Kaplan, J.
On October 3,1973, the defendant Dr. Kenneth Edelin, Chief Resident in obstetrics and gynecology at Boston City Hospital, performed an abortion by hysterotomy on a seventeen year old, unmarried woman, she and her mother having requested an abortion and consented to the operation. For his conduct in connection with the operation, Dr. Edelin was indicted for manslaughter, and convicted after trial. He appeals from the judgment of conviction and from the trial judge’s refusal of a new trial.1
[499]*499All six Justices who heard the appeal, holding that there was error in the proceedings at trial, vote to reverse the conviction.- Five Justices also vote to direct the entry of a judgment of acquittal; the Chief Justice, dissenting in part in a separate opinion, would order a new trial. The five Justices2
3 are agreed that there was insufficient evidence to go to a jury on the overarching issue whether Dr. Edelin was guilty beyond a reasonable doubt of the “wanton” or “reckless” conduct resulting in a death required for a conviction herein and that motions for a directed verdict of acquittal should have been granted accordingly. Three of the five Justices4 would reach the same result of reversal and acquittal on each of the additional, independent grounds (a) that there was insufficient evidence to go to a jury of a live birth, an indispensable element for conviction of manslaughter, (b) that there was prejudicial divergence between the accusation against Dr. Edelin and the instructions to the jury. The two other Justices in a separate opinion explain their concurrence on the issue of wanton or reckless conduct; they decline to accept ground (a) or (b).
I. The Trial
A. The Setting. For many years a criminal abortion statute (G. L. c. 272, § 19) had had the effect in the Commonwealth of punishing as a crime the performance of any abortion except when carried out by a physician “in good faith and in an honest belief that it [was] necessary for the preservation of the life or health of a woman.”5 [500]*500Similar prohibitory legislation existed in other States.6
On January 22, 1973, the Supreme Court of the United States decided the cases of Roe v. Wade, 410 U.S. 113, and Doe v. Bolton, 410 U.S. 179. These decisions not only “rendered inoperative” our criminal abortion statute, as we had occasion to say in Doe v. Doe, 365 Mass. 556, 560 (1974), but introduced a new regime affording constitutional protections as follows (quoting from Wade, 410 U.S. at 164-165):
“ (a) For the stage prior to approximately the end of the first trimester, the abortion decision and its effectuation must be left to the medical judgment of the pregnant woman’s attending physician.
“ (b) For the stage subsequent to approximately the end of the first trimester, the State, in promoting its interest in the health of the mother, may, if it chooses, regulate the abortion procedure in ways that are reasonably related to maternal health.
“(c) For the stage subsequent to viability, the State in promoting its interest in the potentiality of human life may, if it chooses, regulate, and even proscribe, abortion except where it is necessary, in appropriate medical judgment, for the preservation of the life or health of the mother.”
Thus a physician is constitutionally protected in performing an abortion during the nine months of the patient’s pregnancy except as the State may have acted to regulate the matter — more precisely, regulated the matter to promote its “interest in the health of the mother” after the first trimester, or to promote its “interest in the [501]*501potentiality of human life” after the stage of viability. The Wade opinion also states, “Viability is usually placed at about seven months (28 weeks) but may occur earlier, even at 24 weeks,” and that at viability the fetus is “potentially able to live outside the mother’s womb, albeit with artificial aid,” and “presumably has the capability of meaningful life outside the mother’s womb.” Id. at 160, 163.7
Following the Wade-Bolton decisions, our Legislature began to deliberate whether and how it might “choose” to intervene and regulate the subject, and comprehensive legislation was formulated and was finally adopted as G. L. c. 112, §§ 12H-12R, in August, 1974.
The record shows, as could be expected, that requests for abortions multiplied at Boston City Hospital after the Wade-Bolton decisions came down. The practice of the hospital regarding the conditions for accepting patients for abortion was materially liberalized. It was in the interim between the new Wade-Bolton dispensation and the enactment of the State legislation that the patient here was accepted for the abortion procedure.
