Dissenting Opinion by
Mr. Justice Musmanno:On September 26, 1955, the then Majority of this Court handed down a decision (Commonwealth v. Thomas, 382 Pa. 639)* based on one of the most illogical doctrines ever announced in a criminal case. The decision was not a juridical conclusion but, in effect, a legislative fiat; it was not a decision founded on authority and precedent, but an arbitrary edict which declared an act to be murder which was never murder before.
The facts in the Thomas case were very simple. Two men, Henry Thomas and Henry Jackson, entered a grocery store in Philadelphia for the purpose of committing a robbery. While Jackson covered the grocer with his revolver, Thomas gathered up the money from the cash register and then both fled. The grocer, as the bandits ran, seized his own revolver from behind the counter and gave pursuit to Jackson who returned the fire which the grocer poured at him. In this gun battle Jackson was killed. Thomas was later arrested and charged with murder under the so-called felony-murder doctrine.
*555It was agreed by all parties concerned that Thomas had no hand in the killing of his confederate Jackson and that he was not actually present at the moment that the grocer’s bullet cut Jackson down. The killing was the result of an attempt, which succeeded, on the part of the bandits’ victim to prevent the escape of a robber who had just robbed him. Such a killing is recognized by law not only as justifiable but, in some instances, recommendable in the interests of law and order.
On these facts, however, this Court declared on September 26, 1956, that Thomas could be convicted of first degree murder. In all the time which has elapsed since that declaration was made, I have not heard one lawyer speak of it approvingly.* No State in the United States and no jurisdiction in the legal world, so far as I have been able to determine, has supported so Draconian a concept.
The Thomas case came to our Court in the following circumstances. After the Commonwealth had presented its evidence at Thomas’s trial on the charge of murder, his attorney demurred, arguing that, admitting the presented facts to be true, the Commonwealth had still not presented a prima facie case of murder, according to established law. The lower Court agreed, and granted the demurrer. The Commonwealth appealed. This Court heard the appeal, reversed the judgment of the lower Court,** and ordered the case back for re-trial on the murder charge.
*556But when the mandate of this Court arrived in the Court of Oyer and Terminer of Philadelphia County, where Thomas had originally been tried, an extraordinary thing occurred. The District Attorney declined to try Thomas for murder. In fact he did the precise opposite. He entered a nolle prosequi of the murder indictment. He also nol-prossed the voluntary and involuntary manslaughter indictments. He was willing to take, and did take from Thomas a plea of guilty to robbery which, of course, is what Thomas’s crime was, all the time — not murder.
The nolle prosequi was filed, of course, with the approval of the Court of Oyer and Terminer, the same Court which had originally decided, through a brilliant and courageous, opinion written by the distinguished Judge Guerin, that Thomas could not be convicted of murder.
That is the amazing and inglorious history of the Thomas case. The rule laid down in the decision entered in that case endured a little over two years and it is today being overruled by Commonwealth v. Red-line, 391 Pa. 486. In an admirable, scholarly and gripping opinion, Chief Justice Jones, following what he had said in his dissenting opinion in the Thomas case, now authoritatively shatters the bad logic and defective reasoning in the Thomas case, and, speaking for the majority of this Court, repudiates the Thomas decision.
Which is all to the good.
But I am sorry to note that much of the defective lumber which went into the scaffolding of the Thomas case is now being used to support the Bolish decision. The facts in the Bolish case, which involve an application of the same felony-murder rule under discussion, are as uncomplicated as they were in the Thomas cáse. *557A young man in Scranton by tbe name of Bobert Flynn was killed when a container of gasoline exploded as he was attempting to ignite an arsonious fire. Investigation revealed that a Daniel Bolish had planned with young Flynn to set fire to a building known as the Torti house in order to obtain certain financial gains not necessary to relate here. Bolish was not in or near the Torti house when the explosion occurred. The Majority Opinion, therefore, is in error when it says: “There were, however, both direct and circumstantial evidence showing the presence of Bolish in the Torti house at the time of the explosion.” No evidence whatsoever remotely substantiates such an assertion.
Flynn’s death was entirely accidental. No one killed him, no one tried to kill him, no one wanted to kill him.
