This view did not at first command universal acceptance, largely on the practical ground that medical science did not then permit a clear proof of causal connection, but it was adopted in early Victorian times by the First and Second Criminal Law Commissioners (B.P.P., 1839, 19, 266 and B.P.P., 1846, 24, 127) and never substantially doubted since. In R. v. West (1848) 2 Car. & K., a case to which I must briefly return, the rule was extended to a situation such as the present where the assault caused the death, not through injury to the child, but by causing the child to be born prematurely. In R. v. Senior (1832) 1 Mood. C.C. 346 the principle was applied to manslaughter, where death resulted from gross negligence by a midwife before the child had been fully born. Since the principle is not disputed I will not cite the numerous references to it by institutional writers during the past three centuries.

B. Two Arguments for the Crown

1. The First Argument: the Foetus identified with the Mother.

    The decision of the Court of Appeal founded on the proposition that the foetus is part of the mother, so that an intention to cause really serious bodily injury to the mother is equivalent to the same intent directed towards the foetus. This intent could be added to the actus reus, constituted (as I understand it) by the creation of such a change in the environment of the foetus through the injury to the mother that the baby would be born at a time when, as events proved, it would not survive. I must dissent from this proposition for I believe it to be wholly unfounded in fact. Obviously, nobody would assert that once the mother had been delivered of S, the baby and her mother were in any sense "the same". Not only were they physically separate, but they were each unique human beings, though no doubt with many features of resemblance. The reason for the uniqueness of S was that the development of her own special characteristics had been enabled and bounded by the collection of genes handed down not only by M but also by the natural father. This collection was different from the genes which had enabled and bounded the development of M, for these had been handed down by her own mother and natural father. S and her mother were closely related but, even apart from differing environmental influences, they were not, had not been, and in the future never would be "the same." There was, of course, an intimate bond between the foetus and the mother, created by the total dependence of the foetus on the protective physical environment furnished by the mother, and on the supply by the mother through the physical linkage between them of the nutriments, oxygen and other substances essential to foetal life and development. The emotional bond between the mother and her unborn child was also of a very special kind. But the relationship was one of bond, not of identity. The mother and the foetus were two distinct organisms living symbiotically, not a single organism with two aspects. The mother's leg was part of the mother; the foetus was not.

    The only other ground for identifying the foetus with the mother that I can envisage is a chain of reasoning on the following lines. All the case law shows that the child does not attain a sufficient human personality to be the subject of a crime of violence, and in particular of a crime of murder, until it enjoys an existence separate from its mother; hence, whilst it is in the womb it does not have a human personality; hence it must share a human personality with its mother. This seems to me an entire non sequitur, for it omits the possibility that the foetus does not (for the purposes of the law of homicide and violent crime) have any relevant type of personality but is an organism sui generis lacking at this stage the entire range of characteristics both of the mother to which it is physically linked and of the complete human being which it will later become. The argument involves one fiction too far, and I would reject it.

2. The Second Argument: The foetus as a separate organism

    I would, therefore, reject the reasoning which assumes that since (in the eyes of English law) the foetus does not have the attributes which make it a "person" it must be an adjunct of the mother. Eschewing all religious and political debate I would say that the foetus is neither. It is a unique organism. To apply to such an organism the principles of a law evolved in relation to autonomous beings is bound to mislead. I prefer, so far as binding authority permits, to start afresh, and to do so by reference to the second of the arguments advanced by the Attorney-General. This builds on the rules stated above by the following stages. If D struck X intending to cause her serious harm, and the blow, in fact, caused her death, that would be murder (Rule 1). If she had been nursing a baby Y which was accidentally struck by the blow and consequently died, that would also be murder (Rules 1 and 2). So, also, if an evil-doer had intended to cause harm but not death to X by giving her a poisoned substance and the substance was, in fact, passed on by X to the baby, which consumed it and died as a result (Rules 1, 2, and 3). Again, it would have been murder if the foetus had been injured in utero and had succumbed to the wound after being born alive (Rules 1, 2, 4 and 5). It is only a short step to make a new rule, adding together the malice towards the mother, the contemporaneous starting of a train of events, and the coming to fruition of those events in the death of the baby after being born alive.

    My Lords, the attractions of this argument are plain, not least its simplicity. But for my part I find it too dependent on the piling up of old fictions, and too little on the reasons why the law takes its present shape. To look for these reasons is not, to use an expression sometimes met, "legal archaeology" for its own sake. Except in those cases, of which the present is not one, where the rationale of the existing law is plain on its face, the common law must build for the future with materials from the past. One cannot see where a principle should go without an idea of where it has come from.

