In June 1980, Francisco Santoya, Ricardo Esle, Luis Arango, and Roberto Diaz-Gomez arranged to sell undercover Drug Enforcement Administration (DEA) agents approximately ten kilograms of cocaine. The agents arrested Santoya, Esle, and Arango at the delivery point; they arrested Diaz at his home shortly thereafter. The arrestees were indicted and charged, in two counts, with conspiracy to possess cocaine with intent to distribute, in violation of 21 U.S.C. § 846 (1982), and possession of cocaine with intent to distribute, in violation of 21 U.S.C. § 841(a)l (1982) and 18 U.S.C. § 2 (1982). Santoya pled guilty under a plea agreement. Esle was tried separately from and prior to Diaz and Arango. Esle and Arango were convicted on both counts. Diaz was convicted of the conspiracy count. All three defendants appeal. We affirm.
I.
In March 1980, DEA Agent Rubin Prieto, operating undercover, met appellant Santo-*1468ya and attempted to arrange a cocaine buy through him. That attempt failed, but Prieto left a telephone number with Santo-ya in case Santoya could arrange a purchase.
Santoya called Prieto in June and informed him that he had a cocaine supplier willing to deal. Prieto and DEA Agent Zapata, posing as the buyer, met Santoya on June 10 in Miami. Santoya told them that he had set up a meeting for that afternoon in the Sheraton Hotel lounge between the two agents and another person, who would examine their money and deliver the cocaine.
Santoya and appellant Diaz, whom Santo-ya introduced as “Roberto,” arrived at the lounge at 5:20 p.m. and met the agents. Diaz discussed price, quantity, quality and method of delivery of the cocaine with the agents. They reached a bargain and agreed that the delivery would take place that evening. Diaz departed the lounge to arrange for the delivery; the others remained behind. He returned at 8:40 p.m. to say that the transaction could not be consummated because of heavy rain. The parties rescheduled the transaction for the next day, June 11, at noon, at the same place.
On June 11, Santoya appeared, but Diaz did not. The agents, afraid of countersur-veillance, indicated that they were going to check out of their motel and leave town. They said, however, that they would be willing to complete the cocaine buy on their way to the airport the next day, but at a different location, if Santoya could arrange it.
The next morning, the agents called San-toya and told him that they would be available at the Marriott Hotel at around noon. They then contacted DEA Agent Parrish to set up a surveillance of the hotel and its parking lot.
Santoya met Prieto and Zapata at the Marriott at about 12:45 p.m. Shortly thereafter, appellant Esle appeared, stating that he had two packages of cocaine to show and eight more that would be delivered if Prieto agreed. The cocaine, Esle indicated, was in a car at the parking lot. Esle went to get the cocaine. As he walked away, the agents noticed appellant Arango in front of the Marriott. Arango joined Esle, and they entered the parking lot. Prieto followed and saw Esle and Arango get into a blue Audi.
Agent Prieto caught up to them and asked what the trouble was and whether the packages were in the car. Esle responded that the cocaine was in the car. Arango indicated that he would not show Prieto even a gram of cocaine there because DEA agents were all around. Prieto asked him where they were, and Arango pointed out three surveillance vehicles. The two men told Prieto to get in the car and go elsewhere with them if he wanted to complete the transaction. When Prieto declined, Arango told Esle to leave. Esle began to start the car, and Prieto gave the signal for their arrest.
Agent Parrish and a Spanish speaking agent placed Esle and Arango under arrest. Agent Zapata, who had waited with Santoya in the hotel, arrested him. Esle and Arango denied ownership of the car. After looking inside the car and finding nothing, the agents opened the trunk of the car and found a briefcase. They later obtained a warrant to search the briefcase and found the cocaine; it was in a plastic bag contained within a brown paper bag. Arango’s fingerprints were discovered on the brown paper bag. The agents also found the title certificate to the blue Audi; it bore Diaz’ name. Diaz was arrested at his home several days later. On June 23, the four arrestees were indicted.
Prior to trial, the appellants moved the court to suppress the evidence derived from the warrantless search of the car trunk. After a hearing, the district court denied their motion, finding that exigent circumstances justified the search.
Appellants also moved the district court to dismiss the indictment because Hispanics were underrepresented in the grand jury venire (from which the grand jury that indicted appellants had been selected), in *1469violation of the fair cross-section requirement of the sixth amendment to the Constitution.1 The court consolidated appellants’ motion with similar motions in other criminal cases pending before the court and, after an evidentiary hearing, denied it because appellants failed to sustain their burden of proving the violation.2 United States v. Cabrera-Sarmiento, 533 F.Supp. 799 (S.D.Fla.1982). Appellants moved the court for reconsideration or, in the alternative, to reopen the evidence and order the U.S. Census Bureau to give them access to unreleased 1980 census data which, they alleged, would make out their sixth amendment claim. The court denied their motion summarily.