B. Summary Statement,8 1. Basic circumstances. Some days before September 30,1973, the patient appeared with her mother at the outpatient clinic at Boston City Hospital requesting an abortion. Dr. H. R. Holtrop, chief of the outpatient OBS/GYN service, interviewed the patient and advised her about alternatives to abortion, which she did not accept. He then inquired about her last menstrual period. She placed it at a date which would indicate she was seventeen weeks gone. After physical examination, Dr. Holtrop concluded that the gestational age was twenty weeks. He then advised and approved abortion by the saline method (a common method in use for abortions in [502]*502the second trimester), and introduced Dr. Edelin as the surgeon who would carry out the procedure.9
At the patient’s admission to hospital on September 30, a student entering his third year at medical school worked up the history and examined the patient. He estimated a gestational age of twenty-four weeks but, because of his lack of clinical experience, he put that as only a guess. Dr. Enrique Giménez-Jimeno, a junior resident, also examined the patient and estimated twenty-four weeks; he recorded finding a fetal heartbeat. On October 1, Dr. Holtrop, with the two estimates before him, reexamined the patient; he concluded that the period was twenty-one to twenty-two weeks.10
On October 2 the patient was brought to the “saline unit” for abortion by amniocentesis with saline infusion, a process of inducing fetal death and a miscarriage by introducing a salt solution into the amniotic sac containing the fetus. Dr. Edelin, with all prior estimates before him, made his own measurement and estimated the gestational period as twenty to twenty-two weeks. Proceeding in the usual way, Dr. Edelin inserted a long needle through the abdominal skin at a selected, locally anesthetized spot, hoping to reach into the amniotic sac and to drain off clear amniotic fluid; salt solution would then be let into the sac. The needle, however, recovered a bloody tap; and the result was the same on further tries. This indicated that the [503]*503needle had not gone true into the amniotic sac. To introduce salt solution elsewhere than in the sac might endanger the patient. Dr. Edelin surmised that he was dealing with an anterior placenta, i.e., one connected to the front of the uterine wall, a condition that could account for his recovering the bloody taps.
Accordingly, Dr. Edelin discontinued his probes and consulted Dr. James F. Penza who as associate director of the OBS/GYN department was his supervisor. It was agreed that Dr. Penza would himself attempt amniocentesis the following morning and, if that failed, Dr. Edelin would perform a hysterotomy. On October 3, Dr. Penza made the further attempts but they failed. Thereupon Dr. Edelin went forward with the other agreed procedure.
The abortion by hysterotomy involved incision of the uterus to reach and remove the products of conception. In about ten minutes’ time after the patient received general anesthesia, Dr. Edelin commenced incising through the abdominal wall to reach the uterus. As he had diagnosed an anterior placenta, he made a low transverse cut just above the pubic hairline. The process of incising, retracting, and so forth was laborious and took perhaps thirty minutes. Reaching the uterus, Dr. Edelin made a transverse incision of about six to seven centimeters, a relatively small cut: the lower part of the uterus is thick at twenty to twenty-two weeks and excessive bleeding was to be avoided. Dr. Edelin reached into the incision with the index and middle fingers of his left hand, steadying the uterus with his right hand. He swept the uterine cavity with his fingers to detach the placenta from the uterine wall; then he began to peel the amniotic sac, intending to recover the sac intact through the incision, the rest to follow. As he brought his fingers behind the sac and began to ease the sac through the incision, the sac ruptured. Dr. Edelin then sought to take hold of a lower extremity of the fetus in order to draw the fetus through the uterine incision. This was difficult because of the size and location of that incision.
Upon removal of the fetus, Dr. Edelin put his hand on [504]*504its chest wall for a few seconds; finding no heartbeat or other sign of life,11 he placed the fetus in a stainless steel basin held for the purpose by an attending nurse. He turned his attention promptly to the patient with the open uterine incision. After removing any remaining material in the cavity, and swabbing the cavity and taking other indicated steps, he undertook the suturing process and concluded the procedure. The patient recovered without incident.
The fetus and placenta were transferred to the pathology laboratory according to usual practice on the morning of the operation. The resident pathologist weighed the fetus twice and recorded 600 grams (one pound, five ounces). For preservation, the fetus and cord were placed in formalin, a ten per cent solution of formaldehyde.
2. The question of the condition and potentiality of the fetus. A large amount of testimony was admitted on these subjects which on the part of the prosecution ran thus. As against the pathologist’s weigh-in of the fetus at 600 grams just after the operation, the medical examiner, Dr. George W. Curtis, at an autopsy on February 12,1974, found a weight of 693 grams for the fetus, seven grams for the umbilical cord (one pound, eight and one-half ounces total), and it was his opinion that the fact that the fetus had been soaking for more than four months would not have increased the weight and might have reduced it. Individual organs were weighed separately. There was evidence also as to the crown-to-heel and crown-to-rump length of the fetus (respectively 33.5 and 21 centimeters). The weight and length figures were related to sundry published tables in order to support estimates about the gestational age of the fetus.
Evidence was admitted about the condition of the fetal lungs and the possibility of respiratory activity by the fetus. According to Dr. Curtis, the lungs sank in water and [505]*505had a solid appearance, indications of the absence of respiratory movement. Microscopic examination of lung tissue later fixed on slides, however, showed partial expansion of some of the alveoli. This suggested respiratory activity, but left the question where it had taken place. Dr. Curtis’s testimony is fairly read as allowing any of three possibilities: that the fetus had sucked amniotic fluid (as it might have done when distressed), or had taken in room air through the uterine incision, or had done so after delivery clear of the uterus. The last might betoken postnatal “fife” in some sense.
