The defendant was found guilty by a jury of murder in the first degree of his wife. The principal issue at trial was whether the defendant lacked criminal responsibility for his acts. The defendant asks us to reexamine our *221earlier determinations that a judge may instruct the jury on the so called “presumption of sanity.” He requests a new trial under G. L. c. 278, § 33E, arguing that the jury’s finding of criminal responsibility was against the weight of the evidence and that no instruction was given on the possible effect of mental impairment on the defendant’s capacity to commit murder in the first degree. He also asserts error in the judge’s refusal to instruct the jury on manslaughter. We find no error, nor any reason pursuant to G. L. c. 278, § 33E, to disturb the jury’s verdict. We affirm.1
On June 2, 1979, the defendant came home from work, stabbed his wife twenty-seven times, and eventually called the police. There was evidence that, at least since the previous autumn, he had suspected her of infidelity and of trying to kill him. There was also evidence that he said that, on the night of the murder, he had seen a man’s suit hanging in the bedroom closet, an indentation in their bed beside his wife, and a strange car leaving the neighborhood, and that his wife had said something to the effect that she could do anything with her body that she wanted. He presented two psychiatrists as witnesses; the Commonwealth presented none.
1. The defendant asks us to reconsider the propriety of a jury charge on the “presumption of sanity” in the form approved in Commonwealth v. Kostka, 370 Mass. 516, 525-537 (1976).2 There we said that, even where there was uncontradicted expert testimony that the defendant lacked criminal responsibility, a judge may instruct the jury that they may consider the fact that a great majority of people are sane and the resulting probability that any particular person is sane. Id. at 525-526. The judge charged the jury in substantially this language. We have recently approved a like charge in Commonwealth v. O’Brien, 377 Mass. 772, *222780 (1979). Such an instruction was upheld against a constitutional attack in Walker v. Butterworth, 599 F.2d 1074, 1077-1080 (1st Cir.), cert. denied, 444 U.S. 937 (1979). We see no reason to depart from our previous position.
2. The defendant next asserts that, in light of the evidence of his lack of criminal responsibility, we should exercise our powers under G. L. c. 278, § 33E, to grant a new trial because the jury’s finding of criminal responsibility was against the weight of the evidence.
In some cases, although juries returned verdicts of guilty of murder in the first degree, the evidence of lack of criminal responsibility was so compelling that justice demanded a new trial. See, e.g., Commonwealth v. Mutina, 366 Mass. 810, 811-812 (1975); Commonwealth v. Cox, 327 Mass. 609, 615 (1951). In other cases, we have not ordered new trials where lay testimony or the defendant’s actions warranted a finding of his criminal responsibility, although the defendant had a history of mental illness and no expert testified for the Commonwealth. See Commonwealth v. Walker, 370 Mass. 548, 582-583, cert. denied, 429 U.S. 943 (1976); Commonwealth v. Kostka, 370 Mass. 516, 537-539 (1976).3 In Commonwealth v. O’Brien, supra, we declined to grant a new trial to a man who shot his wife, although the Commonwealth presented no expert testimony on criminal responsibility (id. at 781) and there was evidence that, in the weeks prior to the shooting, the defendant underwent a personality change and became concerned about the real and imagined infidelities of his wife (id. at 782). There, the defendant had no prior history of mental illness (id. at 782), and there was lay evidence that he appeared normal at the *223time of the shooting (id. at 781). The opinions of the defendant’s three psychiatric experts tended to undercut each other, and “ [e]ach opinion was based in part on facts which the defendant furnished to the expert and which were not corroborated by the expert or otherwise supported in the evidence.” Id. at 784. Hence, we concluded that justice did not demand a retrial. Here, likewise, the jury’s finding that the defendant was criminally responsible does not contradict the weight of the evidence.
The defendant had no history of prior mental illness. However, two psychiatrists, a Dr. Neal Borenstein and a Dr. Anneliese Pontius, testified that in their expert opinions, the defendant was not criminally responsible for killing his wife. Dr. Borenstein believed that Brown suffered from “a mental illness of a psychotic proportion,” probably a severe depression, while Dr. Pontius diagnosed his illness as “paranoid involutional psychosis.” The experts’ testimony was not strong. Dr. Borenstein’s opinion was based on a one-hour conversation with the defendant two months after the killing. He characterized his opinion as “tentative.” Dr. Pontius saw Brown for a total of two hours, four months after the incident. At the time that Dr. Pontius determined that Brown lacked criminal responsibility for killing his wife, she had no knowledge of his behavior during and around the time of the crime. She testified, on cross-examination, that her opinion would change if the defendant told the police that someone else did it. There was evidence that the defendant told the police that he saw the car of his wife’s boyfriend leaving as he came home. Both psychiatrists seem to have based their diagnoses in large measure on the defendant’s uncorroborated statements to them about his belief that his wife was committing adultery and attempting to kill him.
