TUTTLE, Circuit Judge.

Appellant Street and four other persons were indicted for interstate transportation of forged checks and conspiracy. Street was tried first, convicted, appealed, and lost. The others met the same fate later in two additional trials. All of the convictions were affirmed on appeal, appellant Street’s by an unpublished opinion of this court, Kilcullen’s and Reddall's by decision at 546 F.2d 435 (1st Cir. 1976), Dannon’s in an unpublished opinion, dated November 15, 1976, and Cowden’s by an opinion published at 545 F.2d 257 (1st Cir. 1976).

Now Street has proffered newly discovered evidence which he claims is material on the issue of his knowledge, an essential element of the offense charged. The evidence is the testimony of two co-defendants, who refused to testify at Street’s trial for Fifth Amendment reasons, and information about a witness’ bankrupt financial situation. This witness had claimed to be worth ten million dollars at the time of the criminal acts, in 1973. A motion for new trial asserts that he filed a Chapter XIII bankruptcy petition in 1976.

The motion for new trial was denied by the trial court in an opinion which found the bankruptcy evidence cumulative and impeaching and the co-defendants’ expected testimony “for the most part ... either cumulative or impeaching in nature” as well as consisting “primarily of inadmissible hearsay.” In the final paragraph of the court’s opinion the court said:

Finally, I conclude that for the reasons stated above, the proffered evidence, for the most part, is either cumulative or *2 impeaching in nature and thus does not constitute newly discovered evidence, and, more importantly, having in mind that three different juries convicted the five defendants involved in this case, I rule that, in any event, it is not evidence of a caliber which would “probably produce a different verdict in the event of a new trial.” United States v. Strauss, 443 F.2d 987.

The government seems to concede that even though the trial court seemed clearly to base its decision to overrule the motion for new trial on several grounds, the inclusion of the “probably produce a different verdict” ground must be squarely faced by this court. This results from the fact that the trial court did not definitely finally determine that the motion for new trial failed to meet the other required standards.

The appellant concedes in its brief:

The overwhelming weight of authority has established very definite criteria that must be satisfied in order for a defendant to succeed on a motion for new trial. As set forth in the district court’s memorandum these include:

1. The evidence must be newly discovered or unavailable at the time of trial.

2. Facts must be alleged from which the court may infer diligence on the part of movant.

3. The new evidence relied on must be material.

4. The new evidence must not be merely cumulative or impeaching.

Although the appellant describes these four criteria as being “set forth in the district court’s memorandum,” he does not state them accurately. We prefer to set out the criteria as actually stated by the trial court, since this is what the appellant purports to rely upon in his brief here. The trial court stated:

It is settled law that in order to succeed on a motion for new trial based primarily on the grounds of newly discovered evidence, the movant-defendant must show (a) that the evidence was unknown to him at the time of trial, United States v. Granger, 504 F.2d 585 (9 Cir. 1974); (b) that the failure to discover the evidence was not due to a lack of diligence, United States v. Schwartzbaum, 527 F.2d 249 (2 Cir. 1975); (c) that the new evidence is material, United States v. Giglio, [Giglio v. United States] 405 U.S. 150, 92 S.Ct. 763, 31 L.Ed.2d 104 (1972); (d) that the new evidence is neither cumulative nor impeaching, Mesarosh v. United States, 352 U.S. 1, 9, 77 S.Ct. 1, 1 L.Ed.2d 1 (1956); (e) that the new evidence would probably produce an acquittal, United States v. Agurs, 427 U.S. 97, 111 n. 19, 96 S.Ct. 2392, 49 L.Ed.2d 342 (1976); United States v. Schwartzbaum, supra. Motions for a new trial are directed to the discretion of the trial court. Both the weight of the new evidence and the credibility of witnesses are to be considered by the court. United States, v. Zannino, 468 F.2d 1299, 1303 (1 Cir. 1972), cert. denied, 410 U.S. 954, 93 S.Ct. 1419, 35 L.Ed.2d 687 (1973).

It will be noted that the appellant did not include the fifth criterion as used by the trial court, but as to that the appellant states: “These four criteria are universally cited in almost all cases. A fifth test is also usually applied. That test is the ‘probability test.’ ” The fact that it is usually applied, as stated by the appellant, is fully confirmed by the Supreme Court’s statement in United States v. Agurs, 427 U.S. 96, 111, 96 S.Ct. 2392, 2401, 49 L.Ed.2d 342 (1976). Although the precise question as to the standards to be applied in a motion for new trial was not before the Supreme Court in Agurs, the Court there spoke of “the severe burden of demonstrating that newly discovered evidence probably would have resulted in acquittal.” This test is then described in a footnote in the following terms: “This is the standard generally applied by lower courts in evaluating motions for new trial under Fed.Rule Crim.Proc. 33 based on newly discovered evidence.” Id. at 111 n. 19, 96 S.Ct. at 2401.