Dr. John F. Ward, a pathologist, testified on the basis of his microscopic examination of lung tissue that the fetus “did breathe outside the uterus.” His adoption of the third alternative appears based at least in part on his skepticism about the phenomenon of respiratory movement in útero incident to the sucking of amniotic fluid; in this he seemed to stand somewhat apart from other experts.
Their analysis of post mortem materials or data led experts for the Commonwealth12 to put the gestational age at the time of the operation at twenty-four weeks, excepting Dr. Ward who went somewhat higher to twenty-six weeks. All recognized that the figures were only estimates. Several prosecution experts would answer in the affirmative the inferential question of “viability” of the fetus at the time of the operation.13 The exact import of such judgments turned on the meaning to be ascribed to that term. Here the testimony was diverse, with one Commonwealth witness going so far as to say that a fetus was to be considered viable if it had any chance of surviving, and another expert stating that a fetus was viable if it could survive for one moment outside the uterus, which might [506]*506be the case with a fetus of twelve weeks’ age.14 In this connection (as in other respects later mentioned15) the judge allowed very considerable latitude for the expression by witnesses of medical ideas which he did not relate or channel back to legal standards. For example, the judge did not bring to bear during trial (or, for that matter, in his charge to the jury) the Supreme Court’s definition of viability in the relevant sense of the point at which the State might, if it chose, constitutionally assert its interest in bringing the fetus to full term.
3. The question of homicidal conduct. We can find no dispute in the testimony about the proposition that the established procedure (“protocol”) for abortions by hysterotomy at Boston City Hospital accepted and envisaged detachment of the placenta as the first surgical move after completion of the uterine incision, with subsequent peeling of the amniotic sac. On the part of the prosecution, there was testimony tending to show that a reverse procedure, resembling Caesarean section performed when a fetus approached or reached term, was possible and might have been preferable, especially with a fetus of a gestational age approximating that present here: this procedure would involve first breaching the amniotic sac and removing the fetus, then separating the placenta. There was also testimony critical of the relatively small size of the uterine incision which may have caused the rupture of the amniotic sac.
Dr. Gimenez testified that he walked into the operating room when the operation was under way. He said he saw Dr. Edelin’s sweeping motion, evidently the one detaching the placenta, after which, he said, Dr. Edelin remained motionless for at least three minutes with his hand in the uterus, his eyes fixed on a clock on the wall. He said, further, that all others in the room — the anesthetist, student assistant, and nurses — also remained still with [507]*507their eyes on the clock. Then Dr. Edelin delivered the fetus from the uterus. The interval of immobility as testified to by Dr. Gimenez appeared to start immediately after the detachment of the placenta and to end with the delivery of the fetus; the testimony did not refer to any space of time that might cover detachment of the amniotic sac, rupture of the sac with spill of fluid, and grappling for an extremity of the fetus in order to accomplish the withdrawal.
Dr. Gimenez testified that the fetus appeared dead on delivery. With regard to the testing for heartbeat extra útero, an expert called by the defense stated his opinion that ten seconds would be required to reach certainty about the existence of a heartbeat.
C. The Accusation and Theories of Commonwealth and Defense. We turn to the precise accusation.16 The short-form manslaughter indictment found by the grand jury on April 11, 1974, stated simply “that [Dr. Edelin] did assault and beat a certain person, to wit: a male child described to the ... Jurors as Baby Boy17 and by such assault and beating did kill the said person.” The Commonwealth was required to respond to an order for particulars of the alleged manslaughter. First, as to when the death of the “person” had supposedly taken place, the Commonwealth declined the tendered proposition18 that the death occurred “when the fetus was totally expelled or removed from the [508]*508body of the mother,” and adopted the propositions that death occurred “when Baby Boy was within the mother, albeit detached from the mother and independent of the mother, and [szc] ... when Baby Boy was partially expelled or removed from the body of the mother.”19 As to the defendant’s supposed act claimed to constitute the manslaughter, the Commonwealth particularized that Dr. Edelin’s “act during the course of a hysterotomy, which act constituted manslaughter, was his waiting 3-5 minutes after he manually separated the placenta from the uterine wall and before he removed the person from the abdominal cavity of his mother.”
From the indictment as particularized, it appeared that the Commonwealth’s theory was that, upon the detachment of the placenta, the fetus became a “person” within the manslaughter statute, and was then killed by a wanton and reckless act of Dr. Edelin, all before the birth of the fetus through its complete delivery clear of the mother’s body. The defense asserted by repeated motions that such an accusation would be an untenable usage of the manslaughter statute and would, besides, enable the Commonwealth to evade or subvert the constitutional rule of Wade-Bolton. The defense contended that manslaughter could be made out, if at all, only where a fetus was born alive completely outside the mother’s body and was homicidally destroyed by acts committed at that stage.
