In re Moran, 203 U.S. 96, 27 S.Ct. 25, 51 L.Ed. 105; Knewel v. Egan, 268 U.S. 442, 45 S.Ct. 522, 69 L.Ed. 1036. Although the remedy extends to federal prisoners held in violation of federal law and not merely of the Federal Constitution, many cases have denied relief upon allegations merely of error of law and not of a substantial constitutional denial. E.g., Ex parte Parks, supra, 93 U.S. at 20 21; In re Wight, 134 U.S. 136, 148, 10 S.Ct. 487, 490, 33 L.Ed. 865; Harlan v. McGourin, 218 U.S. 442, 448, 31 S.Ct. 44, 47, 54 L.Ed. 1101; Eagles v. United States ex rel. Samuels, 329 U.S. 304, 67 S.Ct. 313, 91 L.Ed. 308. Such decisions are not however authorities against applications which invoke the historic office of the Great Writ to redress detentions in violation of fundamental law.23

In some of the cases the denial of the remedy on jurisdictional grounds seems to have been chosen in preference to decision of the merits of constitutional claims felt to be tenuous. E.g., In re Moran, supra; Knewel v. Egan, supra; Goto v. Lane, 265 U.S. 393, 44 S.Ct. 525, 68 L.Ed. 1070; United States v. Valante, 264 U.S. 563, 44 S.Ct. 411, 68 L.Ed. 850.24 And doubtless a powerful influence against the allowance of the remedy to state prisoners flowed from the availablity of review of state criminal judgments in this Court as of right. See, e.g., Andrews v. Swartz, 156 U.S. 272, 276, 15 S.Ct. 389, 391, 39 L.Ed. 422. Before 1916 review of such judgments was not discretionary by writ of certiorari but of right by writ of error.25 The occasions on which the extraordinary remedy of habeas corpus was indispensable were therefore few, since the practice of the Court was to put the habeas corpus applicant to his writ of error. E.g., In re Frederich, 149 U.S. 70, 13 S.Ct. 793, 37 L.Ed. 653; Bergemann v. Backer, 157 U.S. 655, 15 S.Ct. 727, 39 L.Ed. 845. And when the Court had no general appellate jurisdiction of federal criminal judgments, which was the case until 1891,26 the writ was sparingly allowed for the reason stated by Chief Justice Marshall in Ex parte Watkins, supra. Thus, in Bigelow the Court said: 'No appeal or writ of error * * * lies to this court. The act of congress has made the judgment of that court (the Supreme Court of the District of Columbia) conclusive, as it had a right to do, and the defendant, having one review of his trial and judgment, has no special reason to complain.' 113 U.S. at 329, 5 S.Ct. at 543. The same view is apparent in Ex parte Parks, supra, 93 U.S. at 20—21; Ex parte Curtis, 106 U.S. 371, 375, 1 S.Ct. 381, 385, 27 L.Ed. 232. Cf. Harlan v. McGourin, supra, 218 U.S., at 448, 31 S.Ct., at 47.

Nevertheless, the possibly grudging scope given the Great Writ in such cases is overshadowed by the numerous and varied allegations which this Court has deemed cognizable on habeas, not only in the last decades, but continuously since the fetters of the Watkins decision were thrown off in Ex parte Lange. E.g., Ex parte Wilson, 114 U.S. 417, 5 S.Ct. 935, 29 L.Ed. 89 (Fifth Amendment grand jury right); In re Converse, 137 U.S. 624, 11 S.Ct. 191, 34 L.Ed. 796 (Due Process Clause of Fourteenth Amendment); Rogers v. Peck, 199 U.S. 425, 26 S.Ct. 87, 50 L.Ed. 256 (same); Felts v. Murphy, 201 U.S. 123, 26 S.Ct. 366, 50 L.Ed. 689 (same); Lott v. Pittman, 243 U.S. 588, 37 S.Ct. 473, 61 L.Ed. 915 (same); Callan v. Wilson, 127 U.S. 540, 557, 8 S.Ct. 1301, 1307, 32 L.Ed. 223 (constitutional right to jury trial in federal criminal cases); Hawaii v. Mankichi, 190 U.S. 197, 23 S.Ct. 787, 47 L.Ed. 1016 (same) (by implication); Arndstein v. McCarthy, 254 U.S. 71, 41 S.Ct. 26, 65 L.Ed. 138 (Self-Incrimination Clause of Fifth Amendment); Morgan v. Devine, 237 U.S. 632, 35 S.Ct. 712, 59 L.Ed. 1153 (double jeopardy); Andersen v. Treat, 172 U.S. 24, 19 S.Ct. 67, 43 L.Ed. 351 (Sixth Amendment right to counsel); and see decisions cited at notes 17, 20, and 21, supra.

And so, although almost 300 years have elapsed since Bushell's Case, changed conceptions of the kind of criminal proceedings so fundamentally defective as to make imprisonment pursuant to them constitutionally intolerable should not be allowed to obscure the basic continuity in the conception of the writ as the remedy for such imprisonments.

It now remains to consider this principle in the application to the present case. It was settled in Brown v. Allen, supra, that the use of a coerced confession in a state criminal trial could be challenged in a federal habeas corpus proceeding. Yet actually the principle had been foreshadowed much earlier—indeed, in the very first case in which this Court reversed a state conviction on the ground that coerced confessions had been used in evidence. 'That complaint is * * * of a wrong so fundamental that it made the whole proceeding a mere pretense of a trial and rendered the conviction and sentence wholly void. Moore v. Dempsey * * *. (A)nd the proceeding thus vitiated could be challenged in any appropriate manner.' Brown v. Mississippi, 297 U.S. 278, 286—287, 56 S.Ct. 461, 465, 80 L.Ed. 682. Under the conditions of modern society, Noia's imprisonment, under a conviction procured by a confession held by the Court of Appeals in Caminito v. Murphy to have been coerced, and which the State here concedes was obtained in violation of the Fourteenth Amendment, is no less intolerable than was Bushell's under the conditions of a very different society; and habeas corpus is no less the appropriate remedy.

But, it is argued, a different result is compelled by the exigencies of federalism, which played no role in Bushell's case.

We can appraise this argument only in light of the historical accommodation that has been worked out between the state and federal courts respecting the administration of federal habeas corpus. Our starting point is the Judiciary Act of February 5, 1867, c. 28, § 1, 14 Stat. 385—386, which first extended federal habeas corpus to state prisoners generally, and which survives, except for some changes in wording, in the present statutory codification. The original Act and the current provisions are set out in an Appendix at the end of this opinion, pp. 441—445. Although the Act of 1867, like its English and American predecessors, nowhere defines habeas corpus, its expansive language and imperative tone, viewed against the background of post-Civil War efforts in Congress to deal severely with the States of the former Confederacy, would seem to make inescapable the conclusion that Congress was enlarging the habeas remedy as previously understood, not only in extending its coverage to state prisoners, but also in making its procedures more efficacious. In 1867, Congress was anticipating resistance to its Reconstruction measures and planning the implementation of the post-war constitutional Amendments. Debated and enacted at the very peak of the Radical Republicans' power, see 2 Warren, The Supreme Court in United States History (1928), 455—497, the measure that became the Act of 1867 seems plainly to have been designed to furnish a method additional to and independent of direct Supreme Court review of state court decisions for the vindication of the new constitutional guarantees. Congress seems to have had no thought, thus, that a state prisoner should abide state court determination of his constitutional defense—the necessary predicate of direct review by this Court—before resorting to federal habeas corpus. Rather, a remedy almost in the nature of removal from the state to the federal courts of state prisoners' constitutional contentions seems to have been envisaged. See Ex parte Bridges, Fed.Cas.No.1,862, 2 Woods 428, 432 (Cir.Ct.N.D.Ga.1875); Ex parte McCready, Fed.Cas.No.8732, 1 Hughes 598 (Cir.Ct.E.D.Va.1874). Compare Rev.Stat., 1874, § 641 (providing for removal to Federal Circuit Court 'When any civil suit or criminal prosecution is commenced in any State court, for any cause whatsoever, against any person who is denied or cannot enforce in the judicial tribunals of the State * * * any right secured to him by any law providing for the equal civil rights of citizens of the United States'); Virginia v. Rives, 100 U.S. 313, 25 L.Ed. 667.

The elaborate provisions in the Act for taking testimony and trying the facts anew in habeas hearings27 lend support to this conclusion, as does the legislative history of House bill No. 605, which became, with slight changes, the Act of February 5, 1867. The bill was introduced in response to a resolution of the House on December 19, 1865, asking the Judiciary Committee to determine 'what legislation is necessary to enable the courts of the United States to enforce the freedom of the wives and children of soldiers of the United States * * * and also to enforce the liberty of all persons under the operation of the constitutional amendment abolishing slavery.' Cong. Globe, 39th Cong., 1st Sess. 87. The terms in which it was described by its proponent, Representative Lawrence of Ohio, leave little doubt of the breadth of its intended scope: 'the effect of * * * (bill No. 605) is to enlarge the privilege of the writ of hobeas (sic) corpus, and make the jurisdiction of the courts and judges of the United States coextensive with all the powers that can be conferred upon them. It is a bill of the largest liberty.' Cong.Globe, 39th Cong., 1st Sess. 4151 (1866). This Court, shortly after the passage of the Act, described it in equally broad terms: 'This legislation is of the most comprehensive character. It brings within the habeas corpus jurisdiction of every court and of every judge every possible case of privation of liberty contrary to the National Constitution, treaties, or laws. It is impossible to widen this jurisdiction.' Ex parte McCardle, 6 Wall. 318, 325 326, 18 L.Ed. 816.

In thus extending the habeas corpus power of the federal courts evidently to what was conceived to be its constitutional limit, the Act of February 5, 1867, clearly enough portended difficult problems concerning the relationship of the state and federal courts in the area of criminal administration. Such problems were not slow to mature. Only eight years after passage of the Act, Mr. Justice Bradley, sitting as Circuit Justice, held that a convicted state prisoner who had not sought any state appellate or collateral remedies could nevertheless win immediate release on federal habeas if he proved the unconstitutionality of his conviction; although the judgment was not final within the state court system, the federal court had the power to inquire into the legality of the prisoner's detention. Ex parte Bridges, supra. Accord, Ex parte McCready, supra. This holding flowed inexorably from the clear congressional policy of affording a federal forum for the determination of the federal claims of state criminal defendants, and it was explicitly approved by the full Court in Ex parte Royall, 117 U.S. 241, 253, 6 S.Ct. 734, 741, 29 L.Ed. 868, a case in which habeas had been sought in advance of trial. The Court held that even in such a case the federal courts had the power to discharge a state prisoner restrained in violation of the Federal Constitution, see 117 U.S., at 245, 250—251, 6 S.Ct., at 739—740, but that ordinarily the federal court should stay its hand on habeas pending completion of the state court proceedings. This qualification plainly stemmed from considerations of comity rather than power, and envisaged only the postponement, not the relinquishment, of federal habeas corpus jurisdiction, which had attached by reason of the allegedly unconstitutional detention and could not be ousted by what the state court might decide. As well stated in a later case:

'* * * While the federal courts have the power and may discharge the accused in advance of his trial, if he is restrained of his liberty in violation of the federal constitution or laws, * * * the practice of exercising such power before the question has been raised or determined in the state court is one which ought not to be encouraged. The party charged waives no defect of jurisdiction by submitting to a trial of his case upon the merits, and we think that comity demands that the state courts, under whose process he is held, and which are, equally with the federal courts, charged with the duty of protecting the accused in the enjoyment of his constitutional rights, should be appealed to in the first instance. Should such rights be denied, his remedy in the federal court will remain unimpaired.'28

