I concur with points I and IV of the majority opinion but dissent from points II and III.
However beguiling the rationale of the majority opinion, its melancholy result is to penalize these defendants for a successful appeal. Had defendants graciously accepted the trial court’s erroneous judgment of guilt of violating Penal Code section 415, or had they appealed and lost, they would have remained simple misdemeanants, enjoying probation upon payment of a modest fine. Because they appealed and won they will be shocked to learn that through legal legerdemain they must stand trial a second time on serious felony charges and may ultimately be sentenced to state prison. Theirs is indeed a Cadmean victory.
In People v. Henderson (1963) 60 Cal.2d 482, 497 [35 Cal.Rptr. 77, 386 P.2d 677], Chief Justice Traynor articulated the principle which should guide us here: “A defendant’s right of appeal from an erroneous judgment is unreasonably impaired when he is required to risk his life [or his liberty] to invoke that right. Since the state has no interest in preserving erroneous judgments, it has no interest in foreclosing appeals therefrom by imposing unreasonable conditions on the right to appeal.” (Accord, People v. Hood (1969) 1 Cal.3d 444, 459 [82 Cal.Rptr. 618, 462 P.2d *768370]; People v. Ali (1967) 66 Cal.2d 277, 281 [57 Cal.Rptr. 348, 424 P.2d 932]; People v. Chavez (1970) 4 Cal.App.3d 832, 839 [84 Cal.Rptr. 783].) No distinction may be drawn, in denying exaction of a price for a successful appeal, between reversals on constitutional and nonconstitutional grounds. A defendant’s exercise of a right of appeal in all cases must be free and unfettered; “vindictiveness” against a defendant for having successfully attacked his first conviction is impermissible. (North Carolina v. Pearce (1969) 395 U.S. 711, 724-725 [23 L.Ed.2d 656, 668-669, 89 S.Ct. 2072].)
The majority opinion purports to devise some neat procedural distinctions, but in eifect it conflicts with substantive principles settled in the leading cases in this field, particularly Gomez v. Superior Court (1958) 50 Cal.2d 640 [328 P.2d 976], In re Hess (1955) 45 Cal.2d 171 [288 P.2d 5], and Green v. United States (1957) 355 U.S. 184 [2 L.Ed.2d 199, 78 S.Ct. 221, 61 A.L.R.2d 1119]. To permit a second felony trial to be conducted in the present circumstances violates the prohibition against double jeopardy. (U.S. Const., 5th Amend.; Cal. Const., art. I, § 13.)
The first principle invoked by the majority under section II of the opinion (ante, pp. 759-760) is well accepted: that a defendant who succeeds in having his conviction set aside impliedly waives any objection to being retired on the same charge of which he was convicted. (Forman v. United States (1960) 361 U.S. 416, 425 [4 L.Ed.2d 412, 419, 80 S.Ct. 481].) The key phrase, which the majority overlook, is that the waiver applies to the same charge of which he was convicted. The charge, and the only charge, of which these defendants stand convicted -today, and which they have now succeeded in having set aside, is violation of Penal Code section 415, a misdemeanor. Any implied waiver of objection to retrial therefore applies only to Penal Code section 415.
The waiver issue was considered in Green v. United States (1957) supra, 355 U.S. 184, 191-192 [2 L.Ed.2d 199, 206-207], and disposed of in this manner for the court by Justice Black: “Nevertheless the Government contends that Green ‘waived’ his constitutional defense of former jeopardy to a second prosecution on the first degree murder charge by making a successful appeal of his improper conviction of second degree murder. We cannot accept this paradoxical contention. ‘Waiver’ is a vague term used for a great variety of purposes, good and bad, in the law. In any normal sense, however, it connotes some kind of voluntary knowing relinquishment of a right. Cf. Johnson v. Zerbst, 304 U.S. 458. When a man has been convicted of second degree murder and given a long term of imprisonment it is wholly fictional to say that he ‘chooses’ to forego his constitu*769tional defense of former jeopardy on a charge of murder in the first degree in order to secure a reversal of an erroneous conviction of the lesser offense. In short, he has no meaningful choice. And as Mr. Justice Holmes observed, with regard to this same matter in Kepner v. United States, 195 U.S. 100, at 135: ‘Usually no such waiver is expressed or thought of. Moreover, it cannot be imagined that the law would deny to a prisoner the correction of a fatal error, unless he should waive other rights so important as to be saved by an express clause in the Constitution of the United States.’ ”
Next, the majority recite, but in fact discard, a second fundamental principle: that where a trier of fact, jury or court, finds a defendant guilty of a lesser offense, there is an implied acquittal of the greater offense; and after such acquittal the double jeopardy prohibition of both the federal and state Constitutions bars retrial on the greater offense. But the majority, as I understand their opinion, reject that doctrine here by honing a razor-sharp distinction between a lesser but nonincluded offense, which will permit retrial on the greater charge, and a lesser included offense, which acts as a bar to that retrial. I can find no authoritative support for such an esoteric concept which, in the final analysis, would penalize the defendant for the judge’s error.