The defense did not persuade the trial judge to limit the case as just indicated and to confine the evidence accordingly. (How far the judge’s ultimate instructions to the jury, which we proceed to describe, represented a shift of position, and whether such a shift was prejudicial, are questions considered in “Additional Views of Three Justices,” at 520 infra.)
D. The Judge’s Instructions to the Jury. The judge charged that the “wanton” or “reckless” conduct that had to be proved for conviction of manslaughter predicated a [509]*509frame of mind more egregious or corrupt than that for “negligence” or “gross negligence”; there must be proof of “indifference to or a disregard of the probable consequences to the rights of others.” All this was consistent with Commonwealth v. Welansky, 316 Mass. 383, 396-401 (1944).
As to the further necessary ingredients of the offense in the particular case, the judge began by saying that manslaughter was “inextricably intertwined” with the Wade-Bolton decisions. He quoted the Supreme Court’s (a)(b) - (c) summary in Wade (set forth at 500 supra). Wade-Bolton, he said, struck down the existing Massachusetts laws on abortion, and on the critical date, October 3,1973, the Commonwealth had not taken any steps to regulate the subject. By abortion was meant the termination of pregnancy at any stage during the nine months, and in the absence of any regulatory statute “the abortion process and its effectuation must be left to the medical judgment of the pregnant woman’s attending physician.”
The manslaughter statute could apply only as follows. Only when a fetus had been bom alive outside its mother could it become a “person” within the meaning of the statute. Thus, to be convicted of the crime, Dr. Edelin must be found — by reckless or wanton acts — to have “caused the death of a person who had been alive outside the body of his or her mother,” “caused the death of a person, once that person became such.” We set out this part of the charge in full text:
“A fetus is not a person, and not the subject of an indictment for manslaughter. In order for a person to exist, he or she must be born. Unborn persons, as I said, are not the subject of the crime of manslaughter. Birth is the process which causes the emergence of a new individual from the body of its mother. Once outside the body of its mother, the child has been born within the commonly accepted meaning of that word.
“Killing or causing the death of a person who is born alive and is outside the body of his or her mother may [510]*510be the subject of manslaughter. In order for the defendant to be found guilty in this case, you must be satisfied beyond a reasonable doubt, as I have defined that term for you, that the defendant caused the death of a person who had been alive outside the body of his or her mother. If you believe beyond a reasonable doubt that the defendant, by his conduct, caused the death of a person, once that person became such as I have defined the word for you, you may find the defendant guilty of the crime of manslaughter, if that death was caused by wanton or reckless conduct on the part of the defendant. If, on the other hand, you do not find beyond a reasonable doubt that the defendant by his conduct caused the death of a person, then you must acquit him of the crime charged.”
Two more points. “Bom alive”: As to what birth meant, the judge said it was “the process which causes the emergence of a new individual from the body of its mother.” The judge, however, did not discuss the meaning of “alive” — how far this connoted an ability to survive, under what conditions, for what length of time.
“Viability”: The judge left it to the jury in their “consideration of the facts of this case” to “determine” whether viability had “any applicability.” If they determined that it did, it should be taken to mean “the ability to live postnatally.” The judge, however, did not discuss what the chances, conditions, or duration of survival must be.
After receiving the charge, the jury returned a verdict of guilty. Dr. Edelin was sentenced by the judge to one year’s probation, execution of the sentence being stayed pending appeal.
II. Insufficient Evidence of “Wanton” or “Reckless” Conduct to Go to a Jury
For purposes of the present point, we assume there was sufficient evidence to entitle the jury to find a “live” birth, in some sense of that term, and subsequent death, as re[511]*511quired for conviction of manslaughter under the judge’s charge.
But what acts of Dr. Edelin could be taken into account in deciding his guilt or innocence of manslaughter? Although the judge’s charge is not crystal clear on the matter, we read his charge as meaning, and certainly capable of being understood to mean, that only acts of Dr. Edelin after the birth could constitute the conduct required for conviction of the crime. Thus: “caused the death of a person who had been alive outside the body of his or her mother,” “caused the death of a person, once that person became such.” As the defense has argued, these phrases are cast in the “past perfect.” The starting point of the duty owed by the defendant under the manslaughter statute was thus set at the stage after delivery from the mother, and attention centers upon the “heartbeat test” evidence. Taking the evidence on this point in a light favorable to the Commonwealth (see Commonwealth v. Kelley, 370 Mass. 147, 149-150 [1976]; G. L. c. 278, § 11), we find nothing of substance in it to permit a submission to a jury as to criminal “recklessness” or the like.