Arango thereafter moved the court to sever his case from that of Esle, contending that if they were tried separately Esle would testify for him. The court granted the motion and tried Esle alone, three weeks before Arango and Diaz. Before his trial with Arango began, Diaz moved the court for a severance of parties on the ground that his defense was irreconcilable with Arango’s. The court denied his motion.
At Esle’s trial, agents Zapata, Prieto, and Parrish testified to the occurrences described above. The parties stipulated that the substance the agents had uncovered in their search of the seized briefcase was cocaine. In his defense, Esle called a witness who testified that Esle had been in New York until late on the night before his arrest, implying that Esle could not have been a knowing participant in a cocaine transaction in Miami the following day. Esle did not take the stand. The jury convicted him.
At Diaz’ and Arango’s trial, the three agents again testified. Esle, also testifying for the prosecution, said that Diaz had lent him the car on the day of the arrests. Finally, two DEA agents and an FBI fingerprint expert established that Arango’s fingerprints had been found on the brown bag containing the package of cocaine.
Diaz’ defense was that he was not the “Roberto” who had met with the DEA agents at the Sheraton Hotel on June 10. He did not testify, but he produced witnesses who said that he had been at a doctor’s office with his brother and then at home at the time the agents claimed to have spoken with him at the Sheraton Hotel. To counter the government’s proof that he owned the blue Audi that had transported the cocaine to the Marriott Hotel parking lot, Diaz presented evidence that he had sold the automobile to Esle sometime earlier.
Arango’s defense was that he was an innocent bystander. Taking the witness stand, he readily admitted that he had been in the blue Audi with Esle but insisted that he knew nothing about a drug deal. According to Arango, the two men were together because Esle wanted a job in Aran-go’s gift shop, and they were on the way to lunch to discuss the job. The agents, he said, were lying when they testified that he said he would not show them any cocaine because the DEA had them under surveillance. Arango attempted to explain the presence of his fingerprints on the brown bag containing the cocaine by stating that the bag had come from the gift shop he ran.
In their closing arguments to the jury, both defendants attacked the credibility of the law enforcement officers. They suggested that, if the agents had been telling the truth, they would have corroborated their stories through “body bug” tape recordings of the transactions. The jury convicted the two men.
II.
In this consolidated appeal, appellants ask us to set aside their convictions, and to *1470direct the district court to dismiss the indictment, on the ground that the grand jury that indicted them was drawn from a jury venire lacking sufficient numbers of Hispanics to satisfy the fair cross-section requirement of the sixth amendment. Alternatively, appellants seek a new trial. All contend that their trials were unduly prejudiced because the cocaine that gave rise to their indictment and convictions was inadmissible in evidence, having been seized by the DEA in violation of the fourth amendment. Diaz and Arango present additional reasons for a new trial. Diaz contends that the district court should have severed his trial from Arango’s, because their defenses were antagonistic, and that the court’s failure to do so denied him a fair trial. Arango contends that prosecu-torial misconduct rendered his trial unfair. We address these arguments in turn.
A.
Appellants claim that the system for choosing grand jury venire members in the Southern District of Florida, Miami Division, violated their right to a grand jury representing a fair cross-section of the community, created by the sixth amendment and the Jury Selection and Service Act of 1968, 28 U.S.C. § 1861 (1982).3 They contend that the district court erred in two respects in addressing this claim. First, the court erred in denying their motion to dismiss the indictment because Hispanics were unconstitutionally underrepresented. Second, the court abused its discretion in denying their motion for reconsideration and for access to unreleased 1980 census data.
1.
In order to establish a prima facie fair cross-section violation, appellants were required to demonstrate:
(1) that the group [i.e., Hispanics] alleged to be excluded [was] a “distinctive” group in the community; (2) that the representation of this group in [the ve-nire] from which [appellants’ grand] jur[y was] selected [was] not fair and reasonable in relation to the number of such persons in the community; and (3) that this underrepresentation [was] due to systematic exclusion of the group in the jury-selection process.
Duren v. Missouri, 439 U.S. 357, 364, 99 S.Ct. 664, 668, 58 L.Ed.2d 579 (1979);4 The district court, after hearing appellants’ evidence, found that Hispanics were a distinctive group in the community, but concluded that appellants had not shown unreasonable underrepresentation of Hispanics in the grand jury venire. We move directly to the court’s conclusion that appellants showed no unreasonable underrepresentation of Hispanics because it is determinative.
