Due to the admission in evidence of the confessions I would reverse the convictions for the reasons set forth in my opinion of November 20, 1967, dissenting from the affirmance then by a division of the court of the conviction of rape. With regard to the convictions of first degree felony murder and manslaughter, affirmed now by the court en banc, I would reverse on the additional ground that it was prejudicial error for the trial court to send the case to the jury in such a way as to permit convictions of two degrees of homicide for the same death, contrary to the decision of this court in Naples v. United States, 120 U.S.App.D.C. 123, 131-132, 344 F.2d 508, 516-517, referred to as Naples II.
In Naples II the court said:
We think the court erred in ref using to instruct the jury that it could not find appellant guilty of both first degree and second degree murder. Since the jury found appellant guilty on both counts, the error must be deemed prejudicial.
*1234Sections 22-2401-02, District of Columbia Code (1961), define first degree murder.16 Section 22-2403 provides: “Whoever with malice aforethought, except as provided in sections 22-2401, 22-2402, kills another, is guilty of murder in the second degree.” 17 (Emphasis supplied.) This makes it clear that a single offense cannot be both first and second degree murder. In Goodall v. United States, we said an instruction on both first and second degree murder.
“is necessary only when from the evidence as a whole the jury might reasonably find the defendant guilty of either first or second degree murder, and therefore must decide which degree had been committed.” [686 U.S.App.D.C. 148, 151, 180 F.2d 397, 400, 17 A.L.R.2d 1070 (1950), cert. denied, 339 U.S. 987, 70 S.Ct. 1009, 94 L.Ed. 1389 (1950), emphasis supplied.]
“At common law there were no degrees of murder. All homicide with malice aforethought, whether express or implied, was murder. And all murder was punished by death.” Clark & Marshall, CRIMES § 10.09 at 608 (6th ed.1958). By statute murder was defined in various degrees in order to treat criminal behavior on a more individual basis. “First degree murder” now stands as the criminal act deserving the most severe sanction. To permit conviction of both first and second degree murder for the same offense would be contrary to the intent of the nineteenth century legal reforms which displaced undifferentiated common law murder. [Footnotes omitted.]
It will be seen from this language that the court did not rule that in rendering a verdict of guilty the jury was required also to render formal verdicts of not guilty of other degrees included in the indictment. It is true, as the majority now points out, the Naples II court was silent in face of the contrary interpretation of its opinion by the dissenting judge; but this is not conclusive. There is no need to go beyond the central position that it was error to permit a guilty verdict of both first and' second degree murder. To go beyond this would give rise to the troublesome double jeopardy contentions referred to in the court’s present opinion. Whether those problems would carry the day against Naples II as construed by its dissenting judge we need not ponder; for the problems can be avoided by requiring a case such as Fuller’s to be submitted to the jury under instructions which treat second degree as a lesser included offense of first degree murder. I understand from Part II of the opinion that this course is approved; but the court does not reverse for failure to follow it since counsel for Fuller did not request the instruction. The court affirms the two convictions on the theory that the omission of such an instruction is not plain error affecting a substantial right to be noted under Rule 52(b), Fed.R.Crim.P., a matter I cover when I reach the question of prejudice.
Laying aside for the moment the question of prejudice, I consider first in more detail whether there was error, though as will appear, error and prejudice sometimes become commingled.
The Naples II opinion discusses the error on the basis of the structure of our Code, considered on the background of the common law. The opinion also relies upon Goodall v. United States, 86 U.S.App.D.C. 148, 151, 180 F.2d 397, 400. More recent support of Naples II could be found in Austin v. United States, 127 U.S.App.D.C. 180, 382 F.2d 129, footnote 27, except that Judge Leventhal, the author of the Austin opinion, now considers such support to have been a mistake. In any event Naples II must rest upon its own merits.
Although the court holds that if request had been made the trial judge should have instructed the jury that it could find Fuller guilty of only one homicide the court upholds the two convictions on the theory that the crimes of first degree felony murder and second degree or manslaughter are not incon*1235sistent. I disagree. I think the error lies not only in not following the course dictated by the lesser included offense doctrine but also because the crimes of first degree felony murder and second degree murder or manslaughter are inconsistent. In Green v. United States, 355 U.S. 184, 78 S.Ct. 221, 2 L.Ed.2d 199, it is pointed out that felony murder and second degree murder are “distinct and different” crimes. While distinct and different crimes may be the subject of a single trial a person cannot be convicted of two such crimes for the same offense.
