Kaplan, J.
The defendants Roberto Medina and Erasmo Soto, after a joint trial in September, 1975, for the murder in December, 1974, of Ana Ausua, were found guilty, respectively, of murder in the first and second degree. Their appeals (G. L. c. 278, §§ 33A-33G; St. 1979, c. 346; Mass. R. A. P. 1B, inserted by 378 Mass. 924 [1979]), bring us questions of the correctness of rulings by the trial judge on motions for directed verdicts of acquittal, of his instructions to the jury, and of his denials of certain claims for postjudgment relief. The defendants’ appeals were briefed and argued separately and at different times, but it is convenient to dispose of them in a single opinion. We should say, however, that no charge of conspiracy or joint venture was made, and the responsibility of each defendant thus depends wholly on his individual acts.
We shall conclude that the judgment of conviction of Medina should stand, but that, on a review of the case under G. L. c. 278, § 33E, Soto’s case should stand for a new trial.
We first describe in some detail the testimony at trial, and then pass to the claims of error and the § 33E review.
1. Narrative, (a) Commonwealth’s evidence. The victim’s sister Ama Lagares gave testimony as follows. Around 4 p.m., December 25, 1974, she accompanied the victim and three of the victim’s children to 26 Walden Street (at a housing project in Jamaica Plain). They were in search of [567]*567Soto: the purpose was to “show” the youngest child, four year old Ana Marisa Soto, to her father, who had lived with the victim (in Puerto Rico and Boston) for more than three years before their recent separation in November, 1974. Arriving at 26 Walden, and calling for Soto, they heard a female voice, presumably from the third-floor apartment of Carmen Amaro, saying that Soto was not there. Then, without provocation, Medina and three others (Medina’s sister Virginia Gomez, Soto’s new companion Juanita Amaro, and Carmen Amaro), came downstairs and into the yard. Medina struck Lagares; Juanita hit her on the head with a shoe. Lagares required treatment that afternoon at Peter Bent Brigham Hospital.
Lagares went on to testify that about 9 p.m. that night Medina, brandishing a knife and looking for the victim, showed up at the apartment of Carmen Rodriguez, where the victim and others were having a Christmas party. He was driven away.
Later that evening, about 11 p.m. , the victim, the witness Lagares, Lydia Perez, and four or five children passed 26 Walden on their way to Natividad Rosa’s house, and were suddenly attacked by the same group that had started the earlier fracas, together with some others. As the victim was pushed to the ground and surrounded, the witness ran and hid behind a parked car nearby. From that haven she saw Medina strike the victim repeatedly with a baseball bat. The bat broke, but Medina evidently struck her with the haft. Somehow, according to Lagares, the victim after this beating managed to get to her feet and tried to cross the street; she fell midway.
Then Soto, whom Lagares had not observed among those attacking the victim, drove up in a black car with another person. He left the car holding a rope, bound the victim with it around her middle, and tied the other end to the rear of the car. He entered the car, and drove it a short distance, dragging the victim along the street. Soto untied the victim, retrieving the rope, put the car in reverse, and “drove the car over her.” Soto walked away; the other person drove off. Lagares ran home, finding Lydia Perez and police already there.
[568]*568The prosecution called the victim’s son Luis Martinez, twelve years old at the time of the incident. He had been living with Lagares since his mother’s death. He was a difficult witness. There was a particularly severe language problem (like most of the other Hispanic witnesses, he had virtually no English); many questions he simply did not respond to. His story may be reconstructed thus. He was one of the children who accompanied the victim that night, and confirmed that she was set upon while passing by 26 Walden. Like Lagares, he hid behind a car. He saw Medina hit his mother once on the head with a bat, and she fell. The boy said he recognized Medina, Soto, Juanita Amaro, Maribella Muniz, and Virginia Gomez in the attacking group (but he did not testify to Soto’s striking the victim). It was Medina who drove up in a car; Soto was his passenger. The car “passed over” the victim, then stopped. The two men left the car, tied the victim to the back of the car, reentered the car, dragged the victim the length of a block, stopped the car, untied her, and drove off.