The district court held that appellants failed to establish the second prong of the Duren test because they had not provided *1471the court with any reliable data from which it could find significant underrepresentation of Latins in the grand jury venire. Appellants challenge this conclusion. They contend that the evidence they produced at the evidentiary hearing was plainly sufficient as a matter of fact to establish this critical number and that the district court erred as a matter of law in concluding that a sixth amendment violation had not occurred. The correctness of the district court’s factual premises and its conclusion becomes apparent when one considers the evidence appellants presented to the court and the court’s findings with respect to that evidence.
At the hearing on the motion to dismiss the indictment, appellants presented testimony by Dr. John B. McConahay, a sociology professor at Duke University, and Dr. Abraham Lavender, a sociology professor at the University of Miami. Dr. Lavender did not testify regarding the underrepre-sentation of Hispanics; he only addressed whether they were a cognizable group, concluding that they were.
Dr. McConahay described the population in the area served by the district court’s Miami division (Dade and Collier Counties, from which appellants’ grand jury had been drawn) and that population’s representation on Miami division grand juries. Having no independent knowledge about the makeup of that population, he relied on and published figures from two sources not introduced into evidence, a Dade Latin Market Survey and some preliminary 1980 census data.5 Dr. McConahay gave his sources’ figures for the total population and the total Latin population and then estimated the eligible Latin population as follows. The preliminary census data contained no breakdown of the population by age and citizenship. The Dade Latin Market Survey did attempt such a breakdown. Using the Dade Latin Market Survey breakdown of age and citizenship, Dr. McConahay adjusted the preliminary 1980 census data for Dade County and drew an estimate of the eligible Latin population in the community. He endeavored to correct the preliminary census data only for two juror qualifications, age and citizenship. He did not attempt to control the data for other juror qualifications, i.e., proficiency in English6 or absence of a criminal record. He concluded that eligible Latins constituted from 24.2 to 26.6 percent of the eligible population.
Dr. McConahay then took a sample of jurors from the district court’s master jury wheel (containing the group of people to whom the court mailed juror forms),7 counted the Latin and non-Latin members of the group, and used these numbers to estimate the percentage of Latins on the master wheel. He estimated that 13.2 percent of those on the master wheel were Latin. From the same sample, he counted the Latins and non-Latins who had returned the juror qualification forms mailed them and were not excused or otherwise unavailable; he used this sample to estimate the percentage of Latins on the court’s qualified jury wheel (containing those actually available for jury service). He estimated *1472that 12.9 percent of those on the qualified wheel were Latin.
As we have stated, Dr. McConahay presented these figures, calculated purely to reflect Dade County,8 at the evidentiary hearing on appellants’ motion to dismiss the indictment. He subsequently presented another set of tables he calculated after the first set.9 In the second set he adjusted the preliminary 1980 census data to reflect both Dade and Collier Counties. He also adjusted the preliminary census data and the Dade Latin Market Survey data to exclude from the earlier calculation people over age 64.10
Dr. McConahay’s initial figures from the preliminary 1980 census data showed absolute disparities between the qualified wheel and the eligible population of 11.3 to 13.7 percent. The Dade Latin Market Survey data indicated a 13.1 percent disparity. When Dr. McConahay, in his second set of tables, adjusted both sets of data to exclude those over age 64, the disparities indicated by the Dade County Market Survey increased to 14.3 percent. The preliminary 1980 census data is more difficult to interpret here because Dr. McConahay adjusted it simultaneously to exclude those over age 64 and to include Collier County. That adjusted data indicated an 11.4 to 13.9 percent disparity.11 In his calculation of all figures, Dr. McConahay did not indicate any sampling error.
The government, in response, first attacked the primary foundation of Dr. McConahay’s testimony, the Dade Latin Market Survey. The government pointed out that the survey had been commissioned by area radio stations that broadcast in Spanish or had a sizeable Latin listening audience, and that the stations had an interest in presenting their listening pool in a light attractive to advertising. Second, the government presented a statistician who testified that a sampling error of plus or minus four percent applied to each of Dr. McConahay’s figures.