Appellant correctly notes that several courts of appeals have established somewhat differing standards as to this fifth criterion, citing United States v. Strauss, *3 443 F.2d 986, 989 (1st Cir. 1971), for the proposition that this court has not yet resolved the issue. In that case we said:

As the Supreme Court has noted, courts have traditionally applied two somewhat different tests for determining whether newly discovered evidence warrants a . new trial. See United States v. Johnson, 327 U.S. 106, 111 n. 5, 66 S.Ct. 464, 90 L.Ed. 562 (1946). The majority rule, derived from the case of Berry v. Georgia, 10 Ga. 511 (1851), requires that the evidence “must be such, and of such nature, as that, on a new trial, the newly discovered evidence would probably produce an acquittal.” Johnson v. United States, 32 F.2d 127, 130 (8th Cir. 1929); accord, United States v. Craft, 421 F.2d 693 (9th Cir. 1970); Hudson v. United States, 387 F.2d 331 (5th Cir. 1967), cert. denied, 393 U.S. 876, 89 S.Ct. 172, 21 L.Ed.2d 147 (1968); Anderson v. United States, 369 F.2d 11 (8th Cir. 1966), cert. denied, 386 U.S. 976, 87 S.Ct. 1171, 18 L.Ed.2d 136 (1967). The Seventh Circuit rule, first announced in Larrison v. United States, 24 F.2d 82, 87-88 (7th Cir. 1928), is somewhat less stringent. Under the Larrison rule a new trial must be granted if, on the basis of the newly discovered evidence, “(a) The court is reasonably well satisfied that the testimony given by a material witness is false [and] (b) That without it the jury might have reached a different conclusion.” Id. at 87; accord, Gordon v. United States, 178 F.2d 896, 900 (6th Cir. 1949).

Id. at 989 (footnote omitted) (emphasis in original).

This court in Strauss decided that there had been no false testimony and that therefore the Larrison rule was “inoperative.” The court held that “the record amply supports both that finding and the Court’s evaluation of the newly discovered evidence in light of the Berry v. Georgia standard.” Id. at 990.

This court has most recently dealt with this subject in our opinion this term in In Re United States, 565 F.2d 173 (1st Cir. 1977). In that case, this court considered the question whether the trial court had properly granted a motion for a new trial. We there stated that one of the standards set by Rule 33 and the pertinent law was that the newly discovered evidence “must be such as to have some effect on the outcome (the degree of such effect being as yet an issue of some disagreement among the circuits).” Thereafter, the opinion stated: “Similarly, respondent failed to establish that the outcome of the trial would likely have been different had the jury been presented with the January 4, 1972 agreement” and then elaborated in the footnote as follows:

There is some uncertainty whether the standard to be applied against Rule 33 motions is the majority rule, that the evidence “would probably produce acquittal,” Johnson v. United States, 32 F.2d 127, 130 (8th Cir. 1929), or a minority position, that it “might” have caused the jury to reach a different result. Larrison v. United States, 24 F.2d 82, 87-88 (7th Cir. 1971). See United States v. Strauss, 443 F.2d 986, 989 (1st Cir. 1971). We need not resolve that question since the evidence here fails under either standard.

Id. 177 n.3.

Thus, it is clear that where the trial court places its denial of a motion for new trial on the absence of probability of a different result, we must face the question of whether that is a correct standard. This court has said:

Motions for a new trial are directed to the trial court’s discretion. Under its broad power, the court may weigh the evidence and consider the credibility of the witnesses. The remedy is sparingly used, the courts usually couching their decisions in terms of “exceptional cases”.

United States v. Leach, 427 F.2d 1107, 1111 (1st Cir. 1970) (citations omitted). This basic standard of review as announced by this court adds weight to the government’s argument that this court should follow the clear majority of the circuits in requiring that a movant for a new trial on the ground of newly discovered evidence must convince *4 the trial court that the new evidence would “probably” bring about a different result.

The appellant contends that “following Strauss the evidence should have been reviewed to determine whether the jury ‘might have reached a different conclusion.’ ” Reference to Strauss indicates clearly that this court has construed the Larrison rule to be that a new trial must be granted if “ ‘(a) the court is reasonably well satisfied that the testimony given by a material witness is false [and] (b) that without it the jury ‘might have reached a different conclusion.’ ” Here, the motion for new trial does not allege perjury or prosecutorial misconduct as possible bases for further application of the Larrison test. In the absence of either of these factors, we find no basis for applying the more lenient test of “might,” since in fact this court in Strauss rejected Larrison as a standard where there was no proof of false swearing. 1

We conclude that the trial court did not err in overruling the motion in this case. The judgment is AFFIRMED.

. For a clear discussion of the two criteria and the extent to which they have been recognized in the several circuits, the case of United States v. Stofsky, 527 F.2d 237 (2d Cir. 1975), is particularly instructive. It also demonstrates that the Court of Appeals for the Second Circuit has abandoned its prior practice of testing the motion for new trial by both standards.