Although the fetus may have been alive in the very narrow sense that there was some postnatal gasping of air as revealed by microscopic analysis of preserved lung tissue long after the event, nothing of the sort was observable by Dr. Edelin as he carried out the operation. To all appearances, the fetus was dead. Dr. Edelin found no heartbeat and saw no other indication that he had a live being in his hands. If we give the prosecution the benefit of a remark by a defense witness that a test for heartbeat, to be absolutely conclusive, should extend longer than the few seconds testified to by Dr. Edelin, we have to note — besides the fact that there was no evidence as to any standard practice in this regard — that the witness did not say, nor could he say plausibly, that to test for the short period would be wanton or reckless behavior in any and all circumstances of a delivery. Here Dr. Gimenez, an eyewitness, joins in Dr. Edelin’s observation. Dr. Gimenez testified: “The baby [sz'c] was dead”; it “had no signs of [512]*512life, such as breathing or movement.” Dr. Gimenez had no criticism of the defendant on the score of his postnatal conduct.
The Commonwealth tries to avoid the evident result by contending that the judge’s charge should be read so as to allow consideration of Dr. Edelin’s conduct during the prenatal stage. As already noted, we do not agree with that attempted interpretation.20 In all events, such an interpretation would go counter to the law by which Dr. Edelin is to be judged, because, as we shall indicate, (i) it is very doubtful whether the manslaughter statute spoke to prenatal conduct, and (ii) if it purported to do so, it would be displaced pro tanto for purposes of the present case by the constitutional doctrine of Wade-Bolton. But the further answer is (iii) that Dr. Edelin’s conduct from start to finish, both prenatal and postnatal, did not provide a basis for submission to a jury of the issue of “recklessness” or the like.
Manslaughter assumes that the victim was a live and independent person. This is well understood,21 and the judge so charged. Destruction of a fetus in útero is not a manslaughter. The further question was debated at common law22 whether manslaughter might rest on a defendant’s injuring a fetus in útero, where the fetus was later bom alive, and then died of the injury without further guilty intervention by the defendant. In this Commonwealth, Holmes, J., seems to intimate a preference for the view that the prenatal acts could not ground a manslaughter despite the later live birth and death. Dietrich v. Northampton, 138 Mass. 14, 15, 17 (1884). (Common [513]*513sense, besides common law, tended to support this position, for if a contrary rule were adopted, a putative defendant might be encouraged to make sure he extinguished the fetus while still in útero.) The effect, be it noted, was not to let the defendant off, but to subject him to the typical criminal abortion statute. See note 5 supra; Dietrich v. Northampton, supra at 17. Cf. Keeler v. Superior Court of Amador County, 2 Cal. 3d 619, 635-636 (1970).23
But if acts of a defendant during the prenatal period could ever be availed of in obtaining a conviction of manslaughter, that was a legal impossibility in the circumstances of this case after the Wade-Bolton decisions. (Thus our reading of the judge’s charge is actually required to keep it within constitutional bounds.) Under Wade-Bolton, the State’s regulation of abortions after viability to promote its interest “in the potentiality of human life” is expected to be expressed in “tailored” legislation. See Wade, 410 U.S. at 164-165. The detail that may be involved, representing an adjustment and accommodation of many considerations, is exemplified by the comprehensive legislation in fact adopted by our Legislature in August, 1974.24 The manslaughter statute is flat and contains no such detail, and it would be not only incongruous but, we think, unconstitutional to attempt to bring it to bear on a physician as he went about the predelivery process of performing an abortion. After Wade-Bolton, even if not be[514]*514fore, the manslaughter statute could take hold only after a live birth and only with respect to acts of the physician in the postnatal period.
If the manslaughter statute conceivably could be utilized to control and punish a physician’s conduct in the prenatal period, then the judge would be bound to say just how the statute should be applied consistently with Wade-Bolton25 The most obvious of the several matters calling for guidance by the trial judge would be a definition or description of viability marking the stage at which a physician would begin to owe a duty to the “potentiality of human life” inherent in the fetus.26 On this point, as our digest of the judge’s charge indicates, the instructions were uninstructive and left a wide-open option to the members of the jury.27
However, that omission from the charge does not embarrass a decision to reverse and acquit because, on any view of viability tenable under Wade-Bolton, there was no sufficient showing of recklessness in Dr. Edelin’s prenatal conduct — any more than in his postnatal conduct28 — ■ to permit reference of the question to a jury. His “quo [515]*515animo” turned on whether he believed in good faith that the fetus was not viable at the time of the operation and was not palpably unreasonable in this belief — a combination of an internal and an external standard of criminality. See Commonwealth v. Welansky, 316 Mass. at 398-399, referring to Commonwealth v. Pierce, 138 Mass. 165 (1884). Of course manslaughter could not be supported by proof merely of a mistake of judgment, even if that was the result of negligence or gross negligence.
