BIGGS, Chief Judge.

This is an appeal from a judgment for the defendant in a personal injury action brought under the Jones Act, 46 U.S.C.A. § 688, and tried by a jury.

I.

The first trial of this case in 1946 resulted in a verdict for the defendant, D.C. W.D.Pa. 1947, 70 F. Supp. 1016, which was reversed by this court because of error in the admission of certain evidence. 3 Cir., 1948, 165 F.2d 777. Before the constitution of the jury for the new trial of this cause in 1950, the plaintiff challenged the array of jurors by a motion to strike alleging that the entire panel had been improperly selected, but this motion was denied by the Chief Judge of the District Court without hearing evidence. W.D.Pa. 1951, 100 F. Supp. 494. Just before commencement of the trial, a supplemental motion to strike the panel because it had allegedly been subjected to surveillance and investigation by unauthorized persons was filed, but this motion also was summarily denied by the District Judge assigned to hear the case. At the trial that followed, a verdict was again rendered for the defendant. The plaintiff then moved for a new trial on the ground, among others, that the court erred in failing to grant a hearing on the plaintiff's original and supplemental motions to strike the jury panel. Relying on the ruling of the Chief Judge on the original motion and on his own previous disposition of the supplemental motion, the District Judge denied the plaintiff a new trial. D.C. W.D.Pa. 1951, 100 F. Supp. 493.

On appeal, after deciding that the other points raised did not constitute reversible error, this court directed that "[t]he judgment of the District Court * * * be vacated in order that the case may go back for hearing on the plaintiff's objections to the selection of jurors and the alleged intimidation of them" inasmuch as a factual basis for the disposition of these objections was necessary. To facilitate the hearing on remand, this court then stated that the plaintiff's case could "be properly developed if * * * [she] shows present conditions of the type * * * [she] claims existed at the time of these motions and then shows continuity between the methods used now and the practices of that time." Dow v. United States Steel Corp., 3 Cir., 1952, 195 F.2d 478, 481, 482.

At the hearing which ensued before the court below on the method of selection of the jury panel, counsel for both parties with the approval of the court agreed that the "[c]ourt * * * take the record in the Nelson case as the basis for its findings and * * * adopt the evidence, the exhibits, the rulings of the [c]ourt, and the decision of Judge Stewart as its own." The Nelson case was a criminal prosecution in the court below for violation of the Smith Act, 18 U.S.C. § 2385, in which there had been a hearing and evidence taken on a similar motion to strike the jury panel, and that motion had been denied by Judge Stewart in an opinion which was filed shortly before the agreement in this case was made. United States v. Mesarosh, D.C.W.D.Pa. 1952, 13 F.R.D. 180. On the issue concerning surveillance of the jury, evidence was taken at the hearing in this case. Thereafter, the court below ruled on the first issue that the method of jury selection was proper, adhering to the conclusion of Judge Stewart without making separate findings, and on the second issue, held that there had been no intimidation of the jurors. D.C.W.D.Pa. 1952, 108 F. Supp. 88. This appeal followed.

II.

We shall first discuss the questions, raised by the plaintiff's motion, concerning the methods of selection of jurors in the Western District of Pennsylvania. The evidence adduced to establish these methods pertains primarily to a period after the trial at which the plaintiff challenged the array assembled to try her case. However, this court in remanding for a hearing on the challenge consented to the use of evidence of selection methods at times after the plaintiff's trial if continuity of practices could be shown. Although the plaintiff has not established such continuity, she has chosen to rest her case on the later practices and the defendant, who also consented, is in no way prejudiced since it is evident that any change that did occur in the selection methods during the interim was a change for the better so far as selection procedures go. Consequently, we will consider the case primarily on the basis of the evidence of the later practices now before us.

In the selection of jurors, the clerk of court, or his deputy, and a commissioner, appointed by the court as a "well known member of the principal political party" opposing the party of the clerk or his deputy, are the responsible officials. 28 U.S.C. § 1864. Jurors, to be qualified for selection, must be twenty-one years old, citizens and residents of the judicial district, and must have the usual physical, mental and linguistic capacities, must not have been convicted of a felony, and must be competent "to serve as a grand or petit juror by the law of the State in which the district court is held." 28 U.S.C. § 1861. The qualifications for jury service under the laws of Pennsylvania, which are pertinent here, are substantially the same as those under the federal law. See 17 Purdon's Pa.Stat. Anno. § 1279 and § 1333. From a list of qualified individuals compiled by the clerk or his deputy and the jury commissioner, these officials draw by chance as prescribed by statute the names of the individuals to constitute particular jury panels. See 28 U.S.C. § 1864.