These decisions fashioned a doctrine of abstention, whereby full play would be allowed the States in the administration of their criminal justice without prejudice to federal rights enwoven in the state proceedings. Thus the Court has frequently held that application for a writ of habeas corpus should have been denied 'without prejudice to a renewal of the same after the accused had availed himself of such remedies as the laws of the state afforded * * *.' Minnesota v. Brundage, 180 U.S. 499, 500—501, 21 S.Ct. 455, 456, 45 L.Ed. 639. See also Ex parte Royall, supra, 117 U.S., at 254, 6 S.Ct., at 741. With refinements, this doctrine requiring the exhaustion of state remedies is now codified in 28 U.S.C. § 2254.29 But its rationale has not changed: 'it would be unseemly in our dual system of government for a federal district court to upset a state court conviction without an opportunity to the state courts to correct a constitutional violation * * *. Solution was found in the doctrine of comity between courts, a doctrine which teaches that one court should defer action on causes properly within its jurisdiction until the courts of another sovereignty with concurrent powers, and already cognizant of the litigation, have had an opportunity to pass upon the matter.' Darr v. Burford, 339 U.S. 200, 204, 70 S.Ct. 587, 590, 94 L.Ed. 761. The rule of exhaustion 'is not one defining power but one which relates to the appropriate exercise of power.' Bowen v. Johnston, 306 U.S. 19, 27, 59 S.Ct. 442, 446, 83 L.Ed. 455. Cf. Stack v. Boyle, 342 U.S. 1, 72 S.Ct. 1, 96 L.Ed. 3; Frisbie v. Collins, 342 U.S. 519, 72 S.Ct. 509, 96 L.Ed. 541; Douglas v. Green, 363 U.S. 192, 80 S.Ct. 1048, 4 L.Ed.2d 1142.

The reasoning of Ex parte Royall and its progeny suggested that after the state courts had decided the federal question on the merits against the habeas petitioner, he could return to the federal court on habeas and there relitigate the question, else a rule of timing would become a rule circumscribing the power of the federal courts on habeas, in defiance of unmistakable congressional intent. And so this Court has consistently held, save only in Frank v. Mangum, 237 U.S. 309, 35 S.Ct. 582, 59 L.Ed. 969. In that case, the State Supreme Court had rejected on the merits petitioner's contention of mob domination at his trial, and this Court held that habeas would not lie because the State had afforded petitioner corrective process. However, the decision seems grounded not in any want of power, for the Court described the federal courts' habeas powers in the broadest terms, 237 U.S. at 330—331, 35 S.Ct., at 588, but rather in a narrow conception of due process in state criminal justice. The Court felt that so long as Frank had had an opportunity to challenge his conviction in some impartial tribunal, such as the State Supreme Court, he had been afforded the process he was constitutionally due.

The majority's position in Frank, however, was substantially repudiated in Moore v. Dempsey, 261 U.S. 86, 43 S.Ct. 265, 67 L.Ed. 543, a case almost identical in all pertinent respects to Frank. Mr. Justice Holmes, writing for the Court in Moore (he had written the dissenting opinion in Frank), said: 'if in fact a trial is dominated by a mob so that there is an actual interference with the course of justice, there is a departure from due process of law; * * * (if) the State Courts failed to correct the wrong, * * * perfection in the machinery for correction * * * can (not) prevent this Court from securing to the petitioners their constitutional rights.' 261 U.S., at 90—91, 43 S.Ct., at 266. It was settled in Moore, restoring what evidently had been the assumption until Frank, see, e.g., Cook v. Hart, 146 U.S. 183, 194—195, 13 S.Ct. 40, 43—44, 36 L.Ed. 934; and cases cited in note 28, supra, that the state courts' view of the merits was not entitled to conclusive weight. We have not deviated from that position.30 Thus, we have left the weight to be given a particular state court adjudication of a federal claim later pressed on habeas substantially in the discretion of the Federal District Court: 'the state adjudication carries the weight that federal practice gives to the conclusion of a court * * * of last resort another jurisdiction on federal constitutional issues. It is not res judicata.' Brown v. Allen, supra, 344 U.S., at 458, 73 S.Ct., at 408 (opinion of Mr. Justice Reed). '* * * (N)o binding weight is to be attached to the State determination. The congressional requirement is greater. The State court cannot have the last say when it, though on fair consideration and what procedurally may be deemed fairness, may have misconceived a federal constitutional right.' 344 U.S., at 508, 73 S.Ct., at 446 (opinion of Mr. Justice Frankfurter). Even if the state court adjudication turns wholly on primary, historical facts, the Federal District Court has a broad power on habeas to hold an evidentiary hearing and determine the facts.31

The breadth of the federal courts' power of independent adjudication on habeas corpus stems from the very nature of the writ, and conforms with the classic English practice.32 As put by Mr. Justice Holmes in his dissenting opinion in Frank v. Mangum, supra, 237 U.S., at 348, 35 S.Ct., at 595: 'If the petition discloses facts that amount to a loss of jurisdiction in the trial court, jurisdiction could not be restored by any decision above.' It is of the historical essence of habeas corpus that it lies to test proceedings so fundamentally lawless that imprisonment pursuant to them is not merely erroneous but void. Hence, the familiar principle that res judicata is inapplicable in habeas proceedings, see, e.g., Darr v. Burford, 339 U.S. 200, 214, 70 S.Ct. 587, 595, 94 L.Ed. 761; Salinger v. Loisel, 265 U.S. 224, 230, 44 S.Ct. 519, 521, 68 L.Ed. 989; Frank v. Mangum, 237 U.S. 309, 334, 35 S.Ct. 582, 589, 59 L.Ed. 969; Church, Habeas Corpus (1884), § 386, is really but an instance of the larger principle that void judgments may be collaterally impeached. Restatement, Judgments (1942), §§ 7, 11; Note, Res Judicata, 65 Harv.L.Rev. 818, 850 (1952). Cf. Windsor v. McVeigh, 93 U.S. 274, 282—283, 23 L.Ed. 914. So also, the traditional characterization of the writ of habeas corpus as an original (save perhaps when issued by this Court33) civil remedy for the enforcement of the right to personal liberty,34 rather than as a stage of the state criminal proceedings or as an appeal therefrom, emphasizes the independence of the federal habeas proceedings from what has gone before. This is not to say that a state criminal judgment resting on a constitutional error is void for all purposes. But conventional notions of finality in criminal litigation cannot be permitted to defeat the manifest federal policy that federal constitutional rights of personal liberty shall not be denied without the fullest opportunity for plenary federal judicial review.

Despite the Court's refusal to give binding weight to state court determinations of the merits in habeas, it has not infrequently suggested that where the state court declines to reach the merits because of a procedural default, the federal courts may be foreclosed from granting the relief sought on habeas corpus.35 But the Court's practice in this area has been far from uniform,36 and even greater divergency has characterized the practice of the lower federal courts.37

For the present, however, it suffices to note that rarely, if ever, has the Court predicated its deference to state procedural rules on a want of power to entertain a habeas application where a procedural default was committed by the defendant in the state courts. Typically, the Court, like the District Court in the instant case, has approached the problem as an aspect of the rule requiring exhaustion of state remedies, which is not a rule distributing power as between the state and federal courts. See pp. 417—420, supra. That was the approach taken in the Spencer and Daniels decisions, the most emphatic in their statement of deference to state rules of procedure. The same considerations of comity that led the Court to refuse relief to one who had not yet availed himself of his state remedies likewise prompted the refusal of relief to one who had inexcusably failed to tender the federal questions to the state courts. Either situation poses a threat to the orderly administration of criminal justice that ought if possible to be averted. Whether in fact the conduct of a Spencer or a Daniels was inexcusable in this sense is beside the point, as is the arguable illogicality of turning a rule of timing into a doctrine of forfeitures. The point is that the Court, by relying upon a rule of discretion, avowedly flexible, Frisbie v. Collins, 342 U.S. 519, 72 S.Ct. 509, 96 L.Ed. 541, yielding always to 'exceptional circumstances,' Bowen v. Johnston, 306 U.S. 19, 27, 59 S.Ct. 442, 446, 83 L.Ed. 455, has refused to concede jurisdictional significance to the abortive state court proceeding.

We have reviewed the development of habeas corpus at some length because the question of the instant case has obvious importance to the proper accommodation of a great constitutional privilege and the requirements of the federal system. Our survey discloses nothing to suggest that the Federal District Court lacked the power to order Noia discharged because of a procedural forfeiture he may have incurred under state law. On the contrary, the nature of the writ at common law, the language and purpose of the Act of February 5, 1867, and the course of decisions in this Court extending over nearly a century are wholly irreconcilable with such a limitation. At the time the privilege of the writ was written into the Federal Constitution it was settled that the writ lay to test any restraint contrary to fundamental law, which in England stemmed ultimately from Magna Charta but in this country was embodied in the written Constitution. Congress in 1867 sought to provide a federal forum for state prisoners having constitutional defenses by extending the habeas corpus powers of the federal courts to their constitutional maximum. Obedient to this purpose, we have consistently held that federal court jurisdiction is conferred by the allegation of an unconstitutional restraint and is not defeated by anything that may occur in the state court proceedings. State procedural rules plainly must yield to this overriding federal policy.

A number of arguments are advanced against this conclusion. One, which concedes the breadth of federal habeas power, is that a state prisoner who forfeits his opportunity to vindicate federal defenses in the state court has been given all the process that is constitutionally due him, and hence is not restrained contrary to the Constitution. But this wholly misconceives the scope of due process of law, which comprehends not only the right to be heard but also a number of explicit procedural rights—for example, the right not to be convicted upon evidence which includes one's coerced confession—drawn from the Bill of Rights. As Mr. Justice Holmes explained in Moore v. Dempsey, see pp. 839—840, supra, a mob-dominated trial is no less a denial of due process because the State Supreme Court believed that the trial was actually a fair one. A fortiori, due process denied in the proceedings leading to conviction is not restored just because the state court declines to adjudicate the claimed denial on the merits.

A variant of this argument is that if the state court declines to entertain a federal defense because of a procedural default, then the prisoner's custody is actually due to the default rather than to the underlying constitutional infringement, so that he is not in custody in violation of federal law.38 But this ignores the important difference between rights and particular remedies. Cf. Douglas v. Jeannette, 319 U.S. 157, 63 S.Ct. 877, 87 L.Ed. 1324; Stefanelli v. Minard, 342 U.S. 117, 72 S.Ct. 118, 96 L.Ed. 138; Wolf v. Colorado, 338 U.S. 25, 69 S.Ct. 1359, 93 L.Ed. 1782. A defendant by committing a procedural default may be debarred from challenging his conviction in the state courts even on federal constitutional grounds. But a forfeiture of remedies does not legitimize the unconstitutional conduct by which his conviction was procured. Would Noia's failure to appeal have precluded him from bringing an action under the Civil Rights Acts against his inquisitors? The Act of February 5, 1867, like the Civil Rights Acts, was intended to furnish an independent, collateral remedy for certain privations of liberty. The conceptual difficulty of regarding a default as extinguishing the substantive right is increased where, as in Noia's case, the default forecloses extraordinary remedies. In what sense is Noia's custody not in violation of federal law simply because New York will not allow him to challenge it on coram nobis or on delayed appeal? But conceptual problems aside, it should be obvious that to turn the instant case on the meaning of 'custody in violation of the Constitution' is to reason in circles. The very question we face is how completely federal remedies fall with the state remedies; when we have answered this, we shall know in what sense custody may be rendered lawful by a supervening procedural default.