The record shows that Dr. Edelin, as Chief Resident in OBS/GYN at Boston City Hospital, was well aware of the Wade-Bolton decisions and accordingly believed (correctly) that abortion was not automatically excluded at the stage of viability — that was for the Legislature to decide. He was concerned about viability, however, because it affected medical judgments and because he had a personal scruple against aborting a viable fetus.
Again considering the evidence in a light favorable to the Commonwealth, there is nothing to impeach the defendant’s good faith judgment that the particular fetus was nonviable, and nothing to suggest that that belief was grievously unreasonable by medical standards.29 An independent judgment would be required of Dr. Edelin as surgeon in the case, and his independent estimate of twenty to twenty-two weeks of gestation was consistent with the double-checked estimate of the experienced chief of clinic. The bona fides of Dr. Edelin’s estimate is perhaps reinforced by the fact that he used a relatively small uterine incision. We find nothing in the course of the operation which might have alerted him to the probability that he had been mistaken in his estimate. It is not shown that he was mistaken in fact. Prosecution and [516]*516defense experts divided on whether the fetus was viable, but a trier’s belief in one set of experts would not begin to show that Dr. Edelin’s judgment the other way was reckless. All the testimony bears comparison with the indication in Wade about a twenty-eight to twenty-four weeks gestational age in relation to viability. It is to be borne in mind, also, that no witness was prepared to state that this fetus had more than the remotest possibility of meaningful survival.
If we accept, as we think we must, that there was nothing to show Dr. Edelin believed the fetus to be viable, or was flagrantly mistaken in believing it to be nonviable, then even a three to five minute wait after detachment of the placenta would not count as recklessness because Dr. Edelin would think it indifferent to the possibility of meaningful survival.
But to follow the operation more closely: The criticisms by Commonwealth witnesses of the protocol of commencing the hysterotomy with detachment of the placenta,30 and of fixing the size of the particular uterine incision, were in the area of debate and, far from showing a criminal intent, did not reflect seriously on Dr. Edelin. Dr. Gimenez’s testimony is fairly reconcilable with Dr. Edelin’s: the pause Dr. Gimenez spoke of may well have been the interval of the delay in delivering the fetus that was due to the rupture of the amniotic sac.31 However that may be, criminal recklessness is not in the picture if one starts with an understanding of Dr. Edelin’s belief and judgment as to viability.
If any doubt remains that there was no case for the jury on the question of recklessness, two special factors in the present situation should be considered.
[517]*517First, we deal here with the professional judgments of a qualified physician acting under stress at the operating table. The Supreme Court has cautioned in the abortion cases against the undue trammeling of judgments of the individual attending physician. In Bolton the Supreme Court struck down a statutory scheme for advance review of the attending physician’s judgment by a panel of other physicians who would be conscious of the need to preserve the independence of the individual physician. (See Bolton, 410 U.S. at 195-200.) In the present case we have an attempted post hoc review of the attending physician’s judgments through a battle of experts before a lay jury with a threat of criminal conviction and professional disgrace.
Second, we deal here with the constitutional right of privacy of the patient and the correlative constitutional right of the physician — rights to which the Supreme Court has assigned very high value.
There should be caution and circumspection in the interpretation and application of a criminal statuté which, as employed here, must necessarily trench on professional practice and constitutional freedoms. This implies a wariness in assessing the strength of a criminal case made against a physician before it is held that there is enough evidence to submit to a jury.32
III. Additional Views of Three Justices
A. Insufficient Evidence of Live Birth to Go to a Jury. Whether Dr. Edelin was reckless or wanton in dealing with the fetus is a question of what he believed and how he exercised judgment. There is a further and different question, whether in fact the fetus was born alive outside the mother’s body.
Under the judge’s charge, there could be no subject of manslaughter unless and until the fetus was live born. The [518]*518charge on this point might be accepted as framing the “law of the case” even if it were erroneous (see Commonwealth v. Graves, 363 Mass. 863, 868 [1973], and cases cited; Commonwealth v. Richards, 363 Mass. 299, 307 [1973]), but it is quite plain that the charge was correct.33 The Commonwealth’s original submission that the manslaughter statute should be read to cover the destruction in útero of a “detached” fetus, amounted merely to a misguided attempt, despite Wade-Bolton, to reconstitute the old criminal abortion statute under the guise of manslaughter.