In this case, it is the method of selection of names making up the master list from which panels are drawn that is important. Evidence on this issue was primarily elicited from five witnesses who had served either as the jury commissioner, the clerk of court, or the clerk's deputy. Edward Snodgrass, Jr. testified that he had served as jury commissioner for thirteen years prior to July 31, 1950, when he left office. During that period, he maintained his lists by soliciting from various sources the names of prospective jurors. Each such prospective juror was sent a form questionnaire, and if that was returned and revealed that the individual was qualified, his name was added to the master list. Initially, the sources for the solicitation of names were largely the personal acquaintances of Snodgrass in the District. He requested names from people he had known in business. Moreover, names were obtained from various organizations where he had contacts, such as the Chamber of Commerce, banks, a post of the American Legion, and an association of the Veterans of Railroads. Snodgrass was a member of the Free and Accepted Masonic Order and was a trustee of the diocese of the Episcopalian Church, and names were obtained from both of these sources. In addition, he procured names from postmasters throughout the District. The number of names that were obtained from all of these sources remained undetermined at the hearing except that it was shown that the names of twenty-six individuals came from the banking contacts. However, the greatest source of names, according to Snodgrass, was the former jurors from whom new names were obtained in the following manner: "If I needed names in Lawrence County, for instance, I would particularly locate a person or persons that seemed very intelligent, that understood what they were doing, and then I would ask that person to give me names * *." The next greatest source was volunteers for jury duty who were of "all types." When he was asked several times as to whether he solicited the names of particular types of individuals, Snodgrass replied: "I don't know that they were. I wasn't looking for any particular kind or type, * * * — just that they were eligible citizens and in the places where we needed names out of the eighteen counties in Western Pennsylvania"; "I was not looking for any one class"; "I didn't ask any organizations as organizations. That wasn't my purpose. I only wanted enough names to make the necessary panel, and when I had them that is all there was to it." Snodgrass testified that he did not solicit names from labor unions and that the only source he knew that submitted the names of union men was two former jurors employed at a local steel plant. In regard to obtaining the names of Negroes, Snodgrass had contacts with members of that race through business, participation on a board of religious education, and through working on a local traffic committee. He stated that he had obtained names of twenty or thirty Negroes from two elevator operators in the federal building and sixty or sixty-five from a Negro to whom he had been referred. As to whether the names of other Negroes were submitted, Snodgrass testified that "there were probably colored persons' names given to me that I didn't know. I wasn't inquiring as to their color * * * color made no difference." If Snodgrass knew that an individual was a Negro either from the answer to a question regarding race on the questionnaire or otherwise, the file card of that individual would be marked "Col'd", and he testified that this was done to insure that there would always be more than one Negro on any panel having Negroes "so that they might be together, if they chose to be, to go out to lunch together." In regard to obtaining the names of individuals of particular religions, Snodgrass testified that he contacted no Jewish or Catholic organizations.

In choosing the names from the master list to go into the jury wheel Snodgrass testified that his selection practices were generally as follows: "Well, after I saw them, perhaps age, perhaps I would see [ sic], `well, this man is a farmer, and this man has a store in Waynesburg, and this man is a miner; that is fairly good; I will take those three men.' That is how I would decide." After a person had served on the jury, his card was sometimes marked by Snodgrass to indicate whether he was a "good juror". Snodgrass testified as to his criteria of a "good juror" as follows: "Regular in attendance, that didn't come up and be drunk, or talk in the halls, or do things like that, * * *." In choosing the names of former jurors to go into the jury wheel, these comments were considered by Snodgrass.

Snodgrass was followed in office by William L. Potts who served as jury commissioner from December 21, 1950 to January 10, 1952. Snodgrass had left approximately three to four thousand names in the file, and in order to bring that file up to date, new questionnaires were sent out to all of these. About thirteen hundred replies were received, and approximately that number were retained in the file by Potts. In obtaining new names Potts in substance followed the system that had been used by Snodgrass. Approximately one to two hundred new names were obtained by personal contact with various acquaintances, and about two hundred names were secured by correspondence with acquaintances. Typical of the sources with which Potts corresponded to get names are the following: two names were obtained from a friend who was the manager of a building; an unknown number were gotten from the manager of the local baseball team; an unspecified number were procured from Potts' brother-in-law who was the superintendent of maintenance at a plant; eight names were given Potts by a college classmate whose husband was an engineer; eleven names were obtained from a farmer known through business; and three names were secured from a friend who was a retired accountant. In addition, names were obtained from the American Legion Post of which Potts was a member, from state and federal court judges in the District, postmasters, and like sources. From one hundred to one hundred and fifty names were also obtained from volunteers. All in all, four to five hundred names were procured by these means. Three hundred of these persons replied to questionnaires sent them, and two hundred and fifty were deemed qualified and were added to the list. This made a total list of approximately fifteen hundred and fifty prospective jurors when the new names were added to the thirteen hundred names carried over from Snodgrass.

Potts testified that he relied on acquaintances as the sources of names because practically it was the most effective way to obtain jurors who were willing and able to serve, and when examined as to whether he sought particular types of jurors, Potts replied: "I think you are trying to put it to me as if I ran around looking for various people, which is wrong. It is just like I would walk out on the street here and some man would say, `Can I get on this jury?' or `How do you do it?' And that is the way these contacts were made. I didn't personally put myself out and run all over the place looking for them. I didn't have the time, to begin with." Once a name was obtained, the decision as to whether the individual was added to the list in no way depended on the source from which the name had been obtained but was based solely on the information in the questionnaire. Potts testified that he did not solicit names from nationality group organizations or from labor unions, but he did say that a majority of the one to two hundred people personally contacted were union members. Likewise, Potts did not ask for names from Negro organizations but did obtain the names of Negroes from individual Negroes he knew.