Police Officer Robert Dunford testified that, on patrol that night, he responded with his partner John Curley to a call for assistance at approximately 11:55 p.m., and found the victim lying in slush in the middle of Walden Street, and a baseball bat broken and stained with blood a short distance away; there was also a mop nearby. The victim was unconscious, bloodied at the head and eyes. A police ambulance was summoned and she was taken to Peter Bent Brigham Hospital. After speaking to persons on the scene, the officers went to Carmen Amaro’s apartment at 26 Walden and spoke there to Maribella Muniz. Medina was also there. The right sleeve of his shirt was ripped from the elbow; his pants and shoes were wet, there was a puddle of water around his shoes. He gave his name falsely as Michael Davi.
A “criminalist” with the Boston police department, Leslie F. Wilson, had examined the corduroy jacket worn by the victim. It had various frayed areas. Besides stains of human blood, the garment showed black residue probably de[569]*569riving from lubricating grease or petroleum products. The witness did not find rope fragments on the jacket. There were human blood stains on the two pieces of the baseball bat. Analysis indicated that stains on the haft came after the stains on the head of the bat.
Dr. Leonard Atkins, associate medical examiner for Suffolk County, performed an autopsy the morning of December 31, just following the death. He found many abrasions on the surface of the victim’s body, some infected and showing pus. Examination of the spine revealed a dislocation of the third and fourth vertebrae at the neck, with a gap between them. The examiner opened the skull and found evidence of hemorrhaging throughout the occipital region (back of head). The undersurface of the brain was covered with a green pus — a bacterial meningitis had set in. The lungs showed grayish granularity, indicating bronchopneumonia.
The primary cause of death, according to Dr. Atkins, was the cervical injury resulting from blunt force injury to the neck or head; the hemorrhaging was consistent with such a blow. The blunt force trauma, with ensuing extensive paralysis of functions, would probably itself have been sufficient to cause death. Pneumonia and meningitis were secondary or contributing causes of death. The victim would not have developed pneumonia except for the injury to the neck, and the same injury made her susceptible to meningitis infection. In the witness’s opinion, bacteria probably entered into the victim’s system through abrasions of the body, first infecting the abrasions, and then spreading to the area of the brain, resulting in the inflammation of the meninges called meningitis. The abrasions were consistent with the victim’s being dragged or scraped, but the condition of her body was not consistent with her being run over by an automobile.
Dr. Clyde Lindquist, a general surgery resident, had the treatment of the victim from about the time of her admission to the hospital. The injury to the cervical vertebrae (he had placed it lower than C3-C4) paralyzed movement of [570]*570her arms and legs, and destroyed her ability to breathe except with mechanical assistance. Dr. Lindquist attributed death to the spinal cord trauma. The victim had been treated generally for pneumonia — rather “congestion” than pneumonia. This would not have developed in the absence of the major difficulty, the neck injury. As to the meningitis discovered by autopsy, the witness said a spinal tap, made when the victim was first placed in cervical traction, was clear, without purulence; thus the meningitis must have arisen later. Undiscovered, it had not been treated. Referring to the abrasions, Lindquist said they were like “mat burns,” with the skin rough and abraded.
(b) Evidence for the defense. Medina testified thus. On the afternoon of December 25, the victim, Lagares, and Lydia Perez appeared at Carmen Amaro’s apartment at 26 Walden in search of Soto and Juanita Amaro. Learning that Soto was not there, they went downstairs and broke windows and did other mischief. Medina, Juanita Amaro, Carmen Amaro, and Virginia Gomez, who had been in the apartment, ran downstairs, and an altercation followed, but Medina said he acted as peacemaker. He agreed Juanita had hit Lagares with a shoe.
Medina said that around 11 p.m. he was attending a Christmas party at the same apartment. Soto, Juanita Amaro, Carmen Amaro, Virginia Gomez, Maribella Muniz, and others were there. The victim and her boyfriend (not otherwise identified) appeared in the hallway and banged on the door. The victim was carrying a small axe, and her boyfriend, a baseball bat. Thereupon those at the party chased the intruders downstairs to the street. The exceptions were Soto and Juanita Amaro, who remained upstairs. (Soto had arrived at the party about 7:30 p.m.)
On the street, the victim struck Maribella Muniz in the leg with the axe, whereupon Muniz hit the victim with a mop she was carrying and pounced on the victim and repeatedly hit her in the face with her hands (she wore many rings). They struggled on the ground. Medina said the boyfriend swung at him with the bat but, missing, hit the [571]*571sidewalk and broke the bat in two. Lydia Perez picked up a piece of the bat and pursued Virginia Gomez; Medina wrested it from her and threw it down. Eventually Maribella was pulled off the victim who was on her back, her face bleeding. She was left lying in the street. Medina said Lagares was not at the melee.
Medina and others returned to Carmen Amaro’s apartment. Hearing Maribella Muniz scream, Medina looked out the living room window. He saw a car pushing the victim as she lay inert. The driver, unidentified, got out, walked to the front and looked at the body, returned, reversed, and left. Thereafter the police came.
Medina called as witnesses Maribella Muniz, Juanita Amaro, Carmen Amaro, and Virginia Gomez. All gave testimony in some measure consistent with or supportive of Medina’s account, but there were contradictions and variations. As to the incident at 11 p.m., Maribella Muniz said she struck the victim in the shoulder with a mop and “rumbled” with her on the ground.1 The victim cut her in the leg with an axe. (She sought treatment the next day, but reported the cut as accidental, resulting from a fall on stairs.) She pictured Medina as having got hold of the bat (originally, in her version, belonging to the victim’s boyfriend) . He struck Lydia Perez on the head with it as she attacked him. Then he struck the victim in the face with the bat perhaps twice, but whether it broke on the second stroke, or thereafter, she could not say. No one else struck the victim with the bat or had it in hand after Medina. She, singularly, described in detail the car that pushed the victim, first as a “ [r]ed, black top convertible” with the top up, then as a ’67 Chevrolet red and white convertible with black tape covering the top. And she could describe the driver as a big colored man with a medium Afro. Carmen Amaro asserted that Medina was not in her apartment at 4 p.m.; Juanita Amaro said Medina had not gone outside during [572]*572that episode. Virginia Gomez, returning to the apartment from the 11 p.m. fight, heard no scream and did not see a car pushing the victim.
Johnny Mae Hinton, unrelated to the street brawl, said that from her kitchen window at 42 Walden she saw a car pushing a body that had been lying in the street. She did not see the driver and did not observe anyone get out of the car with a rope. She had told this only to one Tyrone Smith (see below at point 3).
Soto testified that he had been on a visit to Lawrence on December 25 and did not return to Boston until about 7 p.m., and then went to Carmen Amaro’s party at 26 Walden. About 11 p.m. there was a commotion at the door of the apartment and he heard Carmen say a “gang” was there. (He had learned from Carmen that someone was there during the day and he claimed to have had trouble the day before from the victim and her alleged boyfriend.) Immediately he and Juanita went to a bedroom in the apartment with windows faced away from Walden Street, extinguished the light, and closed and latched the door. This left some children in the apartment. Soto and Juanita stayed in the bedroom for about an hour and a half, and were totally unaware of anything that happened outside the room during that period. For instance, Soto was not aware that policemen had appeared and questioned Maribella Muniz and Medina. When Soto emerged from the bedroom and entered the living room, he found Maribella tending her bleeding leg, and only then did he learn of recent events.
A neurologist, Dr. Donald Green, was offered by Medina as a specialist witness. (Soto offered no medical expert.) The witness had read the hospital record and the autopsy report. Answering a lengthy hypothetical question apparently summarizing a defense theory of the facts — therefore omitting mention of blunt force injury to the head or neck (but including the fact of cervical dislocation) — the witness said the cause of death was pneumonia and bacterial meningitis (the victim’s resistance having been reduced by her high alcoholic measure when she was admitted to the hospi[573]*573tal); and while stating that the infection could enter through cuts of the skin, the witness suggested that the meningitis could have been present earlier: the dislocation of vertebrae could have blocked the passage of fluid from the infected areas above, and only clear fluid would be drawn by the spinal tap. In answer to the judge’s direct question, Dr. Green said he had no opinion whether the pneumonia and meningitis were secondary to blunt force trauma in causing death.
2. Motions for Directed Verdicts. The trial judge denied motions on the part of each defendant for a directed verdict of acquittal made at the close of the Commonwealth’s case and again when all the evidence was in. At each of these testing points the case as then made is to be viewed in a light favorable to the Commonwealth (see Commonwealth v. Sandler, 368 Mass. 729, 740 [1975]), indulging reasonable inferences in that behalf (see Commonwealth v. Dunphy, 377 Mass. 453, 455-456 [1979]). By the conventional standards, with issues of credibility assumed to be resolved within reason for the Commonwealth, the judge’s rulings on the motions must be held correct.
(a) Medina. As to Medina there was evidence of premeditation, illustrated by Medina’s arrival at the apartment of Carmen Rodriguez, knife in hand, looking for the victim, some two hours before the affray of 11 p.m. At this latter encounter Medina was found striking the victim repeatedly with a bat with a force sufficient to dislocate the cervical vertebrae. Medical testimony then connected the injury to the death as principal cause. It did not matter that disease came in as a secondary or contributing factor. See Commonwealth v. Hackett, 2 Allen 136, 142 (1861). See also Commonwealth v. Golston, 373 Mass. 249 (1977); Commonwealth v. Costley, 118 Mass. 1, 27 (1875); R. Perkins, Criminal Law 708-710 & n.61 (2d ed. 1969).
The alternative basis for a first degree conviction, “extreme atrocity or cruelty,” was sufficiently shown if one considered only the beating and subsequent paralytic suffering; but there was evidence, also, that Medina participated [574]*574in the “dragging” incident with the unresisting body of the victim.
(b) Soto. The evidence did not implicate Soto in the beating with the baseball bat but did involve him in the dragging incident which, according to medical testimony, led to the pneumonia and bacterial meningitis. Accepting that these conditions hastened death as a contributing although not primary cause, there was enough to warrant sending the case to the jury on a hypothesis of murder in the second degree. See R. Perkins, Criminal Law 699-700 (2d ed. 1969). See also State v. Weston, 155 Or. 556, 579 (1937); Payne v. Commonwealth, 255 Ky. 533, 541 (1934). There is nothing in the contention that “malice aforethought” was not sufficiently shown. See Commonwealth v. Mangum, 357 Mass. 76, 85 (1970); Commonwealth v. Campbell, 375 Mass. 308, 312 (1978). We need not press an inquiry whether there was an adequate basis for the judge’s submitting the case against Soto to the jury on theories of murder in the first degree, for he was acquitted of that charge, and he could not demonstrate that the submission was itself prejudicial. See Commonwealth v. Amazeen, 375 Mass. 73, 81 n.6 (1978).
3. Withholding of Exculpatory Evidence (Medina). In October and December, 1976, Medina and Soto moved respectively for new trials. Medina’s main claim was that the prosecutor had unlawfully withheld from him exculpatory evidence; Soto, that the evidence to be adduced was newly discovered. The motions were heard together. (Soto evidently does not now assert that the judge erred in denying his motion.) One Perry Hooks testified that shortly after the incident of December 25, 1974, he told Detective John Lydon, investigating the homicide, about certain things that he claimed to have witnessed. The story (combining Hooks’s affidavit and testimony) was that around 11:30 that night he had seen two Spanish women fighting, and a man swinging a small bat at one of them; and, further, that he saw a 1966 or 1967 gray or blue-gray Buick Riviera automobile with a white vinyl top, driven by Gary Beasley (not [575]*575available) of 18 Walden Street, coming down the street toward the figure of a woman. He had not seen the impact, but he heard his companion Tyrone Smith exclaim, “He hit her.” As Hooks turned, he saw that the body had been moved some twelve feet. He did not see any rope. Smith had since died by drowning.
Detective Lydon testified that he told the prosecutor the substance of Hooks’s story but made no record of it; neither did the prosecutor. The prosecutor told the gist to Soto’s attorney who then interviewed Hooks; but Hooks said he would not testify,2 and Soto did not call Hooks as a witness at the trial. The prosecutor believed that Soto’s counsel was cooperating with Medina’s counsel and assumed that what he told Soto’s attorney sufficed for Medina. The belief and assumption were mistaken. To Medina’s attorney the prosecutor mentioned the name of Hooks as having possibly helpful information (and perhaps gave Hooks’s address), but he did not describe the substance of the story. Medina’s attorney made an effort to locate Hooks before trial, but failed. (In fact Hooks lived next door to Soto and was friendly with him.) Hooks gave an affidavit to Soto’s attorney in August, 1976, and at that time Medina’s attorney first learned what Hooks had said. The new-trial motion was filed in December, 1976.
The prosecutor should have conveyed all he knew to Medina’s attorney, rather than rely on an assumed equation between counsel for the two defendants. And we think the prosecutor’s duty was not diminished by the circumstance that the sound practice of making a written record of information received was not followed here. See Commonwealth v. Gilbert, 377 Mass. 887, 892-894 (1979); United States v. Joseph, 533 F.2d 282, 286 (5th Cir. 1976). On the other hand, counsel, having failed to lay hold of Hooks, might have been expected to go back and inquire further of [576]*576the prosecutor. Cf. United States v. Stewart, 513 F.2d 957, 960 (2d Cir. 1975); United States v. Miles, 483 F.2d 1372, 1374 n.2 (8th Cir. 1973), vacated on other grounds, 415 U.S. 970, conformed, 498 F.2d 394 (8th Cir.), cert. denied, 419 U.S. 1021 (1974). Charging the prosecutor, however, with the full responsibilities of Brady v. Maryland, 373 U.S. 83 (1963), as extended in United States v. Agurs, 427 U.S. 97 (1976),3 we think Medina’s motion was still properly denied.
The experienced trial judge heard the motion and ruled that Hooks was not a credible witness and refused to accept his testimony “in any respect.” The judge went on to rule that the testimony was in any event “merely cumulative to other testimony offered at trial” and “[i]n the context of the entire record, the omission did not create a reasonable doubt that did not otherwise exist.” See Agurs, supra at 112-113; Commonwealth v. DeChristoforo, 371 Mass. 26, 35 (1976). We attach much importance to these opinions of one having full first-hand knowledge of the cases. And we are ourselves persuaded that Hooks’s testimony (if he could have been induced to testify) would not have raised “a sufficently significant possibility of a different verdict to justify a retrial.” Commonwealth v. Pisa, 372 Mass. 590, 596, cert. denied, 434 U.S. 869 (1977). See Commonwealth v. Adrey, 376 Mass. 747, 754 (1978). With respect to the beating with the bat, which was central to the case against Medina, Hooks’s story is, if anything, confirming, for he claims to have seen such an action, although he does not name the actor. As to the car, his story is merely cumulative of testimony by other witnesses (see United States v. Agurs, supra at 114; Smith v. United States, 364 F. Supp. 1145, 1150 [E.D. Va. 1973]); in particulars, indeed, it contradicts that testimony: thus his description of the car differs from that offered by Maribella Muniz.
[577]*5774. Instructions. Medina on his appeal has not argued that there was any misdirection by the judge. By motion for a new trial filed in October, 1979 (and denied in November, 1979), Soto attempted to attack the charge on grounds not taken by either defendant at trial; he pointed to claimed errors in the instructions on “malice” and in the use of “finding” language. These claims might have some bearing on Medina as well as Soto. We examine them pursuant to our special duty expressed in G. L. c. 278, § 33E.
The passage on malice (set out in the margin)4 was so phrased, Soto argues, as to engross manslaughter, so that a jury who, properly instructed, might have found manslaughter, could have been misled into finding murder in the second degree. We think the criticism is overdrawn, especially when account is taken of the judge’s remarks specifically directed to manslaughter. See generally Commonwealth v. Vanderpool, 367 Mass. 743, 746-747 (1975); Commonwealth v. Gagne, 367 Mass. 519, 524 (1975).
Next, the judge referred to a “presumption” of malice arising from an unlawful killing and this might be thought to have eased improperly the Commonwealth’s burden of proof, and thus offended Mullaney v. Wilbur, 421 U.S. 684 (1975) , and Commonwealth v. Rodriguez, 370 Mass. 684 (1976) .5 The introduction of language of presumption is regrettable (see Sandstrom v. Montana, 442 U.S. 510 [578]*578[1979]; Commonwealth v. Collins, 374 Mass. 596, 600 n.2 [1978] ), but “whether a defendant has been accorded his constitutional rights depends upon the way in which a reasonable juror could have interpreted the instruction” (Sandstrom, supra at 514), and so error is avoided if the charge, read as a whole, makes clear the Commonwealth’s burden. See Gibson v. Commonwealth, 377 Mass. 539, 540-541 (1979) ; Commonwealth v. McInerney, 373 Mass. 136, 149 (1977); Commonwealth v. Johnson, 372 Mass. 185, 192 (1977). Here the judge did once use presumption in lieu of “inference” but in the same sentence he instructed properly that malice may be inferred. See Commonwealth v. Campbell, 378 Mass. 308, 312 (1978). (In another part of the charge he dwelt on the meaning of inference.) Moreover, the judge by repeated mention left no doubt that the Commonwealth had the burden of proving “every essential element” of the crime charged “beyond a reasonable doubt,” and malice aforethought was plainly stated to be an element of the crime of murder; nowhere was the Commonwealth relieved of its burden.6 The situation resembles Commonwealth v. McInerney, supra, where “presumption” was used interchangeably with “inference,” but it was otherwise evident from the instructions that the Commonwealth carried the burden of proving malice, and the defendant had no onus to disprove it.
Nor is there merit in the suggestion that the judge misplaced the burden of proof by using “finding” language. In Connolly v. Commonwealth, 377 Mass. 527 (1979), and Commonwealth v. Rodriguez, supra, we were critical of phrases such as “if you find” when followed by the defendant’s theory of the case. In the instant case the “finding” language was connected with the elements of the case that were upon the Commonwealth to prove. In the one in[579]*579stance of a dubious use of “finding” language, the judge recovered immediately by insisting on the Commonwealth’s burden.7
The defendant Soto at trial asked the judge to charge on alibi. Over exception, the judge declined to do so. There are jurisdictions in which it is considered reversible error to refuse an instruction on alibi when the facts provide an occasion for it. See United States v. Marcus, 166 F.2d 497, 503-504 (3d Cir. 1948). We have not taken that position. It is surely not an error to give the instruction on suitable evidence, and it may be preferable to do so;8 but it cannot be counted a mistake to omit the charge, if it is otherwise made clear that the burden of showing that the defendant was present at the time and place, and thus capable of committing the crime, remains on the Commonwealth. See Commonwealth v. Rodriguez, supra at 689; Commonwealth v. Leaster, 362 Mass. 407, 416-417 (1972); State v. Garvin, 44 N.J. 268 (1965). In the present case the situation was not altogether auspicious for an alibi instruction, as Soto by his own account was hard by the scene (compare People v. Thomas, 393 Ill. 573 [1946]; State v. Ovitt, 126 Vt. 320 [1967], with State v. Post, 255 Iowa 573, 585-586 [1963]; State v. Malpass, 266 N.C. 753 [1966]); but, granting that such an instruction would not have been inapposite, we think that, under the instructions that were given, with their insistence that the Commonwealth fully prove the offense, no juror could have supposed that it was up to the defendant to establish that he was elsewhere at the time. The situation of alibi is different from self-defense where a [580]*580defendant typically admits, at least arguendo, his presence on the spot at the time and his use of violence, but suggests justification: there the jury can slip easily into the assumption that the defendant carries the burden of demonstrating the excuse, unless they are instructed firmly and carefully to the contrary; asseveration of the general proposition that the Commonwealth must prove its case is not enough. See Commonwealth v. Rodriguez, supra at 689-692.
5. Further § 33E Considerations. In a “capital” case, as therein defined, § 33E casts on this court the duty of considering the “whole case” on the law and the evidence. Upon such consideration we are empowered to order a new trial or reduce the degree of guilt on being satisfied that the verdict was “against the weight of the evidence,”9 “or for any other reason that justice may require.” See Commonwealth v. Seit, 373 Mass. 83, 93-95 (1977); Commonwealth v. Pisa, 372 Mass. 590, 597-598 (1977); Commonwealth v. Baker, 346 Mass. 107, 108-109, 118-119 (1963). See also Commonwealth v. Ellison, 376 Mass. 1, 17 (1978).
We have examined the present record with much care and described it at length at point 1 above so that the difficulties might be appreciated. The prosecution’s evidence supported a general hypothesis of the critical events, the evidence on the part of thé defendants a contradictory general hypothesis, and the jury can be taken to have adopted the former. As indicated at point 2, the condition of the proof was such that the judge rightly held there was enough against the defendants to warrant passing their cases to the jury. It is another question whether on examining the whole case pursuant to § 33E we come out satisfied that the verdicts reached by the jury were well supported and should stand as just.
In the case of the defendant Medina, on reflection we see no reason to interfere with the first degree verdict. Concentrating on his responsibility for the beating, which very [581]*581probably led direct to the death, we find not only the testimony of Ama Lagares, supported by Luis Martinez, that Medina struck with the bat, but like testimony by Maribella Muniz, a witness called by Medina. We are not inclined to discount Maribella’s testimony on a theory that, fearful of being seriously charged herself, she threw off falsely on Medina. Medina’s own testimony has him taking the bat (or part of it) in hand, although then supposedly throwing it down.
The case made against Soto is more dubious (there is irony in this because, of the two, Soto was nearer the quarrel that we can sense to be at the heart of the crime). Soto is not shown to have been present at 4 p.m. at Carmen Amaro’s apartment, or at 9 p.m. at Carmen Rodriguez’s. Except for Luis Martinez, who claims to have seen Soto at the 11 p.m. beating incident (but does not say that Soto participated), no one places Soto there; besides Soto’s own denial, there is testimony flatly or inferentially negative from Medina, Maribella Muniz, Virginia Gomez, Carmen Amaro, and Juanita Amaro.
The testimony of Lagares10 involves Soto in the dragging episode together with a second man, unidentified, who evidently did not emerge from the car. The car is not described in any detail. An unsatisfactory witness, Luis Martinez, who, despite denials, may well have rehearsed the matter with his aunt Lagares, also involves Soto in the dragging, but with the not inconsiderable difference that Medina is the second man, and both are concerned with the rope.* 11 But no rope fragments are found by the criminalist on the victim’s garment. Assuming that abrasions on the body, as well as the condition of the garment, are consistent with dragging or scraping, there is little if anything to suggest that these could not have resulted from the “pushing” by a vehicle under conditions testified to, although variously and with contradictions, by several defense witnesses. The [582]*582“pushing” in fact is also testified to by Hinton, an outsider; but she is not an impressive witness. The “pushing” may be a figment, but there is a fair chance that so is the “dragging.”
Finally, with respect to meningitis as a secondary cause of death, there is indeed testimony that the bacterial infection came through the abrasions, but no medical expert was called on Soto’s behalf who might have suggested an alternative source; the kind of bacteria involved was not identified as no culture was made.
It is not easy to believe that Soto immured himself for an hour and a half and was oblivious of the exciting events outside 26 Walden. But it can be said for Soto that he was not impeached by a previous criminal record, as Medina was. We may add that communications from the jury (who were out for three days and were ultimately encouraged to a decision by the Tuey-Rodriguez instruction) may be interpreted as showing particular concern about Soto.
We leave to the end that Soto moved to have his trial severed from Medina’s and an exception was saved from the denial. Of course there was no error, for the matter is generally discretionary with the trial judge. But it remains possible that, as the trial developed, Soto was hurt by connection with Medina, despite efforts by the judge to have the jury keep the defendants distinct.
In all the circumstances we consider that it is in the interest of justice that Soto be granted a new trial.
As to Medina judgment affirmed.
As to Soto judgment reversed and verdict set aside.