After making the utterly unsupportable statement that there were “both direct and circumstantial evidence showing the presence of Bolish in the Torti house at the time of the explosion,” the Majority Opinion, two paragraphs later, says: “As we have heretofore pointed out, defendant was present at the time of the commission of the crime of arson and the resulting explosion.” I repeat that there is nothing in the record to substantiate this statement. What has happened to the noble Institute of Authentic Fact? Has it crumbled and may facts now be fashioned out of thin air? Does the Majority operate on the theory that a repetition of an erroneous statement cleanses it of the impurity of error? In this respect I must say that two times two does not become five on the second or even two-hundredth reiteration of the arithmetical incongruity.
Bolish was tried and convicted under the felony-murder doctriné, and the jury returned a verdict of *558first degree murder, fixing the penalty at death. The defendant appealed to this Court which awarded a new trial because of trial errors, but held, however, that the facts as unfolded by the Commonwealth and as briefly summarized here, justified a conviction of mur der in the first degree. Commonwealth v. Bolish, 381 Pa. 500. The case was then retried and the jury again convicted the defendant of first degree murder, now fixing, however, the penalty at life imprisonment. The defendant has appealed once more, urging that the facts do not bring this case within the scope of felony-murder. The Majority of this Court has affirmed the conviction, and I dissent.
I dissent because I believe the Majority has misread the statute under which the defendant was tried and convicted. Section 701 of the Penal Code of 1939 says: “All murder which shall be perpetrated by means of poison, or by lying in wait, or by any other kind of willful, deliberate and premeditated killing, or which shall be committed in the perpetration of, or attempting to perpetrate any arson, rape, robbery, burglary or kidnapping, shall be murder in the first degree. All other kinds of murder shall be murder in the second degree.”* It should be as plain as the sun in the sky that the subject of this entire statute is murder. It says “All murder which shall be perpetrated . . .”. What is murder? In its original decision in this case, the Majority defined murder as “the unlawful killing of another,” a proper and correct definition. But where is the killing of another in the case at bar? Flynn came to his death, not at the hands of another but as the result of his own negligence.
As already stated, Section 701 makes mivrder the basis of the criminal charge — “All murder which shall *559be . . . committed,” etc. But Flynn did not die as the result of murder. His death was an accident. No one attempted to shoot him, no one tried to poison him, no one sought to explode him. He placed a jar of gasoline on an electric fire, it detonated, and he died —the result of a mischance, a misfortune, a purely fortuitous circumstance. Where is the murder?
In his ably comprehensive Opinion in the RedUne case, Chief Justice Jones says: “The mere coincidence of homicide and felony is not enough to satisfy the requirements of the felony-murder doctrine. ‘It is necessary ... to show that the conduct causing death was done in furtherance of the design to commit the felony. Death must be a consequence of the felony . . . and not merely coincidence.’ ”
The death of Flynn was not in furtherance of the design to commit the felony of arson. In fact, it was just the opposite. It became a barrier to the successful accomplishment of the arson. Of course, an entirely different situation arises where a bandit purposely or accidentally shoots an innocent bystander as he and his confederates are making their escape from the scene of an accomplished or attempted robbery. There he is doing something in furtherance of the felonious design. He is intimidating the whole locale, he is menacing the forces of law and order so as to facilitate the escape of the whole criminal gang. Thus, his bandit-comrades, who stand to benefit from his murderous act, equally with himself, are responsible for the death which occurs. But there is nothing like that in the case at bar. In the isolation of the non-tenanted Torti house, and without contact with anyone, Flynn suffered a mishap which resulted in his own death. Where is the furtherance of the design to commit felony? His death not only did not further the design of arson but it actually brought the arson plan to a *560standstill and complete failure. And it would be sheer gro tes query to assume that Bolish and Flynn had entered into a design or plan whereby Flynn was to kill himself.
The Bedline decision of today has torn down part of the roof of the felony-murder doctrine which had been extended to cover situations which could never have been visualized by the authors of the doctrine. But the present Bolish decision has brought back the carpenters and the canopy has again been extended, even though not so far as the eaves of the Thomas decision. Even so, the pillars beneath are not prepared to support such an extension. In time the roof will sag and buckle, and eventually, unless brought back to its original size or held up with further legislative columns, it will collapse completely. Any rule of law which becomes brutal in its application eventually becomes suspect in its entirety.