After Potts, Dorothy Kopp Marshall next took office as jury commissioner on March 22, 1952. The names of one hundred and ninety individuals to whom questionnaires had not been sent were left her by Potts, the names having come from lists submitted by postmasters and a firemen's relief association, and she sent questionnaires to these, adding those qualified to the lists. At the time of hearing, she had sent out twenty other questionnaires, and two of these went to individuals suggested by the Chief Judge of the court below, two to individuals who had volunteered, and the remainder to her personal acquaintances. She testified that she had contact with Negroes and members of unions through the department store where she worked.

Maintaining the separate master list required by statute to be kept by the clerk of court or his deputy were Stephen P. Laffey, appointed clerk on July 31, 1950, and Genevieve M. Barr, designated as Laffey's deputy at about the same time. Approximately three thousand names had been inherited from Laffey's predecessor, and new questionnaires were sent to all of these in order to bring the file up to date. These questionnaires were returned by approximately eighteen hundred persons, and after three to four hundred of these were eliminated because of failure to meet the necessary qualifications, there remained from the old list approximately fourteen to fifteen hundred names. This list was subsequently increased to approximately forty-five hundred names by securing new lists of individuals which would be, as Barr said, "generally what I would think was representative of the community." New names were obtained in the usual fashion from personal acquaintances and from volunteers. In addition, she estimated that ten names were obtained from employees in the local United States attorney's office, three to five hundred from state and federal judges and court employees, one hundred and fifty from chambers of commerce and boards of trade, five hundred from postmasters, one thousand from veterans' organizations, and one thousand from former jurors who were usually asked as a group to submit names. In regard to solicitation designed to obtain the names of labor union members, Barr testified that in 1950 a number of form letters requesting names were sent to union leaders but, with the excepion of one answer received a year and a half later, no replies were received. From then until the time of the hearing, thirty-seven unions were solicited and replies containing sixty-seven names were received from six of these unions, nineteen of the names being from a teachers' union and twenty-five from a teamsters' union. Because of the lack of response from unions, two to two hundred and fifty names of union members were obtained from personal acquaintances, in addition to the union people on the lists submitted by postmasters and veterans' organizations. In regard to solicitation from Negro groups, seventy-eight names were obtained from twenty-nine Negro organizations contacted, plus the names of Negroes that were obtained through the other sources utilized. No Jewish or Catholic organizations were solicited. When lists of petit jurors were drawn from the master lists thus compiled, they were submitted to the local United States attorney's office so that the individuals could be checked to determine if they had a criminal record.

In addition to these witnesses administering the jury selection process, a defendant in the Nelson case, William Albertson, testified as to statistics that he had compiled showing a disproportion between the representation of certain groups on particular jury lists and their representation in the population as a whole, as revealed by the census. In evaluating one particular panel of thirty-three members according to classifications that he had devised, Albertson demonstrated that fifty-seven percent of the total work force in the Pittsburgh standard metropolitan area, comprising four of the eighteen counties of the district, were manual workers, whereas only twenty-seven percent of the panel, including those whose spouses were manual workers, were in that group. On the other hand, the employer-management class constituted only thirty-four percent of the work force but had a fifty-four percent representation, including spouses, on the panel. The remaining eight to nine percent of the work force and the eighteen to nineteen percent of the panel were characterized by Albertson as "non-manual" and included sales personnel, retired persons and those whose occupation was unknown. In making his analysis according to other classifications, Albertson showed that seven percent of the area work force were government employees while twenty-four percent of the panel, including spouses, were government employees, that only eighteen percent of the panel came from what he termed the basic industries of that area, and that seven and one-half percent of the area population were Negroes while only six percent of the panel were Negroes. Albertson's analysis of other panels revealed sufficiently similar proportions, so that the foregoing is fairly illustrative of his study and obviates any necessity of setting out here his other statistics. It should be noted, nonetheless, that Albertson's statistics showed substantially lower proportions of Negroes on some of the other panels. In evaluating these statistics along with the other evidence it must be recognized that Albertson was not qualified as an expert. Moreover, the correctness of some of his classifications, samples, and figures may well be questioned. Nevertheless, since he explained his methods and they do not seem to us to be fundamentally faulty, we shall accept his statistics as evidence where pertinent and helpful to the court.

III.

While a litigant seeking a new trial on a challenge to the array of jurors must produce evidence to support the challenge, Glasser v. United States, 1942, 315 U.S. 60, 87, 62 S.Ct. 457, 86 L. Ed. 680, that evidence may be confined, as here, to the general method of jury selection, without showing that the particular jury that tried the case was deficient and, consequently, without establishing prejudice to the particular litigant usually prerequisite for invoking remedial action. Ballard v. United States, 1946, 329 U.S. 187, 195, 67 S.Ct. 261, 91 L.Ed. 181. But cf. Northern Pacific R. Co. v. Herbert, 1886, 116 U.S. 642, 646, 6 S.Ct. 590, 29 L.Ed. 755 . Presumably a particular jury of the most desirable type might be drawn from a list that was on the whole selected by faulty methods. But, as a practical matter, it is so difficult to formulate and administer a system prescribing the composition of individual juries that the law has placed the emphasis on insuring a fair system of general juror selection which in operation will normally result in adequate individual juries. It is consequently the general system of selection that is susceptible to successful attack by a litigant. See Thiel v. Southern Pacific Co., 1946, 328 U.S. 217, 220, 66 S.Ct. 984, 90 L.Ed. 1181. Moreover, by allowing the general method of selection to be questioned in any case, the complaining party serves as a helpful spur to the courts in their supervision of the administration of justice. See Ballard v. United States, supra, 329 U.S. at page 195, 67 S.Ct. at page 265.