It is a familiar principle that this Court will decline to review state court judgments which rest on independent and adequate state grounds, notwithstanding the co-presence of federal grounds. See, e.g., N.A.A.C.P. v. Alabama ex rel. Patterson, 357 U.S. 449, 78 S.Ct. 1163, 2 L.Ed.2d 1488; Fox Film Corp. v. Muller, 296 U.S. 207, 56 S.Ct. 183, 80 L.Ed. 158. Section 25 of the Judiciary Act of 1789, c. 20, 1 Stat. 85—87, denied this Court power to base the reversal of a state court decision on any error other 'than such as * * * immediately respects * * * questions of validity or construction of the said (Federal) constitution, treaties, statutes, commissions, or authorities in dispute.' The deletion of the express restriction by the Judiciary Act of February 5, 1867, c. 28, § 2, 14 Stat. 386—387, did not enlarge this Court's power in that regard. Murdock v. Memphis, 20 Wall. 590, 22 L.Ed. 429. Murdock was a case involving state substantive grounds, but the principle is also applicable in cases involving procedural grounds. See, e.g., Herb v. Pitcairn, 324 U.S. 117, 65 S.Ct. 459, 89 L.Ed. 789; Davis v. Wechsler, 263 U.S. 22, 44 S.Ct. 13, 68 L.Ed. 143; Ward v. Board of County Comm'rs, 253 U.S. 17, 40 S.Ct. 419, 64 L.Ed. 761. Thus, a default such as Noia's, if deemed adequate and independent (a question on which we intimate no view), would cut off review by this Court of the state coram nobis proceeding in which the New York Court of Appeals refused him relief. It is contended that it follows from this that the remedy of federal habeas corpus is likewise cut off.39

The fatal weakness of this contention is its failure to recognize that the adequate state-ground rule is a function of the limitations of appellate review. Most of the opinion in the Murdock case is devoted to demonstrating the Court's lack of jurisdiction on direct review to decide questions of state law in cases also raising federal questions. It followed from this holding that if the state question was dispositive of the case, the Court could not decide the federal question. The federal question was moot; nothing turned on its resolution. And so we have held that the adequate state-ground rule is a consequence of the Court's obligation to refrain from rendering advisory opinions or passing upon moot questions.40

But while our appellate function is concerned only with the judgments or decrees of state courts, the habeas corpus jurisdiction of the lower federal courts is not so confined. The jurisdictional prerequisite is not the judgment of a state court but detention simpliciter. The entire course of decisions in this Court elaborating the rule of exhaustion of state remedies is wholly incompatible with the proposition that a state court judgment is required to confer federal habeas jurisdiction. And the broad power of the federal courts under 28 U.S.C. § 2243 summarily to hear the application and to 'determine the facts, and dispose of the matter as law and justice require,' is hardly characteristic of an appellate jurisdiction. Habeas lies to enforce the right of personal liberty; when that right is denied and a person confined, the federal court has the power to release him. Indeed, it has no other power; it cannot revise the state court judgment; it can act only on the body of the petitioner. In re Medley, Petitioner, 134 U.S. 160, 173, 10 S.Ct. 384, 388, 33 L.Ed. 835.

To be sure, this may not be the entire answer to the contention that the adequate state-ground principle should apply to the federal courts on habeas corpus as well as to the Supreme Court on direct review of state judgments. The Murdock decision may be supported not only by the factor of mootness, but in addition by certain characteristics of the federal system. The first question the Court had to decide in Murdock was whether it had the power to review state questions in cases also raising federal questions. It held that it did not, thus affirming the independence of the States in matters within the proper sphere of their lawmaking power from federal judicial interference. For the federal courts to refuse to give effect in habeas proceedings to state procedural defaults might conceivably have some effect upon the States' regulation of their criminal procedures. But the problem is crucially different from that posed in Murdock of the federal courts' deciding questions of substantive state law. In Noia's case the only relevant substantive law is federal—the Fourteenth Amendment. State law appears only in the procedural framework for adjudicating the substantive federal question. The paramount interest is federal. Cf. Dice v. Akron, C. & Y.R. Co., 342 U.S. 359, 72 S.Ct. 312, 96 L.Ed. 398. That is not to say that the States have not a substantial interest in exacting compliance with their procedural rules from criminal defendants asserting federal defenses. Of course orderly criminal procedure is a desideratum, and of course there must be sanctions for the flouting of such procedure. But that state interest 'competes * * * against an ideal * * * (the) ideal of fair procedure.' Schaefer, Federalism and State Criminal Procedure, 70 Harv.L.Rev. 1, 5 (1956). And the only concrete impact the assumption of federal habeas jurisdiction in the face of a procedural default has on the state interest we have described, is that it prevents the State from closing off the convicted defendant's last opportunity to vindicate his constitutional rights, thereby punishing him for his default and deterring others who might commit similar defaults in the future.

Surely this state interest in an airtight system of forfeitures is of a different order from that, vindicated in Murdock, in the autonomy of state law within the proper sphere of its substantive regulation. The difference is illustrated in the settled principle that if a prisoner is detained lawfully under one count of the indictment, he cannot challenge the lawfulness of a second count on federal habeas. McNally v. Hill, 293 U.S. 131, 55 S.Ct. 24, 79 L.Ed. 238. For the federal court to order the release of such a prisoner would be to nullify a proceeding—that under the first count—wholly outside the orbit of federal interest. Contrariwise the only count under which Noia was convicted and imprisoned is admitted to be vitiated by force of federal law.

Certainly this Court has differentiated the two situations in its application of the adequate state-ground rule. While it has deferred to state substantive grounds so long as they are not patently evasive of or discriminatory against federal rights, it has sometimes refused to defer to state procedural grounds only because they made burdensome the vindication of federal rights.41 That the Court nevertheless ordinarily gives effect to state procedural grounds may be attributed to considerations which are peculiar to the Court's role and function and have no relevance to habeas corpus proceedings in the Federal District Courts: the unfamiliarity of members of this Court with the minutiae of 50 States' procedures; the inappropriateness of crowding our docket with questions turning wholly on particular state procedures; the web of rules and statutes that circumscribes our appellate jurisdiction; and the inherent and historical limitations of such a jurisdiction.

A practical appraisal of the state interest here involved plainly does not justify the federal courts' enforcing on habeas corpus a doctrine of forfeitures under the guise of applying the adequate state-ground rule. We fully grant, see p. 438, infra, that the exigencies of federalism warrant a limitation whereby the federal judge has the discretion to deny relief to one who has deliberately sought to subvert or evade the orderly adjudication of his federal defenses in the state courts. Surely no stricter rule is a realistic necessity. A man under conviction for crime has an obvious inducement to do his very best to keep his state remedies open, and not stake his all on the outcome of a federal habeas proceeding which, in many respects, may be less advantageous to him than a state court proceeding. See Rogers v. Richmond, 365 U.S. 534, 547—548, 81 S.Ct. 735, 742—743, 5 L.Ed.2d 760. And if because of inadvertence or neglect he runs afoul of a state procedural requirement, and thereby forfeits his state remedies, appellate and collateral, as well as direct review thereof in this Court, those consequences should be sufficient to vindicate the State's valid interest in orderly procedure. Whatever residum of state interest there may be under such circumstances is manifestly insufficient in the face of the federal policy, drawn from the ancient principles of the writ of habeas corpus, embodied both in the Federal Constitution and in the habeas corpus provisions of the Judicial Code, and consistently upheld by this Court, of affording an effective remedy for restraints contrary to the Constitution. For these several reasons we reject as unsound in principle, as well as not supported by authority, the suggestion that the federal courts are without power to grant haneas relief to an applicant whose federal claims would not be heard on direct review in this Court because of a procedural default furnishing an adequate and independent ground of state decision.

What we have said substantially disposes of the further contention that 28 U.S.C. § 2254 embodies a doctrine of forfeitures and cuts off relief when there has been a failure to exhaust state remedies no longer available at the time habeas is sought. This contention is refuted by the language of the statute and by its history.42 It was enacted to codify the judicially evolved rule of exhaustion, particularly as formulated in Ex parte Hawk, 321 U.S. 114, 64 S.Ct. 448, 88 L.Ed. 572. See the review of the legislative history in Darr v. Burford, 339 U.S. 200, 211—213, 70 S.Ct. 587, 593—595, 94 L.Ed. 761. Nothing in the Hawk opinion points to past exhaustion. Very little support can be found in the long course of previous decisions by this Court elaborating the rule of exhaustion for the proposition that it was regarded at the time of the revision of the Judicial Code as jurisdictional rather than merely as a rule ordering the state and federal proceedings so as to eliminate unnecessary federal-state friction. There is thus no warrant for attributing to Congress, in the teeth of the language of § 2254, intent to work a radical innovation in the law of habeas corpus. We hold that § 2254 is limited in its application to failure to exhaust state remedies still open to the habeas applicant at the time he files his application in federal court.43 Parenthetically, we note that our holding in Irvin v. Dowd, 359 U.S. 394, 79 S.Ct. 825, 3 L.Ed.2d 900, is not inconsistent. Our holding there was that since the Indiana Supreme Court, Irvin v. State, 236 Ind. 384, 139 N.E.2d 898, had reached the merits of Irvin's federal claim, the District Court, 153 F.Supp. 531, was not barred by § 2254 from determining the merits of Irvin's constitutional contentions.

Noia timely sought and was denied certiorari here from the adverse decision of the New York Court of Appeals on his coram nobis application, and therefore the case does not necessarily draw in question the continued vitality of the holding in Darr v. Burford, supra, that a state prisoner must ordinarily seek certiorari in this Court as a precondition of applying for federal habeas corpus. But what we hold today necessarily overrules Darr v. Burford to the extent it may be thought to have barred a state prisoner from federal habeas relief if he had failed timely to seek certiorari in this Court from an adverse state decision. Furthermore, our decision today affects all procedural hurdles to the achievement of swift and imperative justice on habeas corpus, and because the hurdle erected by Darr v. Buford is unjustifiable under the principles we have expressed, even insofar as it may be deemed merely an aspect of the statutory requirement of present exhaustion, that decision in that respect also is hereby overruled.

The soundness of the decision was questioned from the beginning. See Pollock, Certiorari and Habeas Corpus, 42 J. of Crim.L. 356, 357—358, n. 15, 364 (1951). Section 2254 speaks only of 'remedies available in the courts of the State.' Nevertheless, the Court in Darr v. Burford put a gloss upon these words to include petitioning for certiorari in this Court, which is not the court of any State, among the remedies that an applicant must exhaust before proceeding in federal habeas corpus. It is true that before the enactment of § 2254 the Court had spoken of the obligation to seek review in this Court before applying for habeas. E.g., Baker v. Grice, 169 U.S. 284, 18 S.Ct. 323, 42 L.Ed. 748; Markuson v. Boucher, 175 U.S. 184, 20 S.Ct. 76, 44 L.Ed. 124. But that was at the time when review of state criminal judgments in this Court was by writ of error. Review here was thus a stage of the normal appellate process. The writ of certiorari, which today provides the usual mode of invoking this Court's appellate jurisdiction of state criminal judgments, 'is not a matter of right, but of sound judicial discretion, and will be granted only where there are special and important reasons therefor.' Supreme Court Rule 19(1). Review on certiorari therefore does not provide a normal appellate channel in any sense comparable to the writ of error.