The district court, after hearing the evidence, decided to accord Dr. McConahay’s opinion as to the number of Latins eligible for jury duty in the Miami Division no weight.12 The court noted that this number was not a product of Dr. McConahay’s independent knowledge and expertise, nor was it based on officially recognized figures which broke down the area Latin population by age, citizenship and *1473English language ability. 533 F.Supp. at 804. Bather, his disparity number was merely a reflection of the Dade Latin Market Survey figures. The Dade Latin Market Survey, because it was commissioned by Miami Spanish-language radio stations, was of questionable objectivity. Id. The court took judicial notice that many Latins in the Miami area would be ineligible for jury duty for many conceivable reasons including lack of citizenship and inability to speak adequate English. Id. The court thus treated Dr. McConahay’s numbers as if it were being called upon to make the Fed.B.Evid. 104 threshold factual findings 13 necessary to determine the survey’s admissibility into evidence to show the number of Hispanics in the Miami Division eligible for jury service. The court concluded that the survey had no evidentiary value because it failed to provide the number of eligibles, and moreover because it had been conducted in a biased manner for a purpose unrelated to the issue before the court. The court thus accorded the figures Dr. McConahay derived from the survey no weight. In sum, appellants had no evidence showing the number of eligible Hispanics in the Miami Division. Inasmuch as this was an essential element of their claim, their claim necessarily failed.
Appellants attack the court’s rejection of Dr. McConahay’s testimony and the court’s conclusion that they proved no underrepresentation in three ways. We consider their arguments in turn.
Appellants’ first contention ironically ignores their strenuous attempt to prove, at the evidentiary hearing, the number of eligible Latins in the Miami Division. Appellants now contend, relying on Duren, supra, and Castaneda v. Partida, 430 U.S. 482, 97 S.Ct. 1272, 51 L.Ed.2d 498 (1977), that the relevant community for fair cross-section purposes was the entire community, not just those eligible for jury duty. Accordingly, the Latin population’s citizenship, age, and English language ability characteristics were not relevant to a factual showing of underrepresentation and the trial judge erred in ignoring that portion of Dr. McConahay’s testimony based purely on the preliminary 1980 census data which established the gross population numbers.
Their argument has no merit. We need not decide whether the second prong of Duren, and Castaneda, requires a party challenging grand jury selection procedures to prove that the members of the group alleged not to be fairly represented were eligible for jury duty instead of merely showing the population of the entire group. Neither Duren nor Castaneda speak to this distinction, but Chief Justice Burger in his dissent in Castaneda said: “The decisions of this Court suggest, and common sense demands, that eligible population statistics, not gross population figures, provide the relevant starting point.” 430 U.S. at 504, 97 S.Ct. at 1285.
We decide this case without attempting to distinguish Castaneda and Duren because of the several other obvious reasons for the correctness of the district court’s decision. The appellants do not meet the second prong of the Castaneda test in any event. The majority opinion in that case, 430 U.S. at 494, 97 S.Ct. at 1280, says: “Next, the degree of underrepresentation must be proved, by comparing the proportion of the group in the total population to the proportion called to serve as grand jurors, over a significant period of time.” (Emphasis added). An Eleventh ' Circuit opinion, United States v. Perez-Hernandez, 672 F.2d 1380 at 1386 says: “the group must be substantially underrepresented in the office of grand jury foreman over a significant period of time.” Clearly, appellants fail to meet the second prong of the Castaneda test because statistics for only one year were submitted.
*1474Neither do the appellants meet the third prong found in both Castaneda and Du-ren, the latter holding that such “underrep-resentation [was] due to systematic exclusion of the group in the jury-selection process.” Duren v. Missouri, 439 U.S. 357, 364, 99 S.Ct. 664, 668, 58 L.Ed.2d 579 (1979). Appellants submitted no evidence to show that the court’s grand jury selection procedure was susceptible of abuse or that it was not racially neutral.
Appellants’ second argument is also without merit. Appellants contend that the Dade Latin Market Survey showed the number of Latins eligible for jury service in the Miami Division and that the court thus erred in rejecting Dr. McConahay’s testimony on the theory that the Dade Latin Market Survey on which his testimony was based was inadequate to prove that number. As we have indicated supra, the court in effect determined under Fed.R. Evid. 104 that the survey was inadmissible because it was unreliable and did not show the number of eligible Latins.
Appellants first challenge the court’s rule 104 type preliminary findings that the study was of questionable objectivity and unofficial. We review these factual findings under the clearly erroneous standard.
A study may lack objectivity when the party who makes it has an interest in the outcome. The study in question was commissioned by the area radio stations with Spanish programming. Certainly these stations had an interest in finding a large group of Hispanics: the larger their group of potential listeners, the greater the number and the higher the price of the advertising spots they would be able to sell. Accordingly, the finding that the study’s objectivity was questionable was not clearly erroneous. Regarding the court’s preliminary finding that the study was unofficial, evidence showed the study to havé been created and conducted only by private com-mereial entities. Further, no evidence in the record before the court showed the study to be “official.” Therefore, the finding that the study was “unofficial” is likewise not clearly erroneous.