When the question of jury selection methods has thus been raised in a federal court the standards to be applied in determining the validity of those methods is dependent on the source of the judicial power for the solution of the particular case. In those cases coming from the state courts concerning the administration of the state jury system, the Fourteenth Amendment fixes the limits of federal action, and the issue, as it is usually stated, is whether the state has intentionally and systematically excluded any legally cognizable group from its juries. See, e.g., Fay v. New York, 1947, 332 U.S. 261, 284-285, 291, 67 S.Ct. 1613, 91 L.Ed. 2043; Pierre v. Louisiana, 1939, 306 U.S. 354, 59 S. Ct. 536, 83 L.Ed. 757; Martin v. Texas, 1906, 200 U.S. 316, 26 S.Ct. 338, 50 L. Ed. 497.

In cases involving the jury systems of the federal courts, however, the applicable standards are not necessarily the same, for those standards depend on a different source of power. Brown v. Allen, 1953, 344 U.S. 443, 473, note 23, 73 S.Ct. 397, at page 416, 97 L.Ed. 469. There are of course constitutional provisions fixing the requirements for jury systems of the federal courts. See Article III and Amendments VI and VII of the Constitution. And in interpreting those provisions, the flexible nature of the Constitution permits the federal courts, as the Supreme Court has said, to recognize in some degree that "[o]ur notions of what a proper jury is have developed in harmony with our basic concepts of a democratic society and a representative government." See Glasser v. United States, supra, 315 U.S. at page 85, 62 S.Ct. at page 472. Nevertheless, when we examine the common law at the time of the adoption of the Constitution as an aid in the interpretation of those provisions, it is evident that the concepts of a properly constituted jury were limited at that time. Note, 60 Harv.L.Rev. 613, 616 (1947). Consequently, instead of relying on the relevant constitutional provisions, the Supreme Court in prescribing the principles of jury selection for the federal system has chosen to utilize its supervisory power over the administration of justice in the federal courts so as "with greater freedom to reflect * * * [its] notions of good policy." Fay v. New York, supra, 332 U.S. at page 287, 67 S.Ct. at page 1627. In the exercise of that power, Mr. Justice Murphy stated the guiding principles underlying the standards for jury selection to be as follows: "The American tradition of trial by jury, considered in connection with either criminal or civil proceedings, necessarily contemplates an impartial jury drawn from a cross-section of the community. * * * This does not mean, of course, that every jury must contain representatives of all the economic, social, religious, racial, political and geographical groups of the community; frequently such complete representation would be impossible. But it does mean that prospective jurors shall be selected by court officials without systematic and intentional exclusion of any of these groups. Recognition must be given to the fact that those eligible for jury service are to be found in every stratum of society. Jury competence is an individual rather than a group or class matter. That fact lies at the very heart of the jury system. To disregard it is to open the door to class distinctions and discriminations which are abhorrent to the democratic ideals of trial by jury." See Thiel v. United States, supra, 328 U.S. at page 220, 66 S.Ct. at page 985.

In the Thiel case, as in many of the other decisions involving issues as to jury selection, there was uncontradicted evidence of intentional, systematic exclusion of certain groups by the responsible jury officials so that it was unnecessary in the actual determination of the case to go further than to apply the classic test of "intentional and systematic exclusion." However, in the quotation from the Thiel case Mr. Justice Murphy said that it is "contemplated" that jury lists are to be so compiled as to be representative of a cross-section of the community qualified for service and that "[r]ecognition must be given to the fact that those eligible for jury service are to be found in every stratum of society." If that be so, then it would seem that more must be required of jury officials than merely that they not "intentionally and systematically exclude" any groups. It follows, therefore, that it must be the duty of jury officials, at least in the federal system, not to exclude any cognizable groups from jury lists through neglect as well as through intentional conduct.