It is also true that Ex parte Hawk, 321 U.S. 114, 64 S.Ct. 448, 88 L.Ed. 572, a decision cited in the Reviser's Note to § 2254, intimated in dictum that exhaustion might comprehend seeking certiorari here. 321 U.S., at 116—177, 64 S.Ct., at 449—450. But that passing reference cannot be exalted into an attribution to Congress of a design patently belied by the unequivocal statutory language.

The rationale of Darr v. Burford emphasized the values of comity between the state and federal courts, and assumed that these values would be realized by requiring a state criminal defendant to afford this Court an opportunity to pass upon state action before he might seek relief in federal habeas corpus. But the expectation has not been realized in experience. On the contrary the requirement of Darr v. Burford has proved only to be an unnecessarily burdensome step in the orderly processing of the federal claims of those convicted of state crimes. The goal of prompt and fair criminal justice has been impeded because in the overwhelming number of cases the applications for certiorari have been denied for failure to meet the standard of Rule 19. And the demands upon our time in the examination and decision of the large volume of petitions which fail to meet that test have unwarrantably taxed the resources of this Court. Indeed, it has happened that counsel on oral argument has confessed that the record was insufficient to justify our consideration of the case but that he had felt compelled to make the futile time-consuming application in order to qualify for proceeding in a Federal District Court on habeas corpus to make a proper record. Bullock v. South Carolina, 365 U.S. 292, 81 S.Ct. 686, 5 L.Ed.2d 570. And so in a number of cases the Court has apparently excused compliance with the requirement. See, e.g., Weston v. Sigler, 361 U.S. 37, 80 S.Ct. 141, 4 L.Ed.2d 111; Bailey v. Arkansas, 358 U.S. 869, 79 S.Ct. 101, 3 L.Ed.2d 101; Poret v. Sigler, 355 U.S. 60, 78 S.Ct. 144, 2 L.Ed.2d 107; Massey v. Moore, 348 U.S. 105, 75 S.Ct. 145, 99 L.Ed. 135. Cf. Thomas v. Arizona, 356 U.S. 390, 392, n. 1, 78 S.Ct. 885, 886, 2 L.Ed.2d 863. The same practice has sometimes been followed in the Federal District Courts. See Reitz, Federal Habeas Corpus: Postconviction Remedy for State Prisoners, 108 U. of PaL.Rev. 461, 499 (1960).

Moreover, comity does not demand that such a price in squandered judicial resources be paid; the needs of comity are adequately served in other ways. The requirement that the habeas petitioner exhaust state court remedies available to him when he applies for federal habeas corpus relief gives state courts the opportunity to pass upon and correct errors of federal law in the state prisoner's conviction. And the availability to the States of eventual review on certiorari of such decisions of lower federal courts as may grant relief is always open. Our function of making the ultimate accommodation between state criminal law enforcement and state prisoners' constitutional rights becomes more meaningful when grounded in the full and complete record which the lower federal courts on habeas corpus are in a position to provide.

Although we hold that the jurisdiction of the federal courts on habeas corpus is not affected by procedural defaults incurred by the applicant during the state court proceedings, we recognize a limited discretion in the federal judge to deny relief to an applicant under certain circumstances. Discretion is implicit in the statutory command that the judge, after granting the writ and holding a hearing of appropriate scope, 'dispose of the matter as law and justice require,' 28 U.S.C. § 2243; and discretion was the flexible concept employed by the federal courts in developing the exhaustion rule. Furthermore, habeas corpus has traditionally been regarded as governed by equitable principles. United States ex rel. Smith. v. Baldi, 344 U.S. 561, 573, 73 S.Ct. 391, 397, 97 L.Ed. 549 (dissenting opinion). Among them is the principle that a suitor's conduct in relation to the matter at hand may disentitle him to the relief he seeks. Narrowly circumscribed, in conformity to the historical role of the writ of habeas corpus as an effective and imperative remedy for detentions contrary to fundamental law, the principle is unexceptionable. We therefore hold that the federal habeas judge may in his discretion deny relief to an applicant who has deliberately bypassed the orderly procedure of the state courts and in so doing has forfeited his state court remedies.

But we wish to make very clear that this grant of discretion is not to be interpreted as a permission to introduce legal fictions into federal habeas corpus. The classic definition of waiver enunciated in Johnson v. Zerbst, 304 U.S. 458, 464, 58 S.Ct. 1019, 1023, 82 L.Ed. 1461—'an intentional relinquishment or abandonment of a know right or privilege'—furnishes the controlling standard. If a habeas applicant, after consultation with competent counsel or otherwise, understandingly and knowingly forewent the privilege of seeking to vindicate his federal claims in the state courts, whether for strategic, tactical, or any other reasons that can fairly be described as the deliberate by-passing of state procedures, then it is open to the federal court on habeas to deny him all relief if the state courts refused to entertain his federal claims on the merits—though of course only after the federal court has satisfied itself, by holding a hearing or by some other means, of the facts bearing upon the applicant's default. Cf. Price v. Johnston, 334 U.S. 266, 291, 68 S.Ct. 1049, 1063, 92 L.Ed. 1356. At all events we wish it clearly understood that the standard here put forth depends on the considered choice of the petitioner.44 Cf. Carnley v. Cochran, 369 U.S. 506, 513 517, 82 S.Ct. 884, 888—891, 8 L.Ed.2d 70; Moore v. Michigan, 355 U.S. 155, 162—165, 78 S.Ct. 191, 195—197, 2 L.Ed.2d 167. A choice made by counsel not participated in by the petitioner does not automatically bar relief. Nor does a state court's finding of waiver bar independent determination of the question by the federal courts on habeas, for waiver affecting federal rights is a federal question. E.g., Rice v. Olson, 324 U.S. 786, 65 S.Ct. 989, 89 L.Ed. 1367.

The application of the standard we have adumbrated to the facts of the instant case is not difficult. Under no reasonable view can the State's version of Noia's reason for not appealing support an inference of deliberate by-passing of the state court system. For Noia to have appealed in 1942 would have been to run a substantial risk of electrocution. His was the grisly choice whether to sit content with life imprisonment or to travel the uncertain avenue of appeal which, if successful, might well have led to a retrial and death sentence. See, e.g., Palko v. Connecticut, 302 U.S. 319, 58 S.Ct. 149, 82 L.Ed. 288. He declined to play Russian roulette in this fashion. This was a choice by Noia not to appeal, but under the circumstances it cannot realistically be deemed a merely tactical or strategic litigation step, or in any way a deliberate circumvention of state procedures. This is not to say that in every case where a heavier penalty, even the death penalty, is a risk incurred by taking an appeal or otherwise foregoing a procedural right, waiver as we have defined it cannot be found. Each case must stand on its facts. In the instant case, the language of the judge in sentencing Noia, see note 3, supra, made the risk that Noia, if reconvicted, would be sentenced to death, palpable and indeed unusually acute.

It should be unnecessary to repeat what so often has been said and what so plainly is the case: that the availability of the Great Writ of habeas corpus in the federal courts for persons in the custody of the States offends no legitimate state interest in the enforcement of criminal justice or procedure. Our decision today swings open no prison gates. Today as always few indeed is the number of state prisoners who eventually win their freedom by means of federal habeas corpus.45 Those few who are ultimately successful are persons whom society has grievously wronged and for whom belated liberation is little enough compensation. Surely no fair-minded person will contend that those who have been deprived of their liberty without due process of law ought nevertheless to languish in prison. Noia, no less than his codefendants Caminito and Bonino, is conceded to have been the victim of unconstitutional state action. Noia's case stands on its own; but surely no just and humane legal system can tolerate a result whereby a Caminito and a Bonino are at liberty because their confessions were found to have been coerced yet a Noia, whose confession was also coerced, remains in jail for life. For such anomalies, such affronts to the conscience of a civilized society, habeas corpus is predestined by its historical role in the struggle for personal liberty to be the ultimate remedy. If the States withhold effective remedy, the federal courts have the power and the duty to provide it. Habeas corpus is one of the precious heritages of Anglo-American civilization. We do no more today than confirm its continuing efficacy.

Affirmed.

The Judiciary Act of February 5, 1867, c. 28, § 1, 14 Stat. 385—386:

'* * * (T)he several courts of the United States, and the several justices and judges of such courts, within their respective jurisdictions, in addition to the authority already conferred by law, shall have power to grant writs of habeas corpus in all cases where any person may be restrained of his or her liberty in violation of the constitution, or of any treaty or law of the United States; and it shall be lawful for such person so restrained of his or her liberty to apply to either of said justices or judges for a writ of habeas corpus, which application shall be in writing and verified by affidavit, and shall set forth the facts concerning the detention of the party applying, in whose custody he or she is detained, and by virtue of what claim or authority, if known; and the said justice or judge to whom such application shall be made shall forthwith award a writ of habeas corpus, unless it shall appear from the petition itself that the party is not deprived of his or her liberty in contravention of the constitution or laws of the United States. Said writ shall be directed to the person in whose custody the party is detained, who shall make return of said writ and bring the party before the judge who granted the writ, and certify the true cause of the detention of such person within three days thereafter, unless such person be detained beyond the distance of twenty miles; and if beyond the distance of twenty miles and not above one hundred miles, then within ten days; and if beyond the distance of one hundred miles, then within twenty days. And upon the return of the writ of habeas corpus a day shall be set for the hearing of the cause, not exceeding five days thereafter, unless the party petitioning shall request a longer time. The petitioner may deny any of the material facts set forth in the return, or may allege any fact to show that the detention is in contravention of the constitution or laws of the United States, which allegations or denials shall be made on oath. The said return may be amended by leave of the court or judge before or after the same is filed, as also may all suggestions made against it, that thereby the material facts may be ascertained. The said court or judge shall proceed in a summary way to determine the facts of the case, by hearing testimony and the arguments of the parties interested, and if it shall appear that the petitioner is deprived of his or her liberty in contravention of the constitution or laws of the United States, he or she shall forthwith be discharged and set at liberty. And if any person or persons to whom such writ of habeas corpus may be directed shall refuse to obey the same, or shall neglect or refuse to make return, or shall make a false return thereto, in addition to the remedies already given by law, he or they shall be deemed and taken to be guilty of a misdemeanor, and shall, on conviction before any court of competent jurisdiction, be punished by fine not exceeding one thousand dollars, and by imprisonment not exceeding one year, or by either, according to the nature and aggravation of the case. From the final decision of any judge, justice, or court, inferior to the circuit court, an appeal may be taken to the circuit court of the United States for the district in which said cause is heard, and from the judgment of said circuit court to the Supreme Court of the United States, on such terms and under such regulations and orders, as well for the custody and appearance of the person alleged to be restrained of his or her liberty, as for sending up to the appellate tribunal a transcript of the petition, writ of habeas corpus, return thereto, and other proceedings, as may be prescribed by the Supreme Court, or, in default of such, as the judge hearing said cause may prescribe; and pending such proceedings or appeal, and until final judgment be rendered therein, and after final judgment of discharge in the same, any proceeding against such person so alleged to be restrained of his or her liberty in any State court, or by or under the authority of any State, for any matter or thing so heard and determined, or in process of being heard and determined, under and by virtue of such writ of habeas corpus, shall be deemed null and void.' 28 U.S.C. § 2241:

'(a) Writs of habeas corpus may be granted by the Supreme Court, any justice thereof, the district courts and any circuit judge within their respective jurisdictions. * * *

'(c) The writ of habeas corpus shall not extend to a prisoner unless—

'(3) He is in custody in violation of the Constitution or laws or treaties of the United States * * *.'