Without repeating the evidence on the question of live birth, we may note that the eyewitness testimony of Dr. Gimenez (confirming Dr. Edelin’s testimony) was that the fetus was delivered dead. The only evidence suggestive of postnatal life came from examination of the preserved lungs. Gross examination of the lungs by the medical examiner indicated still birth, not live birth, and it was only on microscopic examination of the tissue that any hypothesis of “breathing” could arise. But the breathing, assuming there was any, must be shown to have occurred outside the mother; support for this could be found in the testimony of Dr. Ward, but that was rather equivocal.34
If full marks should nevertheless be given Dr. Ward’s testimony, that is, if it be assumed that it provided enough [519]*519to entitle a jury to conclude beyond a reasonable doubt that some kind of breathing did occur after the fetus was delivered from the uterus — an assumption that appears doubtful indeed — then the question must be faced whether the breathing, such as it could have been, demonstrated postnatal “life” as required for conviction under the manslaughter statute.
Postnatal life could be variously defined for the purpose, at one end of the spectrum as a capacity of the fetus, when delivered, to sustain existence for an indefinite period without artificial aid, and at the other end as a showing of any sign of activity such as a single gasp of air not observed or observable by the eye at the time, and established, if at all, by later microscopic analysis. The jury needed a standard to go by. But the trial judge did not provide a standard.35
This might leave the case, with respect to the present point, in a condition where a new trial would be justified, but not acquittal. Here, however, we are able to say with confidence that on no acceptable standard was there proof sufficient to go to a jury. The defense had offered a proposed instruction drawing upon the definition of “liveborn infant” as formulated and published by the Committee on Terminology of the American College of Obstetricians and Gynecologists: “LIVEBORN INFANT. Liveborn infant is a fetus, irrespective of its gestational age, that after complete expulsion or extraction from the mother, shows evidence of life — that is, heartbeats or respirations. Heartbeats are to be distinguished from several transient cardiac contractions; respirations are to be distinguished from fleeting respiratory efforts or gasps....”36 This suggested standard, which required a minimal demonstration of [520]*520independent existence beyond “fleeting respiratory efforts or gasps,” was less exacting of proof and therefore more favorable from the Commonwealth’s viewpoint than the standard that has been applied in practice around the country in manslaughter cases involving newborns.37 It commended itself by its relative objectivity, practicality, and moderation. By this standard (as by any variant that might be considered acceptable) the Commonwealth failed in its proof.38
B. Prejudicial Shift from the Accusation as Specified by the Commonwealth to the Crime as Defined in the Judge’s Instructions. There were discrepancies between (i) the indictment as particularized by the Commonwealth in its response to the request for particulars and (ii) the ingredients of the crime of manslaughter as the judge described them in his instructions to the jury. The indictment as particularized asserted that the fetus died either while it was in útero or when it was partially removed; under the judge’s charge, either assertion, if proved, would be incompatible with manslaughter, for conviction depended on a live birth after delivery free of the mother. Further, the indictment as particularized asserted that the homicidal act was the deliberate pause in útero immediately following the detachment of the placenta; the judge’s charge confined the jury to postnatal conduct (as the defense interprets the charge) or at least allowed the jury to consider postnatal conduct as part of the basis of conviction (as the prosecution interprets the charge).
The bill of particulars had the effect of restricting the case that the Commonwealth was entitled to prove. “[T]he specification affected the indictment and restricted the [521]*521proof to the matters specified.” Commonwealth v. American News Co., 333 Mass. 74, 77 (1955), and cases cited. It was open to the Commonwealth within reasonable limits of circumstance and time to amend its bill. See G. L. c. 277, § 40. It did not do so. This was despite the repeated warnings sounded by the defense that on the particulars furnished there was no manslaughter. Hence it is appropriate that the Commonwealth should be held to the ground it selected.
The Commonwealth’s proof, as restricted by the particulars, could not support conviction of the offense having the elements described by the judge in his instructions to the jury. Conversely, conviction must be taken to have rested on elements forming no part of the accusation as particularized. “The offense must not only be proved as charged, but it must be charged as proved.” Commonwealth v. Albert, 307 Mass. 239, 244 (1940), quoting from Commonwealth v. Blood, 4 Gray 31, 33 (1855). The upshot is that the defendant in the extraordinary circumstances of this case was entitled to acquittal. Commonwealth v. American News Co., and Commonwealth v. Albert, both supra. See DeJonge v. Oregon, 299 U.S. 353, 362 (1937); In re Ruffalo, 390 U.S. 544, 550-552 (1968). (Whether, after acquittal on such a ground, the defendant might be reindicted, raises a question of double jeopardy that would take us far afield.)