As the court properly points out the elements of one crime may be so inconsistent with those of another, as larceny and receiving stolen goods, as to preclude convictions of both crimes upon the same evidence. In another situation two crimes will share certain elements but each will have elements not shared by the other, such as an assault with a deadly weapon and an assault with intent to kill. They share the element of assault but one requires in addition the use of a deadly weapon and the other the intent to kill. In Ingram v. United States, 122 U.S.App.D.C. 334, 353 F.2d 872, we did not disapprove convictions of two such crimes at one trial, but required non-consecutive sentences. In another situation the elements of one crime may include all those of another but also additional elements. Rape, for example, includes all elements of an assault, but it also includes additional elements of its own.
Felony murder has been held to include all of the elements of second degree murder and of manslaughter. See, e. g., Jackson v. United States, 114 U.S.App. D.C. 181, 183, 313 F.2d 572, 574. However, the relationship between degrees of homicide differs from the relationship, for example, between assault and rape. Whether a crime is premeditated murder or manslaughter the end result is the same — a death. In the case of rape, however, the result is quite different from an assault. Where both assault and rape occur it would not be factually inconsistent for the jury to return a separate conviction for each, though for some other reason the law might not permit this. Where, however, there is only one result — for example, a death — there is inconsistency in attributing two distinct crimes to the several physical actions and states of mind culminating in a single homicidal result.1
It seems to me a jury is erroneously deprived of its historic function of determining the degree of seriousness of a homicide when it is permitted to decide that it is one degree of gravity, manslaughter, and at the same time that it is another degree of gravity, first degree. An unlawful killing is hot divisible into two such verdicts. For the jury the degrees in the end are mutually exclusive. Otherwise there is not only inconsistency but also uncertainty as to which crime the defendant committed.
Naples II does not require the charge to the jury to be in alternative terms, as the court imputes to that decision. The heart of the decision is in not permitting the two verdicts to be returned, whether or not the case goes to the jury in alternative terms or under the lesser included offense doctrine, though in all substance both methods of submission do post alternatives for the jury.
The court concludes that injustice is avoided in the two verdicts by requiring noncumulative sentences. This bypasses the issue. The injustice — the prejudicial character of the error— occurs before the sentencing is reached. It attaches to the manner in which the verdicts themselves are reached. Noncumulative sentences are never a cure for error. 'They are required not to avoid error in verdicts but to carry out the intention of Congress, considered with a policy of leniency, to prevent two penalties for two crimes which, although different, factually are closely related. *1236See Ingram v. United States, supra. Impermissibility of multiple sentences for separate but related crimes does not render permissible multiple convictions for the same crime.
Nor has the Hirabayashi line of cases application to our situation which involves the vulnerability of both verdicts of guilty. Those cases hold that it is unnecessary where the sentences are concurrent to review for possible error convictions on each count of a multiple count indictment when the conviction on one count is found to be free of error, thus supporting the judgment. Here the question is whether either conviction is free of error; for both are in question, including of course the one carrying the greater sentence.
Part II of the court’s opinion accepts the position that had counsel asked for an instruction under the lesser included offense doctrine the two verdicts of guilty would not have been permitted. The court would distinguish that situation from Naples II by stating that such an instruction does not put the matter to the jury in alternatives, since when the jury finds the greater offense it does not consider the lesser. Yet it is because the court considers the lesser as included in the greater that in Part I it approves verdicts of both the lesser and the greater. Paradoxically, in Part I the court concludes that the two homicide verdicts against Fuller are consistent on the ground that the elements of one are included within the elements of the other, yet in Part II the court agrees that upon appropriate request the trial judge should have instructed the jury to return only one homicide verdict on the very ground that one of the crimes was included within the other. Since Fuller’s case could have been submitted in this manner it should have been, avoiding the error in permitting the two guilty verdicts for the same homicide.
Naples II does not depend upon consideration of the possible verdicts in the alternative; but if it did it would not thereby bar the lesser included offense instruction for under such instruction the verdicts are also considered in the alternative on thg entire evidence. Should the greater offense be agreed upon no verdict is rendered for a lesser; and if the lesser is agreed upon no verdict as to a greater is returned. The rendition of only one verdict does not mean that the jury in its deliberations'does not consider simultaneously both degrees.