Although the defendant did not take the stand, there was evidence as to the reasons that he believed his wife was committing adultery and trying to kill him. Some of his asserted reasons seem irrational, even considering that the defendant had only spent one year in this country, and may have been *224unfamiliar with its customs. For instance, he believed that a piece of white cake came from his wife’s wedding to another man; for twenty-two days he kept track of a bag of candy because he thought it was a present to his wife from her lover; he believed that his wife left him a poisoned piece of cake.
There was, however, other evidence from which the jury could have concluded that, apart from whether it was true, the defendant’s conclusion that his wife was seeing another man was at least rational. There was testimony that Brown had said that his wife traveled to Canada without inviting him, and returned with another man; that one day when he stayed home sick from work she returned with a man; that he found cigarette butts in the house but neither he nor his wife smoked; that she showed him a picture of a man she said she was “playing with”; and that he noticed evidence of sexual intercourse on the bedsheets.
As for the defendant’s behavior on the night of the crime, it was neither clearly rational nor clearly irrational. There was testimony that when he returned from work on the night of June 2, 1979, the defendant believed he saw a car leaving the house, a man’s suit in the closet, and an indentation in the bed beside his wife, and that his wife had said something to the effect that she could do anything with her vagina that she wanted. The rationality or the credibility of his observations is somewhat undercut by testimony of a witness that he saw no other car in the street when Brown entered his residence, and testimony that a women’s suit in the closet could have been mistaken for a man’s suit.
Brown’s behavior just after he stabbed his wife was not clearly irrational. He apparently threw the murder weapon out a window, showered, and changed his clothes, but did not clean up his bloody work clothes from the bathroom floor. He then called the police. He told the investigating officers that he saw a man leaving the apartment, but later admitted to owning the bloody clothes. He was calm during the questioning.
*225The evidence would have supported a jury’s concluding beyond a reasonable doubt that the defendant had a substantial capacity to appreciate the wrongfulness of his conduct and to conform his conduct to the requirements of law. See Commonwealth v. McHoul, 352 Mass. 544, 546-547 (1967). His attempt to dispose of the knife and his references to the presence of his wife’s boyfriend tend to show an appreciation of the wrongfulness of his conduct. Late in May, shortly before his wife’s death, the defendant discussed his marital problems with his foreman. When he showed his foreman a large pocket knife, the foreman said, “You wouldn’t use that, would you?” The defendant replied that he was too old and too wise to do such a thing. This evidence tends to show that the defendant could conform his conduct to the requirements of law as well as appreciate the wrongfulness of using the knife on his wife. Moreover, the jury could well have rejected the conclusions of the psychiatrists. Indeed, if the jury concluded that the defendant tried to place the blame on “his wife’s boyfriend,” as the jury could have inferred, Dr. Pontius’s testimony suggests that a finding of criminal responsibility would be appropriate. Despite the evidence that the defendant had some irrational beliefs and engaged in some irrational actions, the jury’s conclusion that he was criminally responsible was not against the weight of the evidence.
3. We perceive no manifest injustice in the failure of the judge to instruct the jury to consider whether the defendant’s mental impairment precluded him from being able (a) to premeditate his wife’s murder or (b) to commit murder with extreme atrocity or cruelty. The defendant was tried before we issued our opinion in Commonwealth v. Gould, 380 Mass. 672 (1980), where we said that, in the future, when the evidence fairly raises an issue of the defendant’s capacity to commit murder in the first degree by premeditation or with extreme atrocity or cruelty, the defendant may have appropriate instructions. Id. at 683, 686 n.16.
*226Although we stated in the Gould opinion that the rule there announced applied only prospectively, we exercised our powers under G. L. c. 278, § 33E, to order a new trial. Our powers under § 33E are extraordinary, and are exercised only when necessary to achieve justice. In the Gould case, there was uncontradicted evidence that the defendant had a “long-standing, constant delusional belief system.” Id. at 676. During the five years preceding the murder, he had been in and out of various institutions for the mentally ill. Id. The Commonwealth’s expert diagnosed his illness as a “paranoid psychosis.” Id. at 677. This court accepted that “[f]or more than five years preceding the fatal stabbing the defendant had suffered from constant, fixed delusions in which he believed that he had a divine mission on earth and that he was required to kill his girl friend because she was ‘impure.’” Id. at 673. In light of these facts, we took the unusual step of concluding that justice required a new trial at which an instruction on diminished capacity would be given.