    Unfortunately these materials are in short supply in the English law of homicide, as Lord Diplock pointed out in Hyam v. D.P.P. [1975] A.C. 55, 89. If one looks at the sources from which the institutional writers drew their law (Bracton, Fleta and a handful of cases from the Year Books and late medieval manuscripts) they offer an interesting picture of contemporary morals and modes of thought, but are too terse to convey any developed course of reasoning. It is not until the writers beginning with Lord Coke that exposition recognisable to a modern reader begins to emerge, and even in the Institutes we find opinions founded, or ostensibly founded, on the ancient fragments. The later writers were of great authority, but on the present topic were mainly content to reproduce what their predecessors had to say, and to call up the same ancient materials as before. Even the more extended discourse of Blackstone carried the subject little further. Moreover, in contrast with the civil law, when the reporting of criminal cases first began on any scale the absence of a systematic process of criminal appeal meant that the reporters were largely confined to dicta and directions to juries gleaned from cases on Assize or at the Guildhall. Sir James Stephen exclaimed that:

 "It is a matter for regret that decisions, necessarily given with little consideration, and under great pressure should be reported at all." (Digest of the Criminal Law (1877) xv.)

He was particularly severe on Cox's Criminal Cases from which many of the decisions are drawn. It was not until much later that the practical and theoretical workings of the criminal law were exposed to view; and by then many of the principles had become so firmly established as to escape critical analysis.

    The materials for scrutinising the existing rules to see how they could be built upon to answer a problem like the present are therefore quite meagre. But an attempt must at least be made. I begin with the first rule. Three lines of thought, not identical but intertwined, seem to have gone to make it up. First there was a presumption that when one person killed another this was culpable homicide unless, in the words of East: Pleas of the Crown, (1803) Vol. I, cap. V, section 12), the defendant proved that there were circumstances of accident, necessity or infirmity; or in a later formulation, unless justified, excused or alleviated (Blackstone: Commentaries on the Laws of England, 17th ed. (1830), pp. 200-201); and, later still (according to Stephen's Digest of Criminal Law (1877), art. 230) unless there existed excuse, justification or extenuation. This rule survived, perhaps only in terms of an evidential burden of proof, until surprisingly late. It was not until Woolmington v. D.P.P. [1935] A.C. 462 that it was finally expunged. For so long as it was current there was no chance of saying that an intent to cause really serious injury was insufficient to found an indictment for murder, for that high degree of wrongful intent took away the possibility of establishing a recognised excuse for the death which actually ensued.

    The second theme was the obverse of the first. The wrongful intent was a demonstration of a general wickedness of mind which expressed itself in whatever specific intent was necessary to give the act in question a criminal character. In short, the wicked intent showed that the defendant was a bad man with a "depraved inclination to mischief" (Russell on Crime 4th ed., (1865), p. 740, citing Hale, History of the Pleas of the Crown (1736) Vol. I, p. 475 and East, Pleas of the Crown (1803), Cap. V, section 18); and this inclination, if resulting in death, should not go unpunished.

    Finally, there was a concept of risk. The doer of a wicked act took the chance that the consequences would be greater than he could foresee. His narrower subjective intent was not an answer to his responsibility for the unintended wider consequences. As Hawkins, Pleas of the Crown, 7th ed. (1795) Vol. I, section 51, put the matter in relation to deaths happening in the course of tumultuous assemblies, "They must at their peril abide the event of their actions who wilfully engage in such bold disturbances of the public peace." Again, at a later date Russell on Crime 4th ed. p. 742 explains Coke's pronouncement that a beating in anger which causes death is murder by the fact that "what he did was malum in se and he must be answerable for all its consequences."

    It was, I believe, the coalescence of these three concepts which founded a doctrine more extreme than the grievous harm rule, under which an unintended death resulting from the commission or attempted commission of an offence of any kind or degree was treated as murder. This had its most notorious expression in the pronouncement of Coke (Inst., Pt. III, ch. 8, p. 56):

 "If the act be unlawful it is murder. As if A meaning to steal a deer in the park of B, shooteth at the deer, and by the glance of the arrow killeth a boy that is hidden in a bush: this is murder, for that the act was unlawful, although A had no intent to hurt the boy, nor knew not of him . . . [so also if one] had shot at a cock or hen, or any tame fowl of another man's, and the arrow by mischance had killed a man, this had been murder, for the act was unlawful."

    As Sir James Stephen would much later show (History of the Criminal Law (1883) vol. 3, pp. 57-58), this doctrine was never securely founded on the authorities, but it left its mark for more than two centuries. The controversy over it in the institutional writers need not be resumed here. It is sufficient to say that it came to be perceived as morally odious, and subsided without any close analysis into the concept of felony/murder, not perhaps very different from a form of "general malice", where the evil intent required was of a high degree. In this form it survived into modern times, although narrowed still further by confinement to crimes of violence D.P.P. v. Beard [1920] A.C. 479. It is indeed still part of the law in some common law jurisdictions. Finally, it was abolished in England by section 1 of the Homicide Act 1957.

