Full opinion text
WYATT, District Judge: These are separate appeals, consolidated in this court, of four defendants — William G. LaChance, William F. Zimmerli, John Schlagenhauf, and Thomas Ciccaglione— named in a superseding indictment returned against them and many other persons by a grand jury in the United States District Court for the District of Connecticut. Appellant Schlagenhauf was named in the superseding indictment as, and is often called in other papers of record, “John Schlaganauf,” apparently a misspelling of his surname; for convenience, he will usually be referred to herein as “John.” Appellant Ciccaglione will usually be referred to, for convenience, as “Thomas.” This prosecution had its beginnings on September 14, 1983, when the Coast Guard stopped and boarded the sailing vessel “Tho” in Long Island Sound some two miles from the mouth of the Connecticut River in the District of Connecticut. Appellant LaChance was aboard as Captain of the Tho and, after 4,300 pounds of marijuana was found on the ship and seized, La-Chance and the two crew members were turned over by the Coast Guard to Drug Enforcement Administration (DEA) agents who placed them under arrest. Marijuana is a Schedule I controlled substance (21 U.S.C. § 812(c) (Schedule I) (c)(10)), the distribution, possession with intent to distribute, and importation of which is unlawful (21 U.S.C. §§ 841(a)(1) and 952). An indictment was returned on September 21, 1983, by a federal grand jury at Bridgeport in the District of Connecticut. A superseding indictment was returned by the same federal grand jury on March 6, 1984, against the four appellants and many other persons; there were twenty-nine counts in the superseding indictment. The indictment and superseding indictment were assigned to Chief Judge Daly. The superseding indictment covered a time period from January 1976 to the date it was returned. The subject matter was the importation into the United States of marijuana from the Caribbean and Colombia, and its distribution in the United States, principally in the District of Connecticut. There was a charge against Zim-merli and his brother-in-law Francolini, of engaging in a continuing criminal enterprise (21 U.S.C. § 848); two charges of conspiracy against many defendants, one to import marijuana into the United States (21 U.S.C. § 963) and one to possess and distribute marijuana (21 U.S.C. § 846); and many charges of substantive law violations over a seven-year period for importing marijuana (21 U.S.C. § 952) and for possessing marijuana in the United States with intent to distribute it (21 U.S.C. § 841(a)(1)). On September 17, 1984, appellant La-Chance pleaded guilty to the counts against him in the superseding indictment (counts two, fifteen, sixteen, seventeen, and eighteen) and his plea was then accepted by Chief Judge Daly. In this connection, defendant LaChance was permitted to reserve the right on appeal from the judgment to review of the adverse determination of a motion by him and other defendants to dismiss the indictment to the extent that such motion was based on the improper selection of grand jurors (Fed.R.Crim.P. 11(a)(2)). On November 8, 1984, sentence was imposed on LaChance by the district court; we are told by the government (Brief, p. 4) that the total effective sentence was ten years’ imprisonment and a $60,000 fine. On November 15 and 27, 1984, notices of appeal were filed by La-Chance. The notice of appeal filed November 15, stated that “LaChance ... hereby appeals ... from the Court’s denial of his ‘Verified Motion to Dismiss Indictment and Stay Proceedings on Grounds of Substantial Failure to Comply with Law in the Selection of Grand and Petit Jurors’____” The notice of appeal filed November 27, 1984, described the appeal as from the order “entered in this action on May 2, 1984 (Denial of Motion to Dismiss for Failure to Comply with Law in Selection of Grand Jurors).” This appeal was given Docket No. 84-1415 in this court. Trial of the superseding indictment began on September 17, 1984 at Bridgeport before Chief Judge Daly and a jury. The jury returned its verdict on October 17, 1984, against the remaining defendants then on trial. Appellant Zimmerli was found guilty on counts one through fifteen and eighteen, and not guilty on count twenty-nine. On November 30, 1984, sentence was imposed on Zimmerli by the district court; we are told by the government (Brief, p. 4) that the total effective sentence was eighteen years’ imprisonment and a $410,000 fine. On December 6, 1984, a notice of appeal was filed for Zimmerli. This appeal was given Docket No. 84-1435 in this court. We are told in the Brief for appellant Zim-merli (p. 5) that on January 24, 1985, the sentences imposed on counts two and eighteen were vacated, “thereby reducing the 13V2 year consecutive sentence by 3 years and cutting the fines imposed by $125,000.” We are told in the Brief for appellant John (p. 2) that the jury found him “guilty on all counts.” The record on appeal does not seem to include any transcript of the return of the jury verdict, nor does the Defendants’ Joint Appendix. We are told in the Brief for appellant John (p. 2) that he was sentenced, among other counts, on count “twenty-one”; we are puzzled because in the superseding indictment shown in the record on appeal (BV I, document 21; “BV” references are to the two brown volumes in the record on appeal) and in Defendants’ Joint Appendix (A177; “A” references are to pages of the Defendants’ Joint Appendix), count “twenty-one” does not charge Schlagenhauf. There is a possible explanation for the apparent mistake. John was charged in count twenty-two of the superseding indictment, and the jury found him “guilty on all counts” (Brief, p. 2). According to the judgment of conviction (SA 1; “SA” references are to pages of John’s “Supplemental Appendix”), John was not sentenced on count twenty-two. Therefore, the sentence on count twenty-one could have been intended to be on count twenty-two. In any event, we assume that the sentence on count twenty-one does not affect the time to be served in prison because it was made to run “concurrently to Counts Eighteen and Twenty-Three” (SA 1). On November 29, 1984, sentence was imposed on John by the district court; we are told by the government (Brief, p. 4) that the total effective sentence was eight years’ imprisonment. On December 12, 1984, a notice of appeal was filed for John and for Craig Randall, another defendant in the same superseding indictment and in the same trial. This appeal was given Docket No. 84-1451 in this court. On April 11, 1985, an order of this court was filed dismissing the appeal of Craig Randall on his consent. Appellant Thomas was found guilty on count twenty-one and not guilty on count eighteen. On November 28, 1984, sentence was imposed on Thomas by the district court; we are told by the government (Brief, p. 5) that the sentence was two years’ imprisonment and a $15,000 fine. On December 10, 1984, a notice of appeal was filed for Thomas. This appeal was given Docket No. 84-1453 in this court. We are told by the government (Brief, p. 5) that the notices of appeal of the four appellants were all “timely filed.” We affirm the several judgments from which these appeals were taken. A. The Appeal of William G. LaChance On September 17, 1984, LaChance pleaded guilty to all counts in which he was charged. Chief Judge Daly accepted his plea. LaChance had properly reserved