In contrast, in a case decided after the Gould case, where the facts concerning mental impairment were less strong, we perceived no injustice arising from the absence of an instruction on diminished capacity. In Commonwealth v. Shelley, 381 Mass. 340 (1980), the defendant was convicted of murder in the first degree in a trial at which his sole defense was his lack of criminal responsibility. Id. at 342. There was evidence that Shelley used a meat cleaver and a “hot dog” fork to hack to death another man with whom he had been lying in bed, and who had made sexual advances to him. After killing the victim, Shelley telephoned the police. Id. Shelley had no psychiatric history prior to the killing. Two expert witnesses testified to his paranoid personality well after the incident; another expert testified to Shelley’s borderline personality and his “dissociative reaction” during the killing; and a fourth expert gave his opinion that Shelley was not criminally responsible for the murder. Id. at 343-345. The court concluded that the trial judge’s failure to instruct the jury to consider the effect of Shelley’s *227mental impairment on his capacity to commit murder in the first degree did not require a new trial under G. L. c. 278, § 33E, where no such instruction was requested and Shelley had no prior psychiatric history. Id. at 354-355.
The case before us does not differ significantly from Commonwealth v. Shelley, supra. A new trial is not warranted.
4. We come, finally, to the defendant’s contention that the judge erred in declining to instruct the jury on manslaughter. The test for determining whether a manslaughter instruction must be given is whether “any view of the evidence will permit a finding that the offence is manslaughter and not murder.” Commonwealth v. LePage, 352 Mass. 403, 419 (1967). Here, after making all reasonable inferences in favor of the defendant, and treating the defendant’s story, regardless of its credibility, as if it were entirely true (see Commonwealth v. Vanderpool, 367 Mass. 743, 746 [1975]), we conclude that a manslaughter instruction was not required. Voluntary manslaughter is “a killing from a sudden transport of passion or heat of blood, upon a reasonable provocation and without malice, or upon sudden combat.” Commonwealth v. Soaris, 275 Mass. 291, 299 (1931).
The defendant asserts that there was evidence that he stabbed his wife after sudden combat. Granted, there was evidence that the defendant told Dr. Pontius that his wife “choked me on the neck with my shirt.” The fact that his wife did so would not be sufficient provocation. This is especially true when one considers that the unarmed victim was stabbed twenty-seven times. See Commonwealth v. Walden, 380 Mass. 724, 727 (1980) (“[P]hysical contact between a defendant and a victim is not always sufficient to warrant a manslaughter instruction, even when the victim initiated the contact”); Commonwealth v. Rembiszewski, 363 Mass. 311, 321 (1973) (“It is an extravagant suggestion that scratches [inflicted by the victim on the defendant’s face] could serve as provocation for a malice-free but ferocious attack by the defendant with a deadly instrument”).
*228The defendant also claims that there was evidence, sufficient to require a manslaughter instruction, that he stabbed his wife in a fit of passion after she suddenly admitted adultery. In Commonwealth v. Schnopps, 383 Mass. 178, 180-182 (1981), where the evidence warranted a finding that the victim suddenly admitted adultery to her spouse, he was entitled to a manslaughter instruction because “[t]he existence of sufficient provocation is not foreclosed absolutely because a defendant learns of a fact from oral statements rather than from personal observation.” Id. at 181, quoting from Commonwealth v. Bermudez, 370 Mass. 438, 440 (1976).
Here, on no view of the facts did the defendant “suddenly discover” adultery. Any claim of “sudden discovery” is belied by the defendant’s statements that for over six months he had suspected his wife of infidelity, that he had previously discovered supporting “evidence” (such as stained sheets), and that his wife had told him earlier that she was “playing with” another man. According to the defendant, on the night of the killing, he saw a car leaving the premises, a man’s suit in the closet, and an extra indentation in the bed, and his wife told him that she could do what she wanted with her vagina. Assuming the truth of these statements, they do not, in the circumstances, rise to the level of a “sudden provocation” sufficient to reduce murder to manslaughter. Compare Commonwealth v. Schnopps, supra, with Commonwealth v. Bermudez, supra at 439-440 (evidence that defendant’s wife, with whom he was not living, swore and told him that she had another man, did not warrant a finding of sufficient provocation, and the absence of a manslaughter charge was not error).
5. We have reviewed the entire record in fulfilment of our duty under G. L. c. 278, § 33E, and we find no reason to disturb the verdict of murder in the first degree.
Order denying motion for a new trial affirmed.
Judgment affirmed.
The judge denied the defendant’s motion for a new trial, which raised no issue not involved in the defendant’s direct appeal.
The word “presumption” should not be used in any such charge. Commonwealth v. Kostka, 370 Mass. 516, 525 n.5 (1976).
Of course, there may be instances in which the Commonwealth will run “the very real risk of reversal and the granting of a new trial if it chooses to rely on the presumption [of sanity] and the circumstantial evidence of sanity . . . rather than to introduce medical evidence of sanity.” Commonwealth v. Kostka, supra at 539-540 (Hennessey, C.J., dissenting in part). See Commonwealth v. Walker, 370 Mass. 548, 584, cert. denied, 429 U.S. 943 (1976) (Hennessey, C.J., dissenting in part); id. at 584-585 (Liacos, J., concurring). This risk is particularly intense when the defendant has a history of mental disease.