'A court, justice or judge entertaining an application for a writ of habeas corpus shall forthwith award the writ or issue an order directing the respondent to show cause why the writ should not be granted, unless it appears from the application that the applicant or person detained is not entitled thereto.

'The writ, or order to show cause shall be directed to the person having custody of the person detained. It shall be returned within three days unless for good cause additional time, not exceeding twenty days, is allowed.

'The person to whom the writ or order is directed shall make a return certifying the true cause of the detention.

'When the writ or order is returned a day shall be set for hearing, not more than five days after the return unless for good cause additional time is allowed.

'Unless the application for the writ and the return present only issues of law the person to whom the writ is directed shall be required to produce at the hearing the body of the person detained.

'The applicant or the person detained may, under oath, deny any of the facts set forth in the return or allege any other material facts.

'The return and all suggestions made against it may be amended by leave of court, before or after being filed.

'The court shall summarily hear and determine the facts, and dispose of the matter as law and justice require.'

Mr. Justice CLARK, dissenting.

I agree fully with and join the opinion of my Brother HARLAN. Beyond question the federal courts until today have had no power to release a prisoner in respondent Noia's predicament, there being no basis for such power in either the Constitution or the statute. But the Court today in releasing Noia makes an 'abrupt break' not only with the Constitution and the statute but also with its past decisions, disrupting the delicate balance of federalism so foremost in the minds of the Founding Fathers and so uniquely important in the field of law enforcement. The short of it is that Noia's incarceration rests entirely on an adequate and independent state ground—namely, that he knowingly failed to perfect any appeal from his conviction of murder. While it may be that the Court's 'decision today swings open no prison gates,' the Court must admit in all candor that it effectively swings closed the doors of justice in the face of the State, since it certainly cannot prove its case 20 years after the fact. In view of this unfortunate turn of events, it appears important that we canvass the consequences of today's action on state law enforcement.

First, there can be no question but that a rash of new applications from state prisoners will pour into the federal courts, and 98% of them will be frivolous, if history is any guide.1 This influx will necessarily have an adverse effect upon the disposition of meritorious applications, for, as my Brother Jackson said, they will 'be buried in a flood of worthless ones. He who must search a haystack for a needle is likely to end up with the attitude that the needle is not worth the search.' Brown v. Allen, 344 U.S. 443, 537, 73 S.Ct. 397, 425, 97 L.Ed. 469 (1953) (concurring opinion). In fact, the courts are already swamped with applications which cannot, because of sheer numbers, be given more than cursory attention.2

Second, the effective administration of criminal justice in state courts receives a staggering blow. Habeas corpus is in effect substituted for appeal, seriously disturbing the orderly disposition of state prosecutions and jeopardizing the finality of state convictions in disregard of the States' comprehensive procedural safeguards which, until today, have been respected by the federal courts. Essential to the administration of justice is the prompt enforcement of judicial decrees. After today state judgments will be relegated to a judicial limbo, subject to federal collateral attack—as here—a score of years later despite a defendant's willful failure to appeal.

The rights of the States to develop and enforce their own judicial procedures, consistent with the Fourteenth Amendment, have long been recognized as essential to the concept of a healthy federalism. Those rights are today attenuated if not obliterated in the name of a victory for the 'struggle for personal liberty.' But the Constitution comprehends another struggle of equal importance and places upon our shoulders the burden of maintaining it—the struggle for law and order. I regret that the Court does not often recognize that each defeat in that struggle chips away inexorably at the base of that very personal liberty which it seeks to protect. One is reminded of the exclamation of Pyrrhus: 'One more such victory * * *, and we are utterly undone.'

These considerations have been of great concern to the Judicial Conference of the United States, which has frequently sought to have Congress repair the judicial loopholes in federal habeas corpus for state prisoners.3 Likewise, the Conference of Chief Justices at its annual meeting has officially registered its dismay,4 as has the National Association of Attorneys General.5 Proposed legislation sponsored by one or more of these groups has passed in the House in three separate sessions, but inaction by the Senate caused each bill to die on the vine.6 Those proposals apparently were sparked by our decision in Brown v. Allen, supra,7 but the Court today goes far beyond that decision by negating its companion case, Daniels v. Allen, 344 U.S. 443, 482—487, 73 S.Ct. 397, 422, 97 L.Ed. 469 (1953). While I have heretofore opposed such legislation, I must now admit that it may be the only alternative in restoring the writ of habeas corpus to its proper place in the judicial system. That place is one of great importance—a remedy against illegal restraint—but it is not a substitute for or an alternative to appeal, nor is it a burial ground for valid state procedures.

Mr. Justice HARLAN, whom Mr. Justice CLARK and Mr. Justice STEWART join, dissenting.

This decision, both in its abrupt break with the past and in its consequences for the future, is one of the most disquieting that the Court has rendered in a long time.

Section 2241 of the Judicial Code, 28 U.S.C. § 2241, entitled 'Power to grant writ,' which is part of the federal habeas corpus statute, provides among other things:

'(c) The writ of habeas corpus shall not extend to a prisoner unless—

'(3) He is in custody in violation of the Constitution or laws or treaties of the United States.'

I dissent from the Court's opinion and judgment for the reason that the federal courts have no power, statutory or constitutional, to release the respondent Noia from state detention. This is because his custody by New York does not violate any federal right, since it is pursuant to a conviction whose validity rests upon an adequate and independent state ground which the federal courts are required to respect.

A full exposition of the matter is necessary, and I believe it will justify the statement that in what it does today the Court has turned its back on history and struck a heavy blow at the foundations of our federal system.

The history of federal habeas corpus jurisdiction, I believe, leaves no doubt that today's decision constitutes a square rejection of long-accepted principles governing the nature and scope of the Great Writ.1

Habeas corpus ad subjiciendum is today, as it has always been, a fundamental safeguard against unlawful custody. The importance of this prerogative writ, requiring the body of a person restrained of liberty to be brought before the court so that the lawfulness of the restraint may be determined, was recognized in the Constitution,2 and the first Judiciary Act gave the federal courts authority to issue the writ 'agreeable to the principles and usages of law.'3 Although the wording of earlier statutory provisions has been changed, the basic question before the court to which the writ is addressed has always been the same: in the language of the present statute, on the books since 1867, is the detention complained of 'in violation of the Constitution or laws or treaties of the United States'? Supra, p. 448.

Detention can occur in many contexts, and in each the scope of judicial inquiry will differ. Thus a child may be detained by a parent, an alien excluded by an immigration official, or a citizen arrested by a policeman and held without being brought to a magistrate. But the custody with which we are here concerned is that resulting from a judgment of criminal conviction and sentence by a court of law. And the question before us is the circumstances under which that custody may be held to be inconsistent with the commands of the Federal Constitution. What does history show?

1. Pre-1915 period.—The formative stage of the development of habeas corpus jurisdiction may be said to have ended in 1915, the year in which Frank v. Mangum, 237 U.S. 309, 35 S.Ct. 582, 59 L.Ed. 969, was decided. During this period the federal courts, on applications for habeas corpus complaining of detention pursuant to a judgment of conviction and sentence, purported to examine only the jurisdiction of the sentencing tribunal. In the leading case of Ex parte Watkins, 3 Pet. 193, 7 L.Ed. 650, the Court stated:

'An imprisonment under a judgment cannot be unlawful, unless that judgment be an absolute nullity; and it is not a nullity if the court has general jurisdiction of the subject, although it should be erroneous.' 3 Pet., at 203.

Many subsequent decisions, dealing with both state and federal prisoners, and involving both original applications to this Court for habeas corpus and review of lower court decisions, reaffirmed the limitation of the writ to consideration of the sentencing court's jurisdiction over the person of the defendant and the subject matter of the suit. E.g., Ex parte Parks, 93 U.S. 18, 23 L.Ed. 787; Andrews v. Swartz, 156 U.S. 272, 15 S.Ct. 389, 39 L.Ed. 422; In re Belt, 159 U.S. 95, 15 S.Ct. 987, 40 L.Ed. 88; In re Moran, 203 U.S. 96, 27 S.Ct. 25, 51 L.Ed. 105.

The concept of jurisdiction, however, was subjected to considerable strain during this period, and the strain was not lessened by the fact that until the latter part of the last century, federal criminal convictions were not generally reviewable by the Supreme Court.4 The expansion of the definition of jurisdiction occurred primarily in two classes of cases: (1) those in which the conviction was for violation of an allegedly unconstitutional statute, and (2) those in which the Court viewed the detention as based on some claimed illegality in the sentence imposed, as distinguished from the judgment of conviction. An example of the former is Ex parte Siebold, 100 U.S. 371, 25 L.Ed. 717 in which the Court considered on its merits the claim that the acts under which the indictments were found were unconstitutional, reasoning that '(a)n unconstitutional law is void, and is as no law,' and therefore 'if the laws are unconstitutional and void, the Circuit Court acquired no jurisdiction of the causes.' 100 U.S., at 376—377.5 An example of the latter is Ex parte Lange, 18 Wall. 163, 21 L.Ed. 872, in which this Court held that if a valid sentence had been carried out, and if the governing statute permitted only one sentence, the sentencing judge lacked jurisdiction to impose further punishment:

'(W)hen the prisoner * * * by reason of a valid judgment, had fully suffered one of the alternative punishments to which alone the law subjected him, the power of the court to punish further was gone.' 18 Wall., at 176.6

It was also during this period that Congress, in 1867, first made habeas corpus available by statute to prisoners held under state authority. Act of February 5, 1867, c. 28, § 1, 14 Stat. 385. In this 1867 Act the Court now seems to find justification for today's decision, relying on the statement of one of its proponents that the bill was 'coextensive with all the powers that can be conferred' on the courts and judges of the United States. Cong.Globe, 39th Cong., 1st Sess. 4151. But neither the statute itself, its legislative history, nor its subsequent interpretation lends any support to the view that habeas corpus jurisdiction since 1867 has been exercisable whether or not the state detention complained of rested on decision of a federal question.

First, there is nothing in the language of the Act—which spoke of the availability of the writ to prisoners 'restrained of * * * liberty in violation of the constitution * * *'—to suggest that there was any change in the nature of the writ as applied to one held pursuant to a judgment of conviction. The language was that typically employed in habeas corpus cases, and, as we have seen, it was not believed that a person so held was restrained in violation of law if the sentencing court had personal and subject matter jurisdiction. Rather, the change accomplished by the language of the Act related to the classes of prisoners (in particular, state as well as federal) for whom the writ would be available.