Appellants next question the court’s rule 104 type conclusion that the survey did not permit it to infer the number of eligible Hispanics. We agree with the district court The court took judicial notice under Fed.R.Evid. 201(b)(1) that many Hispanics within the area either did not speak English, were not citizens, or lacked some other prerequisite for jury duty.14 The survey only adjusted the data for age and citizenship. Thus, even assuming that the survey was objective, the survey did not permit the calculation of the number of eligible Latins.
Appellants’ third argument is that, even assuming the court correctly concluded that the Dade Latin Market Survey was not entitled to any evidentiary weight, the court erred in disregarding Dr. McConahay’s opinion. This argument borders on the frivolous. We accord the district court, sitting as a fact finder, great discretion in considering the weight, if any, it shall give expert testimony. In making such a determination, a trial court as fact finder “need not be bound by expert testimony even if all of the witnesses are presented only by one side.” United States v. Pitts, 428 F.2d 534, 536 (5th Cir.), cert. denied, 400 U.S. 910, 91 S.Ct. 154, 27 L.Ed.2d 149 (1970). Where, as here, the basis of the expert’s opinion has been thoroughly impeached, the court, as fact finder, is thus plainly authorized to reject the opinion entirely. Accordingly, we find no error in the court’s rejection of Dr. McConahay’s testimony, and thus its conclusion that appellants failed to establish a factual disparity number to which the court could apply sixth amendment law.
*1475Appellants emphasize to this court that their expert witness used the “best available” figures in establishing his opinion evidence and that the figures were “totally unrebutted.” The fact that these figures were the best he could find does not mean that they automatically were sufficient to show unreasonable underrepresentation. Nor did the government’s failure to rebut those figures make them sufficient to show underrepresentation; the government had no obligation to rebut insufficient figures in order to prevail. The court heard the appellants’ presentation and the government’s cross-examination and evidence pointing out the weaknesses in appellants’ opinion evidence. The court then concluded that the appellants had not shown underrepresentation. The court correctly applied the law to the evidence before it; accordingly, we affirm its ultimate conclusion.
2.
After the court denied the appellants’ motion to dismiss the indictment, appellants moved the court to reconsider or, in the alternative, to order that they be provided access to a set of 1980 census figures not publicly available, which appellants alleged existed, could be processed, and would yield support for their position. This was in the nature of a motion to reopen the evidence and was addressed to the sound discretion of the trial judge; we must not reverse this determination unless we find an abuse of discretion.
We find that the trial judge did not abuse his discretion for several reasons. First, appellants did not make any showing concerning the availability of the evidence, the nature of the evidence, the time it would take to process it, or what the ultimate value of the figures might be. They could have presented such information to the court in support of their motion to reopen the evidence; they could have obtained affidavits from U.S. Census Bureau personnel regarding all these matters, or could have requested a hearing on their motion and subpoenaed such personnel. Second, appellants gave no excuse for not presenting the evidence earlier; they stated only that they had considered their case strong enough so that they chose not to try to obtain it. The court thus faced a request for access to figures which might or might not exist, which might take substantial time for processing and hold up the disposition of any other affected cases, and which, ultimately, might add nothing to appellants’ argument, all because they had considered their case so strong that they did not choose to investigate the evidence earlier. The denial of such a request plainly was not an abuse of discretion.
We would so conclude even if we assumed that appellants demonstrated that the 1980 census data were not available earlier; for to grant appellants relief would have been to give them an indefinite open-ended continuance for whatever time it might take to obtain and assess the data. The interests of all parties in the speedy processing of criminal cases counsels against the granting of such a continuance. At some point the trial judge must draw a line, allowing the case to proceed to a conclusion. We cannot say that the judge drew that line improperly here.
B.
Appellants next argue that the district court erroneously denied their motion to suppress the cocaine found in the briefcase in the trunk of the car. They concede that, if the arresting officer had probable cause to believe that contraband was in the car, he had the right to search all parts of the car, including the trunk, under the principles enumerated in United States v. Ross, 456 U.S. 798, 102 S.Ct. 2157, 2162-64, 72 L.Ed.2d 572 (1982), (recognizing that “[gjiven the nature of an automobile in transit ... an immediate intrusion is necessary if police officers are to secure the illicit substance [which they have probable cause to believe is carried-in the car],” and *1476Carroll v. United States, 267 U.S. 132, 45 S.Ct. 280, 69 L.Ed. 543 (1925).15
Appellants argue that while Agent Prieto may have had probable cause to believe the cocaine was in the car, Agent Parrish, who made the arrests and searched the car, had no such probable cause. Accordingly, they argue, the search violated their fourth amendment rights. We disagree.