    My Lords, since the original concepts are no longer available to explain why an intent to cause grievous bodily harm will found a conviction for murder the reason must be sought elsewhere: for reason, in regard to such a grave crime, there must surely be. The obvious recourse is to ascribe this doctrine to the last vestiges of the murder/felony rule, and to see in it a strong example of that rule, for unlike the more extravagant early manifestations it offers at least some resemblance in nature and degree between the intended act and its unintended consequences. It would follow, therefore, that when the murder/felony rule was expressly abolished by section 1 of the Homicide Act 1957 the only surviving justification for the "grievous harm" rule fell away, with nothing left. This proposition was indeed advanced soon after the 1957 Act in R. v. Vickers [1957] 2 Q.B. 664, where it was dismissed out-of-hand. The same concept was developed in Hyam v. D.P.P. [1975] A.C. 55, where after close analysis it was adopted by Lord Diplock, and in a concurring speech by Lord Kilbrandon. The majority in the House did not agree. The question was raised again in Rex v. Cunningham [1982] A.C. 566, and this time a decisive answer was given. The "grievous harm" rule had survived the abolition of the murder/felony principle. The speeches show that it did so because a solid and long-lasting line of authority had decreed that this was the law, and the House saw no need to change a rule which answered practical needs.

    My Lords, in a system based on binding precedent there could be no ground for doubting a long course of existing law, and certainly none which could now permit this House even to contemplate such a fundamental change as to abolish the grievous harm rule: and counsel rightly hinted at no such idea. But when asked to strike out into new territory it is, I think, right to recognise that the grievous harm rule is an outcropping of old law from which the surrounding strata of rationalisations have weathered away. It survives but exemplifies no principle which can be applied to a new situation.

    I turn to the second rule, of "transferred malice." For present purposes this is more important and more difficult. Again, one must look at its origins to see whether they provide a theme which can be applied today. Three of them are familiar. Taking Lord Coke's example of the glancing arrow we have seen how one explanation of the poacher's responsibility founded on the notion of risk. The person who committed a crime took the chance that the outcome would be worse than he expected. Amongst many sources one can find the idea in Russell on Crime, 4th ed. (1845), p. 739:

 "If an action, unlawful in itself, be done deliberately, and with the intention of mischief or great bodily harm to particular individuals, or of mischief indiscriminately, fall where it may, and death ensue or beside the original intention of the party, it will be murder."

In a later edition (1855: p. 759) this was exemplified by cases of particular malice to one individual falling by mistake upon another. In support are cited R. v. Saunders (1573) 2 Plowd 473 (a poisoned apple intended for the mother but given to the child) and Gore 9 Co. Rep. 81 (medicine poisoned by the wife to kill her husband and consumed by the apothecary to prove his innocence); also 1 Hawkins P.C., c. 31, 545 and 1 Hale 436. As already suggested, this doctrine does survive in some small degree today, but as the foundation of a modern doctrine of transferred malice broad enough to encompass the present case it seems to me quite unsupportable.

    Secondly, there is the reversed burden of proof whereby the causing of death is prima facie murder, unless it falls within one of the extenuating categories recognised by the institutional writers. Again, this concept is long out-of-date. Nobody could seriously think of using it to make new law.

    Third, there was the idea of "general malice", of an evil disposition existing in the general and manifesting itself in the particular, uniting the aim of the offender and the result which his deeds actually produced. According to this theory, there was no need to "transfer" the wrongful intent from the intended to the actual victim; for since the offender was (in the words of Blackstone, supra, pp. 198-200) "an enemy to all mankind in general", the actual victim was the direct object of the offender's enmity. Plainly, this will no longer do, for the last vestiges of the idea disappeared with the abolition of the murder/felony doctrine.

    What explanation is left: for explanation there must be, since the "transferred malice" concept is agreed on both sides to be sound law today? The sources in more recent centuries are few. Of the two most frequently cited the earlier is R. v. Pembliton (1874) L.R. 2 C.C.R. 119. In the course of a fight the defendant threw a stone at others which missed and broke a window. He was indicted for that he "unlawfully and maliciously did commit damage, injury, and spoil upon a window . . . " The jury found that he did not intend to break the window. On a case stated to the Court for Crown Cases Reserved it was argued for the prosecution that "directly it is proved that he threw the stone . . . without just cause, the offence is established." The ancient origins of this argument need no elaboration, and indeed the report of the argument as it developed showed that it was based on a conception of general malice. The interventions in argument are instructive. After the prosecutor had relied on the fact that the prisoner was actuated by malice, Blackburn J. responded: "But only of a particular kind, and not against the person injured." Later, in reply to a reliance on a passage from Hale the same judge said:

 "Lord Coke, 3 Inst., p. 56, puts the case of a man stealing deer in a park, shooting at the deer, and by the glance of the arrow killing a boy that is hidden in a bush, and calls this murder; but can anyone say that ruling would be adopted now?"

This most learned of judges continued:

 "I should have told the jury that if the prisoner knew there were windows behind, and that the probable consequence of his act would be to break one of them, that would be evidence for them of malice.".

The conviction was quashed. It is sufficient to quote briefly from the judgment of Blackburn J.: