delivered the opinion of the court:
After a jury trial John Smith and George Ross were found guilty of the crime of murder. They appeal from that conviction on the grounds that their guilt was not proven beyond a reasonable doubt and that certain comments by the Assistant State’s Attorney during closing argument denied them a fair trial.
The following facts in this case have not been disputed. During the late evening of February 23, 1969, defendant, George Ross, hosted a party in his second floor apartment at 5936 South State Street, Chicago, Illinois, and John Smith, his half-brother, attended. Approximately twenty or thirty people, male and female, were at the party. There apparently was a considerable amount of drinking and dancing during the evening of the 23rd and early morning of the 24th. After the party ended, approximately eight or ten people remained to sleep at Ross’ apartment. At 5:30 A.M. on February 24th a Chicago police patrolman was directed by one Margaret Guest to an alley behind 5936 South State Street, where he found the dead body of a black male, later identified as Maurice Parrett. Parrett had suffered four stab wounds in the chest area. A few minutes after the policeman arrived on the scene, he was approached by defendant, John Smith. They had a conversation during which Smith said he had discovered the body earlier while walking his dogs and was then enroute to a public phone to call the police. Detectives questioned Smith again two hours later at his mother’s first floor apartment at 5936 South State Street and later took his statement at the police station. Detectives also entered George Ross’ apartment and questioned the occupants. One of those questioned was Melvin Crumpton, who stated that he knew nothing about Maurice Parrett’s murder. There were apparently no suspects until two months later, when Melvin Crumpton gave the police a statement materially different from the one given on February 24th. In his later statement he claimed to have witnessed the defendants, between 3:00 A.M. and 4:00 A.M., February 24th, in the act of beating a man and dragging him toward the alley where the deceased was subsequently found. Smith and Ross were then arrested, indicted and convicted for the murder of Maurice Parrett.
The defendants’ primary contention on appeal is that they were not *66proven guilty beyond a reasonable doubt. In support for this contention they claim that the testimony of the prosecutions sole occurrence witness was contradictory and impeached sufficiently to severely undermine its credibility. They further claim the benefit of an evidentiary inference and conclude that this inference, combined with the lack of credible evidence adduced to support the States circumstantial case, required a not guilty verdict as a matter of law.
The State’s case in chief consisted almost entirely of the testimony of Melvin Crumpton. He was twenty years old at the time of the occurrence and had known both defendants for ten years. On direct examination Crumpton testified that he went to a party at George Ross’ apartment at approximately 9:30 P.M., February 23, 1969. During the course of the evening and early morning he left the party four times, returning each time. He left for the fourth time at 3:00 A.M., February 24th, for the purpose of purchasing liquor. As he left, he observed the defendants holding “a man # # on the floor,” slapping him and going through his pockets. Crumpton returned from his errand in ten minutes and proceeded to the back stairs where he observed the defendants “tugging and puffing this man down the stairs” and toward the back of the house. A crowd of people was following the defendants. Crumpton returned upstairs where, a short time later, he again saw the defendants. They were in “the back room on the way upstairs” and were arguing. Smith was holding a knife on which there appeared to be blood. Crumpton later overheard the defendants say they threw the knife out the living room window. He then went to sleep, as the party was breaking up.
On cross-examination Crumpton said that he did not know Parrett on the night in question, but further cross-examination revealed that he played pool with Parrett and others that night and returned to the party with them at 1:00 A.M. He estimated that eighteen boys from the party had foHowed the defendants as they dragged Parrett down the back stairs. He admitted that he told the police who arrived on the morning of the 24th that he knew nothing and offered as an explanation for this silence the fact that he was scared. He further admitted to having consumed approximately ten drinks during the night in question.
Defendants have emphasized several weaknesses in this testimony. First, Melvin Crumpton’s testimony at trial is totally at variance with the story he told investigating detectives on February 24th. Although he sought to explain this inconsistency by the fact that he was “scared,” he did not elaborate, and the record is unrevealing as to Crumpton’s motive for changing his story a full two months later. Second, Crumpton testified on direct examination that he had never before seen the man who was being assaulted by the defendants, but he stated on cross-examination *67that he had played pool earlier that night with Maurice Parrett. Third, Crumpton testified that he, Parrett and others arrived at the party together at 1:00 A.M. However, during the prosecutions case in chief Detective Francis Healy said that his investigation had revealed that Maurice Parrett did not leave his place of employment on February 24th until 3:00 A.M. Fourth, Crumpton admitted consuming approximately ten drinks during the night in question.