the right to review the denial of his motion to dismiss the indictment to the extent that such motion was based on the improper selection of grand jurors (Fed.R.Crim.P. 11(a)(2)). As noted earlier, LaChance filed two notices of appeal. One notice states, and the other indicates, that the appeal is from the district court’s order denying his motion to dismiss the indictment for improper selection of jurors. These notices of appeal raise a question not mentioned by the government, but which should be addressed. By stating or indicating in the notices of appeal that his appeal is from the district court’s order of May 2 denying his motion, LaChance failed to comply technically with Fed.R.Crim.P. 11(a)(2). Rule 11(a)(2) allows a defendant who has entered “a conditional plea of guilty” to reserve the right to review of an “adverse determination of any specified pretrial motion” on “appeal from the judgment” (emphasis supplied). Therefore, rather than appealing from the May 2 order, LaChance should have stated that his appeal was from the judgment of conviction entered against him November 19, 1984. Nevertheless, in light of Sanabria v. United States, 437 U.S. 54, 98 S.Ct. 2170, 57 L.Ed.2d 43 (1978), we conclude that LaChance’s appeal is properly before us. In discussing the effect of mistakes in a notice of appeal, the Sanabria Court stated: “A mistake in designating the judgment appealed from is not always fatal, so long as the intent to appeal from a specific ruling can fairly be inferred by probing the notice and the other party was not misled or prejudiced.” Id. at 67 n. 21, 98 S.Ct. at 2180 n. 21 (citing Daily Mirror, Inc. v. New York News, Inc., 533 F.2d 53, 56 (2d Cir.) (per curiam), cert. denied, 429 U.S. 862, 97 S.Ct. 166, 50 L.Ed.2d 140 (1976)). Under his reserved right to review of an adverse determination of a pretrial motion, appellant LaChance makes a single argument for reversal of his conviction: that the trial judge was in error in denying, without an evidentiary hearing, his pretrial motion to dismiss the indictment to the extent that such motion was based on the improper selection of grand jurors. This motion had alleged a “substantial and statistically significant underrepresentation” of blacks and women selected as grand and petit jurors, and women selected as grand jury forepersons, in violation of the “fair cross section requirement” of the Jury Selection and Service Act of 1968 (28 U.S.C. § 1861 and following; the “Act”) and of the sixth amendment. The motion had also alleged violations of the Act “which affect the random nature and objectivity of the selection process.” 1 On February 2, 1984, LaChance and other then co-defendants filed the “Verified Motion to Dismiss Indictment and Stay Proceedings on Grounds of Substantial Failure to Comply with Law in the Selection of Grand and Petit Jurors” (A71), described above. This motion also sought an evidentiary hearing. A supporting affidavit (A78) by Dr. John Lamberth, a statistics expert, accompanied the motion. On April 27, LaChance and others filed an “Amended Verified Motion” to dismiss the indictment, etc., to correct an error contained in the original motion of February 2. On May 2, 1984, Judge Daly denied the “Amended Verified Motion” of LaChance for “lack of a sufficient showing” and without an evidentiary hearing, except that he heard argument on the grand jury foreperson issue and reserved decision on that (A136). It is this ruling, to the extent that it denied the motion to dismiss the indictment for improper selection of grand jurors, which LaChance asserts was error, requiring reversal of his conviction. By endorsed order, filed July 10, 1984, Judge Daly denied the motion of LaChance as to the grand jury foreperson issue (A141). This was on the basis of the then recent Supreme Court decision in Hobby v. United States, 468 U.S. 339,104 S.Ct. 3093, 82 L.Ed.2d 260 (1984), settling the issue. In the same order, Judge Daly described his May 2 ruling as follows (A141): [T]he Court ruled that, applying the statistical analysis used by the Second Circuit in United States v. Jenkins, 496 F.2d 57, 65-66 (1974), cert. denied, 420 U.S. 925, 95 S.Ct. 1119, 43 L.Ed.2d 394 (1975), the defendants had failed to make an adequate showing on their claims of discrimination in the overall selection of grand and petit jurors. The contentions of LaChance as to impropriety in the selection of petit juries and of grand jury forepersons ultimately were abandoned, in part because the right to review of an adverse determination of these issues was not reserved at the time LaChance entered his guilty plea, in part (it seems fair to assume) because the conviction of LaChance was not by a petit jury, and in part because the Hobby decision had settled the grand jury foreperson issue. In consequence, these contentions are not now before us, and will not be referred to hereafter. 2 We take note of jury selection procedures as prescribed by the Act and by the District’s plan thereunder, so as to place in context the claims of LaChance of underre-presentation of women and blacks as grand jurors. The District of Connecticut’s Plan for Random Selection of Grand Jurors Pursuant to Jury Selection and Service Act of 1968 (1968 as amended) (“Plan”), was adopted by the judges of the District of Connecticut and approved by the Judicial Council of this Circuit, as required by Section 1863(a) of the Act. Although the Plan was not introduced below and is not part of the Record on Appeal, it is a proper subject for judicial notice. For the purpose of jury selection, the District of Connecticut is divided into three divisions: the Hartford Division, which draws jurors from the counties of Hartford, Litchfield, Windham, and Tolland; the New Haven Division, which draws jurors from the counties of New Haven, New London, and Middlesex; and the Bridgeport Division, which draws jurors from Fairfield County. Plan at 2; see 28 U.S.C. §§ 1863(b)(3), 1869(e). Jurors in the Bridgeport Division are drawn from the voter registration lists of Fairfield County. See Plan at 3; see also 28 U.S.C. § 1863(b)(2). After obtaining the voter registration list, the source list, the first step in the jury selection process is to form the “master wheel.” The master wheel is composed of the names of persons selected at random from the source list; these constitute one percent of the source list. 28 U.S.C. § 1863(b)(4); Plan at 6. The master wheel is emptied and refilled once every four years. 28 U.S.C. § 1863(b)(4); Plan art. X. The data available to LaChance included the master wheels formed in 1977 and 1981 only. Dr. Lamberth’s affidavit accompanying the motion states that “[ajccording to information supplied by [Mr. Kevin F. Rowe, then Chief Deputy Clerk in charge of jury matters], all data prior to the 1977 wheel has been destroyed” (A79). See 28 U.S.C. § 1868 (permitting destruction after four years). Whenever it is anticipated that jurors will be needed for a pool for service on grand and petit juries, names are drawn at random from the master wheel. 28 U.S.C. § 1864(a); Plan at 6. Each person whose name is so drawn is sent a “juror qualification form,” sometimes called a “questionnaire.” 28 U.S.C. § 1864(a); Plan at 6-7. According to LaChance’s Brief (p. 8), those persons who receive and return a questionnaire constitute the “venire.” Neither the Act nor the Plan defines “venire,” but, for convenience, we shall use the word as defined by LaChance. The questionnaire requires the prospective juror to answer enumerated questions regarding qualifications for jury service. He or she must state, among other things, whether an exemption or excusal from juror service is claimed. Qualifications (Art. VII), exemptions (Art. VIII), and excuses (Art. IX) are set forth in the Plan and in the questionnaire. When the questionnaire is returned, a specified district judge “shall determine solely on the basis of information furnished on [the questionnaire] and other competent evidence whether a person is unqualified for, or exempt, or to be excused from jury service.” Plan at 3. Those persons who receive and return a questionnaire and who are not exempted, excused or otherwise disqualified from jury service constitute the “qualified wheel.” 28 U.S.C. § 1866(a); Plan at 8. The Plan provides that from the qualified wheel shall be separately drawn the names of persons to be summoned for service on grand juries and on petit juries. Separate lists are prepared of those summoned for service on grand juries and on petit juries. We are here concerned solely with the procedure for selecting a grand jury in the Bridgeport Division. The Plan (art. XV) provides in relevant part as follows: Grand jurors shall be selected, drawn, summoned and impaneled on a “divisional” basis at such time as the public interest requires. The names of all grand jurors summoned for service at each seat of Court will be placed in a Special Jury Wheel from which twenty-three names will be drawn by lot for service on the grand jury panel. Each grand jury shall serve for a period of eighteen months at each seat of Court unless earlier discharged by the Court. The grand jury which returned the indictment and superseding indictment against LaChance and the other appellants was selected under this procedure, beginning with the master wheel formed in 1981. This means that the names of a pool of grand jurors were drawn at random from the qualified wheel (formed from the 1981 master wheel), and those persons were summoned to appear at Bridgeport at some time earlier than September 21, 1983, when the first indictment was returned. The names of the grand jurors were put in a Special Jury Wheel from which were drawn “twenty-three names ... for service on the grand jury panel.” We are not informed by the government or in any of the statistics of LaChance how many names were in the Special Jury Wheel from which the twenty-three members of the LaChance grand jury were drawn. We do know that grand juries normally serve for eighteen months and that there are no peremptory challenges. In consequence, it would seem to us that fewer grand jurors than petit jurors would be needed from which twenty-three names are to be drawn. Master wheels and qualified wheels for selecting grand and petit juries were formed by the Bridgeport Division in 1977 and in 1981 (Plan art. X, as amended; A79). It is undisputed that the grand jury which returned the superseding indictment against LaChance and the other appellants was derived from the master wheel formed in 1981. 3(a) The argument to this court for LaChance is that the District Court erred in denying his motion to dismiss the superseding indictment without affording him the opportunity to present evidence at a hearing. Although the Brief for LaChance concludes by asking simply for reversal of the judgment against him, we believe that his intent is to ask for reversal and for a remand to the District Court for an evidentia-ry hearing on his claims of improper selection in the Bridgeport Division of the grand jury derived from the 1981 master wheel, which returned the indictment and superseding indictment against him. 3(b) Section 1867(a) of the Act states: In criminal cases, before the voir dire examination begins, or within seven days after the defendants discovered or could have discovered, by the exercise of diligence, the grounds therefor, whichever is earlier, the defendant may move to dismiss the indictment or stay the proceedings against him on the ground of substantial failure to comply with the provisions of this title in selecting the grand or petit jury. Under Section 1867(d), if the defendant files a motion pursuant to Section 1867(a) “containing a sworn statement of facts which, if true, would constitute a substantial failure to comply with the [Act],” then the defendant is “entitled to present in support of such motion the testimony of the jury commission [sic] or clerk, if available, any relevant records and papers not public or otherwise available used by the jury commissioner or clerk, and any other relevant evidence.” In his motion under Section 1867(a), La-Chance submitted a sworn statement that there had been a “substantial failure to comply” with the Act in two ways. First, he stated that the “fair cross section requirement of [the Act] and the Sixth Amendment has been breached”; second, he stated that there had been violations of procedures prescribed by the Act which “affect the random nature and objectivity of the selection process.” We consider these allegations in turn. 3(c) The sixth amendment affords every criminal defendant entitled to a jury trial the right to trial “by an impartial jury.” The Supreme Court has interpreted this right to mean, among other things, that the pool from which the petit jury is drawn must represent a “fair-cross-section” of the community in which the defendant is tried. Duren v. Missouri, 439 U.S. 357, 363, 99 S.Ct. 664, 668, 58 L.Ed.2d 579 (1979). The Act extends this fair cross section requirement of the sixth amendment to the pool from which federal grand jurors are selected: “It is the policy of the United States that all litigants in Federal courts entitled to trial by jury shall have the right to grand and petit juries selected at random from a fair cross section of the community in the district or division wherein the court convenes.” 28 U.S.C. § 1861. To establish a prima facie violation of the sixth amendment’s fair cross section requirement, the Supreme Court applies a three-pronged test, under which the defendant must show: (1) that the group alleged to be excluded is a “distinctive” group in the community; (2) that the representation of this group in venires from which juries are selected is not fair and reasonable in relation to the number of such persons in the community; and (3) that this underre-presentation is due to systematic exclusion of the group in the jury-selection process. Duren, 439 U.S. at 364, 99 S.Ct. at 668. This test for a prima facie case under the sixth amendment is also applied in determining whether a motion under the Act states “facts which, if true, would constitute a substantial failure to comply with” the fair cross section requirement of the Act. United States v. Clifford, 640 F.2d 150, 154-55 (8th Cir.1981); see Taylor v. Louisiana, 419 U.S. 522, 528-30, 95 S.Ct. 692, 696-98, 42 L.Ed.2d 690 (1975) (“Recent federal legislation [the Act] governing jury selection within the federal court system has a similar thrust [to the sixth amendment’s representative cross section requirement]”); United States v. Test, 550 F.2d 577, 584-85 (10th Cir.1976) (en banc) (Act’s fair cross section standard is “functional equivalent of the constitutional ‘reasonably representative’ standard”). LaChance bases his claim of a fair cross section violation on the Act and on the fifth and sixth amendments. He has spelled out no separate argument on his fifth amendment challenge; in the district court he merely quoted from Justice Powell’s dissenting opinion in Castaneda v. Partida, 430 U.S. 482, 97 S.Ct. 1272, 51 L.Ed.2d 498 (1977), which states that “[t]he right to a ‘representative’ grand jury is a federal right that derives ... from the Fifth Amendment’s explicit requirement of a grand jury. That right is similar to the right — applicable to state proceedings — to a representative petit jury under the Sixth Amendment.” Id. at 509-10, 97 S.Ct. at 1287-88 (Powell, J., dissenting). We will not, therefore, distinguish between the fifth amendment claim and that based on the sixth amendment. Further, because the Duren test governs fair cross section challenges under both the Act and the sixth amendment, our discussion of the statutory challenge also disposes of his constitutional claim. We now consider the Duren test as applied to LaChance’s statement that blacks and women are underrepresented on grand juries in the Bridgeport Division. 