In other decisions, there seems to have been some recognition of the validity of requiring such a standard of officials selecting prospective jurors. In Glasser v. United States, supra, 315 U.S. at pages 85-86, 62 S.Ct. at page 472, it was said: "[The exercise of the duty of selection] must always accord with the fact that the proper functioning of the jury system, and, indeed, our democracy itself, requires that the jury be a `body truly representative of the community', and not the organ of any special group or class. If that requirement is observed, the officials charged with choosing federal jurors may exercise some discretion to the end that competent jurors may be called. But they must not allow the desire for competent jurors to lead them into selections which do not comport with the concept of the jury as a cross-section of the community. Tendencies, no matter how slight, toward the selection of jurors by any method other than a process which will insure a trial by a representative group are undermining processes weakening the institution of jury trial, and should be sturdily resisted. That the motives influencing such tendencies may be of the best must not blind us to the dangers of allowing any encroachment whatsoever on this essential right. Steps innocently taken may one by one lead to the irretrievable impairment of substantial liberties." In some Fourteenth Amendment cases, the Supreme Court has held it to be a wrongful exclusion for jury officials to choose names only from their acquaintances when it was demonstrated that they had no acquaintances among a group that should have been included. Cassell v. Texas, 1950, 339 U.S. 282, 70 S.Ct. 629, 94 L.Ed. 839. See Hill v. Texas, 1942, 316 U.S. 400, 62 S. Ct. 1159, 86 L.Ed. 1559 and Smith v. Texas, 1940, 311 U.S. 128, 61 S.Ct. 164, 85 L.Ed. 84. Thus, although there was apparently no basis for finding that a particular group had been intentionally excluded in those cases, new trials were ordered because of the failure of the officials to obtain an adequate cross-section representation. In connection with this point, it should be stated that the Manual for Clerks of the United States District Courts (1950), paragraph 9.03, provides: "In order that grand and petit jurors to serve in United States district courts may be truly representative of the community, the sources from which they are selected should include all economic and social groups of the community."

Thus, the present-day desideratum of jury selection is the obtaining of lists of individuals representative of a cross-section of the qualified community and it seems clear that it is the duty of the jury officials to use methods reasonably designed to attain that end. Undoubtedly, these officials have discretion in their choice of methods, but it must be exercised with this purpose in view. When in any court of this circuit there has been a failure to do this, either intentionally or through neglect, this court in the exercise of its supervisory power must require a new trial so that the error will be obliterated. It is not significantly more difficult to make a finding of negligent practices resulting in exclusion than to make a finding of intentional exclusion; the same reasons for requiring a retrial when there has been intentional exclusion exist when there has been an exclusion by reason of neglect; and any greater interference in the course of litigation caused by enforcing such a requirement is justifiable because it is necessary to insure fair trials.

In order for any such failure of duty to be remediable, the group excluded must of course be a cognizable one, and in the Thiel case, supra, 328 U.S. at page 220, 66 S.Ct. at page 985, it was recognized that there might be classifications along economic, social, religious, racial, political, or geographical lines. In the challenge to the jury here the plaintiff complains of the exclusion of "foreign born citizens, Negroes, veterans, labor people and many others." Insofar as foreign born citizens are concerned, there was testimony that names were not solicited from nationality organizations; otherwise, there was no evidence adduced to demonstrate that it was not reasonable to assume that the names of foreign born citizens were obtained from the other sources solicited and that there was not a fair representation of these citizens. Consequently, this lack of evidence of exclusion makes it unnecessary to decide whether our foreign born citizenry constitutes a separate classification for which the selection systems must be designed to obtain representation. Regarding veterans, there was evidence which clearly established that there was no exclusion of them, so that it is also unnecessary to determine if they are a cognizable category.

In regard to the Negroes and "labor people" referred to in the motion and the religious groups about which there was considerable testimony at the hearing, the evidence presents much greater problems. There is no doubt that each is a classifiable category of our society which cannot be wrongfully excluded. Section 1863, 28 U.S.C. specifically prohibits exclusion on account of race or color. In the Thiel case, supra, 328 U.S. at page 220, 66 S.Ct. at page 985, Mr. Justice Murphy stated that there could be no exclusion along economic lines. He also referred to religious groups as a category that could not be lawfully excluded. Consequently, we must determine if the plaintiff has established any neglect of duty on the part of the officials charged with jury selection in not using reasonable methods designed to obtain the representation of these groups.

IV.

In the case of Negroes, the testimony of the jury officials revealed that each had solicited the names of Negroes in one way or another. Although some officials may have used more effective or vigorous methods of solicitation than others, each official was aware of this significant racial segment of our population, and each, in the proper exercise of his discretion, devised his own methods to insure their representation. The effectiveness of these methods is shown by the statistical data. Albertson's study revealed a variation from approximately six to two percent in the representation of Negroes on the panels used by him and the plaintiff's reply brief concedes that the Negro population in the whole Western District was only four and four-tenths percent of the total population as against the seven and one-half percent of the Pittsburgh sample area used by Albertson. Therefore, on the basis of the evidence before us we cannot say that these methods, considering the practical problems involved, were unreasonable.

Snodgrass testified that he marked his file cards to identify the Negroes on his list. In Avery v. Georgia, 1953, 345 U.S. 559, 562, 73 S.Ct. 891, 892, 97 L.Ed. 1244, where cards of a distinctive color were used for Negroes, it was held that discrimination could be inferred in the absence of other evidence, since "that practice makes it easier for those to discriminate who are of a mind to discriminate." However, in different circumstances, the contrary inference, that the practice was utilized to insure a fair representation of Negroes, has been held to be proper. United States v. Dennis, 2 Cir., 1950, 183 F.2d 201, 223. In this case, Snodgrass explained that the cards were marked to enable him to be sure that more than one Negro served on any panel having Negroes in case they might desire the companionship of one another. Therefore, it cannot be inferred here that the practice was followed for the purpose of exclusion.

In regard to the claim concerning the representation of those from the lower economic levels of our society, there is no evidence of exclusion of poor people, manual workers, or wage earners, whatever be the proper designation of that group. In fact, each jury official denied any discrimination against this class, and each did in some way directly solicit the names of "manual workers" as they were called by Albertson. Moreover, there has been no evidence to show that the assumption made by some of the jury officials, that the names of manual workers were obtained through other sources of solicitation, was erroneous. Labor unions would indeed be an obvious source of such names but the evidence demonstrates the meager results of soliciting names from unions. Although Albertson's statistics show some disproportion between the number of manual workers in the population and their number on the jury lists, there was nevertheless a substantial representation of manual workers on the lists. And what disproportion there was may well be due to fewer manual workers returning their questionnaires because of the greater financial hardships of jury duty on them. Consideration of all the evidence demonstrates that it is insufficient to show that the methods employed for obtaining the representation of this class were unreasonable.

At the hearing inquiry was made as to whether there had been any solicitation from Catholic or Jewish organizations, and the jury officials replied that there had been none, though the names of Catholics and Jews were contained on the lists. Snodgrass did solicit names from the Episcopal Church, which solicitation presumably resulted in the addition of Episcopalians to the lists. However, there is no evidence that the effect of this was to slant significantly the lists along religious lines, since the number of names thus obtained is unknown, and there is nothing to indicate that the other officials solicited from sources that submitted lists weighted with members of one religion rather than another. Thus, on the whole, the showing is insufficient to establish that the methods were unreasonable.

In the course of the hearing, there was also evidence adduced concerning the solicitation of names from government sources and the existence of a large proportional representation of government employees on the jury lists. It must be recognized, of course, that the effect of including too many of one group may well be the exclusion of another, so that a large representation of government workers might result in the exclusion of economic or political groups. In this case the cooperation of public officials in submitting the names of their employees and the availability of government workers were shown as a practical matter to be important factors in the inclusion of so many government employees in the lists. Convenience of course cannot excuse deficient methods. Practical considerations nonetheless are relevant in a determination of the reasonableness of methods and there is insufficient evidence to show that the inclusion of so many government workers resulted from unreasonable means operating to exclude other cognizable groups.

V.

We must now comment on the arguments made by the plaintiff concerning the practices that were very generally followed by all of the jury officials in this case and concerning the burden imposed on the plaintiff in sustaining an attack on those practices. First, as we have said, the burden of proving wrongful exclusion is upon the party attacking the validity of the system, and this of course means that the plaintiff must show that the methods used were not reasonably designed to obtain representative lists. In many instances, such as those involving religious groups, the plaintiff in substance has shown only that the officials did not solicit religious sources directly for religious representation. However, to establish neglect of duty on the part of the jury officials, the plaintiff must show that a proper representation of cognizable religious groups could not reasonably be expected to have been obtained by a random solicitation from the other sources utilized.

Likewise, other general practices of which the plaintiff complains fail to require a reversal because the plaintiff has not shown them to be unreasonable in the light of the objectives of jury selection. This is true of the plaintiff's argument that it was wrong for the jury officials to solicit from organizations and from among their friends. The jury officials could in the exercise of their discretion resort to others for the suggestion of names so long as it could reasonably be expected that a cross-section would thereby be obtained, and using the names taken from such a variety of sources does not result in an unlawful delegation of duty by the officials. The Supreme Court in the Glasser case, supra, did condemn the taking of names of any group from only one private organization. But here many sources were utilized, and there is no showing that any one source was so relied on that it improperly weighted the list. See United States v. Dennis, supra, 183 F.2d at page 218. The plaintiff, however, goes further in her argument and maintains that the only proper method for compiling jury lists is by taking names in some set system from the voting lists. But, as we have already said, the choice of the methods to be used lies with the jury officials. In arguing that the use of voting lists is the only reasonable method, the plaintiff has not even shown that the desideratum of a cross-section would be attained thereby. It may be that some cognizable groups generally fail to register to vote, so that such a method would be inherently faulty. Consequently we cannot adhere to the plaintiff's contention.

The plaintiff also complains of the use of the names of volunteers. On occasion, this practice has been criticized because of the chance that certain persons may volunteer for improper purposes. See the Report of the Committee on Juror Selection to the Judicial Conference of the United States, pp. 25-26 (1942), and the action of the Conference thereon, Report of Chief Justice Stone, Sept. Sess. 1943, p. 15. However, here there has been no evidence of this or that taking volunteers operated to exclude any cognizable group. Consequently, this contention must also be dismissed, although we state that the volunteer method is not a desirable one and should no longer be employed in this circuit.

In addition, the plaintiff argues the unlawfulness of the solicitation of names from former jurors on the ground that this "intensifies the discrimination originally practiced and confines the list to this inbred class of persons and their friends." But, we have ruled that in initially obtaining names from outside sources no exclusion was practiced. If it can be assumed, as the plaintiff suggests, that the new names reflect the types of the former jurors who submitted the names, then the new names obtained by this method would surely be representative of a satisfactory cross-section. But we do not think that our decision should rest on any such assumption. The plaintiff's contention must be dismissed because she has failed to demonstrate the unreasonableness of the method. In United States v. Dennis, supra, 183 F.2d at page 217, Judge Learned Hand, on the basis of the Glasser case, indicated that the use, as a "principal source," of names submitted by a Federal Grand Jury Association which recommended individuals solely on the basis of social and educational standing was unlawful. But here the use of names suggested by former jurors was not the same since the jurors submitted the names individually and since the individuals whose names were submitted were not shown to have been screened to be of only a particular type, thereby excluding other types. Most of the jury officials asked the panels in a group to submit individually the names of prospects. It is true that Snodgrass did solicit names only from former jurors whom he thought "seemed intelligent." Nonetheless he was warranted in using some discretion in choosing his sources and such an exercise of discretion by him has not been shown to have resulted in any exclusion.

The plaintiff also contends that it was improper for Snodgrass to exercise any choice in picking names from the master list to be put into the jury wheel. Snodgrass testified that he had to examine the information concerning the individuals whose names were put in the wheel to determine whether they were from the proper localities or had served before recently. By statute, jurors may be apportioned geographically within a district and may be subject to challenge if they have served previously within one year. 28 U.S.C. § 1865 and 1869. Snodgrass testified that while ascertaining this information as was thus appropriate he did discretionarily choose names so that he would get a diverse representation. Section 1864, 28 U.S.C. prescribes that "[t]he jury box shall from time to time be refilled by the clerk of court, or his deputy, and a jury commissioner * * *" and that "[t]he jury commissioner and the clerk, or his deputy, shall alternatively place one name in the jury box without reference to party affiliations * * *" But the method for picking the names to be placed in the wheel is not otherwise specified, and it would seem that a good means for obtaining a cross-section is the very method used by Snodgrass. The plaintiff's argument thus fails.

The plaintiff also contends that it was improper for Barr to send the master lists of names to the United States attorney's office, but this argument is completely without merit. Barr testified that this was done only to check the criminal record of the individuals, which directly pertained to the statutory disqualification of convicted felons, and using the United States attorney's office for this purpose was not an improper procedure requiring reversal.

Aside from the arguments as to specific methods of selection, the plaintiff makes much of some disproportions between the composition of the jury lists and the population as a whole, as shown by the statistics of Albertson. Such statistics comparing the proportion of cognizable groups on the jury lists to their proportion in the population may be useful as a basis for drawing inferences as to whether or not there has been unlawful exclusion. United States v. Flynn, 2 Cir., 1954, 216 F.2d 354, 384. When appropriate, we have thus considered these statistics along with all the other evidence. However, this is the limit of their usefulness, since proportional representation is clearly not required. See, e.g., Fay v. New York, supra, 332 U.S. at page 276, 67 S.Ct. at page 1622 and United States v. Dennis, supra, 183 F.2d at page 222. Proportional representation is a desideratum, but the reach of our review is only to inquire into the reasonableness of the methods employed to achieve it. We find the methods employed not to have been unreasonable.

VI.

There need be comment only on one other objection of the plaintiff to the selection methods used. In the questionnaires sent to persons whose names had been obtained, there was the following question: "Are you opposed to the American form of government?" The use of a questionnaire is not in itself improper, since the jury officials must have some means of acquiring the necessary information to carry out their functions. Wilson v. United States, 5 Cir., 1939, 104 F.2d 81. The plaintiff, however, in effect complains that the inclusion of this question was improper since one may lawfully oppose the form of our government and, therefore, that the selection process was faulty in excluding all those who do oppose the form of government of the United States.

The meaning of the question is far from clear. Its content seems not to have been thought through. The term "form" is broad indeed. Opposition to the "form" of our government may have either a lawful or unlawful semblance. For example, one who seeks to change the form of our government by constitutional means, as for example does the proponent of a proposed constitutional amendment, is entitled to do so. On the other hand those who seek to overthrow the government of the United States by force and violence may incur the criminal sanctions of the Smith Act as did the defendants in United States v. Mesarosh, decided by this court concurrently with the instant case. As phrased the question is so broad as to be objectionable. But no matter what its meaning is, here it has resulted in the exclusion of such an infinitesimally small group that it cannot serve as the basis for a new trial. See United States v. Dennis, supra, 183 F.2d at page 217. At the hearing, Snodgrass testified that an affirmative answer to the question would have been considered in the selection process but that he did not remember ever seeing a "yes" answer. Potts testified that he did exclude one or two persons who submitted such an answer, though there may have been as many as ten. Barr said that she would have eliminated any person answering yes and that she may have had one such answer. Thus, at best only eleven persons out of the many thousands who were sent questionnaires were excluded on this basis. Nevertheless, the plaintiff argues that many more people may have been excluded because those who would have answered in the affirmative may have been deterred from returning the questionnaires. Yet this is an assumption for which the plaintiff has shown no convincing basis, and therefore we do not consider it a sufficient ground for setting aside a jury verdict.