The policy of the statute on “variance,” G. L. c. 277, § 35,39 tends to support acquittal in the present case. It says that a defendant “shall not be acquitted on the ground of variance between the allegations and the proof,” but this is conditioned on the “essential elements of the [522]*522crime” being “correctly stated.” That was not the case here. The statute indicates, further, that even when the condition is met, acquittal is called for where the defendant was “thereby prejudiced [i.e., by the variance] in his defence.” As will be indicated below, that was the case here.
It has been suggested that the particulars in the present case should be taken to have been informally amended to correspond more or less to the definition of the crime as finally given in the instructions, because the prosecution let it be known early on, by the time of the opening of trial, that it intended to attempt to prove “breathing.” But especially in the face of the particulars furnished that the homicidal act — the pause — occurred just following detachment of the placenta, this slipshod prosecution “notice” could well be understood to refer to proof of intrauterine respiratory activity bearing on “viability,” not to proof of a live birth outside the mother. Quite apart from this, however, the “notice” did not provide any warning that the prosecution was shifting ground from the claimed homicidal pause in útero to a supposed homicidal act or omission after birth clear of the mother.
A switchover such as occurred here may be thought inherently prejudicial (as § 35 indeed suggests). But if there must be a demonstration of probable prejudice, the defense has shown it sufficiently.
The defense puts an example of prejudice through confusion of the target. As noted earlier (note 31 supra), the defense attacked Dr. Gimenez’s credibility as a means of challenging his testimony (which stood alone) regarding the three to five minute wait. But the attack naturally tended also to discredit Dr. Gimenez’s further testimony, as an eyewitness, that the fetus was dead at delivery. Yet that testimony, if believed, would destroy the Commonwealth’s case, assuming the judge’s instructions to the jury. This is an example of how the switchover could unfairly confound the defense.
In another aspect, prejudice is shown. During the long course of the case (twenty-eight trial days) a great welter [523]*523of testimony was admitted, over repeated objections, which had the effect of drenching the jury with the prosecution’s theory of the case — that the fetus became a person at detachment of the placenta and was then destroyed by the surgeon’s deliberate pause. Commonwealth experts expounded medical doctrine, largely of a definitional character, tending in one way or another to support that hypothesis: for example, that there was “personhood” when, after separation from the placenta, the fetus could maintain itself for any length of time on its own systems; that a fetus was to be considered viable if it was possible to project that, were it delivered at that moment, it would have any chance to survive for even an instant; that “abortion” was a procedure confined to the first twenty weeks of pregnancy; that “abortion” was related only to termination of pregnancy, and, notwithstanding any right of a woman in that regard, the physician was duty bound at all times to preserve the fetus.
It is true that the judge indicated, both during the trial and in his charge, that testimony of this kind was being received in order to orient the jury to the witness’s professional manner of thought or speaking, and was not to be taken as laying down the controlling legal precepts. The first trouble with regarding this as a corrective is that there was no attempt during trial to state the critical divergence of the definitional propositions from their proper legal counterparts. An attempt of this sort was made in the final instructions but the judge’s remarks (i) were very brief in relation to the voluminous and accumulative testimony, (ii) did not point out wherein the testimony differed from the proper legal norms, and (iii) left important concepts empty of definition so that the received testimony would rush in to fill the spaces, since nothing else was being supplied. This is strikingly true of “viability” which was not only left in substance undefined by the judge, but was, according to him, to be availed of by the jury “if... you determine that the ... [term has] any applicability.”
There is, then, a strong probability of prejudice in that, notwithstanding the switchover, the jury may have con[524]*524victed Dr. Edelin on the basis of the Commonwealth’s erroneous view of the case.40
IV. Conclusion
This opinion does not seek an answer to the question when abortions are morally justifiable and when not. That question is wholly beyond our province. Rather we have dealt with a question of guilt or innocence under a particular statute on a particular state of facts. We are conscious that the significance of our decision as precedent is still further reduced by the fact that the case arose in an interregnum between the Supreme Court’s abortion decisions of 1973 and the adoption of legislation intended to conform to those decisions — a kind of interval not likely to be repeated.
Legal relationships during that interval were hard to define with certainty. Understandably, the doubts infected the theory on which the Commonwealth went. The judge tried the case with skill and careful attention to detail, but faced unprecedented problems which he had to resolve on the spot.
In the comparative calm of appellate review, the essential proposition emerges that the defendant on this record had no evil frame of mind, was actuated by no criminal purpose, and committed no wanton or reckless acts in carrying out the medical procedures on October 3, 1973. A larger teaching of this case may be that, whereas a physician is accountable to the criminal law even when performing professional tasks, any assessment of his responsibility should pay due regard to the unavoidable difficulties and dubieties of many professional judgments.41
[525]*525The judgment is reversed and the verdict set aside. Judgment of acquittal is to be entered.
So ordered.