I have said that no State in the Union has interpreted the doctrine as Pennsylvania has. Attempts have been made to cover situations like the Bolish case but they have failed. In the cage of People v. Ferlin, 265 Pac. 230, the prosecution attempted to hold a co-arsonist guilty of murder because his accomplice was killed at the time of the fire. The Supreme Court of California rejected the theory: “It cannot be said from the record in the instant case that defendant and deceased had a common design that deceased should accidentally kill himself. Such an event was not in furtherance of the conspiracy but entirely opposed to it.” As I have already indicated, the same is true of the Bolish case: the death of Flynn was not in furtherance of the conspiracy, but in opposition to it.
In the New York case of People v. LaBarbera, 287 N. Y. S. 257, the prosecution sought to convict the two defendants of murder because they had engaged one *561Gagliano to burn down a building but who, in that attempt, lost his own life. The New York Penal Code defines homicide as “the killing of one human being by act, procurement or omission of another.” The New York Court held: “The killing of Gagliano occurred by his own act, and therefore was not by the act of another, and so his death does not come within the definition of criminal homicide. The liability of the conspirators for a death occurring during a felony may be described as the liability of one for the acts of his agent ... If Gagliano had killed another . . ., then the proof before the grand jury would have been warranted in bringing the accusation of murder in the first degree against the defendants. However, in view of the fact that he killed himself, Tie did not commit homicide as defined in the statute of this state, and if he did not commit homicide, certainly his associates in crime cannot be held for homicide
While, of course, Pennsylvania is not bound by decisions of other States, the inherent respect we bear toward our sister sovereignties in the federation of the law, usually invites us to a study and consideration of decisions on facts similar to those before us for review. The problem in the New York and California cases cited is so identical to the one before us that I am surprised that the Majority has not attempted to distinguish those cases or to renounce them as unworthy precedent. I do not understand why the Majority should insist that Pennsylvania commit itself to a weed-infested road of unreasoning punishment when our sister States have all avoided it. No State has made arson a capital offense. Why should Pennsylvania do so, especially through the medium of judicial interpretation?
I would say, however, that the most serious fault in the Majority’s decision lies in its complete miscon*562ception of the Act under which Bolish has been tried and convicted, and his sentence here affirmed. We know that murder has never been defined by statute in Pennsylvania, but we know also that, according to common law, murder was the unlawful killing of another with malice aforethought, express or implied. In his Opinion in the Redline case, Chief Justice Jones has pointed out that: “The so-called murder statute of this state is but a categorizing of common law murder into two degrees — a dichotomy still unrecognized in England whence the definition of murder as known and applied in Pennsylvania was derived. In fact, the General Assembly of this State was the first legislative body in America to divide the crime of murder into degrees (see Section 2 of the Act of 1794, supra) ... Indeed, in enacting the murder degree statute of 1794, supra, the Pennsylvania legislature constricted the penalty for felony-murder by imposing capital punishment only for such felony-murders as occurred in the perpetration of arson, rape, robbery or burglai'y.”
It will thus be seen that the purpose of the felony-murder statute Avas to sIioav Avhat types of murder committed in the perpetration of felonies became first degree murder. Anyone who participated in the perpetration of arson which resulted in murder Avas guilty of first degree murder. In developing his excellent historical development of the felony-murder doctrine, Chief Justice Jones said further: “All felony-murder in Pennsylvania other than such as is committed in the perpetration of one of the common laAV felonies specified in our degree statute is murder of the second degree by virtue of the express terms of that statute respecting ‘All other kinds of murder.’ It is plain enough that neither the Act of 1794, supra, nor any of its subsequent re-enactments made all homicides occurring in the perpetration of felonies murder of the first degree.” (Italics in original.)
*563It must be as plain as a pikestaff from this irrefutable exposition that if there is any murder when arson occurs, all those who participate in the arsonious act are guilty of first degree murder, not second degree murder. But it must be murder to begin with, not manslaughter, not an accidental killing, not a justifiable killing, but murder. Where is the murder in the Bolish case? Bolish, of course, did plan the arson, he participated in the perpetration of the arson, and he can be convicted of arson, but he cannot be convicted of murder, because there was no murder. An accident occurred, and a confederate died, but there was no murder.