In seeking to distinguish the treatment of lesser but nonincluded offenses, the majority repeatedly assert (ante, p, 762) that in ruling on a motion for new trial the function of the trial court “is to grant a new trial if it finds the evidence insufficient,” and the court “has no authority to acquit the defendant expressly, impliedly or inadvertently.” Obviously, however, the court does have the authority to acquit the defendant “impliedly”—the majority so -recognize earlier in the opinion (ante, p. 761), where they concede that “If in lieu of granting a new trial the court decides to modify the verdict to a lesser included offense, and this modified verdict ultimately ripens into a final judgment of conviction, the conviction bars further prosecution of either the offense charged or the lesser offense. This follows because of the rule that a conviction of a lesser offense is a bar to prosecution of another offense of which the lesser is a part.” The majority imply that this rule is applicable only to cases in which the trial court reduces the conviction to a lesser included offense. No reason is given, other than the majority’s ipse dixit, to explain why it would not be equally applicable when the court reduces the conviction to a lesser but nonincluded offense.
Perhaps in an attempt to provide such a reason, the majority state (ante, p. 762) that “When a trial judge reduces a conviction to a lesser included offense, it is reasonably inferable that the judge had a belief that the evi*770dence proved the lesser offense and not the greater.” Why is this not equally “reasonably inferable” when the judge reduces the conviction to a lesser but nonincluded offense? The majority then imply that in the latter case the trial court desired to exercise “leniency.” Why, in turn, could this not be also true in cases in which the court reduces the conviction to a lesser included offense? In short, it is not the defendants but the majority who are being “illogical” in this case.
Unfortunately this result appears to be a retrogression to pre-1958 days when a comparable unnatural categorization was accepted. Prior to the decision in Gomez v. Superior Court (1958) supra, 50 Cal.2d 640, there was an anomalous distinction between a conviction of a lesser included offense and a conviction of a lesser degree of a crime which is divided into degrees. A conviction of a lesser included offense was deemed an acquittal of the greater offense (In re Hess (1955) supra, 45 Cal.2d 171, 176) but a conviction of a lesser degree would, after reversal, allow subsequent conviction of the greater degree of the same crime (People v. Keefer (1884) 65 Cal. 232, 235 [3 P. 818]). In Gomez this court found the distinction to be logically indefensible and held that the double jeopardy clause precluded convicting a defendant of a higher degree of a crime after reversal of a lower-degree conviction.
The fact-finding process involved here is relatively simple. It was described long ago in People v. Gilmore (1854) 4 Cal. 376, 377: ‘“The jury, in such a case, in contemplation of law, render two verdicts; the one acquitting him of the higher crime; the other convicting him of the inferior. They must first determine his guilt or innocence upon the charge made by the indictment, before proceeding to inquire whether he is guilty of an inferior crime.’ ” (See also People v. Gordon (1893) 99 Cal. 227, 230 [33 P. 901].) An acquittal is equally valid whether express or implied. (Price v. Georgia (1970) 398 U.S. 323, 329 [26 L.Ed.2d 300, 305, 90 S.Ct. 1757].) And it is clear that the Gilmore process applies to motions for new trial considered by the court. (In re Ferguson (1965) 233 Cal.App.2d 79, 82 [43 Cal.Rptr. 325].)