But the plaintiff contends further that the very fact that the question was asked placed jurors under such a restraint, or so intimidated them, that they were unable to exercise free and independent judgment in arriving at a verdict. This argument goes to the specific issue of intimidation discussed below. The plaintiff places reliance on the case of Gideon v. United States, 8 Cir., 1931, 52 F.2d 427, 428-429, in which it was held that certain questions on a questionnaire mailed to prospective jurors with the summons for jury duty, as set out in the margin, were "reprehensible" and constituted "unauthorized" "[s]ecret preliminary questioning" which, trenching on the "broad ground of fair trial", constituted adequate ground for reversal in a case in which the defendant, among others, was charged with a conspiracy to violate the National Prohibition Act and certain Internal Revenue statutes.

One of the questions in the Gideon case went to the religion of the prospective juror; another, as we have seen, asked him whether he was in favor of "Prohibition". The Court of Appeals stated that "* * * they [the jurors] must have been led to believe that the government had some purpose in asking questions about their beliefs, and was keeping a record of the answers for future use; they doubtless were led to believe also that, in the minds of the government officials, at least, their usefulness as jurors was in some way affected by the beliefs about which inquiry was made; and it is not impossible that they were led to think that the government intended to influence them in their beliefs." In the Gideon case the impact of the questions was immediate to the trial for the questionnaire went with the summons to jury duty. A court may impose penalties on a juror for failing to obey a jury summons and it is quite possible that the average juror receiving a questionnaire with a summons might conclude, not unreasonably however incorrectly, that if he failed to answer the questionnaire sanctions similar to those which could be imposed by the court on his failure to respond to the summons might also be imposed upon him. This element is lacking in the instant case for the questionnaires preceded any jury summons by a substantial amount of time. Moreover, whatever may have been the situation in the Gideon case the court below in the instant trial, as in the trial in United States v. Mesarosh, fully and properly informed the jury as to their sworn duty and we cannot conclude, in the absence of any proof whatsoever, that any juror would act in such abnegation of his oath as to permit his judgment to be swayed by the single question objected to by the plaintiff. Moreover, under such circumstances the existence of prejudicial error may not be presumed. It follows that this contention of the plaintiff cannot serve as a basis for setting aside a jury verdict.

VII.

In the supplemental motion to strike the jury panel, the plaintiff alleged that "some of the members of the jury panel have been under surveillance and investigation by unauthorized persons, and jury investigation lists have been used by counsel." To substantiate this allegation, evidence was adduced that a jury investigation service was in existence in the Western District. When jury lists were made up, the investigator obtained copies from various lawyers who gave them to him. The investigator then proceeded to converse, usually by telephone, with the various jurors' friends and neighbors, whose names were obtained from street directories. The investigator testified as to the extent of his questioning of those whom he contacted: "I ask the neighbor if they know this particular person that is called, and if it is a man where he is employed, if he is married, if he has any children, whether they'd be old enough to work, and if they did work if the neighbor would know where they worked; if they were younger, were they of school age or still younger than that; about how old the man was; if he owned his own property there; if he knew if he ever had any cases in court litigation; if he had ever been hurt in any accident; and if they knew their politics or religion. That's practically everything." The information thus obtained was compiled in lists that were sold to lawyers.

In support of the argument that this type of activity is unlawful and vitiates any trial where it has occurred, the plaintiff relies on Sinclair v. United States, 1929, 279 U.S. 749, 49 S. Ct. 471, 73 L.Ed. 938. That case involved a criminal contempt conviction of a litigant who had hired a corps of private detectives to keep under surveillance every member of the jury during the trial of the litigant's case. When the plaintiff in this case obtained a remand on the prior appeal, it was said: "This court quite realizes the importance of protecting jurors from intimidation. The Sinclair case was cited to us. It is not in point here on the facts. But the general proposition that a jury must be free to decide cases submitted to it on the law and the facts without outside pressure is one to which we heartily subscribe." Dow v. United States Steel Corp., supra, 195 F.2d at

We now repeat that the Sinclair case is not directly in point but that outside pressure on jurors is of course improper. As was said in Sinclair, supra, 279 U.S. at pages 764-765, 49 S.Ct. at page 476, activities which "destroy the equilibrium of the average juror and render impossible the exercise of calm judgment upon patient consideration" tend "to obstruct the honest and fair administration of justice" and cannot be tolerated. And as was held in Sinclair, there does not have to be direct contact with the jurors for the activity to be improper so long as the "reasonable tendency of the acts done" is to intimidate.

In this case, however, it has not been shown that the "reasonable tendency" of the activities of the investigator service was to intimidate. Its purpose was not to pressure jurors but to inform lawyers. Litigants are given the right to challenge jurors. In order to exercise that right intelligently in densely populated areas where the realm of one's own acquaintances is necessarily limited, a reasonable investigation of jurors is helpful, and consequently such investigatory procedures are widely used. It has not been demonstrated that such investigations, whereby a prospective juror may know that someone working for lawyers has asked questions about him, would normally have a tendency to intimidate the juror, even if it could be assumed that the knowledge by a juror of such an investigation would work for the benefit of one party rather than another in a case such as this. And it has not been shown in this case that the effect of these procedures was to intimidate or to "pressure" jurors. There is an obvious and determinative difference between the investigations conducted here and the constant surveillance of jurors in the course of trial that took place in the Sinclair case.

Other grounds asserted as a challenge to the array do not require discussion. We find no merit in the challenge.

Accordingly, the judgment of the court below will be affirmed.