I do not know why the Majority and I do not find the same printed words in Section 701. The words which come before my vision are: “All murder which shall be . . . committed in the perpetration of, or attempting to perpetrate any arson . . . shall be murder in the first degree.” However, when the Majority reads the same section the print which seems to eome before its eyes is: “All deaths which shall occur in the perpetration of, or attempting to perpetrate any arson shall be murder in the first degree.”
In the face of that phenomenon one might hazard the conclusion that either the Majority is using the wrong glasses or some one is shifting the law books. But if we are reading the same book and our respective visions are good enough to detect the words. as they actually appear, it cannot be that Bolish is guilty, of murder in the first degree when there was no murder. There was a killing, a death, but no murder.
Law is supposed to be the embodiment of reason, and it should not be used to play verbal tricks. The whole object of the felony-murder doctrine is to place in their proper, gravé settings such major. crimes as. robbery,' burglary,, rape,' arson, -kidnapping.-., .If. the *564man who robs a bank kills the cashier, he has committed murder, and it is proper and just that the lookout sitting in the car outside should also be held for murder even though he did not pull a trigger or even have a gun. He was part of the enterprise which resulted in a proved murder. But it does not follow that if one of the bandits should die from a heart attack the others should be held for murder. In such a situation there would be no murder at the outset. It does not follow that if an escaping bandit falls over a cliff and is killed, his co-bandits should be held for murder. In such a state of affairs there would be no murder in the first place. It does not follow that if a person intending arson drops the dynamite and kills himself, without injuring any one else, the fellow-arsonists should be held for murder. Under those circumstances there would be no murder to begin with.
In the original Bolish decision the Majority Opinion expounded at length on the need to protect society from criminals. No honest person doubts that need. Whether it be a murder or a traffic law violation, the violator must answer to the law, but he can only be made to answer to the law as written. This Court cannot, and should not, simply because it has unlimited power, punish a traffic law violator with 10-years imprisonment if the offended law only provides for a fine and a suspension of motor license. Neither can it punish an arsonist with death or life imprisonment if the statute only provides for a 20-year maximum imprisonment. This Court cannot, by flying the flag of a punitive expedition, amend the criminal code. The criminal code can be altered only by the Legislature. When this Court infringes upon the prerogatives of the Legislature, which it certainly does, by creating a crime not heretofore recognized, it violates the Constitution which has spécifically assigned legislative duties to the Legislature and judicial functions to the judiciary.
*565It was because the unwritten criminal code of English Common Law was so barbaric — some 200 offenses were punishable by hanging — that the transplanted English colonists in America spelled out the crimes which would be punishable with death. Hence arose the statute of 1794 and the re-enactments which have followed. The Legislature has jealously guarded the confines of murder; it has built a wall around the deeds which are to be classified as murder. Acts committed outside those walls cannot be considered murder. But this Court is doing the very thing which the Legislature has prohibited. It is, by forced interpretation, adding to the list of offenses which are punishable by death or life imprisonment. This it may not do under the Constitution, even though by its own interpretation of the Constitution it may not be called to account for this usurpation of power.
Bolish committed arson. As bad as that crime is, Is it is not murder, which, of course, with treason, is the most heinous of all crimes. When the people of Pennsylvania through their duly elected representatives believe that arsonists should be electrocuted or incarcerated for life, a duly enacted statute to that effect will be placed on the statute books. But until that happens, this Court has no right to fit the penalty for murder to that for arson or any other crime unless so designated by the sovereign power of the L Commonwealth. —
By increasing the punishment for an offense after it was committed, this Court is doing what even the Legislature may not do. It is enacting an ex post facto law. And so, for the third time, I dissent against this Court’s unprecedented, unwarranted, unjustified, unnecessary, uncalled-for, untenable creation of ex post facto law, the infliction of punishment not called for by the criminal code, and the laying down of ex*566pansion rules which can only tend to bring about disrespect for, and damage confidence in, the certainty, the impartiality, and the fairness of the law.
The following Justices dissented from the Majority Opinion; JóNes <now ¿hief Justice) r' ChidseY; and- MusuanNó. '
In the summer of 1956 while I was taking some courses at Oxford University, the distinguished and venerable scholar A. L. Ooodhart, Queen’s Counsel and Master of the University College, expressed to me great surprise over the doctrine announced in the Thomas case.
The Court was divided 4 to 3.
Italics throughout, mine..