Appellants have not considered the well-recognized principle that, where a group of officers is conducting an operation and there is at least minimal communication among them, we look to the collective knowledge of the officers in determining probable cause. United States v. Astling, 733 F.2d 1446, 1460 (11th Cir.1984); United States v. Agostino, 608 F.2d 1035, 1037 (5th Cir.1979). Given Prieto’s knowledge that the cocaine was somewhere in the car, his contact with Parrish in setting up the search, his presence in the vicinity at the time the trunk was opened, and Parrish’s testimony that another agent told him that the cocaine was in the ear, we find ample communication between Prieto and Parrish to apply this principle. Probable cause to search the entire automobile existed.
C.
Diaz argues that the trial court abused its discretion in denying his motion for a severance under Fed.R.Crim.P. 14 because his defense was antagonistic to that of Arango, with whom he was tried. We first note that, generally, coconspirators should be tried jointly. United States v. Astling, 733 F.2d 1446, 1454 (11th Cir. 1984). The trial judge’s decision to sever is discretionary; to demonstrate an abuse of discretion, Diaz must demonstrate that he “suffered compelling prejudice against which the trial court was unable to afford protection.” United States v. Russell, 703 F.2d 1243, 1247 (11th Cir.1983) (quoting United States v. Berkowitz, 662 F.2d 1127, 1132 (5th Cir. Unit B 1981). To show prejudice resulting from antagonistic defenses, Diaz must convince us that the defenses were antagonistic “to the point of being mutually exclusive ... or irreconcilable.” United States v. Riola, 694 F.2d 670, 672 (11th Cir.) (quoting United States v. Berkowitz, 662 F.2d 1127, 1133 (5th Cir.1981), cert. denied, 460 U.S. 1073, 103 S.Ct. 1532, 75 L.Ed.2d 953 (1983)). This threshold is reached, and severance is compelled, “if the jury, in order to believe the core of testimony offered on behalf of [a] defendant, must necessarily disbelieve the testimony offered on behalf of his co-defendant.” United States v. Berkowitz, 662 F.2d at 1134.
The evidence in this case did not create such an irreconcilable conflict. Diaz’ and Arango’s strategies both acknowledged they had never met or conspired together. Diaz tried to show that, in effect, he was the wrong “Roberto”. Diaz put on witnesses who established an alibi for him and a witness who testified that he looked like a Roberto Reduello, the owner of the house Diaz rented, in whose name the agents mistakenly had made out the initial arrest warrant. Arango’s strategy *1477was to show that the scheme was complete without him, that he was an innocent bystander. Arango did not know whether Diaz was the correct “Roberto” or not. In closing argument to the jury, he merely referred to the seller as “Roberto”, as had the agents in their testimony. We do not find these defenses to have been irreconcilable enough to have shown the compelling prejudice necessary for us to conclude that the trial judge abused his discretion in denying the motion to sever.
D.
Arango asks us to conclude that prosecutorial misconduct denied him a fair trial and thereby reverse his conviction. In evaluating a prosecutorial misconduct claim, we will only reverse if, viewed in the context of the entire trial, the misconduct may have prejudiced substantial rights of the accused. United States v. Johns, 734 F.2d 657, 661 (11th Cir.1984); United States v. Bosby, 675 F.2d 1174, 1185 (11th Cir.1982).
Arango points to three specific instances of misconduct; he objected to all of them. First, he alleges prosecutorial misconduct in a question posed to his wife on cross-examination, seeking Arango’s birthplace. The defense promptly objected, arguing that the question was not relevant to any issue in the case. The court agreed and, after excusing the jury, chastised the prosecutor. When the jury returned, the court gave the following instruction:
Now, ladies and gentlemen you will remember that the Judge told you earlier that the testimony that you consider comes through the witness or from documents or stipulations and questions of lawyers are not evidence.
As far as this lady’s testimony goes, it is absolutely immaterial where this defendant was born. I hope you understand that. I cannot, for the life of me at this stage in this case see what difference it would make where he was born, so I want you to understand that because I believe that was the question that was asked which received no answer. It does not matter where he was born as far as the evidence in this case, so far, is concerned. That is all I need to tell you at the moment.
Little, if any, prejudice resulted from the unanswered question; whatever lingering prejudice may have existed was cured by the instruction. See United States v. Capo, 693 F.2d 1330, 1335 (11th Cir.1982), cert. denied, 460 U.S. 1092, 103 S.Ct. 1793, 76 L.Ed.2d 359 (1983).
The second instance of misconduct Arango complains of was likewise cured by instruction. Defense counsel argued that, although Esle was called by the government at Arango’s trial, the government did not question Esle about Arango. Therefore, counsel concluded, the government must have known that Esle would testify that Arango was innocent. Counsel then argued that, “[i]f Santoya would have been able to , say [Arango] was involved, the Government would have put him on the stand.”