3(c)(i) Appellant LaChance properly stated in his motion (A74) that both blacks and women constitute distinctive groups in the community. See Taylor, 419 U.S. at 531, 95 S.Ct. at 698 (women); United States v. Jenkins, 496 F.2d 57, 65 (2d Cir.1974) (blacks), cert. denied, 420 U.S. 925, 95 S.Ct. 1119, 43 L.Ed.2d 394 (1975). Therefore, the first prong of the Duren test is satisfied. mm LaChance attempted to satisfy the second prong of the Duren test through a statistical presentation in his amended motion, which we show in the following chart: Blacks (6.98% Black Voting Age Population in District of Connecticut According to 1980 Census) Number of Standard Deviations Probability Comparative From Expected of Occurring Disparity Number_by Chance 1977 and 1981 Venires (COMBINED) (3.52% Black) 49.57% 8.46 less than two in 1,000,000 1977 and 1981 Qualified Wheels (COMBINED) (4.83% Black) 30.8 % 3.69 less than four in 10,000 Feriales (53.17% Female Voting Age Population in District of Connecticut According to 1980 Census) 1977 and 1981 Qualified Wheels (COMBINED) (49.91%’ Female) 8.01%- 3.77 less than two in 10,000 The District Court denied LaChanee’s motion without affording him the opportunity to present evidence at a hearing because the court had applied to the statistics of underrepresentation stated for La-Chance the “absolute numbers” analysis used by this court in United States v. Jenkins, 496 F.2d 57, 66 (1974), cert. denied, 420 U.S. 925, 95 S.Ct. 1119, 43 L.Ed.2d 394 (1975) (A141). Appellant La-Chance, in effect, asks us (Brief, p. 17) to overrule Jenkins because Castenada v. Partida, 430 U.S. 482, 496-97 n. 17, 97 S.Ct. 1272, 1281 n. 17, 51 L.Ed.2d 498 (1977), in an equal protection challenge to Texas’ grand jury selection process, made reference to the statistical approach used in the LaChance motion — the standard deviation analysis. Jenkins involved a challenge under the Act to juror selection in the New Haven Division of the District of Connecticut. In appealing from their criminal convictions, the appellants in Jenkins argued that the percentage of blacks in the “New Haven jury pool” (determined from “questionnaires sent out”) (496 F.2d at 64) — 3.3%— was so disproportionate to the percentage of blacks in the adult population — 5.45%— as to constitute a substantial violation of the Act. We rejected their contention that whether the disparity was substantial should be determined by “the ratio of the Negro percentage of the adult population to the Negro percentage of those adults available for jury service under the district court’s plan.” Id. at 65. Although we recognized that the approximate 5 to 3 ratio “by itself appears substantial indeed,” we were not persuaded by that ratio, but rather held {id. at 66): The test of fairness intended by Congress is the more practical one of the difference in absolute numbers rather than a difference in percentages. Judged by this standard, a difference of one (1) Negro in a panel of 60 jurors is not substantial. It follows, then, that under the absolute numbers analysis, it is, as the first step, determined how many members of the allegedly underrepresented group would be expected to appear on the jury panel in order to reflect the percentage of the group in the total population eligible for jury service. It is next, as the second step, determined how many members of the group would be expected to appear on the jury panel in order to reflect the percentage of the group in the venire or qualified wheel or both, depending on the claim being made. A successful challenge to selection procedures depends on whether the difference between the numbers determined in the two steps is “substantial.” The following chart represents the numbers determined by applying the “absolute numbers” analysis to the 1977 and 1981 combined data presented in the LaChance motion. (Some of the figures have been rounded off.) The Jenkins court applied the analysis to the panel derived from the last step in the petit jury selection process accomplished by a random drawing — the petit jury panel of 60. Similarly, we apply the same analysis to the panel derived from the last step in the grand jury selection process accomplished by a random drawing — the grand jury panel of 23 (see section A, part 2 of opinion). Blacks (6.98% of eligible population) Venires (3.52%) Expected From, Expected Difference In Panel Eligible From Absolute Size_Population_Venires_Numbers 23 1.6 .8 .8 Qualified Wheels (4.83%) 23 1.6 1.11 .49 Females (53.17% of eligible population) Qualified Wheels (49.91%) 23 12.23 11.48 .75 The Jenkins court found that the effect of the underrepresentation of blacks in the selection process there considered, measured by the difference in absolute numbers, would be one juror (that is, a correction of the process would add one black to a panel of 60), and that this effect was not “substantial.” Application of the absolute numbers analysis to the LaChance statistics shows an effect of the claimed underre-presentation of blacks and women in the selection process here of less than the one juror effect held to be not “substantial” in Jenkins (“substantial” comes from “substantial failure to comply with the provisions of this title” in Section 1867(d) of the Act). It was for this reason that Judge Daly denied the motion below without an evidentiary hearing. On this appeal, counsel for LaChance (Brief, pp. 17-23) ask us to abandon and disregard Jenkins and to apply a “standard deviation” analysis in determining whether the effect of the claimed violation here was “substantial.” The standard deviation analysis measures the statistical significance of fluctuations from an expected number in a random sampling. For example, if a box were filled with 1,000 slips of paper, 600 of which were marked “X” and 400 of which were marked “Y”, and someone randomly selected 100 slips, the “expected number” of “Y” slips selected would be 40; that is, because the ratio of “X” slips to “Y” slips is three to two, it would be expected that a random selection of 100 slips would yield 60 “X” slips and 40 “Y” slips. However, a statistician would not be surprised if the number of “Y” slips fluctuated, or “deviated,” from the expected value of 40. See Castaneda, 430 U.S. at 496 n. 17, 97 S.Ct. at 1281 n. 17 (“Of course, in any given drawing some fluctuation from the expected number is predicted.”) Rather, the statistician would predict a fluctuation, and that predicted fluctuation from the expected number is the