The court denies to Fuller the benefit of the lesser included offense procedure, under which he could have been convicted of but one degree of homicide, because his counsel omitted to alert the court to its availability. The court seeks to justify this by suggesting several advantages in permitting two guilty verdicts for the same homicide. None of these was sought by Fuller, and none overrides the prejudice to him which has occurred, more fully discussed in a moment. Indeed the court now recognizes that submitting two guilty verdicts as permissible might be prejudicial if the jury had an “unrestricted function in determining whether its verdict should reflect a conclusion of guilt as to the lesser or greater offense.” Yet this is truly the function the jury does have. For though it is true as the court states that the jury in the proper performance of its duty may not return a verdict on the lesser offense unless it has failed to agree on the greater, in its deliberations the jury considers all the evidence bearing on the greater and the lesser offenses and decides which has been proved to its satisfaction. It then renders a single verdict accordingly. In so deliberating the jury acts rationally in the performance of its duty under the law, not irrationally or inconsistently with its duty as the court suggests.
Considering now more fully the issue of prejudice, this is an elusive concept. To decide the issue of prejudice requires a court to reach a conclusion about the effect of error upon the decisional processes of a jury. The governing standard is not whether prejudice was “actual” or even “probable.” Where there is error, especially in a capital *1237case, the question is “whether there is reasonable possibility that [the defendant was prejudiced.” United States ex rel. Hetenyi v. Wilkins, 348 F.2d 844, 864 (2d Cir.), cert. denied, Mancusi v. Hetenyi, 383 U.S. 913, 86 S.Ct. 896, 15 L.Ed.2d 667 (Emphasis in original.) The Supreme Court has adopted the “reasonable possibility” formulation to determine prejudice to an accused in other contexts. See, e. g., Fahy v. State of Connecticut, 375 U.S. 85, 86-87, 84 S.Ct. 229, 11 L.Ed.2d 171; Stoner v. State of California, 376 U.S. 483, 490, n. 8, 84 S.Ct. 889, 11 L.Ed.2d 856. But the problem need not be confined to an idea so expressed. Here we simply do not know what'the verdict would have been had Naples II been followed. We do know that the jury, if properly instructed, might have found the defendant (1) guilty only of second degree murder, (2) guilty only of manslaughter as a lesser included offense, or (3) not guilty of either of the two murder counts. Each of these was submitted to the jury, but the jury was not confined to rendering a guilty verdict as to but one. It is significant that under the second count the jury returned a verdict of manslaughter rather than second degree murder. If, as is not unlikely, the jury then concluded that having also found the defendant guilty of rape it was required to find him guilty of first degree felony murder, the prejudice is clear. If confined to one guilty homicide verdict they might have found the rape independent of the homicide, and limited the homicide to manslaughter or second degree. Or the homicide might have been thought to have been unintentional, for which reason the jury might have rendered a second degree verdict, as the trial judge authorized. The jury might have been unwilling on the evidence of the confessions alone to reconstruct the crime beyond a reasonable doubt as requiring a verdict of first degree felony murder,
For me to say the result would have been the same as it was, that is, first degree, with manslaughter added as superfluous as it were, would be for me to impose a verdict of my own. To do this would be for me as an appellate judge to substitute my speculation for what should be the jury’s determination. I could as well impose verdicts of manslaughter and rape, with no first degree. Though such a result would be unlikely at the hands of a jury on the evidence against Fuller, I cannot say it would not be reasonably possible.
The prejudice becomes even clearer when it is remembered that the trial court quite properly instructed the jury that a verdict of second degree was permissible without the necessity of finding any other guilty verdict of homicide.
The court recognizes the incongruity of Fuller having two sentences for the same homicide. Though counsel made no point of it, this relatively insignificant matter is deemed to require correction while the failure to request the lesser included offense instruction, which might have resulted in no verdict of first degree, is not. The court vacates the less serious sentence. If either sentence is to be vacated it should be the more serious one.
The imposition of the two sentences evidences the error in the two verdicts, but the prejudice lies elsewhere, it is in the two verdicts where only one should have been rendered, leaving uncertainty as to which would have been rendered if the jury had been confined to one. To paraphrase Dozier v. United States, 127 U.S.App.D.C. 266, 382 F.2d 482, 483, although the jury under erroneous instructions decided that all elements of first degree were proved, it does not follow that under proper instructions they would have done so rather than resting their verdict upon manslaughter and, separately, rape.
I respectfully dissent.
. I lay aside as not presently presented any question involved in verdicts for first degree felony and at the same time first degree murder of deliberate and premeditated malice,