In rebuttal, the prosecutor tried to argue that witness subpoenas were available to both sides. He stated: “Both sides have subpoena powers and they kept asking why didn’t the Government subpoena ... the natural question [is] why didn’t the defense subpoena.” When defense counsel objected, the court instructed the jury:
You ladies and gentlemen must bear in mind that the Government has the entire burden of proof. The defendant has no obligation to call any witnesses or produce any evidence because of the presumption of innocence.
Mr. Weinstein wants me to have you recall that the Government has the entire burden of proof.
What Counsel [for the Government] says is true. There have been witnesses testifying for the defendants and subpoena power is available to the defendants, and that is entirely correct, but you must bear in mind that the defendant has no burden of proof. The Government has the entire burden of proof.
*1478Arango argues that this instruction, far from removing prejudice, compounded it by telling the jury that the defense had subpoena power. Arango misconstrues the nature of the prejudice threatened by the argument. It is not prejudicial for the jury to know that witnesses may be called by either side. See United States v. Vincent, 648 F.2d 1046, 1051 (5th Cir. Unit A 1981). The prejudice threatened by the argument was that the jury might interpret the statement, “why didn't the defense subpoena,” as shifting the burden of proof on the guilt/innocence issue to the defense. The jury was promptly disabused of any such notion by the above instruction.
Arango’s final assertion of prosecutorial misconduct is based on another rebuttal argument. The prosecutor was trying to respond to arguments by Arango that none of the witnesses had even heard of him until the final delivery of the cocaine and that he thus could not have been involved in the transaction. The prosecution stated:
Lastly, Mr. Arango who never appeared until the last few seconds, and the government never tried to tell you otherwise, he only appeared the last few seconds, but in a conspiracy situation you are never going to know everybody in a conspiracy. We do not know who owned the cocaine. We probably will never know.
We are never going to know exactly what role that anyone was playing, but that is not important.
The fact that they were there and they were involved and the man you see is the least is usually the man who is on top.
[Arango’s attorney]: Objection, Your Honor, totally improper argument.
THE COURT: Complete your argument, please, sir.
[The prosecutor]: The man who has put layers of people between him, the man who has exposed himself the least. (Emphasis added.)
This was the only argument made on the subject. Only the last two sentences (emphasized) were questionable. They were, however, merely common sense inferences that the jury was entitled to draw from the evidence. The prosecutor did not make an elaborate argument that Arango himself was the kingpin of the operation or that he had committed other crimes. .The prosecutor was simply illustrating a common sense idea that, even though a member of a conspiracy is not seen frequently, he can still be an active member. Nowhere did the prosecution argue that Arango was in charge of the conspiracy. We do not find these two statements, in the context of the entire trial, to have been sufficiently prejudicial to warrant reversal.
AFFIRMED.
. Const, amend. VI states in relevant part, "In all criminal prosecutions, the accused shall enjoy the right to a ... trial ... by an impartial jury of the State and district wherein the crime shall have been committed____” The Supreme Court has interpreted this amendment to require that juries be composed of a fair cross-section of the community. Taylor v. Louisiana, 419 U.S. 522, 530, 95 S.Ct. 692, 697-98, 42 L.Ed.2d 690 (1975).
. See infra note 4.
. 28 U.S.C. §§ 1861-1869 (1982) provides for the random selection of grand and petit jurors from a fair cross section of the community in the district or division wherein the court convenes. Pursuant to a written plan promulgated under 28 U.S.C. § 1863 (1982), the names of potential jurors are randomly selected from voter lists and placed in a master jury wheel. The court may adopt other sources of names in addition to voter lists where necessary to secure a fair cross section of the community. From time to time the court publicly draws at random from the master jury wheel the names of as many persons as may be required for jury service. The people selected in the second drawing receive juror qualification forms which they must complete and return to the clerk of the court. The court determines solely on the basis of information provided on the juror qualification form and other competent evidence whether a person is unqualified for, or exempt, or to be excused from jury service. Reasons for disqualification include the lack of citizenship, conviction of a felony, and inability to understand the English language. The test for a sixth amendment fair cross-section violation applies in determining a violation of the statutory fair cross-section requirement. United States v. Perez-Hernandez, 672 F.2d 1380 (11th Cir.1982).
. The burden of proving a sixth amendment violation by a preponderance of the evidence was on the appellants. The High Court has broken down this burden into two conceptual phases, the "prima facie case” quoted above, and the government’s rebuttal of that case by evidence that the disparity was justified. Duren, supra.
