dissenting:

This eighteen day trial covering an entire month, with a transcript of testimony consisting of nineteen volumes, in addition to a clerk’s record of 490 pages, should not be overturned on what I believe to be harmless error, if any.

I.

It is my considered judgment that the procedural prohibition against a trial judge’s asking jurors their numerical standing, as stated in Brasfield v. United States, 272 U.S. 448 , 47 S.Ct. 135 , 71 L.Ed. 345 (1926), did not survive the adoption by the Supreme Court of the Federal Rules of Criminal Procedure.

Brasfield was decided by the United States Supreme Court on November 22, 1926, long prior to September 1, 1945, the effective date of F.R.Crim.P. These rules were authorized by the Act of Congress of June 29, 1940, 54 Stat. 688 , and promulgated by the United States Supreme Court by order dated December 26, 1944. They were promulgated and adopted to govern all criminal proceedings thereafter commenced and insofar as just and practicable all proceedings then pending. Rule -59, F.R. Crim.P.

Rule 52(a), F.R.Crim.P., provided and now provides:

My belief that Brasfield can no longer be considered as authoritative and that the act of a judge asking the jury its numerical standing should, at best, be viewed as harmless, is based principally on the ground: (1) that Rule 52(a), F.R.Crim.P., was not in existence at the time of the decision in Brasfield , and (2) that the Ninth Circuit, time after time, has approved the Allen 1 instruction which is far more coercive than the simple inquiry involved in Brasfield . As recently as June 28, 1978, in United States v. Seawell, 583 F.2d 416 , 417—418 (CA9 1978), rehearing and rehearing in banc denied September 27, 1978, cert. denied November 27, 1978, - U.S. -, 99 S.Ct. 591 , 58 L.Ed.2d 666 , we again gave approval to the substance of the Allen instruction. Indeed, the Allen instruction is cited with approval by the Supreme Court as recently as 1972. See Johnson v. Louisiana, 406 U.S. 356, 361-362, 380 , 92 S.Ct. 1620 , 32 L.Ed.2d 152 (1972).

The Allen instruction approved in Sea-well reminded the jury: (1) of the expense in time, effort and money to both the defense and the prosecution in the event a new trial was necessary; (2) that the case, like all cases, had to be disposed of some time and that there was no reason to believe that another trial would not be as costly to both sides as was the present trial; (3) that a future jury must be selected in the same manner as it was and that there was no reason to believe that the case would ever be submitted to twelve men and women more conscientious, more impartial, or more competent to decide the case; (4) that it was their duty to consult with one *1306 another and deliberate with a view of reaching a verdict if they could do so without doing violence to their individual conscience; (5) that each of the jurors must decide the case for himself, but that he should do so only after consideration of the evidence in the case with his fellow jurors; (6) that he should not hesitate to reexamine his own views or to change his mind if he is convinced it was originally erroneous; (7) that in order to bring twelve minds to an unanimous verdict, each juror should examine the questions submitted with candor and frankness and with a proper deference to and regard for the opinions of others. More important, on the issue of numerical divisions, the judge told the jury that “if much the greater number of you are for a conviction, each dissenting juror ought to consider whether a doubt in his or her own mind is a reasonable one, since it makes no effective impression upon the minds of so many equally conscientious fellow jurors who bear the same responsibility under the same oath and have heard the same evidence.”

I do not believe the majority would argue that the instruction in Brasfield was as coercive as the one given and approved in Seawell . The contrary is obvious. The instruction given in Seawell , if coercive at all, was far more so than that given in Bras-field. One thing is certain, the decision in Brasfield is diametrically opposed to the decisions in Allen, Seawell and those cited in Seawell . Logically, they cannot be reconciled. I would hold that Allen, approved by the Supreme Court as recently as Johnson v. Louisiana, supra (1972), is the law of the land and that the simple question propounded to the jurors in Brasfield and by the court below should no longer be viewed as reversible error. See Rule 52(a), (b).

True enough, our circuit in Spaugh v. United States, 77 F.2d 720 (CA9 1935), and Jordan v. United States, 22 F.2d 966 (CA9 1927), reversed two criminal convictions in reliance upon Brasfield . Nonetheless, each of these cases was decided prior to the adoption of the Federal Rules of Criminal Procedure and prior to our decisions in Marsh v. Cupp, 536 F.2d 1287 (CA9 1976), cert. denied 429 U.S. 981 , 97 S.Ct. 494 , 50 L.Ed.2d 590 (1976), and United States v. Williams, 444 F.2d 108 (CA9 1971).

In Marsh , the Oregon state trial «judge, prior to giving a second Allen instruction, inquired as to the numerical split of the jury. Marsh v. Cupp, 392 F.Supp. 1060, 1063 (D.C.Or.1975). The judge was informed that the jury was divided nine-three, the same division as is before us. The Marsh court concluded that Brasfield was not controlling despite the fact that the jury was given the Allen charge on two different occasions. Marsh, supra, 536 F.2d at 1291 . In Williams , the court held that the trial court was not precluded from giving the Allen instruction by reason of the fact that the jury had made an unsolicited disclosure of its numerical division. In United States v. Rao, 394 F.2d 354 (CA2 1968), cert. denied 393 U.S. 845 , 89 S.Ct. 129 , 21 L.Ed.2d 116 (1968), and Bowen v. United States, 153 F.2d 747 (CA8 1946), cert. denied 328 U.S. 835 , 66 S.Ct. 980 , 90 L.Ed. 1611 (1946), the numerical split was volunteered by the jury also. The Williams, Rao and Bowen courts distinguished Bras-field on that ground. In Marsh, however, the inquiry was made by the court.

Logically, I am unable to find a valid distinction between an inquiry by the court as to the numerical standing of a jury and the volunteering of such information by the foreman of the jury. The assumed coercive effect would be upon the minority jurors and this effect would be felt whether the foreman volunteered it or the court requested it. In each instance, the information is revealed in open court. However, in neither instance does the information reveal the standing of the jury on the issue of guilt or innocence. I completely agree with the Marsh v. Cupp district court that the decision in Brasfield was based on the court’s supervisory powers in the lower courts and not on constitutional interpretations. The Brasfield court condemned the inquiry as to numerical division on the ground that it, “. . . affects the proper relations of the court to the jury . . ” This is clearly a supervisory ruling on a *1307 procedural subject. I repeat that Bras field, as a procedural rule, was modified or overturned by the enactment and adoption of the Federal Rules of Criminal Procedure and that the Harmless Error Rule 52(a), F.R.Crim.P., should be applied to these facts.

Nor was the Bras field court confronted with the plain error provision of Rule 52(b). It is well established in this circuit that this Rule will be invoked only if it is highly probable that the error materially affected the jury’s verdict, United States v. Dixon, 562 F.2d 1138, 1143 (CA9 1977), cert. denied 435 U.S. 927 , 98 S.Ct. 1494 , 55 L.Ed.2d 521 ; United States v. Segna, 555 F.2d 226 (CA9 1977), or if it appears to be necessary in order to prevent the miscarriage of justice or to preserve the integrity and reputation of the judicial process. United States v. Larson, 507 F.2d 385, 387 (CA9 1974); United States v. Trejo, 501 F.2d 138, 141 (CA9 1974). 2

Rules 1 and 2, F.R.Crim.P., provide that the rules will govern the procedure in all criminal proceedings in the courts of the United States and were designed to simplify existing procedure and to eliminate outmoded technicalities. United States v. Debrow, 346 U.S. 374, 376 , 74 S.Ct. 113 , 98 L.Ed. 92 (1953); United States v. Bickford, 168 F.2d 26, 27 (CA9 1948).

Not only did counsel for appellants fail to make an objection, 3 they failed to ask the court to recall the jury and instruct the members to disregard and put out of their minds all questions by the court with reference to the numerical standing of any one of appellants. It would seem that counsel and defendants were as anxious to know the numerical standing of the jury as was the court. In these circumstances, our decisions interpreting the plain error provisions of Rule 52(b), United States v. Dixon, supra; United States v. Trejo, supra; and the Fifth Circuit case of United States v. Jacquillon, 469 F.2d 380 (CA5 1972), cert. denied 410 U.S. 938 , 93 S.Ct. 1400 , 35 L.Ed.2d 604 (1973), should be followed. It appears from the record that trial counsel for appellants were exceptionally able and resourceful lawyers and they should not be permitted to gamble for a favorable verdict and then resort on appeal to errors that might have easily been corrected by objection at trial. United States v. Jacquillon, supra, at 386.

II.

Appellants’ seriously contend that the district judge violated the provisions of Rule 43, F.R.Crim.P., when he assembled the jury in the absence of counsel and defendants and asked the jury foreman to report the jury’s numerical division as to each de *1308 fendant without indicating which party was in the majority. After supplying the court with this information, the jury retired and deliberated two additional hours before reconvening the following morning. On the third day of deliberations, the jury was again assembled, this time in the presence of defendants and counsel. The court asked if any verdicts had been reached and the foreman responded in the negative. When asked whether the jury should be dismissed, however, the foreman stated that he believed the jury should continue with its deliberations. The judge then asked if there had been any change in the numerical standing as to each defendant. When the foreman advised that there was such a change, the court asked whether the numerical standing is “. . . more now toward a verdict or more further away from a verdict.” The foreman responded that the jury was closer to a verdict. Counsel made no objection to the exchange between the judge and the jury foreman. Instead counsel asked the court to reveal in open court the numerical standing of the jury on each defendant as outlined to the court on the previous occasion when defendants and counsel were not present. (See Footnotes 2 & 3, supra.) No Allen -type instruction was given.

Inasmuch as I would hold that Brasfield is no longer controlling, I am led to the conclusion that the absence of the appellants and their counsel at the time when the court first inquired as to the numerical standing of the jury as to each appellant, while a technical violation of Rule 43, was harmless and did not affect the substantial rights of the appellants under Rule 52, F.R. Crim.P.

It is uniformly held that Rule 43, F.R. Crim.P., requiring the presence of defendants at all stages of the trial, must be read in the light of the harmless error rule as stated in Rule 52(a), F.R.Crim.P. United States v. Gradsky, 434 F.2d 880, 884 (CA5 1970), cert. denied 401 U.S. 925 , 91 S.Ct. 884 , 27 L.Ed.2d 828 (1971); United States v. Schor, 418 F.2d 26, 30 (CA2 1969). If, as in United States v. Florea, 541 F.2d 568, 573 (CA6 1976), cert. denied 430 U.S. 945 , 97 S.Ct. 1579 , 51 L.Ed.2d 792 (1977), rehearing denied 431 U.S. 925 , 97 S.Ct. 2201 , 53 L.Ed.2d 240 , defendants were not prejudiced because they were absent when the district judge excused one juror, substituted an alternate, and sequestered the jury, then certainly the appellants were not prejudiced by their absence on the occasion of the judge asking simple questions with reference to the jury’s numerical division.

Moreover, I would hold that appellants did not have a constitutional right, as distinguished from a procedural right, to be present at the time and under the circumstances then and there existing. The error of the court, if any, in making the inquiries in my opinion was harmless.

III.

Prior to the beginning of the voir dire examination, appellant Noah’s counsel made a statutory challenge to the racial composition of the group of prospective jurors. After a hearing lasting in the neighborhood of seven minutes, the court denied the motion. On the morning of the hearing, the court heard certain motions and then recessed until 1:30 P.M. At 11:00 A.M. the court resumed the bench at the request of attorneys for Noah. They requested the court to reconvene, but did not notify the Marshals to return the appellants to the courtroom, nor did the attorneys make any objection to the absence of appellants or request their presence. The court did not know the purpose for which the attorneys asked it to reconvene. After reconvening, the court, for the first time, learned that the purpose was to make a motion challenging the method of selecting the entire jury panel.

Appellant Noah argues (1) that this challenge constitutes part of the empanelling process of the jury, (2) that Rule 43 thus guarantees him a right to be present, and (3) that the error was not harmless. I am convinced that this and other pretrial motions do not fall within the scope of Rule 43(a), but are governed by Rule 43(c)(3), which does not require the presence of a *1309 defendant at a conference or argument upon a question of law. Irregardless, the error, if any, was harmless under Rule 52, F.R.Crim.P.

Moreover, I feel that the appellants’ absence was, in fact, due to the failure of their attorneys to arrange for their presence.

Furthermore, the Western District of Washington has an approved plan for random jury selection made pursuant to the provisions of 28 U.S.C. § 1863 . There is no challenge to the plan nor any effort on behalf of appellant to establish that a substantial deviation from the plan with reference to a cognizable group took place. In such case the court did not commit error in denying the motion. United States v. Brady, 579 F.2d 1121 (CA9 1978), rehearing and rehearing in banc denied August 8, 1978, cert. denied -U.S.-, 99 S.Ct. 849 , 59 L.Ed.2d 41 (1979) [Montana Plan]; United States v. Potter, 552 F.2d 901 (CA9 1977) [Nevada Plan]. In the former case, we said, among other things, that a defendant challenging a grand or petit jury must prove either a systematic exclusion of an identifiable class or a substantial deviation between the identifiable groups and that before a court can make a meaningful comparison for determining a substantial deviation in a plan for grand and petit jury selections in federal court, those challenging the make up of the jury must offer substantial proof on the deviation or division basis. Brady at 1134 . No such showing was made in the record before us. This assignment of error is meritless.

Our circuit in United States v. Calaway, 524 F.2d 609 (CA9 1975), in commenting on a procedure utilized prior to the commencement of the selection of the trial jury to eliminate jurors who were obviously disqualified emphasized this eminently practicable procedure as employed by the trial judge and held that it could not reasonably be considered a part of the criminal trial.

IV.

Noah suggests that there is insufficient evidence connecting him with the conspiracy as charged in the indictment. True enough, the evidence linking him to the conspiracy is mostly circumstantial and not as clear as the evidence against the other appellants. However, it can be said that the evidence of a conspiracy is substantial. The conspiracy is established by independent evidence. Circumstantial evidence is not inherently less probative than direct evidence. United States v. Turner, 528 F.2d 143, 162 (CA9 1975), cert. denied sub nom. Grimes v. United States, 423 U.S. 996 , 96 S.Ct. 426 , 46 L.Ed.2d 371 (1975). Time and time again, we have said that we have no power to disturb a jury’s verdict where it could rationally conclude from the evidence presented and the inferences drawn therefrom that guilt was established beyond a reasonable doubt. Turner, supra, at 162 ; United States v. Irion, 482 F.2d 1240, 1248 (CA9 1973), cert. denied 414 U.S. 1026 , 94 S.Ct. 454 , 38 L.Ed.2d 318 (1973); United States v. Nelson, 419 F.2d 1237, 1241 (CA9 1969). In such a case evidence establishing beyond a reasonable doubt a connection of a defendant with the conspiracy, even though the connection is slight, is sufficient to convict him with knowing participation in the conspiracy. United States v. Dunn, 564 F.2d 348, 357 (CA9 1977).

That a conviction may be based on the uncorroborated testimony of an accomplice is settled law. United States v. Turner, supra, at 161 , United States v. Daniel, 459 F.2d 1029, 1030 (CA9 1972). Noah’s principal contention on this point is that since coconspirator Vernice Garrett was acquitted, her testimony linking him to the conspiracy should be disregarded and, consequently, he should be acquitted. Appellant cites no authority to support his contention. For that matter, the available authority is to the contrary. Kamanosuke Yuge v. United States, 127 F.2d 683 (CA9 1942), cert. denied sub nom. Mateus v. United States, 317 U.S 648, 63 S.Ct. 43 , 87 L.Ed. 522 , a case in point, holds that the fact that a jury fails to convict one defendant on the conspiracy charges does not in and of itself render testimony of that person’s acts and declarations inadmissible as against other *1310 alleged coconspirators. 127 F.2d at 689 . The principle stated in Kamanosuke is fully supported by the United States Supreme Court decisions in Hamling v. United States, 418 U.S. 87, 101 , 94 S.Ct. 2887 , 41 L.Ed.2d 590 (1974), and Dunn v. United States, 284 U.S. 390, 393 , 52 S.Ct. 189 , 76 L.Ed. 356 (1932). See also United States v. Martorano, 557 F.2d 1, 8 (CA1 1977), cert. denied 435 U.S. 922 , 98 S.Ct. 1484 , 55 L.Ed.2d 515 (1978).

My review of the entire record convinces me that there is substantial evidence to sustain Noah’s conviction. I am satisfied that the jurors reasonably could decide that Noah was guilty as charged. In determining the sufficiency of the circumstantial evidence in a case such as this, the question is not whether the evidence excludes every hypothesis of guilt except that of guilt, but rather whether the trier of fact could reasonably arrive at its conclusion. United States v. Eaglin, 571 F.2d 1069, 1076 (CA9 1977).

I need not here cite the authorities supporting the well established rule that in reviewing the evidence to determine whether it was sufficient to allow a jury to convict, we must view it in the light most favorable to the government.

The following quotation from United States v. Dunn, supra, 564 F.2d at 357 (CA9 1977), is worthy of mention on Noah’s claim. We quote:

V.

The government, in its case in chief, introduced evidence showing that appellant Lamb induced Ms. Emery to participate in the conspiracy by telling her that he owed money to another penitentiary inmate for narcotics and that if he did not pay the debt, he would be killed. Lamb also told her that with some of the left over money, they could purchase additional narcotics for distribution. In order to corroborate Emery’s testimony, whose credibility was under severe attack, the government offered evidence that Lamb was involved in previous narcotics transactions. This evidence was offered to explain Lamb’s motive and intent in participating in the conspiracy to defraud the government. Rule 404(b), F.R.E., permits the use of this type of evidence. See Spencer v. Texas, 385 U.S. 554, 560-61 , 87 S.Ct. 648 , 17 L.Ed.2d 606 (1967); United States v. Brashier, 548 F.2d 1315 , 1325—26 (CA9 1976); United States v. Nichols, 534 F.2d 202, 204 (CA9 1976). This contention is groundless.

CONCLUSION

I would affirm the judgment of the lower court and hold: (1) that Brasfield is not controlling on the record before us; (2) that the provisions of Rule 52(a) and (b), F.R. Crim.P., control on this record; (3) that the action of the court in interrogating the jury on its numerical division in the absence of defendants and their attorneys was at most harmless error which was cured by the later interrogation along the same line in the presence of counsel, to which counsel not only failed to object but asked the court to reveal the numerical division of the jury as ascertained on the first interrogation. Alternatively, the actions of the district court would not qualify for reversal under the plain error provisions of Rule 52(b), F.R. Crim.P.; (4) that the other assignments of error are groundless.

. Allen v. United States, 164 U.S. 492 , 17 S.Ct. 154 , 41 L.Ed. 528 (1896).

. It should be noted that here the jury was not deadlocked, but was making definite progress toward verdicts at the time of the last interrogation:

“THE COURT: Are you of the opinion, Mr. Israel, [foreman] that the jury should continue to deliberate or should I dismiss the jury?

JUROR ISRAEL: I think we should continue.

THE COURT: Would you prefer to continue tonight or come back tomorrow?

JUROR ISRAEL: Tomorrow, I think.

THE COURT: All right; now, last night at about this time you were in some numerical standing as to each defendant.

Has there been any change in the numerical standing as to the Defendant Wiley?

JUROR ISRAEL: Yes.

THE COURT: Any change in the numerical standing as to the Defendant Collins?

JUROR ISRAEL: Yes.

THE COURT: Noah?

JUROR ISRAEL: Yes.

THE COURT: Kinsey?

JUROR ISRAEL: Yes.

THE COURT: And Lamb?

JUROR ISRAEL: Yes.

THE COURT: All right, would you indicate whether the numerical standing is towards more now towards a verdict or more further away from a verdict?

JUROR ISRAEL: Close to a verdict.

THE COURT: On each of the defendants?

JUROR ISRAEL: Yes.”

[R.T., Vol. XX, pp. 3162-3163.] [Emphasis supplied.]

. “MR. KESSLER: (Defense Counsel) Could you give us the numberical [sic] standings?

THE COURT: As to Mr. Wiley it was seven-five. The same for Mr. Noah and the same as to Mr. Lamb. As to Mr. Kinsey it was nine-one-two. Mr. Collins was nine-zero^- and three undecided.

MR. KESSLER: Thank you.”

[R.T., Vol. XX, p. 3164.]