standard deviation. For jury selection purposes, the standard deviation equals the square root of the product of the total number in the sample times the probability of drawing a member of the allegedly underrepresented group times the probability of drawing a nonmember. Id. at 497 n. 17, 97 S.Ct. at 1281 n. 17. This formula yields a different result depending on the size of the sample. “One of the principal reasons for using a standard deviation analysis and hypothesis testing is that it is axiomatic in statistical analysis that the precision and dependability of statistics is directly related to the size of the sample being evaluated.” Moultrie v. Martin, 690 F.2d 1078, 1083 (4th Cir.1982); see generally Castaneda, 430 U.S. 482, 97 S.Ct. 1272, 51 L.Ed.2d 498; Villafane v. Manson, 504 F.Supp. 78 (D.Conn.1980); Finkelstein, The Application of Statistical Decision Theory to the Jury Discrimination Cases, 80 Harv.L.Rev. 338 (1966). Recalling the above sample of a box filled with “X” and “Y” slips, if a sampling from the box resulted in selection of a number of “Y” slips substantially fewer than the expected value, and hence more than a certain number (depending on the sample size) of standard deviations below the expected value, a statistician would question if the sampling was random. In Castaneda, the Supreme Court stated that “[a]s a general rule for such large samples [870], if the difference between the expected value and the observed number is greater than two or three standard deviations, then the hypothesis that the jury drawing was random would be suspect to a social scientist.” 430 U.S. at 497 n. 17, 97 S.Ct. at 1281 n. 17. Relying on Castaneda, La-Chance argues that the second prong of the Duren test (group representation in jury venires not “fair and reasonable”) is satisfied by his allegations of differences between the expected values and observed numbers of 8.46 standard deviations as to blacks on the venires, 3.69 standard deviations as to blacks on the qualified wheels, and 3.77 standard deviations as to females on the qualified wheels. The government argues (Brief, p. 18), however, that Jenkins is still controlling because Castaneda decided an equal protection challenge, which involves a showing of discriminatory intent. Castaneda, 430 U.S. at 494-95, 97 S.Ct. at 1280-81. Because LaChance’s statistical presentation does not, as will be seen later, establish the degree of underrepresentation on the 1981 venire and qualified wheel, we need not decide whether the standard deviation analysis should apply to the cross section challenge here made or, if it should, whether the number of standard deviations alleged would satisfy the second prong of the Duren test. See Castaneda, 430 U.S. at 497 n. 17, 97 S.Ct. at 1281 n. 17 (disapproving differences of 29 and 12 standard deviations between expected and observed numbers in two samples — greater differences than claimed by LaChance here). This is because LaChance has not stated facts as to the 1981 sources, from which the grand jury which indicted him was derived, so as to enable us or the court below to apply either the “absolute numbers” analysis of Jenkins or the “standard deviation” analysis which he urges on us. La-Chance has, therefore, not made a “sufficient showing,” as the court below properly ruled. Before it can be determined, under the absolute numbers analysis, the standard deviation analysis, or any other statistical analysis, whether the claimed underrepre-sentation is “substantial” (28 U.S.C. § 1867(d)), two percentages must be shown. First, the defendant “ ‘must demonstrate the percentage of the community made up of the group alleged to be underrepresented.’ ” United States v. Goodlow, 597 F.2d 159,162 (9th Cir.) (quoting Duren, 439 U.S. at 364, 99 S.Ct. at 668), cert. denied, 442 U.S. 913, 99 S.Ct. 2830, 61 L.Ed.2d 280 (1979). Second, the defendant “must show the percentage representation of the groups in question on jury venires.” Id. The statistical presentation made for La-Chance has two striking features: First, it purports to compare statewide population data with Fairfield County jury selection data; and second, it combines jury selection data for 1977 and 1981, whereas the grand jury which indicted LaChance and the other defendants was derived from the master wheel of 1981 only. 3(c)(ii)(A) LaChance attempted to “demonstrate the percentage of the community made up of the group alleged to be underrepresented” by comparing the percentage of blacks on the venires and qualified wheels, and the percentage of women on the qualified wheels, in the Bridgeport Division (Fairfield County only) with district wide (all of Connecticut) black and female voting-age populations. The government argues (Brief, pp. 12-13) that because “[t]here is no basis for concluding that Fairfield County [from which juries in the Bridgeport Division are drawn] has the same percentage of blacks in its voting-age population as Connecticut as a whole____ [t]he entire basis for comparison is suspect.” (This argument applies equally to LaChance’s allegations with regard to the underrepresentation of women.) In response to this argument, counsel for La-Chance asserts that the district court’s denial of a hearing precluded defendants “from explaining their use of district-wide voter data” (Reply Brief, p. 2.). Further, at oral argument, counsel for LaChance argued that these were the “best available statistics” and that “many courts have used district-wide statistics when they are measuring a smaller division.” Our research shows, however, that the voting-age populations of blacks and females, as stated by LaChance in his motion below (BV II, LaChance Document No. 10), mistakenly identified by counsel for LaChance (in the motion and also Brief, pp. 8, 9, 10, and Reply Brief, p. 2) as “in the District of Connecticut” or “district-wide,” are in fact the voting-age populations for Fairfield County only. According to I United States Bureau of Census, United States Department of Commerce, 1980 Census of Population 8-200 (1982), the black voting-age population of Fairfield County (blacks age 18 and over: 41,176) is 6.98% of the total voting-age population of Fairfield County (all persons age 18 and over: 589,-669). The female voting-age population of Fairfield County (females age 18 and over: 313,520) is 53.168% of the total voting-age population of Fairfield County. By contrast, the statewide black and female voting-age population percentages are, respectively, 5.97% and 52.8%. See id. at 8-24 to 8-27. Thus, the voting-age population percentages stated by LaChance of blacks (6.98%) and females (53.17%) were in fact percentages of Fairfield County’s voting-age population and, therefore, served as proper bases of comparison for the percentages of blacks on the Bridgeport Division venires and qualified wheels and of females on the Bridgeport Division qualified wheels. We conclude, therefore, that LaChance has adequately stated “the percentage of the community made up of the group[s] alleged to be underrepresented [blacks and women].” Duren, 439 U.S. at 364,99 S.Ct. at 668, Goodlow, 597 F.2d at 162. 