Thus before the trial court, acting as the fact-finder here, could consider defendants’ guilt of an inferior crime;—whether a properly included or an inappropriate lesser offense—it necessarily first determined the defendant’s innocence or guilt of the charge contained in the pleading. Only by an implied finding of not guilty of that charge could the judge reach the second plateau, i.e., a determination of the inferior charge of which he believed the defendants guilty. His subsequent error in stage two cannot in any way *771affect his prior necessarily implied, and legally controlling, findings and conclusion on the initial charge.
No matter how the matter is analyzed or rationalized, and regardless of whether the trial judge scrupulously weighed the evidence or acted out of compassion, it is apparent to me beyond the slightest doubt that the judge intended to—and did—find the defendants not guilty of a felony, and guilty of a misdemeanor. That he chose the wrong misdemeanor in no way alters the validity and finality of the underlying acquittal of the felony charge. Failure to understand this point is the primary cause of the majority’s error.
I find it impossible to read Penal Code section 1181, subdivision 6, as do the majority, to conclude that modifying the judgment is not an acquittal of the greater offense. The code language is unequivocal on that subject: “if the evidence shows the defendant to be not guilty of the degree of the crime of which he was convicted ... the court may modify the verdict, . . (Italics added.) Once again, the process is elementary: the court must first determine the defendants not guilty of the charge of which the jury convicted them, and then it proceeds to modify by finding guilt of a lesser offense. Error in the latter step cannot affect the prior not guilty determination. As stated in People v. Baca (1966) 247 Cal.App.2d 487, 497 [55 Cal.Rptr. 681]: “subdivision 6 of section 1181 . . . applies only where the evidence shows defendant to be not guilty of the offense of which he was found guilty.” (Italics added.)
The majority, through “a tangled variety of legal fictions” (Van Alstyne, In Gideon’s Wake (1965) 74 Yale L.J. 606, 626), blithely sidestep the impact of Justice Traynor’s impeccably reasoned opinion in In re Hess (1955) supra, 45 Cal.2d 171. There the defendant was charged with the felony of forcible rape; he was found guilty of contributing to the delinquency of a minor, a misdemeanor. After analysis of the nature of the crimes, this court held contributing was not a necessarily included offense, and “that the court therefore acted in excess of its jurisdiction in entering a judgment of conviction of that offense against him.” (Id. at p. 175.)
Here the majority make the identical determination: since Penal Code section 415 is not a necessarily included offense within section 452, subdivision (b), the court acted in excess of its jurisdiction (ante, p. 765). Nevertheless, the majority detour around the uncomplicated manner in which Hess disposed of the problem: the rule is that a finding of guilt of a lesser offense implies a finding of innocence of the greater offense, “and that any error affecting the express verdict of guilty does not affect the *772conclusiveness of the implied verdict of acquittal.” The trier of fact must be presumed to have rejected guilt of the greater offense before considering conviction of what it erroneously believed to be a lesser included offense. (45 Cal.2d at p. 176.)
The United States Supreme Court has made it abundantly clear that the Fifth Amendment prohibition is not against being twice punished but against being twice put in jeopardy. (United States v. Ball (1896) 163 U.S. 662, 669 [41 L.Ed. 300, 302, 16 S.Ct. 1192].) Since the trial court found these defendants guilty of a misdemeanor, rather than the felony for which they were originally tried, upon reversal they may not be put in jeopardy a second time on a felony charge. The misdemeanor operates “as a ceiling whether or not the original error affected the fundamental fairness of the trial” (Van Alstyne, op. cit. supra, 74 Yale L.J. at p. 636). “Such a result flows inescapably from the Constitution’s emphasis on a risk of conviction, and the Constitution’s explication in prior decisions . . . .” (Price v. Georgia (1970) supra, 398 U.S. 323, 327 [26 L.Ed.2d 300, 304].)
Tobriner, J., concurred.
Appellants’ petition for a rehearing was denied August 29, 1973. Tobriner; J., and Mosk, J., were of the opinion that the petition should be granted.