Second, what little legislative history there is does not suggest any change in the nature of the writ. The extremely brief debates indicated only a lack of understanding as to what the Act would accomplish, coupled with an effort by the proponents to make it clear that the purpose was to extend the availability of the writ to persons not then covered; there was no indication of any intent to alter its substantive scope.7 Thus, less than 20 years after enactment, a congressional committee could say of the 1867 Act that it was not 'contemplated by its framers or * * * properly * * * construed to authorize the overthrow of the final judgments of the State courts of general jurisdiction, by the inferior Federal judges * * *.'8

Third, cases decided under the Act during this period made it clear that the Court did not regard the Act as changing the character of the writ. In considering the lawfulness of the detention of state prisoners, the Court continued to confine itself to questions it regarded as 'jurisdictional.' See, e.g., In re Rahrer, 140 U.S. 545, 11 S.Ct. 865, 35 L.Ed. 572; Harkrader v. Wadley, 172 U.S. 148, 19 S.Ct. 119, 43 L.Ed. 399; Pettibone v. Nichols, 203 U.S. 192, 27 S.Ct. 111, 51 L.Ed. 148. And the Court repeatedly held that habeas corpus was not available to a state prisoner to consider errors, even constitutional errors, that did not go to the jurisdiction of the sentencing court. E.g., In re Wood, 140 U.S. 278, 11 S.Ct. 738, 35 L.Ed. 505; Andrews v. Swartz, 156 U.S. 272, 15 S.Ct. 389, 39 L.Ed. 422; Bergemann v. Backer, 157 U.S. 655, 15 S.Ct. 727, 39 L.Ed. 845.

At the same time, in dealing with applications by state prisoners the Court developed the doctrine of exhaustion of state remedies, a doctrine now embodied in 28 U.S.C. § 2254. In Ex parte Royall, 117 U.S. 241, 6 S.Ct. 734, 29 L.Ed. 868, the prisoner had brought federal habeas corpus seeking release from his detention pending a state prosecution, and alleging that the statute under which he was to be tried was void under the Contract Clause. The power of the federal court to act in this case, if the allegations could be established, was clear since under accepted principles the State would have lacked 'jurisdiction' to detain the prisoner. But the Court observed that the question of constitutionality would be open to the prisoner at his state trial and, absent any showing of urgency, considerations of comity counseled the exercise of discretion to withhold the writ at this early stage. In subsequent decisions, the Court continued to insist that state remedies be exhausted, even when the applicant alleged a lack of jurisdiction in state authorities which, if true, would have enabled the federal court to act on the application immediately. E.g., Ex parte Fonda, 117 U.S. 516, 6 S.Ct. 848, 29 L.Ed. 994; Cook v. Hart, 146 U.S. 183, 13 S.Ct. 40, 36 L.Ed. 934; New York v. Eno, 155 U.S. 89, 15 S.Ct. 30, 39 L.Ed. 80. As stated in Cook v. Hart, 146 U.S., at 195, 13 S.Ct., at 44, 'The party charged waives no defect of jurisdiction by submitting to a trial of his case upon the merits * * *. Should (his) rights be denied, his remedy in the federal court will remain unimpaired.' (Emphasis added.) The question whether the Constitution deprived the State of jurisdiction, in other words, would remain open under traditional doctrine, on collateral as well as direct attack.

There can be no doubt of the limited scope of habeas corpus during this formative period, and of the consistent efforts to confine the writ to questions of jurisdiction. But the cardinal point for present purposes is that in no case was it held, or even suggested, that habeas corpus would be available to consider any claims by a prisoner held pursuant to a state court judgment whose validity rested on an adequate nonfederal ground. Indeed, so long as the writ was confined to claims by state prisoners that the State was constitutionally precluded from exercising its jurisdiction in the particular case, it is difficult to conceive of a decision to detain in such cases resting on an adequate state ground. Even when the concept of jurisdiction was expanded, as in Ex parte Siebold, 100 U.S. 371, 25 L.Ed. 717, and other decisions, the matters open on habeas were still limited to those which were believed to have deprived the sentencing court of all competence to act, and which therefore could always be raised on collateral attack. It is for this reason that the Royall line of 'exhaustion' cases, relied on so heavily by the Court, has no real bearing on the problem before us. For those cases dealt only with the discretion of the court to take action which, if the allegations of lack of state jurisdiction were upheld, it would have had power to take either before or after state consideration. The issue here, on the other hand, is one of power, and wholly different considerations are involved.

In those few instances during this early period when the Court discussed questions it did not regard as jurisdictional, it occasionally went so far as to suggest that a constitutional claim could not be raised on habeas even if the state decision to detain rested on an inadequate state ground—that the only avenue of relief was direct review. Thus in Andrews v. Swartz, 156 U.S. 272, 15 S.Ct. 389, 39 L.Ed. 422, where the claim made on federal habeas was the systematic exclusion of Negroes from a state jury, the Court held it 'a sufficient answer to this contention that the state court had jurisdiction both of the offence charged and of the accused.' Id., at 276, 15 S.Ct., at 391. It continued:

'Even if it be assumed that the state court improperly denied to the accused * * * the right to show by proof that persons of his race were arbitrarily excluded * * * it would not follow that the court lost jurisdiction of the case within the meaning of the well-established rule that a prisoner under conviction and sentence of another court will not be discharged on habeas corpus unless the court that passed the sentence was so far without jurisdiction that its proceedings must be regarded as void.' Ibid.

2. 1915—1953 period.—The next stage of development may be described as beginning in 1915 with Frank v. Mangum, 237 U.S. 309, 35 S.Ct. 582, 59 L.Ed. 969, and ending in 1953 with Brown v. Allen, 344 U.S. 443, 73 S.Ct. 397, 97 L.Ed. 469. In Frank, the prisoner had claimed before the state courts that the proceedings in which he had been convicted for murder had been dominated by a mob, and the State Supreme Court, after consideration not only of the record but of extensive affidavits, had concluded that mob domination had not been established.9 Frank then sought federal habeas, and this Court affirmed the denial of relief. But in doing so the Court recognized that Frank's allegation of mob domination raised a constitutional question which he was entitled to have considered by a competent tribunal uncoerced by popular pressures. Such 'corrective process' had been afforded by the State Supreme Court, however, and since Frank had received 'notice and a hearing, or an opportunity to be heard' on his constitutional claims (237 U.S., at 326, 35 S.Ct., at 586, 59 L.Ed. 969), his detention was not in violation of federal law and habeas corpus would not lie.

It is clear that a new dimension was added to habeas corpus in this case, for in addition to questions previously thought of as 'jurisdictional,' the federal courts were now to consider whether the applicant had been given an adequate opportunity to raise his constitutional claims before the state courts. And if no such opportunity had been afforded in the state courts, the federal claim would be heard on its merits. The Court thus rejected the views expressed in Andrews v. Swartz, supra, 372 U.S., p. 455, 83 S.Ct., p. 857, by holding, in effect, that a constitutional claim could be heard on habeas if the State's refusal to give it proper consideration rested on an inadequate state ground. But habeas would not lie to reconsider constitutional questions that had been fairly determined. And a fortiori it would not lie to consider a question when the state court's refusal to do so rested on an adequate and independent state ground.

In this connection, it is important to note the section of the opinion relating to Frank's separate constitutional claim that his involuntary absence from the courtroom at the time the verdict was rendered invalidated the conviction. Frank had failed to raise this point in his motion for a new trial; the state court held that it had been 'waived'; and this Court decided that the state rule barring assertion of the point after failure to raise it in a motion for new trial was reasonable and did not violate due process.10 Clearly, the significance of the Court's ruling was that as to this constitutional claim, whatever its merits if the point had been properly preserved, there was an adequate nonfederal ground for the detention.

In no case prior to Brown v. Allen, I submit, was there any substantial modification of the concepts articulated in the Frank decision. In Moore v. Dempsey, 261 U.S. 86, 43 S.Ct. 265, 67 L.Ed. 543, this Court did require a hearing on federal habeas of a claim similar to that in Frank, of mob domination of the trial, even though the state appellate court had purported to pass on the claim, but only by refusing to 'assume that the trial was an empty ceremony.'11 The decision of this Court is sufficiently ambiguous that it seems to have meant all things to all men.12 But I suggest that the decision cannot be taken to have overruled Frank; it did not purport to do so, and indeed it was joined by two Justices who had joined in the Frank opinion. Rather, what the Court appears to have held was that the state appellate court's perfunctory treatment of the question of mob domination, amounting to nothing more than reliance on the presumptive validity of the trial, was not in fact acceptable corrective process and federal habeas would therefore lie to consider the merits of the claim. Until today, the Court has consistently so interpreted the opinion, as in Ex parte Hawk, 321 U.S. 114, 118, 64 S.Ct. 448, 450, 88 L.Ed. 572, where Moore was cited as an example of a case in which 'the remedy afforded by state law proves in practice unavailable or seriously inadequate.' See also Jennings v. Illinois, 342 U.S. 104, 111, 72 S.Ct. 123, 127, 96 L.Ed. 119.

Certainly, there is no basis in the Moore opinion, whatever it may fairly be taken to mean, for concluding that the Court required consideration on federal habeas of a question which the state court had had an adequate state ground for refusing to consider. The claim of mob domination was considered, although apparently inadequately, by the state court, and it was only on this premise that the claim was required t be heard on habeas.

Subsequent decisions involving state prisoners continued to indicate that the controlling question on federal habeas—apart from matters going to lack of state jurisdiction in light of federal law—was whether or not the State had afforded adequate opportunity to raise the federal claim. If not, the federal claim could be considered on its merits. See, e.g., Mooney v. Holohan, 294 U.S. 103, 55 S.Ct. 340, 79 L.Ed. 791; White v. Ragen, 324 U.S. 760, 65 S.Ct. 978, 89 L.Ed. 1348; Woods v. Nierstheimer, 328 U.S. 211, 66 S.Ct. 996, 90 L.Ed. 1177; cf. Jennings v. Illinois, 342 U.S. 104, 72 S.Ct. 123, 96 L.Ed. 119.13

A development paralleling that in Frank v. Mangum took place during this period with regard to federal prisoners. The writ remained unavailable to consider questions that were or could have been raised in the original proceedings, or on direct appeal, see Sunal v. Large, 332 U.S. 174, 67 S.Ct. 1588, 91 L.Ed. 1982, but it was employed to permit consideration of constitutional questions that could not otherwise have been adequately presented to the courts. E.g., Johnson v. Zerbst, 304 U.S. 458, 58 S.Ct. 1019, 82 S.Ct. 1461; Walker v. Johnston, 312 U.S. 275, 61 S.Ct. 574, 85 L.Ed. 830; Waley v. Johnston, 316 U.S. 101, 62 S.Ct. 964, 86 L.Ed. 1302. This limited scope of habeas corpus, and its statutory substitute 28 U.S.C. § 2255, in relation to federal prisoners may have survived Brown v. Allen and may still survive today. See, e.g., Franano v. United States, 8 Cir., 303 F.2d 470, cert. denied, 371 U.S. 865, 83 S.Ct. 125, 9 L.Ed.2d 102. Compare Jordan v. United States, 352 U.S. 904, 77 S.Ct. 151, 1 L.Ed.2d 114.

To recapitulate, then, prior to Brown v. Allen, habeas corpus would not lie for a prisoner who was in custody pursuant to a state judgment of conviction by a court of competent jurisdiction if he had been given an adequate opportunity to obtain full and fair consideration of his federal claim in the state courts. Clearly, under this approach, a detention was not in violation of federal law if the validity of the state conviction on which that detention was based rested on an adequate nonfederal ground.

3. Post-1953, Brown v. Allen, period.—In 1953, this Court rendered its landmark decisions in Brown v. Allen, 344 U.S. 443, 73 S.Ct. 397, 97 L.Ed. 469, and Daniels v. Allen, reported therewith, 344 U.S., at 482—487, 73 S.Ct., at 420—422.14 Both cases involved applications for federal habeas corpus by prisoners who were awaiting execution pursuant to state convictions. In both cases, the constitutional contentions made were that the trial court had erred in ruling confessions admissible and in overruling motions to quash the indictment on the basis of alleged discrimination in the selection of jurors.

In Brown, these contentions had been presented to the highest court of the State, on direct appeal from the conviction, and had been rejected by that court on the merits, State v. Brown, 233 N.C. 202, 63 S.E.2d 99, after which this Court had denied certiorari, 341 U.S. 943, 71 S.Ct. 997, 95 L.Ed. 1369. At this point, the Court held, Brown was entitled to full reconsideration of these constitutional claims, with a hearing if appropriate, in an application to a Federal District Court for habeas corpus.

It is manifest that this decision substantially expanded the scope of inquiry on an application for federal habeas corpus.15 Frank v. Mangum and Moore v. Dempsey had denied that the federal courts in habeas corpus sat to determine whether errors of law, even constitutional law, had been made in the original trial and appellate proceedings. Under the decision in Brown, if a petitioner could show that the validity of a state decision to detain rested on a determination of a constitutional claim, and if he alleged that determination to be erroneous, the federal court had the right and the duty to satisfy itself of the correctness of the state decision.

But what if the validity of the state decision to detain rested not on the determination of a federal claim but rather on an adequate nonfederal ground which would have barred direct review by this Court? That was the questionin Daniels. The attorney for the petitioners in that case had failed to mail the appeal papers on the last day for filing, and although he delivered them by hand the next day, the State Supreme Court refused to entertain the appeal, ruling that it had not been filed on time. This ruling, this Court held, barred federal habeas corpus consideration of the claims that the state appellate court had refused to consider. Language in Mr. Justice Reed's opinion for the Court appeared to support the result alternatively in terms of waiver,16 failure to exhaust state remedies,17 and the existence of an adequate state ground.18 But while the explanation may have been ambiguous, the result was clear: habeas corpus would not lie for a prisoner who was detained pursuant to a state judgment which, in the view of the majority in Daniels, rested on a reasonable application of the State's own procedural requirements. Moreover, the issue was plainly viewed as one of authority, not of discretion. 344 U.S., at 485, 73 S.Ct., at 421, 97 L.Ed. 469.

I do not pause to reconsider here the question whether the state ground in Daniels was an adequate one; persuasive arguments can be made that it was not. The important point for present purposes is that the approach in Daniels was wholly consistent with established principles in the field of habeas corpus jurisdiction. The problem, however, had been brought into sharper focus by the result in Brown. Once it is made clear that the questions open on federal habeas extend to such matters as the admissibility of confessions, or of other evidence, the possibility that inquiry may be precluded by the existence of a state ground adequate to support the judgment is substantially increased.

Issues similar to those in Daniels next came before the Court in Irvin v. Dowd, 359 U.S. 394, 79 S.Ct. 825, 3 L.Ed.2d 900. In that case, the state court's decision affirming Irvin's conviction for murder was ambiguous and it could have been interpreted to rest on a state ground even though Irvin's federal constitutional claims were considered. Irvin v. State, 236 Ind. 384, 139 N.E.2d 898; see also the dissenting opinion of this writer in Irvin v. Dowd, supra, 359 U.S. 412, 79 S.Ct. 835, 3 L.Ed.2d 900. This Court, in reversing a dismissal of an application for federal habeas corpus, concluded that the state court decision had rested on determination of Irvin's federal claims, and held that those claims could therefore be considered on federal habeas. The majority appeared to approach the problem as one of exhaustion,19 but the basic determination was that the state court judgment, pursuant to which Irvin was detained, did not rest on an application of the State's procedural rules.

This brings us to the present case. There can, I think, be no doubt that today's holding—that federal habeas will lie despite the existence of an adequate and independent nonfederal ground for the judgment pursuant to which the applicant is detained—is wholly unprecedented. Indeed, it constitutes a direct rejection of authority that is squarely to the contrary. That the result now reached is a novel one does not, of course, mean that it is necessarily incorrect or unwise. But a decision which finds virtually no support in more than a century of this Court's experience should certainly be subject to the most careful scrutiny.

The true significance of today's decision can perhaps best be laid bare in terms of a hypothetical case presenting questions of the powers of this Court on direct review, and of a Federal District Court on habeas corpus.

1. On direct review.—Assume that a man is indicted, and held for trial in a state court, by a grand jury from which members of his race have been systematically excluded. Assume further that the State requires any objection to the composition of the grand jury to be raised prior to the verdict, that no such objection is made, and that the defendant seeks to raise the point for the first time on appeal from his conviction. If the state appellate court refuses to consider the claim because it was raised too late, and if certiorari is sought and granted, the initial question before this Court will be whether there was an adequate state ground for the judgment below. If the petitioner was represented by counsel not shown to be incompetent, and if the necessary information to make the objection is not shown to have been unavailable at the time of trial, it is certain that the judgment of conviction will stand, despite the fact the indictment was obtained in violation of the petitioner's constitutional rights.20

What is the reason for the rule that an adequate and independent state ground of decision bars Supreme Court review of that decision—a rule which, of course, is as applicable to procedural as to substantive grounds? In Murdock v. Memphis, 20 Wall. 590, 632—636, 22 L.Ed. 429, it was concluded that under the governing statute (i) the Court did not have jurisdiction, on review of a state decision, to examine and decide 'questions not of a Federal character,' id., at 633, and (ii) an erroneous decision of a federal question by a state court could not warrant reversal if there were:

'any other matter or issue adjudged by the State court, which is sufficiently broad to maintain the judgment of that court, notwithstanding the error in deciding the issue raised by the Federal question.' Id., at 636.

But as the Court in Murdock so strongly implied, and as emphasized in subsequent decisions, the adequate state ground rule has roots far deeper than the statutes governing our jurisdiction, and rests on fundamentals that touch this Court's habeas corpus jurisdiction equally with its direct reviewing power. An examination of the alternatives that might conceivably be followed will, I submit, confirm that the rule is one of constitutional dimensions going to the heart of the division of judicial powers in a federal system.

One alternative to the present rule would be for the Court to review and decide any federal questions in the case, even if the determination of nonfederal questions were adequate to sustain the judgment below, and then to send the case back to the state court for further consideration. But it needs no extended analysis to demonstrate that such action would exceed this Court's powers under Article III. As stated in Herb v. Pitcairn, 324 U.S. 117, 126, 65 S.Ct. 459, 463, 89 L.Ed. 789:

'(O)ur power is to correct wrong judgments, not to revise opinions. We are not permitted to render an advisory opinion, and if the same judgment would be rendered by the state court after we corrected its views of federal laws, our review could amount to nothing more than an advisory opinion.'

Another alternative, which would avoid the problem of advisory opinions, would be to take the entire case and to review on the merits the state court's decision of every question in it. For example, in our hypothetical case the Court might consider on its merits the question whether the state court correctly ruled that under state law objections to the composition of the grand jury must be made prior to the verdict.

To a limited extent, of course, this procedural ruling of the state court raises federal as well as state questions. It is clear that a State may not preclude Supreme Court review of federal claims by discriminating against or evading the assertion of a federal right, and indeed that state procedural grounds for refusal to consider a federal claim must rest on a 'fair or substantial basis.'21 Occasionally this means that a state procedural rule which may properly preclude the raising of state claims in a state court cannot thwart review of federal claims in this Court.22 These principles are inherent in the concept that a state ground, to be of sufficient breadth to support the judgment, must be both 'adequate' and 'independent.'

But determination of the adequacy and independence of the state ground, I submit, marks the constitutional limit of our power in this sphere. The reason why this is so was perhaps most articulately expressed in a different but closely related context by Mr. Justice Field in his opinion in Baltimore & O.R. Co. v. Baugh, 149 U.S. 368, 401, 13 S.Ct. 914, 37 L.Ed. 772. He stated, in a passage quoted with approval by the Court in the historic decision in Erie R. Co. v. Tompkins, 304 U.S. 64, 78—79, 58 S.Ct. 817, 822, 82 L.Ed. 1188:

'(T)he constitution of the United States * * * recognizes and preserves the autonomy and independence of the states, independence in their legislative and independence in their judicial departments. Supervision over either the legislative or the judicial action of the states is in no case permissible except as to matters by the constitution specifically authorized or delegated to the United States. Any interference with either, except as thus permitted, is an invasion of the authority of the state, and, to that extent, a denial of its independence.'

For this Court to go beyond the adequacy of the state ground and to review and determine the correctness of that ground on its merits would, in our hypothetical cae, be to assume full control over a State's procedures for the administration of its own criminal justice. This is and must be beyond our power if the federal system is to exist in substance as well as form. The right of the State to regulate its own procedures governing the conduct of litigants in its courts, and its interest in supervision of those procedures, stand on the same constitutional plane as its right and interest in framing 'substantive' laws governing other aspects of the conduct of those within its borders.

There is still a third possible course this Court might follow if it were to reject the adequate state ground rule. The Act of 1867, which in § 1 extended the habeas corpus jurisdiction to state prisoners detained in violation of federal law, in § 2 gave the Supreme Court the authority, in cases coming from the state courts, to order execution directly without remanding the case. 14 Stat. 385, 386—387. That authority, which has been exercised at least once,23 remained unimpaired through the modifications of appellate and certiorari jurisdiction,24 and exists today.25 Acting pursuant to that authority in our hypothetical case, this Court might grant certiorari, 'ignore' the state ground of decision, decide the federal question and, instead of merely remanding the case, issue a writ requiring the petitioner's release from custody. By this simple device, the Court, it might be argued, would avoid problems of advisory opinions while at the same time refraining from consideration of questions of state law.

But apart from the unseemliness of such a dispoition, it is apparent that what the Court would actually be doing would be to decide the state law question sub silentio and to reverse the state court judgment on that question. For if the petitioner is detained pursuant to the judgment, and his detention is to be terminated, that must mean that the state ground is not adequate to support the only purpose for which the judgment was rendered. The judgment, in other words, becomes a nullity.

Moreover, the future effect of such a disposition is precisely the same as a reversal on the merits of the question of state law. If noncompliance with a state rule requiring a particular constitutional claim to be raised before verdict does not preclude consideration of the claim by this Court, then the rule is invalid in every significant sense, since no judgment based on its application can ever be effective.

In short, the constitutional infirmities of such a disposition by this Court are the same as those inherent in review of the state question on its merits. The vice, however, is greater because the Court would, in actuality, be invalidating a state rule without even purporting to consider it.

2. On habeas corpus.—The adequate state ground doctrine thus finds its source in basic constitutional principles, and the question before us is whether this is as true in a collateral attack in habeas corpus as on direct review. Assume, then, that after dismissal of the writ of certiorari in our hypothetical case, the prisoner seeks habeas corpus in a Federal District Court, again complaining of the composition of the grand jury that indicted him. Is that federal court constitutionally more free than the Supreme Court on direct review to 'ignore' the adequate state ground, proceed to the federal question, and order the prisoner's release?

The answer must be that it is not. Of course, as the majority states, a judgment is not a 'jurisdictional prerequisite' to a habeas corpus application, ante, p. 430 but that is wholly irrelevant. The point is that if the applicant is detained pursuant to a judgment, termination of the detention necessarily nullifies the judgment. The fact that a District Court on habeas has fewer choices than the Supreme Court, since it can only act on the body of the prisoner, does not alter the significance of the exercise of its power. In habeas as on direct review, ordering the prisoner's release invalidates the judgment of conviction and renders ineffective the state rule relied upon to sustain that judgment. Try as the majority does to turn habeas corpus into a roving commission of inquiry into every possible invasion of the applicant's civil rights that may ever have occurred, it cannot divorce the writ from a judgment of conviction if that judgment is the basis of the detention.

Thus in the present case if this Court had granted certiorari to review the State's denial of coram nobis, had considered the coerced confession claim, and had ordered Noia's release, the necessary effects of that disposition would have been (1) to set aside the conviction and (2) to invalidate application of the New York rule requiring the claim to be raised on direct appeal in order to be preserved. It is, I think, beyond dispute that the Court does exactly the same thing by affirming the decision below in this case. In doing so, the Court exceeds its constitutional power if in fact the state ground relied upon to sustain the judgment of conviction is an adequate one. See pp. 472—476, infra. The effect of of the approach adopted by the Court is, indeed, to do away with the adequate state ground rule entirely in every state case, involving a federal question, in which detention follows from a judgment.

The majority seems to recognize at least some of the consequences of its decision when it attempts to fill the void created by abolition of the adequate state ground rule in state criminal cases. But the substitute it has fashioned—that of 'conscious waiver' or 'deliberate bypassing' of state procedures is, as I shall next try to show, wholly unsatisfactory.

ATTEMPTED PALLIATIVES.

Apparently on the basis of a doctrine analogous to that of 'unclean hands,' the Court states that a federal judge, in his discretion, may deny relief on habeas corpus to one who has understandingly and knowingly refused to avail himself of state procedures. But such a test, if it is meant to constitute a limitation on interference with state administration of criminal justice, falls far short of the mark. In fact, as explained and applied in this case, it amounts to no limitation at all.

First, the Court explains that the test is one calling for the exercise of the district judge's discretion, that the judge may, in other words, grant relief even when a conscious waiver has been shown. Thus the Court does not merely tell the States that, if they wish to detain those whom they convict, they must revamp their entire systems of criminal procedures so that no forfeiture may be imposed in the absence of deliberate choice; the States are also warned that even a deliberate, explicit, intelligent choice not to assert a constitutional right may not preclude its assertion on federal habeas.

Second, the Court states (as it must if it is to adhere to its definition) that '(a) choice made by counsel not participated in by the petitioner does not automatically bar relief.' Ante, p. 439. It is true that there are cases in which the adequacy of the state ground necessarily turns on the question whether the defendant himself expressly and intelligently waived a constitutional right. Foremost among these are the cases involving right to counsel, for the Court has made it clear that this right cannot be foregone without deliberate choice by the defendant. See Johnson v. Zerbst, 304 U.S. 458, 58 S.Ct. 1019, 82 L.Ed. 1461; Carnley v. Cochran, 369 U.S. 506, 82 S.Ct. 884, 8 L.Ed.2d 70. But to carry this principle over in full force to cases in which a defendant is represented by counsel not shown to be incompetent is to undermine the entire representational system. We have manifested an ever-increasing awareness of the fundamental importance of representation by counsel, see Gideon v. Wainwright, 372 U.S. 335, 83 S.Ct. 792, and yet today the Court suggests that the State may no more have a rule of forfeiture for one who is competently represented than for one who is not. The effect on state procedural rules may be disastrous.

Third, when it comes to apply the 'waiver' test in this case, the Court then in effect reads its own creation out of existence. Recognizing that Noia himself decided not to appeal, and that he apparently made this choice after consultation with counsel, the Court states that his decision was nevertheless not a 'waiver.' Since a new trial might have resulted in a death sentence, Noia was, in the majority's view, confronted with a 'grisly choice,' and he quite properly declined to play 'Russian roulette' by appealing his conviction. Ante, pp. 439—440.

Does the Court mean by these colorful phrases that it would be unconstitutional for the State to impose a heavier sentence in a second trial for the same offense? Apparently not, since the majority assures us that there may be some cases in which a risk of a heavier sentence must be run. What distinguishes this case, we are told, is that the risk of the death sentence on a new trial was substantial in view of the trial judge's statement that Noia's past record and his involvement in the crime almost led the judge to disregard the jury's recommendation against a death sentence.

What the Court seems to be saying in this exercise in fine distinctions is that no waiver of a right can be effective if some adverse consequence might reasonably be expected to follow from exercise of that right. Under this approach, of course, there could never be a binding waiver, since only an incompetent would give up a right without any good reason, and an incompetent cannot make an intelligent waiver. The Court wholly ignores the question whether the choice made by the defendant is one that the State could constitutionally require.

Looked at from any angle, the concept of waiver which the Court has created must be found wanting. Of gravest importance, it carries this Court into a sphere in which it has no proper place in the context of the federal system. The true limitations on our constitutional power are those inherent in the rule requiring that a judgment resting on an adequate state ground must be respected.

ADEQUACY OF THE STATE GROUND HERE INVOLVED.

It is the adequacy, or fairness, of the state ground that should be the controlling question in this case.26 This controlling question the Court does not discuss.

New York asserts that a claim of the kind involved here must be raised on timely appeal if it is to be preserved, and contends that in permitting an appeal it has provided a reasonable opportunity for the claim to be made. The collateral post-conviction writ of coram nobis, the State has said, remains a remedy only for the calling up of facts unknown at the time of the judgment. See People v. Noia, decided sub nom. People v. Caminito, 3 N.Y.2d 596, 601, 170 N.Y.S.2d 799, 804, 148 N.E.2d 139, 143. In other words, the State claims that it may constitutionally detain a man pursuant to a judgment of conviction, regardless of any error that may have led to that conviction, if the relevant facts were reasonably available and an appeal was not taken.

Under the circumstances here—particularly the fact that Noia was represented by counsel whose competence is not challenged—is this a reasonable ground for barring collateral assertion of the federal claim? Certainly the State has a vital interest in requiring that appeals be taken on the basis of facts known at the time, since the first assertion of a claim many years later might otherwise require release long after it was feasible to hold a new trial. And although in Daniels v. Allen it might have been argued that the State's refusal to entertain an appeal actually received on time amounted to an evasion of the federal claim. no such argument can be made here, since no appeal was ever sought.

Moreover, we should be slow to reject—as an invalid barrier to the raising of a federal right—a state determination that one forum rather than another must be resorted to for the assertion of that right. A far more rigid restriction of federal forums was upheld in Yakus v. United States, 321 U.S. 414, 64 S.Ct. 660, 88 L.Ed. 834. In that case, the Court sustained a federal statute permitting an attack on the validity of an administrative price regulation to be made only on timely review of the administrative order, and precluding the defense of invalidity in a later criminal prosecution for violation of the regulation. What the Court there said bears repetition here:

'No procedural principle is more familiar to this Court than that a constitutional right may be forfeited in criminal as well as civil cases by the failure to make timely assertion of the right before a tribunal having jurisdiction to determine it.' 321 U.S., at 444, 64 S.Ct., at 677.

But is there some special circumstance here that operates to invalidate the nonfederal ground? Certainly it cannot be that the claim of a coerced confession is of such a nature that a State is constitutionally compelled to permit its assertion at any time even if it could have been, but was not, raised on appeal. Many federal decisions have held that a federal prisoner held pursuant to a federal conviction may not assert such a claim in collateral proceedings when it was not, but could have been, asserted on appeal. E.g., Davis v. United States, 7 Cir., 214 F.2d 594, cert. denied, 353 U.S. 960, 77 S.Ct. 870, 1 L.Ed.2d 911; Smith v. United States, 88 U.S.App.D.C. 80, 187 F.2d 192, cert. denied, 341 U.S. 927, 71 S.Ct. 792, 95 L.Ed. 1358; see Hodges v. United States, 108 U.S.App.D.C. 375, 282 F.2d 858, cert. dismissed, 368 U.S. 139, 82 S.Ct. 235, 7 L.Ed.2d 184.

Is it then a basis for invalidating the nonfederal ground that Noia's two codefendants are today free from custody on facts which Noia says are identical to those in his case? Does the nonfederal ground fall when the federal claim appears to have obvious merit? There may be some question whether the facts in Noia's case and those in Bonino's and Caminito's are identical,27 but assuming that they are, I think it evident that the nonfederal ground must still stand.

Again, there is highly relevant precedent dealing with federal prisoners. In Sunal v. Large, 332 U.S. 174, 67 S.Ct. 1588, 91 L.Ed. 1982, Sunal and Kulick had been prosecuted for violation of the Selective Service Act, and both had sought to raise a defense the court had refused to consider. Both were convicted and sentenced to imprisonment but took no appeal, quite evidently because such an appeal would have been to no avail under the existing state of the law. Subsequently, in another case, this Court held on comparable facts that the defense in question must be permitted. Estep v. United States, 327 U.S. 114, 66 S.Ct. 423, 90 L.Ed. 567. Sunal and Kulick then sought relief on habeas corpus, and this relief was denied. The opinion of the Court observed that there had been no barrier to the perfection of appeals by these prisoners and no facts which were not then known. That an appeal may have appeared futile at the time (indeed, far more futile than was the case here) was held not a sufficient basis for collateral relief. The present case, I submit, would be less troublesome than Sunal even had it involved a federal prisoner.

Surely, the state ground is not rendered inadequate because on a new trial for the same offense, Noia might have received the death sentence. The State is well within constitutional limits in permitting such a sentence to be imposed. Of particular relevance here is the decision in Larson v. United States, 5 Cir., 275 F.2d 673. Two criminal defendants had been tried and sentenced to imprisonment by a federal court. One defendant, Juelich, had moved for a continuance or a change of venue, on the ground of community prejudice, and his motion had been denied. Both defendants were convicted; Juelich appealed from his conviction; and the Court of Appeals reversed, Juelich v. United States, 5 Cir., 214 F.2d 950, holding that the constitutional requirement of a fair trial had been violated by the refusal to grant a change of venue or a continuance. Larson, the other defendant, had chosen not to appeal, apparently because he feared that the death sentence might be imposed in a new trial, but after his codefendant's success, he sought collateral relief under § 2255. That relief was denied by the District Court, and the Court of Appeals affirmed, stating:

'We do not say * * * that in every instance, before resort can be had to Section 2255 there must be an appeal. We say only that, in the circumstances of this case, Larson, taking a calculated risk, made a free choice not to jeopardize his life, and he is bound by that decision. * * * Whatever errors there were in his trial were known to Larson and to his counsel—for the same errors formed the basis for Juelich's appeal. Manifest Justice to an accused person requires only that he have an opportunity to correct errors that may have led to an unfair trial. The orderly administration of justice requires that even a criminal case some day come to an end.' 275 F.2d at 679—680.

This Court denied certiorari. 363 U.S. 849, 80 S.Ct. 1627, 4 L.Ed.2d 1732.

Decisions such as Sunal and Larson are reasoned expressions by the federal judiciary of its views on the fair and proper administration of federal criminal justice. We cannot turn around and tell the State of New York that it is constitutionally prohibited from being governed by the same considerations.

I recognize that Noia's predicament may well be thought one that strongly calls for correction. But the proper course to that end lies with the New York Governor's powers of executive clemency, not with the federal courts.28 Since Noia is detained pursuant to a state judgment whose validity rests on an adequate and independent state ground, the judgment below should be reversed.