3(c)(ii)(B) The percentage of the community represented by blacks and women is to be compared, under the absolute numbers analysis and under the standard deviation analysis, with the percentage of blacks and women in the jury pool (venire, qualified wheel) from which the 23 member grand jury was derived. The statistical presentation made for LaChance did not contain any data from which this percentage could be calculated because it combined jury selection data for the 1977 and 1981 venires, and combined data from 1977 and 1981 qualified wheels, whereas the grand jury which indicted LaChance was derived from the venire and qualified wheel of 1981 only. In an effort to establish the degree of underrepresentation of blacks and women on grand juries in the Bridgeport Division, LaChance examined data from the 1977 and 1981 venires and qualified wheels. Rather than presenting the data from each year separately, however, LaChance combined the 1977 and 1981 data. He thereby prevented the court below, and now prevents us, from determining the degree of underrepresentation on the 1981 venire and qualified wheel. Thus, LaChance has failed to allege “facts which, if true, would constitute a substantial failure to comply” with the Act’s fair cross section requirement. LaChance was indicted by a grand jury derived from the 1981 master wheel (Brief for LaChance, p. 7) which was emptied and refilled in 1981 pursuant to the Act and the Plan. (See section A, part 2 of this opinion.) Section 1861 of the Act entitled La-Chance to a grand jury “selected____from a fair cross section of the community.” Under the Act, “[defendants, of course, may challenge only improprieties affecting the particular grand jury which indicted them.” United States v. Bearden, 659 F.2d 590, 601 (5th Cir.1981) (emphasis in original), cert. denied, 456 U.S. 936, 102 S.Ct. 1993, 72 L.Ed.2d 456 (1982). Therefore, although the 1977 data may be relevant in determining whether underrepresentation on the 1981 sources is due to systematic exclusion (“that is, inherent in the particular jury-selection process utilized,” Duren, 439 U.S. at 366, 99 S.Ct. at 669), LaChance would not be entitled to any relief based solely on claimed underrepresentations in the 1977 venire and qualified wheel. By combining the 1977 data with the 1981 data, LaChance made it impossible for the court below, and now for us, to determine the degree of underrepresentation, if any, in the 1981 sources. For example, it is possible that the underrepresentation was very high on the 1977 venire and master wheel but nonexistent or very low on the 1981 venire and master wheel. These circumstances make meaningless the number of standard deviations claimed by La-Chance for the 1977 and 1981 data taken together. Combining the data prevents discovery of any absence of, or insubstantial nature of, any underrepresentation in the 1981 venire and qualified wheel. La-Chance argues that, had he been afforded an evidentiary hearing, “Dr. Lamberth could have testified that combining the 1977 and 1981 data allows for a larger sample which in the discipline is considered more stable and produces more reliable results” (Reply Brief, p. 2). This argument is without merit. The burden was on La-Chance to state “facts which, if true, would constitute a substantial failure to comply” with the Act’s fair cross section requirement. LaChance was entitled to an eviden-tiary hearing only if his motion made out a prima facie case. He was not entitled to a hearing to explain why he did not make out a prima facie case. This is not a case such as Duren, in which the Supreme Court accepted use of six year old census data in the absence of “evidence ... in the record to suggest that [the old census data] significantly distorted” the statistical presentation, 439 U.S. at 365, 99 S.Ct. at 669. Here, through extensive discovery (see Brief for LaChance, pp. 6-7), appellants had available to them the data to make a meaningful statistical presentation, but did not do so. We rejected the use of overbroad statistics in United States v. Newman, 549 F.2d 240 (2d Cir. 1977). That case involved a claim that the government used its peremptory challenges to exclude blacks from petit juries in the New Haven Division of the District of Connecticut. In examining the claim, the district court had combined data regarding petit juries in the New Haven and Hartford Divisions. We stated (id. at 244): It is only the New Haven Division and its procedures for selecting the jury in this case which are relevant. By merging the New Haven Division statistics with those of the Hartford Division, the district court has ... greatly distorted and arbitrarily altered the statistics of the New Haven Division which are, as thus changed, made useless. Similarly, the merging by LaChance of the 1977 and 1981 data may have “greatly distorted and arbitrarily altered” the 1981 data. Even if LaChance’s statements of fact regarding the 1977 and 1981 combined data are true, those facts would not “constitute a substantial failure to comply” with the Act’s fair cross section requirement as to the 1981 venire and qualified wheel, from which the indicting grand jury was derived. LaChance was, therefore, not entitled to an evidentiary hearing on his fair cross section challenge, and the denial of his motion below, without a hearing, must be sustained. 3(c)(iii) Because we have concluded that La-Chance failed to satisfy the second prong of the Duren test, we need not consider whether he stated facts sufficient to satisfy the further aspect of that test, that the claimed underrepresentation was due to “systematic exclusion.” 3(d) LaChance also stated in his motion what were claimed to be violations of the Act “which affect the random nature and objectivity of the selection process” (A75). Such violations were alleged to have included: (1) “Erroneous permanent disqualifications, exemptions, excusáis or exclusions, based upon insufficient medical documentation,” “previous jury duty,” “occupation,” “child care,” and “student status”; (2) “Usurpation of judicial functions by jury clerks”; (3) “Temporary or permanent disqualifications, exemptions, excusáis, or exclusions, exceeding one per centum of the number of persons who return executed juror qualification forms during the period specified in the Plan between two consecutive filling[s] of the master jury wheel”; and (4) “Erroneous inclusion of permanently excused and exempted jurors in the qualified wheel.” Finally, it was alleged that “[o]ut of 2040 disqualified juror questionnaires examined, 232 were disqualified inaccurately, constituting an error rate of 11.37%” (A76). These claimed violations of the Act’s requirement of selection “at random” are insufficient for the same reason as is the LaChance statistical presentation on the claimed underrepresentation of women and blacks. It was impossible for the court below, and now is impossible for us, to determine the extent of the “at random” violations in the selection of grand jurors derived from the master wheel formed in 1981. For example, we are not told how many, if any, of the “[ejrroneous permanent disqualifications, exemptions, excusáis or exclusions” were issued to jurors drawn from the 1981 master wheel; we are not told when the alleged “[usurpation of judicial functions by jury clerks” occurred; nor are we told how many, if any, of the 232 jury questionnaires “disqualified inaccurately” had been sent to persons drawn from the 1981 master wheel. The Act entitles a defendant to present evidence at a hearing if the facts stated in the motion would, “if true, constitute a substantial failure to comply” with the Act. 28 U.S.C. § 1867(d). Mere “technical” violations of the procedures prescribed by the Act do not constitute “substantial failure to