Our own analysis of the record has revealed other weaknesses in the testimony of Melvin Crumpton. The only evidentiary link between the defendants and the deceased was Crumptons testimony regarding their alleged assault on Parrett. However, at no time during his direct examination did Crumpton positively identify the victim of defendants’ assault as the deceased, and it was only through his responses on cross-examination that the name “Maurice Parrett” became associated with the “man” referred to on direct. To the extent that these responses can be deemed as “identification,” they present a weak identification with a total lack of foundation. Defendants took the stand in their defense and contradicted or denied each material element of Crumpton’s testimony.
The defendants have also urged that, in the course of weighing the legal sufficiency of the evidence, this court should consider an inference stemming from the evidence which was presented. They suggest that the State’s failure to call other occurrence witnesses to corroborate the chief witness gives rise to an inference that the other testimony would have been damaging to the State’s case. Clearly, such an inference has been recognized in Illinois under some circumstances. (People v. Williamson, 78 Ill.App.2d 90; People v. Gonzales, 125 Ill.App.2d 225; People v. Di Vito, 66 Ill.App.2d 282.) However, the rule is equally well settled that the State is not obligated to produce every witness to a crime. (People v. Jones, 30 Ill.2d 186; People v. Graham, 127 Ill.App.2d 272.) The Graham case is illustrative of those cases which have cited the limited obligation of the prosecution. The primary witness for the prosecution in Graham had actually witnessed the shooting for which defendant had been charged, and he testified directly as to that event. His testimony was “clear and convincing” and “complete in all respects.” Further, it was corroborated by the testimony of several police officers. By contrast, the prime witness in the case at bar was not a direct occurrence witness, but testified only to events which circumstantially implicated defendants in the stabbing death of Maurice Parrett. The testimony was contradictory, impeached and hardly “complete in all respects.” The substance of Crumpton’s testimony was uncorroborated. In addition, the record reveals the existence of approximately eighteen other witnesses to the events related by Melvin Crumpton; it also reveals the State’s *68knowledge of many of their names and addresses. Not one of these witnesses was called to support the evidence provided by a lone witness who had consumed ten drinks on the night in question. We therefore conclude that the totality of the circumstances in this case warrants invoking the inference that the witnesses available to the State, if called, would have testified adversely to the prosecution.
It is our duty in the case at bar to carefully examine the evidence, while giving due consideration to the fact that the court and jury saw and heard the witnesses. If, after such consideration we are of the opinion that the evidence is insufficient to establish the defendants’ guilt beyond a reasonable doubt, it is our duty to reverse the judgment of guilty. (People v. Bartley, 25 Ill.2d 175, 180.) Even where the credibility of witnesses is the primary issue, we will nonetheless disturb the jury’s finding of guilty if the evidence is so unsatisfactory as to leave a reasonable doubt as to defendants’ guilt. People v. Glover, 49 Ill.2d 78, 84, 85; People v. Pellegrino, 30 Ill.2d 331.
The only evidence tending to support the conviction of John Smith and George Ross was the testimony of Melvin Crumpton. That testimony constituted circumstantial evidence. It was weak, contradictory, impeached and refuted by the defendants. Further, it was uncorroborated, even though the State was aware of numerous other occurrence witnesses.
The weakness of this evidence compels us to comment at this juncture on the conduct of both prosecution and defense attorneys. From the beginning, the transcript is rife with personal attacks and appeals to passion totally unwarranted by the circumstances and the evidence adduced at trial. This conduct was particularly manifest during the course of closing arguments, and has, in fact, been cited in defendants’ alternative request for a new trial. However, we need not decide that issue. We feel that the weakness in the State’s case, coupled with exceptionally impassioned presentations during closing arguments requires the same disposition as found in People v. White, 347 Ill. 576, 585, 586:
“While this court is committed to the doctrine that the jury in their deliberation are judges of the facts and of the weight of the evidence in criminal cases, still this court will reverse a judgment of conviction where the evidence on which it is based was of such unsatisfactory character that after a full consideration there remains such grave and serious doubt of the guilt of the accused as leads to the conclusion that the verdist of the jury was the result of passion or prejudice.” Accordingly, we reverse.
Judgment reversed.
LEIGHTON, P. J., and SCHWARTZ, J., concur.