. Dr. McConahay testified from a report he had prepared for the evidentiary hearing. The population figures and disparity calculations contained therein were drawn from the Dade Latin Market Survey and some preliminary 1980 census data. The report was admitted into evidence. We note that the government could have objected on best evidence grounds to the appellants' attempt to bring the Dade Latin Market Survey data and the preliminary census data into evidence by having Dr. McConahay publish them. The government could also have objected on hearsay grounds to the Dade Latin Market Survey data because appellants put no witness with personal knowledge of the survey’s compilation on the stand and because the survey was not a public record or report, see Fed.R.Evid. 803(8), a market report or commercial publication, see id. at (17), or otherwise a document independently admissible under the rules of evidence. The government made none of these objections, nor did it move to strike Dr. McCo-nahay’s publication of the data.
. While some proficiency in English is required for citizenship, the jury service statute does not necessarily incorporate the same proficiency requirement.
. See supra note 3 for the statutory procedure for making up the master jury wheel.
. As we have indicated in the text supra, Dade and Collier Counties constitute the Miami Division of the Southern District of Florida. Collier County constituted three percent of the population; Dade County made up the other 97 percent.
. These tables were based on the same sources Dr. McConahay used in preparing his first report. See supra note 5.
. The appellants give no explanation of why, in the revised figures, they chose age sixty-four as a cut-off point. We cannot determine any rational basis for so narrowing the data.
. The Dade County data, adjusted for age, showed a much greater jump in disparity than did the preliminary 1980 census data, with Collier County included and adjusted for age. This suggested that the Collier County data, without the age adjustment, would have actually reduced Dr. McConahay’s initial estimate of the disparity between the percent of Latins qualified in the population and those on the qualified wheel.
. The court concluded that the figures Dr. McConahay provided were not sufficiently reliable to establish the second prong of the Duren test; i.e., unreasonable underrepresentation. This was a mixed question of law and fact; the court’s determination of whether any disparity has been shown, and if so its extent, is a question of fact, and the sufficiency of that disparity, once its value has been determined, to show a sixth amendment violation is a question of law.
The district court stated regarding the evidence before it.on the Latin population, “I find, however, that the figures presented by the defendants to show underrepresentation by Latins on grand juries ... are deficient ... [and] that the defendants have failed to prove underrepre-sentation____” Cabrera-Sarmiento, 533 F.Supp. at 804 (emphasis added). We believe this formulation shows the district court to have concluded in its fact-finding capacity that no under-representation was shown by Dr. McConahay’s opinion because the figures on which it was based were too unreliable. It does not show the court to have concluded that some underrepre-sentation occurred but that as a matter of law it was not unreasonable. We review the court’s determination accordingly.
. Fed.R.Evid. 104 provides in relevant part: "(a) Questions of admissibility generally. Preliminary questions concerning ... the admissibility of evidence shall be determined by the court____” Had the appellants offered the survey into evidence the court would have been required to determine whether the survey had been conducted in such a manner as to produce a reliable estimate of each number appellants sought to prove.
. Appellants contend that the court erred in so doing. This contention is meritless. Fed.R. Evid. 201(b)(1) provides that an adjudicative fact may be judicially noticed if it is "generally known within the territorial jurisdiction of the trial court.” It was certainly common knowledge within the Southern District of Florida that many Hispanics in the area, like many Caucasians in the area, would have failed to meet one or more of the qualifications for jury duty, such as age, citizenship, or ability to speak English.
. We note that the fact that the men driving the car were arrested in the parking lot did not reduce the danger involved in leaving contraband in a vehicle capable of motion. The owner of the car, as well as any other unknown players in the scheme, could have tried to retrieve the cocaine while the officers awaited issuance of a warrant.
In addition, the Supreme Court noted in Ross: [The] decision in Carroll was not based on the fact that the only course available to the police was an immediate search____ The Court in Chambers v. Maroney, 399 U.S. 42, 90 S.Ct. 1975 [26 L.Ed.2d 419] ... —with only Justice Harlan dissenting — refused to adopt a rule that would permit a warrantless seizure but prohibit a warrantless search. The Court held that if police officers have probable cause to justify a warrantless seizure of an automobile on a public roadway, they may conduct an immediate search of the contents of that vehicle. “For constitutional purposes, we see no difference between on the one hand seizing and holding a car before presenting the probable cause issue to a magistrate and on the other hand carrying out an immediate search without a warrant. Given probable cause to search, either course is reasonable under the Fourth Amendment.” Chambers v. Maroney, 399 U.S., at 52, 90 S.Ct., at 1981.
Id. 102 S.Ct. at 2163 n. 9.