comply” with its provisions. United States v. Carmichael, 685 F.2d 903, 911 (4th Cir.1982), cert. denied, 459 U.S. 1202, 103 S.Ct. 1187, 75 L.Ed.2d 434 (1983); United States v. Capone, 683 F.2d 582, 589 (1st Cir.1982); Bearden, 659 F.2d at 601. Whether a violation is “substantial” or merely “technical” depends upon the nature and extent of its effect on the wheels and venire from which a defendant’s grand jury was derived. Because the nature and extent of the claimed violations as to the 1981 wheels and venire cannot be determined from the LaChance motion, there was not statement of “facts, which, if true, would constitute a substantial failure to comply” with the Act’s requirement of selection “at random.” In this aspect also, Chief Judge Daly properly denied La-Chance’s motion without an evidentiary hearing, and his ruling must be sustained. B. The Appeal of William F. Zimmerli Appellant Zimmerli was found guilty on October 17, 1984, on all counts in which he was named, save for count twenty-nine. He was found guilty on count one, engaging in a continuing criminal enterprise (21 U.S.C. § 848); on count two, conspiracy to import marijuana into the United States (21 U.S.C. § 963); on counts three, five, seven, nine, eleven, thirteen, and fifteen, importation of marijuana into the customs territory of the United States and into the District of Connecticut (21 U.S.C. § 952); on counts four, six, eight, ten, twelve, fourteen, and seventeen, possession of marijuana with intent to distribute (21 U.S.C. § 841(a)(1)); and on count eighteen, conspiracy to possess with intent to distribute, and to distribute, marijuana (21 U.S.C. § 846). Zim-merli was found not guilty on count twenty-nine, bribery of a witness (18 U.S.C. § 201(h)). Counsel for appellant Zimmerli make four arguments for reversal of his conviction. The first and second arguments are based on a search for, and seizure of, items under a search warrant for the “family dwelling” of Zimmerli at Glastonbury, Connecticut on September 15, 1983. While not expressly stated, presumably the claim is that the search and seizure violated the constitutional rights of Zimmerli, that the motion to suppress seized evidence should have been granted, and that the conviction should be reversed for admission of any of this evidence. The first and second arguments are presented in the brief to this Court for Zimmerli. The third argument is based on the denial by the trial judge of a pretrial motion for Zimmerli to dismiss the superseding indictment because the grand jury which returned it had been selected in violation of the Constitution and of the Jury Selection Act in that women and blacks were not proportionately represented. This argument is not discussed in the brief for Zimmerli, who relies (Brief, p. 49) in this respect on the brief for appellant LaChance. The fourth argument is based on the denial by the trial judge of a motion by Zimmerli, after the completion of the jury voir dire at the beginning of the trial, to strike the petit jury venire and for other relief. The ground for the motion was that the petit jury venire did not contain a fair cross section of young adults. The fourth argument is not discussed in the brief for Zimmerli, who relies (Brief, p. 49) in this respect on the brief for appellant John. 1(a) On Friday, September 14,1983, 308 bales (4,300 pounds) of marijuana were seized on the sailing vessel “Tho” in Long Island Sound; the three men on board — La-Chance, Stanko, and Busby — were turned over by the Coast Guard to Drug Enforcement Administration (DEA) agents, who placed them under arrest. Busby was interrogated and gave information to DEA, which since 1979 had been receiving information considered reliable linking Zimmerli to imports of large amounts of marijuana in sailing vessels from the Caribbean to New England, mostly to Connecticut. Further information implicating Zimmerli was obtained by DEA on September 14 and 15. Zimmerli had his home in Glastonbury, Connecticut, a town on the Connecticut River, a few miles south of Hartford. On Saturday, September 15, 1983, shortly before 5:03 p.m., DEA applied to Magistrate Thomas P. Smith in the District of Connecticut for a search warrant. An affidavit of Michael W. Meyrick, a DEA agent, was submitted in support of the application. The search was asked for the “single family dwelling” of Zimmerli at 37 Ledge-wood Drive, Glastonbury. The affidavit recited the seizure aboard the “Tho” (of German registry) of 308 bales of marijuana the day before (the 4,300 pounds); that Tho was boarded about 2 miles from the mouth of the Connecticut River; that at the time of boarding “charts on board were opened to navigation charts for that river”; that Busby, a member of the Tho crew, had told Meyrick that Tho had picked up the contraband marijuana two weeks before in Jamaica and had sailed for Connecticut for offloading along the Connecticut River; that Busby was to be paid $50,000 upon delivery offloading; that he was told by the other crew members to make no statement if arrested but to call “attorney John Mark-Flowers”; that Stanko called an attorney in Montana, who referred him to Goldfarb, an attorney in Hartford, and told him to advise Goldfarb of his arrest; that since 1979 the DEA had received reliable information that Zimmerli was “a large-scale marijuana smuggler” running “several boatloads a year”; that Zimmerli’s lawyer was Alexander Goldfarb of Hartford; that at about 2 p.m. on September 15 an agent had called the home of Zimmerli and claimed to be a “friend” of LaChance, advising that there was “trouble”; and that Joy Zimmerli, wife of William, replied that her husband had suspected trouble and was off in an airplane “looking” (GA 2-4; “GA” references are to pages of the Appendix to the Government’s Brief). The affidavit further stated that about 3 p.m. on September 15, agents “observed at least two Vans in the driveway and approximately eight persons in the yard” at the Zimmerli house in Glastonbury; and that vans are used “as offshore loading vehicles by marijuana smugglers.” The Magistrate issued a search warrant at 5:03 p.m. on September 15. The warrant authorized a search of the premises in Glastonbury and of “two vans” parked on the premises. The property which the warrant authorized to be searched for and seized was described as follows: “money intended to be furnished in exchange for controlled substances; records of dealing in controlled substances; and documentary evidence of the involvement of William and Joy Zim-merli in the conspiracy to smuggle approximately 4,300 pounds of marijuana into Connecticut aboard the ‘Tho’.” The search warrant was executed on the same day issued, Saturday, September 15, 1983, beginning about 6:30 p.m. (A247). Many items were seized. A motion was filed for Zimmerli on November 7, 1983, to suppress as evidence all of the items seized in the search of the home in Glastonbury. The grounds were said (A43) to be (1) no “probable cause” shown for issuance of the search warrant; (2) the supporting affidavit had “reckless and/or deliberate untruths and significant omissions”; (3) the warrant was “overly broad, general and vague as to what property the agents were empowered to search for and seize”; and (4) the “scope of the search far exceeded the limits set ... by the warrant____” On December 6, the trial judge fixed December 19, 1983 for hearing of the motion by Zimmerli to suppress the evidence seized at his home. On December 14, an affidavit of counsel for Zimmerli, sworn to December 13, was filed in support of his motion to suppress (A45-48). On December 16, the government filed a “response to amended motions to suppress” (A49-50). Apparently, the government treated the December 13 affidavit of counsel as an amendment to the Zimmerli motion to suppress. On December 19, 1983, evidence was presented on the motion to suppress before Chief Judge Daly; there was a further hearing on January 9, 1984, which completed the presentation of evidence. A docket entry for that date states that decision was reserved (A4). On September 17, 1984, when the trial began, Chief Judge Daly rendered his decision on the motion to suppress from the Bench (T104-06; “T” references are to pages of the stenographic transcript). He found that the warrant was “supported by probable cause” (T104), citing Illinois v. Gates, 462 U.S. 213, 103 S.Ct. 2317, 76 L.Ed.2d 527 (1983). As to whether the affidavit on which the warrant was granted was false in any respects, Chief Judge Daly ruled that, to the extent that there were any material omissions or misstatements in the affidavit, the evidence did not establish that “they were made deliberately or in reckless disregard for the truth” (T104), citing Franks v. Delaware, 438 U.S. 154, 98 S.Ct. 2674, 57 L.Ed.2d 667 (1978). As to the description of the search items in the warrant, the trial judge ruled that the description “was sufficient and more specificity could not properly be expected,” citing United States v. Balsamo, 468 F.Supp. 1363, 1391 (D.Me.1979). As to the execution of the warrant in seizing “non-financial items” (T106), the trial judge upheld the seizure. As to “financial or expenditure evidence____ seized pursuant to Agent Hoyt’s theory of plain view tax violation” (T105), the trial judge deferred a ruling pending disclosure by the government of which documents it planned to offer (T105). On September 19, 1984, Chief Judge Daly suppressed as evidence “those financial items ... seized pursuant to the Government’s theory that they were evidence in plain view of a tax violation” (T153). The suppressed items were those “listed in the government’s submission of yesterday’s date [September 18, 1984]” (T154). The “submission” is a part of the record (A54). The reasoning of Chief Judge Daly in reaching this result was set out by him orally (T153-54). During the trial, on October 2, the government filed a “Motion to Reopen Hearing on Motion to Suppress and for Clarification And/or Reconsideration of Ruling” (A61-63). The government urged reconsideration of the ruling on the “plain view exception to the warrant requirement” and also to consider, for the period after discovery (during the search) of the “Zimmerlis’ tax returns,” the theory of seizure “to document the ... expenditures in excess of income.” The trial court did reopen the hearing and, in the absence of the jury, took further evidence on October 2 and 3 (T1861-1931, 2075-2194). At the end of the hearing on October 3, the trial judge made his ruling. He found that the searching officers were properly on the premises and, of course, he knew that the cause of the application for a search warrant was the Tho importation, which had been halted and prevented the day before the issuance of the warrant. The trial judge found that, in the search for the Tho items, “the other items were uncovered and as to at least some of the specified documents under the plain view doctrine they were properly seizable as evidence of criminal activity” (T2193-94). He indicated that, as to two or three of the documents, he had “some problems” and that he would “rule with more specificity in the morning.” We have not found the further ruling thus anticipated, and we assume that those two or three documents were not admitted into evidence. 1(b) The argument for appellant Zimmerli seems in large part directed more to a motion to suppress evidence than to reversal of a criminal conviction. There is a general discussion of a “prohibited general search” (e.g., Brief, p. 13 and following) and of how “the ‘plain view’ exception does not apply” (e.g., Brief, p. 40 and following). It would, however, be difficult for us to see why, in view of the overwhelming evidence aside from that seized in the search at issue, a conviction of Zimmerli for numerous serious drug' offenses should be reversed because of a generalized claim of misconduct in the execution of a search warrant. Zimmerli points (Brief, p. 12) to some evidence seized in the search which was received in evidence. These are (a) pictures and letters establishing ownership by Zim-merli of an expensive home in St. Barthele-my in the French West Indies; (b) documents showing a financial interest of Zim-merli in sailboats used in some of the drug importations; (c) tax returns of Zimmerli and his wife, and associated records; and (d) notes of receipts from imports of marijuana and payments to participants in the smuggling. The items in (b) and (d) seem clearly within the warrant as records of dealing in controlled substances. The items in (a) seem within the warrant because they show associations between Zim-merli and his wife and others engaged in their marijuana operation; for example, the photographs show Anne Taylor with Mrs. Zimmerli and others (T2500; Anne Taylor was indicted and was alleged in the indictment to have been a crew member on sailboats used in the drug smuggling (A152)). The items in (c), principally the tax returns, are not so clearly within the warrant, but we understand from the record (T303) that they were found within an hour to an hour and a half after the search began and after some $280,000 in currency had been found in the house. They seem to us, therefore, to be evidence of a crime found by officers making a search under a valid warrant; the income tax returns showed income of the Zimmerlis for 1980 of $14,910, for 1981 of $31,900, and for 1982 of $51,330 (T304); to the officers, therefore, the expenditures obviously made by the Zimmerlis for houses, trips, automobiles, sailboats, and the like, plus a sum of $280,000 in currency found in the house, showed “expendi-tures____far in excess of [the income on their tax returns], a pretty good indication they filed a fraudulent return” (T305-06). The financial records would also be relevant in showing “substantial income or resources,” part of the definition of a “continuing criminal enterprise” (21 U.S.C. § 848(b)(2)(B)). Before discovery of the tax returns, Supervising Agents Hoyt and DiCarlo required that they approve any item seized by other agents; after discovery of the tax returns, all items in the financial records which showed expenditures by the Zimmer-lis were authorized to be seized by any agent as evidence of an offense within plain view (T303-07). We are told by the government, however, that “documents from this category were not offered at trial” (Brief, p. 36). We realize that, under Chapman v. California, 386 U.S. 18, 26, 87 S.Ct. 824, 829, 17 L.Ed.2d 705 (1967), it may well be that the burden of proof was on the government to show that any error in the admission of evidence seized in the search was “harmless” beyond a reasonable doubt, but we believe that the government, if it had such a burden, met the burden. 1(c) Appellant Zimmerli makes no argument here that the warrant was not supported by a showing of probable cause, believing that United States v. Leon, 468 U.S. 897, 104 S.Ct. 3405, 82